JNB2-Simulations.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "879c6463-39c1-42a5-9c8c-23ba1ce0ec23",
   "metadata": {},
   "source": [
    "# JNB2 - Run DFT simulations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Welcome to this Jupyter Notebook, where we will run the DFT simulation. \n",
    "\n",
    "This notebook represents the second step of the protocol discussed in the associated article (see the README.md file) as illustrated in the Figure 2 below.\n",
    "\n",
    "The notebook is organized into sections based on the type of manipulations applied to the crystals. Here is an overview of the sections:\n",
    "\n",
    "* [Unit Cells `Single Point`](#unit_cells_sp) \n",
    "* [Unit Cell `Geometry Optimization` ](#unit_cells_go) \n",
    "* [Pure Species `Single Point`](#pure_species_sp)\n",
    "* [Pure Species `Geometry Optimization`](#pure_species_go)\n",
    "* [Supercells `Single Point`](#supercells_sp)\n",
    "* [Supercells `Geometry Optimization`](#supercells_go)\n",
    "* [Strain `Single Poin`](#strain_sp)\n",
    "* [Vacancies `Single Poin`](#vacancies_sp)\n",
    "* [Substitution `Single Poin`](#substitution_sp)\n",
    "* [Interstitial `Pre Optimization`](#interstitial_sp) \n",
    "* [Interstitial `Geopmetry Optimization` ](#interstitial_go)\n",
    "* [Slab `PRE Optimization`](#slab_po) \n",
    "* [Slabs `Geopmetry Optimization` ](#slab_go)\n",
    "* [Amorphous `ab initio MD`](#amorphous_aimd_) \n",
    "* [Amorphous `Single Point`](#amorphous_sp) \n",
    "* [Bulk 300 K `ab initio MD`](#300_aimd) \n",
    "* [Bulk 300 K `Single Point`](#300_sp)\n",
    "* [Bulk 500 K `ab initio MD`](#500_aimd) \n",
    "* [Bulk 500 K `Single Point`](#500_sp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<figure>\n",
    "    <img src=\"assets/img/wf-JNB2.png\" alt=\"JNB2 workflow\">\n",
    "    <figcaption><strong>Figure 2</strong>: Protocol workflow and part covered by this Jupyter Notebook highlighted in red.<figcaption>\n",
    "<figure> "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "eedbeed9",
   "metadata": {},
   "source": [
    "## Settings"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ca235714",
   "metadata": {},
   "source": [
    "In this section, we define the constants and generators that will be used throughout the notebook:\n",
    "\n",
    "* `MAIN_DIR_JNB2`: The main folder where the results of this Jupyter Notebook will be stored.\n",
    "* `CHEM_FORMULA`: The chemical formula of the SEI component.\n",
    "\n",
    "These constants and generators play a crucial role in ensuring reproducibility and providing the necessary inputs for various computations and simulations in this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "72a5e2bf-e3e3-4be1-9397-a070d3ab4f03",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tools.misc import MAIN_DIR_JNB1, MAIN_DIR_JNB2\n",
    "\n",
    "CHEM_FORMULA = \"LiF\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "904118a7-4cd2-4f7d-8012-adb331fca768",
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "import os\n",
    "import pickle\n",
    "\n",
    "import matplotlib as mp\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scm.plams as scm\n",
    "from ase.visualize import view\n",
    "from pymatgen.core import Structure\n",
    "from pymatgen.io.ase import AseAtomsAdaptor\n",
    "from scm.plams.interfaces.molecule.ase import fromASE, toASE\n",
    "\n",
    "from tools import make_dir\n",
    "from tools.misc import mtx2str\n",
    "\n",
    "matget2ase = AseAtomsAdaptor()\n",
    "\n",
    "plt.style.use(\"default\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dba2252d",
   "metadata": {},
   "source": [
    "Making the output folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b060831d",
   "metadata": {},
   "outputs": [],
   "source": [
    "make_dir(os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d315bfda-2754-4dd5-8dc6-53fa00c6138c",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Unit Cells `Single Point` <a id='unit_cells_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "83080e52-7d0c-4373-8e21-78a1a84a4dce",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"0-UnitCell_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7dcc2caf-4161-4571-9e9a-10543054915f",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "8807a6ad-f46f-46eb-a061-0f6a9f861b96",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tyes\n",
      "                     Hessian: \tyes\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           KSpace: \t\n",
      "                  Quality: \tGood\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_0_SP\n",
    "\n",
    "print(settings_0_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "497feee2-f96d-41a9-9650-5cbb8b242cc4",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "1f268c8e-bf68-4b5c-a2ea-2a71c147fd9c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# HPC Job settings\n",
    "NODES = 2\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "008a37ac-e658-4df1-8166-8dad92ee4e84",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = r\"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3596a0e9-710d-4912-a8a5-209fa358877d",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_0_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d507f940-3542-4616-8d8c-e05f540ac9b5",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7bf3a6a0-8fc4-4f09-bf40-1975bfc03965",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "8509081c-1100-45c3-a306-beedc88e5449",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f1e24d3c-bcf7-49ad-8b5a-5c4a685bcb62",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "369f3e52-091e-430e-974a-63c308dc32f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"SP-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_0_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6f197d29-0335-4350-9995-5adf51d1fb24",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c659f33c-5fd6-485c-a5c3-73d679d06e45",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[05.07|14:45:18] JOB jobs STARTED\n",
      "[05.07|14:45:18] JOB jobs/SP-0-Li2CO3_P6_3mcm_-2.19 STARTED\n",
      "[05.07|14:45:18] JOB jobs/SP-1-Li2CO3_C2c_-2.26 STARTED\n",
      "[05.07|15:57:57] JOB jobs/SP-1-Li2CO3_C2c_-2.26 SUCCESSFUL\n",
      "[05.07|16:02:01] JOB jobs/SP-0-Li2CO3_P6_3mcm_-2.19 SUCCESSFUL\n",
      "[05.07|16:02:01] JOB jobs SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "568a48b2-927a-4457-b746-5fcb95f4b1d4",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9af7b355-89c3-44d1-84ac-b411fd99f94c",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "8f431ab7-fdbd-4868-8fb6-c658925c40cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "d4dc5d2f-118d-4a97-bd26-59138cb72f63",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_0_SP).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom_SP\"] = energy_per_atom\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "a4f2acfd-12aa-44d7-b77a-8f6434ad01e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>scm_energy_SP</th>\n",
       "      <th>scm_energy_per_atom_SP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-LiF_Fm-3m_-3.18.cif</th>\n",
       "      <td>-4.845146</td>\n",
       "      <td>-3.180409</td>\n",
       "      <td>-9.622935</td>\n",
       "      <td>-4.811468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-LiF_P6_3mc_-3.17.cif</th>\n",
       "      <td>-4.838474</td>\n",
       "      <td>-3.173737</td>\n",
       "      <td>-19.208830</td>\n",
       "      <td>-4.802207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-LiF_Pm-3m_-2.89.cif</th>\n",
       "      <td>-4.557603</td>\n",
       "      <td>-2.892867</td>\n",
       "      <td>-9.023544</td>\n",
       "      <td>-4.511772</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        energy_per_atom  ...  scm_energy_per_atom_SP\n",
       "0-LiF_Fm-3m_-3.18.cif         -4.845146  ...               -4.811468\n",
       "1-LiF_P6_3mc_-3.17.cif        -4.838474  ...               -4.802207\n",
       "2-LiF_Pm-3m_-2.89.cif         -4.557603  ...               -4.511772\n",
       "\n",
       "[3 rows x 4 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "a51233d1-e769-4b63-9b96-453ff235e528",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x280 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 2.8], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    idx = np.arange(len(results_df))\n",
    "    ax.bar(idx - 0.1, results_df[\"scm_energy_per_atom_SP\"], width=0.20, label=\"simulated\")\n",
    "    ax.bar(idx + 0.1, results_df[\"energy_per_atom\"], width=0.20, label=\"materials project\")\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(results_df.index, rotation=25, ha=\"right\")\n",
    "    # ax.legend(loc='center left', bbox_to_anchor=(1., 0.5))\n",
    "    ax.legend(loc=\"lower center\")\n",
    "    all_E = results_df[\"energy_per_atom\"].values.tolist() + results_df[\"scm_energy_per_atom_SP\"].values.tolist()\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.05, np.max(all_E) + np.abs(np.mean(all_E)) * 0.05)\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"0-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7a9c0386-8074-4e66-ae27-7f33ab133502",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "4c1b0b88-5349-4ba1-9b23-298b6954f186",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4ee6311d-d60f-407c-8c2e-0978718b0541",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Unit Cell `Geometry Optimization` <a id='unit_cells_go'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "3412c32d-d8bd-4ef3-abad-24b7264dcb38",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"0-UnitCell_GO\"\n",
    "TASK = \"GO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "61fecff7-741a-4bd8-af36-647668a0a563",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "19560fc0-4118-48f4-9d2a-56112dc58478",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tGeometryOptimization\n",
      "          Properties: \t\n",
      "                     Gradients: \tno\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tyes\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t500\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           SCF: \t\n",
      "               Iterations: \t150\n",
      "           RadialDefaults: \t\n",
      "                          NR: \t3000\n",
      "           CPVector: \t256\n",
      "           KGRPX: \t10\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_0_GO\n",
    "\n",
    "print(settings_0_GO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "844962d6-c43a-4dfd-9d42-6a284df33f40",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "a8169564-9fcd-4cce-a852-dba8f3aadbac",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d08248ee-2c7c-41fd-a932-f71d9141f5a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a641f306-e6e8-4a1f-b992-ab548ecbf2d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_0_GO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fd8bee4a-49a8-4054-b5c7-eb83d1b88d0f",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "be983f36-445f-47b4-a7d9-34d1f6ff5a6d",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "a7720f5f-1bad-4edb-8cc7-dce7d3705ac5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a243deb4-e528-4fdc-a579-e3677e4ca3ef",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e07980e9-1394-4864-abc1-e8a5639d791b",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_0_GO, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "export SCM_SHAR_NONUMA=yes\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1e75292f-2470-4db3-b917-b7a3db8ea44e",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9051c217-b75a-4038-9478-3e8af9cca4b7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[25.12|23:06:46] JOB jobs STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-0-C_R-3m_na4 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-1-Li_R-3m_na3 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-2-O_C2m_na8 STARTED\n",
      "[25.12|23:21:32] JOB jobs/GO-1-Li_R-3m_na3 SUCCESSFUL\n",
      "[26.12|15:45:03] Job jobs/GO-2-O_C2m_na8 reported warnings. Please check the the output\n",
      "[26.12|15:45:03] JOB jobs/GO-2-O_C2m_na8 SUCCESSFUL\n",
      "[26.12|23:13:11] JOB jobs/GO-0-C_R-3m_na4 FAILED\n",
      "[26.12|23:13:14] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d71fb9c7-6f69-4acd-9cde-e29d78a25d1a",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3ef3e7e2-c97f-4bf1-b647-453ac3ef9b2d",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "df83b517-9d7d-43b5-9a87-796676d0cd65",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "prev_sim = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, \"0-UnitCell_SP\")\n",
    "data_path = os.path.join(prev_sim, \"results.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "3a522423-b6c9-4858-81d0-f346d8f3fe37",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_0_GO).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_GO\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom_GO\"] = energy_per_atom\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "c5d7b8ae-39fe-4208-815d-91dccb259120",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>scm_energy_SP</th>\n",
       "      <th>scm_energy_per_atom_SP</th>\n",
       "      <th>scm_energy_GO</th>\n",
       "      <th>scm_energy_per_atom_GO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-LiF_Fm-3m_-3.18.cif</th>\n",
       "      <td>-4.845146</td>\n",
       "      <td>-3.180409</td>\n",
       "      <td>-9.622935</td>\n",
       "      <td>-4.811468</td>\n",
       "      <td>-9.625820</td>\n",
       "      <td>-4.812910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-LiF_P6_3mc_-3.17.cif</th>\n",
       "      <td>-4.838474</td>\n",
       "      <td>-3.173737</td>\n",
       "      <td>-19.208830</td>\n",
       "      <td>-4.802207</td>\n",
       "      <td>-19.214337</td>\n",
       "      <td>-4.803584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-LiF_Pm-3m_-2.89.cif</th>\n",
       "      <td>-4.557603</td>\n",
       "      <td>-2.892867</td>\n",
       "      <td>-9.023544</td>\n",
       "      <td>-4.511772</td>\n",
       "      <td>-9.024396</td>\n",
       "      <td>-4.512198</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        energy_per_atom  ...  scm_energy_per_atom_GO\n",
       "0-LiF_Fm-3m_-3.18.cif         -4.845146  ...               -4.812910\n",
       "1-LiF_P6_3mc_-3.17.cif        -4.838474  ...               -4.803584\n",
       "2-LiF_Pm-3m_-2.89.cif         -4.557603  ...               -4.512198\n",
       "\n",
       "[3 rows x 6 columns]"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "4dc9a993-c1f5-4dfe-971d-7416de516e99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x250 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 2.5], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    idx = np.arange(len(results_df.index))\n",
    "    ax.bar(idx - 0.2, results_df[\"energy_per_atom\"], width=0.20, label=\"Materials Project\")\n",
    "    ax.bar(idx + 0.0, results_df[\"scm_energy_per_atom_SP\"], width=0.20, label=\"BAND (SP)\")\n",
    "    ax.bar(idx + 0.2, results_df[\"scm_energy_per_atom_GO\"], width=0.20, label=\"BAND (GO)\")\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(results_df.index, rotation=25, ha=\"right\")\n",
    "    # ax.legend(loc='center left', bbox_to_anchor=(1., 0.5))\n",
    "    ax.legend(loc=\"lower center\")\n",
    "    all_E = (\n",
    "        results_df[\"energy_per_atom\"].values.tolist()\n",
    "        + results_df[\"scm_energy_per_atom_SP\"].values.tolist()\n",
    "        + results_df[\"scm_energy_per_atom_GO\"].values.tolist()\n",
    "    )\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.05, np.max(all_E) + np.abs(np.mean(all_E)) * 0.05)\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"0-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7344c5ff-78d7-4298-9c07-56c5423e3643",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "421124a7-6168-43ac-a890-075e995b6474",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ae7de68d",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Pure Species `Single Point` <a id='pure_species_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "67353e62",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"0.1-PureSpeices_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b12a72c8",
   "metadata": {},
   "source": [
    "#### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "03e8ae2b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tyes\n",
      "                     Hessian: \tyes\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           KSpace: \t\n",
      "                  Quality: \tGood\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_0_SP\n",
    "\n",
    "print(settings_0_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c80ac01f",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "d9112215",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 1\n",
    "CORES_PER_NODE = 128\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "adff0331",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "396e04e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_0_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "515bc9ff",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "be983f36-445f-47b4-a7d9-34d1f6ff5a6d",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a7720f5f-1bad-4edb-8cc7-dce7d3705ac5",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a243deb4-e528-4fdc-a579-e3677e4ca3ef",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e07980e9-1394-4864-abc1-e8a5639d791b",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_0_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "export SCM_SHAR_NONUMA=yes\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1e75292f-2470-4db3-b917-b7a3db8ea44e",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9051c217-b75a-4038-9478-3e8af9cca4b7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[25.12|23:06:46] JOB jobs STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-0-C_R-3m_na4 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-1-Li_R-3m_na3 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-2-O_C2m_na8 STARTED\n",
      "[25.12|23:21:32] JOB jobs/GO-1-Li_R-3m_na3 SUCCESSFUL\n",
      "[26.12|15:45:03] Job jobs/GO-2-O_C2m_na8 reported warnings. Please check the the output\n",
      "[26.12|15:45:03] JOB jobs/GO-2-O_C2m_na8 SUCCESSFUL\n",
      "[26.12|23:13:11] JOB jobs/GO-0-C_R-3m_na4 FAILED\n",
      "[26.12|23:13:14] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d71fb9c7-6f69-4acd-9cde-e29d78a25d1a",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3ef3e7e2-c97f-4bf1-b647-453ac3ef9b2d",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "df83b517-9d7d-43b5-9a87-796676d0cd65",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0.1-pure_species\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "3a522423-b6c9-4858-81d0-f346d8f3fe37",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_0_SP).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom_SP\"] = energy_per_atom\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "c5d7b8ae-39fe-4208-815d-91dccb259120",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>element</th>\n",
       "      <th>N</th>\n",
       "      <th>scm_energy_SP</th>\n",
       "      <th>scm_energy_per_atom_SP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-F_C2c_na4.cif</th>\n",
       "      <td>-1.911456</td>\n",
       "      <td>0</td>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "      <td>-7.119855</td>\n",
       "      <td>-1.779964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-Li_R-3m_na3.cif</th>\n",
       "      <td>-1.908923</td>\n",
       "      <td>0</td>\n",
       "      <td>Li</td>\n",
       "      <td>3</td>\n",
       "      <td>-5.469886</td>\n",
       "      <td>-1.823295</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   energy_per_atom  ...  scm_energy_per_atom_SP\n",
       "0-F_C2c_na4.cif          -1.911456  ...               -1.779964\n",
       "1-Li_R-3m_na3.cif        -1.908923  ...               -1.823295\n",
       "\n",
       "[2 rows x 6 columns]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "4dc9a993-c1f5-4dfe-971d-7416de516e99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x280 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 2.8], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    idx = np.arange(len(results_df))\n",
    "    ax.bar(idx - 0.1, results_df[\"scm_energy_per_atom_SP\"], width=0.20, label=\"simulated\")\n",
    "    ax.bar(idx + 0.1, results_df[\"energy_per_atom\"], width=0.20, label=\"materials project\")\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(results_df.index, rotation=25, ha=\"right\")\n",
    "    # ax.legend(loc='center left', bbox_to_anchor=(1., 0.5))\n",
    "    ax.legend(loc=\"lower center\")\n",
    "    all_E = results_df[\"energy_per_atom\"].values.tolist() + results_df[\"scm_energy_per_atom_SP\"].values.tolist()\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.05, np.max(all_E) + np.abs(np.mean(all_E)) * 0.05)\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"0.1-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7344c5ff-78d7-4298-9c07-56c5423e3643",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "421124a7-6168-43ac-a890-075e995b6474",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4ee6311d-d60f-407c-8c2e-0978718b0541",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Pure Species `Geometry Optimization` <a id='pure_species_go'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "3412c32d-d8bd-4ef3-abad-24b7264dcb38",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"0.1-PureSpeices_GO\"\n",
    "TASK = \"GO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "61fecff7-741a-4bd8-af36-647668a0a563",
   "metadata": {},
   "source": [
    "#### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "19560fc0-4118-48f4-9d2a-56112dc58478",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tGeometryOptimization\n",
      "          Properties: \t\n",
      "                     Gradients: \tno\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tyes\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t500\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           SCF: \t\n",
      "               Iterations: \t150\n",
      "           RadialDefaults: \t\n",
      "                          NR: \t3000\n",
      "           CPVector: \t256\n",
      "           KGRPX: \t10\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_0_GO\n",
    "\n",
    "print(settings_0_GO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "844962d6-c43a-4dfd-9d42-6a284df33f40",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "a8169564-9fcd-4cce-a852-dba8f3aadbac",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d08248ee-2c7c-41fd-a932-f71d9141f5a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "a641f306-e6e8-4a1f-b992-ab548ecbf2d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_0_GO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fd8bee4a-49a8-4054-b5c7-eb83d1b88d0f",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "be983f36-445f-47b4-a7d9-34d1f6ff5a6d",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "a7720f5f-1bad-4edb-8cc7-dce7d3705ac5",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0.1-pure_species\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a243deb4-e528-4fdc-a579-e3677e4ca3ef",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e07980e9-1394-4864-abc1-e8a5639d791b",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_0_GO, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "export SCM_SHAR_NONUMA=yes\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1e75292f-2470-4db3-b917-b7a3db8ea44e",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9051c217-b75a-4038-9478-3e8af9cca4b7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[25.12|23:06:46] JOB jobs STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-0-C_R-3m_na4 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-1-Li_R-3m_na3 STARTED\n",
      "[25.12|23:06:46] JOB jobs/GO-2-O_C2m_na8 STARTED\n",
      "[25.12|23:21:32] JOB jobs/GO-1-Li_R-3m_na3 SUCCESSFUL\n",
      "[26.12|15:45:03] Job jobs/GO-2-O_C2m_na8 reported warnings. Please check the the output\n",
      "[26.12|15:45:03] JOB jobs/GO-2-O_C2m_na8 SUCCESSFUL\n",
      "[26.12|23:13:11] JOB jobs/GO-0-C_R-3m_na4 FAILED\n",
      "[26.12|23:13:14] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d71fb9c7-6f69-4acd-9cde-e29d78a25d1a",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3ef3e7e2-c97f-4bf1-b647-453ac3ef9b2d",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "df83b517-9d7d-43b5-9a87-796676d0cd65",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "prev_sim = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, \"0.1-PureSpeices_SP\")\n",
    "data_path = os.path.join(prev_sim, \"results.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "3a522423-b6c9-4858-81d0-f346d8f3fe37",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_0_GO).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_GO\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom_GO\"] = energy_per_atom\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "c5d7b8ae-39fe-4208-815d-91dccb259120",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>element</th>\n",
       "      <th>N</th>\n",
       "      <th>scm_energy_SP</th>\n",
       "      <th>scm_energy_per_atom_SP</th>\n",
       "      <th>scm_energy_GO</th>\n",
       "      <th>scm_energy_per_atom_GO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-F_C2c_na4.cif</th>\n",
       "      <td>-1.911456</td>\n",
       "      <td>0</td>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "      <td>-7.119855</td>\n",
       "      <td>-1.779964</td>\n",
       "      <td>-7.182324</td>\n",
       "      <td>-1.795581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-Li_R-3m_na3.cif</th>\n",
       "      <td>-1.908923</td>\n",
       "      <td>0</td>\n",
       "      <td>Li</td>\n",
       "      <td>3</td>\n",
       "      <td>-5.469886</td>\n",
       "      <td>-1.823295</td>\n",
       "      <td>-5.476674</td>\n",
       "      <td>-1.825558</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   energy_per_atom  ...  scm_energy_per_atom_GO\n",
       "0-F_C2c_na4.cif          -1.911456  ...               -1.795581\n",
       "1-Li_R-3m_na3.cif        -1.908923  ...               -1.825558\n",
       "\n",
       "[2 rows x 8 columns]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "4dc9a993-c1f5-4dfe-971d-7416de516e99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x250 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 2.5], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    idx = np.arange(len(results_df.index))\n",
    "    ax.bar(idx - 0.2, results_df[\"energy_per_atom\"], width=0.20, label=\"Materials Project\")\n",
    "    ax.bar(idx + 0.0, results_df[\"scm_energy_per_atom_SP\"], width=0.20, label=\"BAND (SP)\")\n",
    "    ax.bar(idx + 0.2, results_df[\"scm_energy_per_atom_GO\"], width=0.20, label=\"BAND (GO)\")\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(results_df.index, rotation=25, ha=\"right\")\n",
    "    # ax.legend(loc='center left', bbox_to_anchor=(1., 0.5))\n",
    "    ax.legend(loc=\"lower center\")\n",
    "    all_E = (\n",
    "        results_df[\"energy_per_atom\"].values.tolist()\n",
    "        + results_df[\"scm_energy_per_atom_SP\"].values.tolist()\n",
    "        + results_df[\"scm_energy_per_atom_GO\"].values.tolist()\n",
    "    )\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.05, np.max(all_E) + np.abs(np.mean(all_E)) * 0.05)\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"0.1-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7344c5ff-78d7-4298-9c07-56c5423e3643",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "421124a7-6168-43ac-a890-075e995b6474",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ffad3842-1433-4dc4-964d-96993ab62fc7",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Supercells `Single Point` <a id='supercells_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "4842b982-cf29-4dbe-8946-eb57a51eb1b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"1-Supercells_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a36b1699-fb05-44b8-80ac-bbda12c70f53",
   "metadata": {},
   "source": [
    "#### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "dd45eb4f-128f-4251-be82-eade3b20d1c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           KSpace: \t\n",
      "                  Quality: \tGood\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_1_SP\n",
    "\n",
    "print(settings_1_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5ee37e5d-32cb-4f4f-899c-f0efac6c71c0",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "bd22c841-1f06-4955-8936-a55c107e4648",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 3\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2fad9903-b5cf-4033-a6b4-2408bf3ce9d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "d907cbae-a816-469c-98d7-b7486b4712b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_1_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d10c7620-c6eb-439d-8dd4-e9b46550108d",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "43a7bb10-fb4f-4c24-b800-4aa4d96cb329",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "4182ee02-416a-4f97-8a04-9d72c2e6607d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"1-supercell\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ee75e680-05b2-43b6-aece-e9c3f358bc0a",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e58daec6-1896-427e-aee1-6641af3fe9b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_1_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "\n",
    "export SCM_SHAR_NONUMA=yes\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2069ae67-3d41-4879-b309-b8a167434767",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f9d3d541-5b2e-46e1-8260-87bc877f536f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[05.07|19:49:58] JOB jobs STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.0-0-Li2CO3_P6_3mcm_-2.19_2x1x1 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.0-1-Li2CO3_C2c_-2.26_2x1x1 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.1-0-Li2CO3_P6_3mcm_-2.19_2x2x1 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.1-1-Li2CO3_C2c_-2.26_2x2x1 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.2-0-Li2CO3_P6_3mcm_-2.19_2x2x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.2-1-Li2CO3_C2c_-2.26_2x2x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.3-0-Li2CO3_P6_3mcm_-2.19_3x2x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.3-1-Li2CO3_C2c_-2.26_3x2x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.4-0-Li2CO3_P6_3mcm_-2.19_3x3x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.4-1-Li2CO3_C2c_-2.26_3x3x2 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.5-0-Li2CO3_P6_3mcm_-2.19_3x3x3 STARTED\n",
      "[05.07|19:49:58] JOB jobs/SP-1.5-1-Li2CO3_C2c_-2.26_3x3x3 STARTED\n",
      "[05.07|19:55:26] JOB jobs/SP-1.0-0-Li2CO3_P6_3mcm_-2.19_2x1x1 SUCCESSFUL\n",
      "[05.07|19:56:21] JOB jobs/SP-1.0-1-Li2CO3_C2c_-2.26_2x1x1 SUCCESSFUL\n",
      "[05.07|20:01:31] JOB jobs/SP-1.1-1-Li2CO3_C2c_-2.26_2x2x1 SUCCESSFUL\n",
      "[05.07|20:04:33] JOB jobs/SP-1.1-0-Li2CO3_P6_3mcm_-2.19_2x2x1 SUCCESSFUL\n",
      "[05.07|20:20:42] JOB jobs/SP-1.4-1-Li2CO3_C2c_-2.26_3x3x2 SUCCESSFUL\n",
      "[05.07|20:21:50] JOB jobs/SP-1.2-0-Li2CO3_P6_3mcm_-2.19_2x2x2 SUCCESSFUL\n",
      "[05.07|20:23:03] JOB jobs/SP-1.2-1-Li2CO3_C2c_-2.26_2x2x2 SUCCESSFUL\n",
      "[05.07|20:26:26] JOB jobs/SP-1.3-0-Li2CO3_P6_3mcm_-2.19_3x2x2 SUCCESSFUL\n",
      "[05.07|20:27:33] JOB jobs/SP-1.3-1-Li2CO3_C2c_-2.26_3x2x2 SUCCESSFUL\n",
      "[05.07|20:29:58] JOB jobs/SP-1.4-0-Li2CO3_P6_3mcm_-2.19_3x3x2 SUCCESSFUL\n",
      "[05.07|20:48:45] JOB jobs/SP-1.5-1-Li2CO3_C2c_-2.26_3x3x3 SUCCESSFUL\n",
      "[05.07|21:15:03] JOB jobs/SP-1.5-0-Li2CO3_P6_3mcm_-2.19_3x3x3 SUCCESSFUL\n",
      "[05.07|21:15:05] JOB jobs SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c903cfd7-45da-488a-b2b3-0f1701438b88",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0bc3a67a-2b7e-4654-ad98-05020184ae8e",
   "metadata": {},
   "source": [
    "Load crystel data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "90d395f2-a3d4-4c58-a576-97a047fe5805",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"1-supercell\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "0f02e883-0d45-4d44-b9bf-01bf5b3ee53d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition\n",
       "1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif                     2\n",
       "1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif                    2\n",
       "1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif                     2\n",
       "1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif                     4\n",
       "1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif                    4"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "708bfe27-b7a4-4cf9-bd78-4193462ad168",
   "metadata": {},
   "outputs": [],
   "source": [
    "crystal_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "data_path = os.path.join(crystal_dir, \"info.json\")\n",
    "crystals_df = pd.read_json(data_path)\n",
    "\n",
    "for job_dir in np.sort(glob.glob(os.path.join(dir_jobs, \"*\"))):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_1_SP).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_per_atom_SP\"] = energy_per_atom\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "64b16fbf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>energy_SP</th>\n",
       "      <th>energy_per_atom_SP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-19.251315</td>\n",
       "      <td>-4.812829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-38.292625</td>\n",
       "      <td>-4.786578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-18.076721</td>\n",
       "      <td>-4.519180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-38.502611</td>\n",
       "      <td>-4.812826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-76.653339</td>\n",
       "      <td>-4.790834</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition  ...  energy_per_atom_SP\n",
       "1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif                     2  ...           -4.812829\n",
       "1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif                    2  ...           -4.786578\n",
       "1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif                     2  ...           -4.519180\n",
       "1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif                     4  ...           -4.812826\n",
       "1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif                    4  ...           -4.790834\n",
       "\n",
       "[5 rows x 3 columns]"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "900a5429-5b33-437f-a4cd-e709621c0408",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x250 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[4, 2.5])\n",
    "n_sc = len(results_df[\"unitcell_repetition\"].unique())\n",
    "idx = np.arange(len(crystals_df))\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    Set2 = mp.colormaps[\"Set2\"]\n",
    "    for j, n_uc in enumerate(results_df[\"unitcell_repetition\"].unique().tolist()):\n",
    "        ax.bar([np.nan], [np.nan], width=0.8 / n_sc, label=f\"# unitcell = {n_uc}\", color=Set2.colors[j])\n",
    "        sub_set = results_df.loc[results_df[\"unitcell_repetition\"] == n_uc, :]\n",
    "        for i, name in enumerate(sub_set.index):\n",
    "            ax.bar(\n",
    "                i + 0.8 * (j + 0.5) / n_sc - 0.4,\n",
    "                results_df[\"energy_per_atom_SP\"][i],\n",
    "                width=0.8 / n_sc,\n",
    "                color=Set2.colors[j],\n",
    "            )\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(crystals_df.index, rotation=25, ha=\"right\")\n",
    "    ax.legend(loc=\"center left\", bbox_to_anchor=(1.0, 0.5))\n",
    "    all_E = results_df[\"energy_per_atom_SP\"].values.tolist()\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.02, np.max(all_E) + np.abs(np.mean(all_E)) * 0.02)\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"1-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4fac93fe-133a-428f-98e7-6f60b780b5de",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "d056ef3e-329d-4fa9-905d-97ee76587f9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "75e61cb0-e87f-48f5-8c1b-f8a5e8a0a02a",
   "metadata": {},
   "source": [
    "## Supercells `Geometry Optimization`  <a id='supercells_go'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "id": "3c684aac-7a51-4768-b50d-8a407a4abe6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"1-Supercells_GO\"\n",
    "TASK = \"GO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c0b606af-0bb3-433a-a191-23bbbd9c90ab",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "id": "cd431661-ba72-4fa6-8b3d-e6df274a4a0e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tGeometryOptimization\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t150\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           KSpace: \t\n",
      "                  Quality: \tGood\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_1_GO\n",
    "\n",
    "print(settings_1_GO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bd8959e1-257a-41b5-94df-ce772a9c7b00",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "id": "6a868ad2-1043-428f-a655-49d1e7e5b18e",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c176c806-03b9-4978-ae92-cb62decbb76a",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "9f31caa0-91f1-4594-ba6d-9b10ec06a7f9",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_1_GO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9b4fb122-9324-4474-a4e9-ab08cd4b189a",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "94e764b5-3079-4d1b-98c4-9f633c1f979d",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "b4d83d57-5155-4401-8ad8-0c5b6ada529d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"1-supercell\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dee5ce4c-c4a9-4fb5-8342-39c8a0a33f90",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "7ca9dde5-6712-46a1-a6bb-681eed3514f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_1_GO, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "59a44b53-2e44-461e-9995-d21ddf17e743",
   "metadata": {},
   "source": [
    "### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8312c2c3-b0d5-465f-97f3-181d581d081b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[05.07|21:50:03] JOB jobs STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.0-0-Li2CO3_P6_3mcm_-2.19_2x1x1 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.0-1-Li2CO3_C2c_-2.26_2x1x1 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.1-0-Li2CO3_P6_3mcm_-2.19_2x2x1 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.1-1-Li2CO3_C2c_-2.26_2x2x1 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.2-0-Li2CO3_P6_3mcm_-2.19_2x2x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.2-1-Li2CO3_C2c_-2.26_2x2x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.3-0-Li2CO3_P6_3mcm_-2.19_3x2x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.3-1-Li2CO3_C2c_-2.26_3x2x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.4-0-Li2CO3_P6_3mcm_-2.19_3x3x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.4-1-Li2CO3_C2c_-2.26_3x3x2 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.5-0-Li2CO3_P6_3mcm_-2.19_3x3x3 STARTED\n",
      "[05.07|21:50:03] JOB jobs/GO-1.5-1-Li2CO3_C2c_-2.26_3x3x3 STARTED\n",
      "[05.07|21:53:26] JOB jobs/GO-1.0-0-Li2CO3_P6_3mcm_-2.19_2x1x1 SUCCESSFUL\n",
      "[05.07|21:54:40] JOB jobs/GO-1.0-1-Li2CO3_C2c_-2.26_2x1x1 SUCCESSFUL\n",
      "[05.07|21:55:18] JOB jobs/GO-1.1-0-Li2CO3_P6_3mcm_-2.19_2x2x1 SUCCESSFUL\n",
      "[05.07|22:00:59] JOB jobs/GO-1.3-0-Li2CO3_P6_3mcm_-2.19_3x2x2 SUCCESSFUL\n",
      "[05.07|22:01:06] JOB jobs/GO-1.2-0-Li2CO3_P6_3mcm_-2.19_2x2x2 SUCCESSFUL\n",
      "[05.07|22:05:39] JOB jobs/GO-1.1-1-Li2CO3_C2c_-2.26_2x2x1 SUCCESSFUL\n",
      "[05.07|22:07:29] JOB jobs/GO-1.4-0-Li2CO3_P6_3mcm_-2.19_3x3x2 SUCCESSFUL\n",
      "[05.07|22:10:03] JOB jobs/GO-1.3-1-Li2CO3_C2c_-2.26_3x2x2 SUCCESSFUL\n",
      "[05.07|22:12:50] JOB jobs/GO-1.2-1-Li2CO3_C2c_-2.26_2x2x2 SUCCESSFUL\n",
      "[05.07|22:24:56] JOB jobs/GO-1.5-0-Li2CO3_P6_3mcm_-2.19_3x3x3 SUCCESSFUL\n",
      "[05.07|22:30:16] JOB jobs/GO-1.4-1-Li2CO3_C2c_-2.26_3x3x2 SUCCESSFUL\n",
      "[05.07|23:21:56] JOB jobs/GO-1.5-1-Li2CO3_C2c_-2.26_3x3x3 SUCCESSFUL\n",
      "[05.07|23:21:59] JOB jobs SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e56dafe7-265a-4a47-8b83-aeb5080a9c86",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "27c73618-040e-4e2c-b565-530b5e9fecf8",
   "metadata": {},
   "source": [
    "Load crystel data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "id": "e9725096-d9a0-4b0b-89b4-e54dbb0aa121",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "prev_sim = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, \"1-Supercells_SP\")\n",
    "data_path = os.path.join(prev_sim, \"results.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "id": "491d94ef-ad35-4a8a-8d9e-a99e76c842ce",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>energy_SP</th>\n",
       "      <th>energy_per_atom_SP</th>\n",
       "      <th>energy_GO</th>\n",
       "      <th>energy_per_atom_GO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-19.251315</td>\n",
       "      <td>-4.812829</td>\n",
       "      <td>-19.251315</td>\n",
       "      <td>-4.812829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-38.292625</td>\n",
       "      <td>-4.786578</td>\n",
       "      <td>-38.292625</td>\n",
       "      <td>-4.786578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-18.076721</td>\n",
       "      <td>-4.519180</td>\n",
       "      <td>-18.076721</td>\n",
       "      <td>-4.519180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-38.502611</td>\n",
       "      <td>-4.812826</td>\n",
       "      <td>-38.502611</td>\n",
       "      <td>-4.812826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-76.653339</td>\n",
       "      <td>-4.790834</td>\n",
       "      <td>-76.653339</td>\n",
       "      <td>-4.790834</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition  ...  energy_per_atom_GO\n",
       "1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif                     2  ...           -4.812829\n",
       "1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif                    2  ...           -4.786578\n",
       "1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif                     2  ...           -4.519180\n",
       "1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif                     4  ...           -4.812826\n",
       "1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif                    4  ...           -4.790834\n",
       "\n",
       "[5 rows x 5 columns]"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "id": "d14351bd-f033-47f5-b553-fc191f8fa874",
   "metadata": {},
   "outputs": [],
   "source": [
    "crystal_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"0-materialproject\")\n",
    "data_path = os.path.join(crystal_dir, \"info.json\")\n",
    "crystals_df = pd.read_json(data_path)\n",
    "\n",
    "for job_dir in np.sort(glob.glob(os.path.join(dir_jobs, \"*\"))):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_1_GO).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_GO\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_per_atom_GO\"] = energy_per_atom\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "id": "8aba3d24",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>energy_SP</th>\n",
       "      <th>energy_per_atom_SP</th>\n",
       "      <th>energy_GO</th>\n",
       "      <th>energy_per_atom_GO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-19.251315</td>\n",
       "      <td>-4.812829</td>\n",
       "      <td>-19.053002</td>\n",
       "      <td>-4.763251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-38.292625</td>\n",
       "      <td>-4.786578</td>\n",
       "      <td>-38.003627</td>\n",
       "      <td>-4.750453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "      <td>-18.076721</td>\n",
       "      <td>-4.519180</td>\n",
       "      <td>-17.868644</td>\n",
       "      <td>-4.467161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-38.502611</td>\n",
       "      <td>-4.812826</td>\n",
       "      <td>-38.105975</td>\n",
       "      <td>-4.763247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>-76.653339</td>\n",
       "      <td>-4.790834</td>\n",
       "      <td>-76.077089</td>\n",
       "      <td>-4.754818</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition  ...  energy_per_atom_GO\n",
       "1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif                     2  ...           -4.763251\n",
       "1.0-1-LiF_P6_3mc_-3.17_2x1x1.cif                    2  ...           -4.750453\n",
       "1.0-2-LiF_Pm-3m_-2.89_2x1x1.cif                     2  ...           -4.467161\n",
       "1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif                     4  ...           -4.763247\n",
       "1.1-1-LiF_P6_3mc_-3.17_2x2x1.cif                    4  ...           -4.754818\n",
       "\n",
       "[5 rows x 5 columns]"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "id": "d3873d20-5f5d-49a3-802f-a59bd629d56c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[4, 2])\n",
    "n_sc = len(results_df[\"unitcell_repetition\"].unique())\n",
    "idx = np.arange(len(crystals_df))\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    Set2 = mp.colormaps[\"Set2\"]\n",
    "    for j, n_uc in enumerate(results_df[\"unitcell_repetition\"].unique().tolist()):\n",
    "        ax.bar([np.nan], [np.nan], width=0.8 / n_sc, label=f\"# unitcell = {n_uc}\", color=Set2.colors[j])\n",
    "        sub_set = results_df.loc[results_df[\"unitcell_repetition\"] == n_uc, :]\n",
    "        for i, name in enumerate(sub_set.index):\n",
    "            ax.bar(\n",
    "                i + 0.8 * (j + 0.5) / n_sc - 0.4,\n",
    "                results_df[\"energy_per_atom_GO\"][i],\n",
    "                width=0.8 / n_sc,\n",
    "                color=Set2.colors[j],\n",
    "            )\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks(idx)\n",
    "    ax.set_xticklabels(crystals_df.index, rotation=25, ha=\"right\")\n",
    "    ax.legend(loc=\"center left\", bbox_to_anchor=(1.0, 0.5))\n",
    "    all_E = results_df[\"energy_per_atom_GO\"].values.tolist()\n",
    "    ax.set_ylim(np.min(all_E) - np.abs(np.mean(all_E)) * 0.02, np.max(all_E) + np.abs(np.mean(all_E)) * 0.02)\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"1-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fb86ebc9-604d-4773-91dc-004508d63848",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "baf93f80-4e2c-491d-92d9-5b4b70c9bc79",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1f9ab8c1-5c5c-43de-a3c0-f145dae0b96b",
   "metadata": {},
   "source": [
    "## Strain `Single Poin` <a id='strain_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "id": "3a414d65-4267-472b-94f2-40e3b7ca0889",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"2-Strain_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "cf4ea157-d460-422f-9768-ffb4d94123c5",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "id": "9eedaa07-700d-419d-9542-41cc28c46b35",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_2_SP\n",
    "\n",
    "print(settings_2_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2aaff2a4-9e77-4238-999c-caf5ccafb580",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "id": "2d883fca-1470-4f66-a7d7-a4f351766f4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "id": "bb6b70c1-c7d2-47d3-bc2a-4c7a28e1715a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[19.06|00:41:10] PLAMS working folder: /home/paolo/Enhancing_ReaxFF/step02-sim/LiF/2-Strain_SP.002\n"
     ]
    }
   ],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "RUNNER.settings.special.cpu_freq = \"--cpu-freq=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "id": "7c26d4a2-5230-40cd-83c5-c2d784f87a94",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_2_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "19a3f9fd-8e1f-47db-a241-492e4d9b3fd0",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bd8e3c76-2094-4e1f-b26c-e33e4ab3d081",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "id": "9749a6d7-a918-4103-9669-a26f7499845b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"2-strain\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dea7ae61-77e9-4a73-b1d9-6253f4352d76",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ebccd063-2fa7-4da5-b297-a8c6ef636554",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job_dir = os.path.join(RUN_DIR, \"jobs\", sim_name)\n",
    "    if os.path.exists(job_dir):\n",
    "        job = scm.AMSJob.load_external(job_dir, settings=settings_2_SP)\n",
    "        if not job.check():\n",
    "            print(f\"sim {sim_name} FAILE will rerun it\")\n",
    "        else:\n",
    "            continue\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_2_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "export SCM_SHAR_NONUMA=yes\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d55a54b5-5a62-40e0-b84c-b389661546b2",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b90030af-eb1d-436a-94d0-43c86b0a2326",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    "    cpu_freq=2250000,\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "47d07a96-f913-4e45-845f-f39b4c24147f",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b4ba44c8-2ad6-4600-8bc2-4fcf28e8f782",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "id": "09a76ff4-e238-46d1-9156-8d23131aaf25",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"2-strain\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "id": "98b01833-a469-4f4c-b0d4-d93183cb5ea2",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_2_SP)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    mol = toASE(job_result.get_molecule(\"Molecule\"))\n",
    "    Na = len(mol)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    strain_mtx = results_df.loc[sim_name[3:] + \".cif\", \"strain_matrix\"]\n",
    "    simple_str, latex_str = mtx2str(strain_mtx)\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"strain_str\"] = simple_str\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"strain_str_latex\"] = latex_str\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "    strain_mtx = np.array(strain_mtx)\n",
    "    strain_mtx_sq = np.power(strain_mtx, 2)\n",
    "    von_mises_strain = (\n",
    "        2\n",
    "        / 3\n",
    "        * np.sign(strain_mtx.sum())\n",
    "        * np.power(\n",
    "            3 / 2 * np.sum((strain_mtx - 1 / 3 * strain_mtx.diagonal(0).sum() * np.eye(3)).sum(axis=0) ** 2)\n",
    "            + 3 / 4 * np.sum(np.tril(2 * strain_mtx, -1) ** 2),\n",
    "            1 / 2,\n",
    "        )\n",
    "    )\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"von_mises_strain\"] = von_mises_strain\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"vol\"] = mol.get_volume()\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "id": "28d3b66c-7841-4ef5-a369-4f94c8bfc9c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>strain_matrix</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>strain_str</th>\n",
       "      <th>strain_str_latex</th>\n",
       "      <th>system_id</th>\n",
       "      <th>von_mises_strain</th>\n",
       "      <th>vol</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2.0-0-LiF_Fm-3m_-3.18_xx_-0.125.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.125, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0....</td>\n",
       "      <td>-75.961595</td>\n",
       "      <td>-4.747600</td>\n",
       "      <td>e_11</td>\n",
       "      <td>$\\varepsilon_{ 1 1 }$</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-8.333333e-02</td>\n",
       "      <td>117.932997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.1-0-LiF_Fm-3m_-3.18_xx_-0.095.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.095, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0....</td>\n",
       "      <td>-76.494122</td>\n",
       "      <td>-4.780883</td>\n",
       "      <td>e_11</td>\n",
       "      <td>$\\varepsilon_{ 1 1 }$</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-6.333333e-02</td>\n",
       "      <td>122.559565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.2-0-LiF_Fm-3m_-3.18_xx_-0.065.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.065, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0....</td>\n",
       "      <td>-76.823927</td>\n",
       "      <td>-4.801495</td>\n",
       "      <td>e_11</td>\n",
       "      <td>$\\varepsilon_{ 1 1 }$</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-4.333333e-02</td>\n",
       "      <td>127.017725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.3-0-LiF_Fm-3m_-3.18_xx_-0.035.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.035, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0....</td>\n",
       "      <td>-76.981781</td>\n",
       "      <td>-4.811361</td>\n",
       "      <td>e_11</td>\n",
       "      <td>$\\varepsilon_{ 1 1 }$</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-2.333333e-02</td>\n",
       "      <td>131.324627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.4-0-LiF_Fm-3m_-3.18_xx_-0.005.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.005, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0....</td>\n",
       "      <td>-77.010916</td>\n",
       "      <td>-4.813182</td>\n",
       "      <td>e_11</td>\n",
       "      <td>$\\varepsilon_{ 1 1 }$</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-3.333333e-03</td>\n",
       "      <td>135.494698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.34-2-LiF_Pm-3m_-2.89_xxyyzz_0.115.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[0.115, 0.0, 0.0], [0.0, 0.115, 0.0], [0.0, 0...</td>\n",
       "      <td>-69.102913</td>\n",
       "      <td>-4.318932</td>\n",
       "      <td>e_11 e_22 e_33</td>\n",
       "      <td>$\\varepsilon_{ 1 1 } \\varepsilon_{ 2 2 } \\vare...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>182.991053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.35-2-LiF_Pm-3m_-2.89_xxyyzz_0.145.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[0.145, 0.0, 0.0], [0.0, 0.145, 0.0], [0.0, 0...</td>\n",
       "      <td>-68.165433</td>\n",
       "      <td>-4.260340</td>\n",
       "      <td>e_11 e_22 e_33</td>\n",
       "      <td>$\\varepsilon_{ 1 1 } \\varepsilon_{ 2 2 } \\vare...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.925231e-17</td>\n",
       "      <td>196.542625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.36-2-LiF_Pm-3m_-2.89_xxyyzz_0.175.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[0.17500000000000002, 0.0, 0.0], [0.0, 0.1750...</td>\n",
       "      <td>-67.159394</td>\n",
       "      <td>-4.197462</td>\n",
       "      <td>e_11 e_22 e_33</td>\n",
       "      <td>$\\varepsilon_{ 1 1 } \\varepsilon_{ 2 2 } \\vare...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.925231e-17</td>\n",
       "      <td>210.413131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.37-2-LiF_Pm-3m_-2.89_xxyyzz_0.205.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[0.20500000000000002, 0.0, 0.0], [0.0, 0.2050...</td>\n",
       "      <td>-66.106027</td>\n",
       "      <td>-4.131627</td>\n",
       "      <td>e_11 e_22 e_33</td>\n",
       "      <td>$\\varepsilon_{ 1 1 } \\varepsilon_{ 2 2 } \\vare...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.925231e-17</td>\n",
       "      <td>224.595400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.38-2-LiF_Pm-3m_-2.89_xxyyzz_0.235.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[0.23500000000000001, 0.0, 0.0], [0.0, 0.2350...</td>\n",
       "      <td>-65.030647</td>\n",
       "      <td>-4.064415</td>\n",
       "      <td>e_11 e_22 e_33</td>\n",
       "      <td>$\\varepsilon_{ 1 1 } \\varepsilon_{ 2 2 } \\vare...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>239.082721</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>117 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         unitcell_repetition  ...         vol\n",
       "2.0-0-LiF_Fm-3m_-3.18_xx_-0.125.cif                        8  ...  117.932997\n",
       "2.1-0-LiF_Fm-3m_-3.18_xx_-0.095.cif                        8  ...  122.559565\n",
       "2.2-0-LiF_Fm-3m_-3.18_xx_-0.065.cif                        8  ...  127.017725\n",
       "2.3-0-LiF_Fm-3m_-3.18_xx_-0.035.cif                        8  ...  131.324627\n",
       "2.4-0-LiF_Fm-3m_-3.18_xx_-0.005.cif                        8  ...  135.494698\n",
       "...                                                      ...  ...         ...\n",
       "2.34-2-LiF_Pm-3m_-2.89_xxyyzz_0.115.cif                    8  ...  182.991053\n",
       "2.35-2-LiF_Pm-3m_-2.89_xxyyzz_0.145.cif                    8  ...  196.542625\n",
       "2.36-2-LiF_Pm-3m_-2.89_xxyyzz_0.175.cif                    8  ...  210.413131\n",
       "2.37-2-LiF_Pm-3m_-2.89_xxyyzz_0.205.cif                    8  ...  224.595400\n",
       "2.38-2-LiF_Pm-3m_-2.89_xxyyzz_0.235.cif                    8  ...  239.082721\n",
       "\n",
       "[117 rows x 9 columns]"
      ]
     },
     "execution_count": 215,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "e9198ac9-4e43-4361-93aa-1c46d9217321",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArMAAAEUCAYAAAA4Fr7QAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAABJ0AAASdAHeZh94AACNkElEQVR4nO3dd3iT1dvA8W/SvVto6aALKHsoe8jeoDIVUByI86cv7sVQcC9UFByIAxBUVJaCbNmUKXsX6F5075Hkef+ojS1NaZumTdPen+vqddFn5U44Te6c55z7qBRFURBCCCGEEMICqc0dgBBCCCGEEMaSZFYIIYQQQlgsSWaFEEIIIYTFkmRWCCGEEEJYLElmhRBCCCGExZJkVgghhBBCWCxJZoUQQgghhMWSZFYIIYQQQlgsSWaFEEIIIYTFkmRWCCGEEEJYLElmhRBCCCGExZJkVgghhBBCWCxJZoUQQgghhMWSZFYAcPXqVaZNm0bTpk2xtbWlSZMmDBo0iMLCQnOHJuqpmmxz0p5FeSy13UmbFqJ81uYOQJhfWloat912Gz179uSrr77Cw8ODlJQUYmJisLGxMXd4oh6qyTYn7VmUx1LbnbRpIW5OklnBgQMHiI+PZ/LkyQwdOhRbW1usraVpiJpTk21O2rMoj6W2O2nTQlRAEQ1eYmKi0rJlSwVQAMXDw6PU/jfffFNp2bKlolKplLVr11Z6nxDlMbbNpaSkKLfffrvSsmVLpVOnTsqwYcOUy5cvV/ralTlf2nT9VZPvddVpdxXtryhuRVGUSZMmKR07dlRuvfVWpXv37sr27dsrtU+I+kDGzAqys7MZNWoUX331Ffv372ffvn2l9g8bNozNmzfTv3//MufebJ8Q5TG2zalUKp599lkuXbrEyZMnueOOO3jkkUcqfe3KnC9tuv6qyfe66rS7ivZXFDfA4sWLOXXqFMePH2fx4sXcfffd6HS6CvcJUS+YO5sW5pWenq4EBgYqf/zxR4XHDhgwoNyeqpvtE6IkU7U5RVGUI0eOKEFBQUZd29D5VXlsYVlq8r3OlO3uxv1VvbaiKMrOnTsVDw8PRavVVmmfEJZKBt00cHv27CEyMpLWrVubOxTRQJiyzS1YsICxY8cafe0bzxf1V02+15m63ZXcX5VrP/fcc6xfv5709HRWr16NWq2u1D4hLJ205gbOy8sLgAceeIDVq1eza9cuVqxYwfTp080cmaivTNXm3njjDa5evcp7771n1LUNnS/qr5p8rzNlu7txf1Wu/emnn3L16lVWrlzJyy+/TEFBQaX2CWHpJJlt4Hr27MmPP/5IYWEhDz30EKNHj+bdd9+lefPm5g5N1FOmaHNvv/02f/31F5s2bcLR0bHK1y7vfFF/1eR7nananaH9xsQ9cuRIUlNTOX36dJX2CWGpZJiB4L777uO+++4zdxiiAalOm3vjjTf466+/2Lp1K25ublW+dkXni/qrJt/rqtvubra/omvn5uYSHx9Ps2bNAAgNDSU5OZnmzZvfdJ8Q9YVKURTF3EGIum3evHl8++23XL9+HRcXF+zt7Tl48CD+/v433SeEscprV+np6XTo0IEWLVrg7OwMgLW1NUePHq3Udc+ePVvh+dKmG66aeq+rqN1Vpl3eTEpKCrfffjuZmZlYW1vj5OTEW2+9xeDBg2+6T4j6QpJZIYQQQghhsWTMrBBCCCGEsFiSzAohhBBCCIslyawQQgghhLBYkswKIYQQQgiLJcmsEEIIIYSwWJLMCiGEEEIIiyWLJlRBWloau3fvJiAgADs7O3OHI2pBfn4+UVFRDBgwAHd3d3OHUyFpow2TtFNhCSytnQrLIclsFezevZtx48aZOwxhBuvWrWPs2LHmDqNC0kYbNmmnwhJYSjsVlkOS2SoICAgAiv4QQ0JCzByNqA1hYWGMGzdO/39f10kbbZiknQpLYGntVFgOSWaroPh2WEhICO3btzdzNKI2WcqtUGmjDZu0U2EJLKWdCsshE8CEEEIIIYTFkmRWCCGEEEJYLBlmIASQHR5O9Op15ERG4hgYiP/EcTgFB5s7rHpNXnMh6h/5uxbmIMmsaPCyw8M59cpsdHl5AOSER5By+AidPnjHzJFZroo+0G72mt/4wScfjkJYhqr8XQthSpLMigYvevU6/ZtvMV1eHjFr1sOIoWaKynIZ+kBLDg3FZ8RwUKkpTE8j/dRpg6/56Vmv4xgYgJWdHVYO9ug0WtL+OY6i1eqvlXLoMJ0+fLdMciwJrxDmdbP30lbPP2OmqERDIMmsaPCyr141uD0nMhKbWo6lrisvaVS0WrKvhZNx/gIx6/4o84GmFGqI2/BXhdfXZmeTef7CTY/R5edzeuZreHTtjFNwMFaODoQv/RFdfj4gvUFCmEtORKTh7ZGGtwthKlVKZs+dO2fUg4SEhGBra2vUuULUJEWnQ5OZZXCfY2AghbUcT11msMf1wAGcmjUjJyq6TAJ7I5W1FU7NmmPj7kZuVDR58fFljnEMDMCtY0e0eXlo8/JI++c42tzcMsdpc3JI2rufpL37DT6W9AYJYQYqlcHNjoGBtRyIaGiqlMx26NABVTmN1RBFUVCpVBw5coQuXbpUOTghalrsnxsoTE8vs11tb0/TCWMJz842Q1R1U+RPq8r2uGq0ZF0O0/+utrfHyt6OwrSyr6nnbbfpk8sbE+Pic1u98Gyp3tSLHy8gac/eMtdy8G+Knacn2deuUZieYTDe5IMHif49gEa9euDo769/XBmOIITpZV68RE5ERJntxe+lQtSkKg8zWLRoEe3atavUsRqNhuHDh1c5KCFqQ3Z4BBHLVwLg0q4tdl6e5EZG4RgYSNMJY4uSnLNnzRukmRWmp5N0IJSkPfvIOHfe4DFWTk4E3jsF17ZtcAoOIicqymCiWvIDzSk4mE4fvEPMmvX6xFL/mpfgP3EcKYePlLlW65eeLxreoChceP8jUg4eKhOXLr+AiB9XEvHjShya+uHSpjVJe/ejKygAZDiCEKaiycnh4sefgqJg7eaKW/v25MXFlft3LYSpVTmZ7dq1Kz169KjUsVqtFkVRqhyUEDVNV1jIpU8/Q9FosHZ1pc3LL2Dr4WHusMxK32sZHo6VgwOo1GReugQ63U3Pa9StK353jNb/XtlE1Sk4uMJhABVdS6VSEXjPJNJOnCyV8KpsbHBt35bMC5fQ5eWRGxNLbkxsmevLcIS6Q3rNLdfVr5eQn5AIKhWtX3we904dzR2SaGCqlMzqKvhQu5GVlVWVzxGiNkSu/Jmc8KJbYiFPPdHgE9msa9c4/fJMdAUGRgmr1bjfeguubVoTvXqtfqIVlH8LsTKJamVVdK2bJby6ggLSTp0m5dBhErb/bTAxz7hwAUWrRWVlZZJ4RdVJSSfLlbhzF9d37wHAf+J4SWSFWVQpmT1y5Ajdu3evqViEqBXpZ84Ss+4PAJoMHUzjXj3NHJH56AoLSdq7n6vffmcwkXVuGUK712Zh4+YGQKOe3SvscTWH8hJeta0tjbp1pVG3rmjz8kjas6/MMfkJiRz73//hM3IE3kOHUJCSLD2EtUxKOlmm3Lg4rny9BADnVi0JuGeymSMSDVWVktmePXvSokUL7rnnHqZMmVLpsbNC1BWa7GwuL/gcFAV7H2+aPTzd3CGZhSYrm/gtW4nb8BcFKSnlHqdoNPpEFkzb41rb/CeOJ+Xw0dJJk0oFikJ+QiIRy34kcuXPoCil69pKD2GNK690k5R0qrt0hYVc+ngBurw8rBwdaf3ic6itpdqnMI8qtbxff/2VX375hY8//ph33nmHDh06cO+99zJ58mSC5Y1eWICr33xH/vUkUKtp+ezTWDs6mDukGldyLKJ9kyaoHRxIOXS4VFJn7eaGxkBVh/pUUsfQcAS/8WPQ5uQS/9dmkkMPomg0Zc6THsKa5xgYqB/2c+N2UTdF/vSLvpJJi/89hr23t5kjEg1ZlZLZu+66i7vuuousrCzWrVvHqlWreP3115k1axY9e/bk3nvv5e6778ZbGrWog5L2H+D6rt0A+E8Yh2vbNmaOqOYZGouop1bj2bcPTceOQWVtVWEFgvqgvJ5lt/btKEhJ5cRzLxgsK5Z58aK+1KAwPf+J40g5eEhfaQLqZ/urL9JOnCRmzToAmgweiFf/fuYNSDR4amNOcnZ25r777uPPP/8kISGBb775BmdnZ5577jn8/f0ZNmwYP/zwg6ljFcJo+ckpXPlqMQBOLZoTMGWSmSOqHYbGIkJRndaui7+g9QvP4RzSQt9r6TWgP07NgvEa0L/B3Vq3beSBW6dOBvflxSdwZvbrpDfwUm01xSk4mA7vvqX/3aVN6wbX/ixFYXo6lxZ8DoC9rw/NHn3EzBEJYYLlbN3d3Xn44Yd5+OGHuX79Om+++SZffvklf//9Nw899JApYhTCaEW32Nfqx0qqrK1p9dwzqG0axkK15Y05VNvYYN+kSaltljwe1lQM1bUtHlebcfYcZ2a9jlunjgROvQcrezuZKGZCLi1DsPfzIy82FreOHeS1rIMUReHy519QmJpW9F76wnMNYqiWqPtMMlo7PDycX375hV9++YVTp07h6urK+PHjTXFpIYxmaJUpAEVbdlxkfSVjEavG4LjasXeSExVF1C+/khcXT/qp05w+dRrUan2pL5koZhqO/k3Ji40lNzra3KGIEorH3WecOaufMBo49R5cWoaYOTIhihidzMbGxrJq1Sp++eUXjh49ip2dHXfccQdz585l9OjR2NnZmTJOIarM0C12RaNpUJN5yltBS8Yils9QD7Vzi+Z49etL4s7dRK36jfzExDI1a2WiWPU5+DeFw0fIiY4xdyjiXwY7Bf6tPS1EXVGlZDYpKYnffvuNX375hf3796NWqxk2bBjLli1j3LhxODs711ScQlSZoR5JaFjlfiq7GpeomMrKCu+hg/Ea0I9jjz9JQXLZkmZZ166ZIbL6wzHAH4C8uHhZyKKOMDjuXqcjdt0f8sVN1BlVSmZ9fX3R6XT069ePL774grvuuovGjRvXVGxCVIu2xEpVJTW0W+wyFta01DY2uLZvT9KevWX25cXEEvXbavzG3IGV3J2qMoemTYGiOyi5cfE4+jc1c0RCagALS1ClZPaDDz5g8uTJNG0qbzCibkv8exf5CQlltsstdmEKBieKAYpWS+SKn4jftIWg++7BMSiImHV/yCSxSnIokbzmxsRIMlsHWDs6Gtze0DoFRN1WpWT2+eefL7Pt3LlzHDt2jKioKKZPn46Pjw9hYWF4e3vj4uJiskCFqKyc6GiufP0NAI5BQTgGBZAbFS232IXJGBq+4TmwH0m79nJ99x4KkpO5/NkifSUEkElilWHt5ISNhweFqankRkVDzx7mDqlBy7+eRPa18DLbpVNA1DVGTwDLyspi+vTprF69GmtrazQaDSNHjsTHx4dZs2YRGBjI/PnzTRmrEBXS5udz8cOP0eXnY+XkRNvZr2Lv3aTiE4WoIkPDNxp16YLfmDsIX7qc9NNn9IlsMZkkVjHHAH/SU1PJjZFJYOakaLVcWvA52txcVNbWuHe+hYKkZOkUEHWSUYsmQFEv7YEDB9i+fTuZmZkoJd60R48ezebNm00SoBBVce27H8iJKBrL1fLppySRFbXOOaQF7d+ah10Tw20v68qVWo7IshSPm82JkmTWnKJXryXjTNEiIc0fe5h2c2Zx64KPafX8M5LIijrH6GR2zZo1fPDBBwwaNAirG2acBgUFERFheCa5EDXl+p59JGzZBoDvHaNp3KunmSMSDZVKpcKlTWuD+3JjYrny9RIKUlNrOSrL4BhQlMzmxsSU6iQRtSfjwkUif14FQOPePfEePszMEQlxc0Yns7m5ueVWMsjMzCyT4ApRk3JjYwn74isAnFq0IHjaA2aOSDR0/hPHoba3L7tDUYjftJljjz9FxIqf0GRn135wdVhxz6w2J4eCFEn4a5smO5tLHy8AnQ7bxo1p8dT/UKlU5g5LiJsyesxs9+7dWb58OSNHjiyz7/fff6dPnz7VCkyIytIVFHDxo0/Q5eVh5ehI65eebzDL1Yq6y9AkMZ+Rw0g5+g9xf25El59P9G+rid+8hSaDBpGfkkxudEyDr3rg8G+tWSjqnbVr3MiM0TQsiqJw5avFRYuCqNW0euEZbGQit7AARiezb731FsOGDWPo0KHcfffdqFQq/vrrLz799FN+//139uzZY8o4hShX+NLlZF8tKlYf8tQTOPj6mDkiIYoYmiTm2q4dvrePJmrVbyRs244mM4vYP/7U72/oVQ9sGzXCysEBbW4uuVHRuHfqaO6QGozEv3eStHc/AP53TcCtfXszRyRE5Rg9zKBfv37s2LGD/Px8/u///g9FUZg7dy5Xr15l+/btdO/e3ZRxClFGdng4p2bOIW7jJgAa39Ybz763mTkqISpm17gRIU8+TpdFn2Hr6Vlmvy4vj/AfljfIMaMqlUpfbzYnOtrM0TQcubGxXP3mOwBc2rQmcMokM0ckROUZ3TMLcNttt7F3715yc3NJTU3F3d0dx3IKLAthStnh4Zx6eRa6Eqt8pR79h+zw8AbZmyUsk0NTP6ydnShISiqzL+3ESY7/37P4jBiG18ABFKQkE716XYNYgMHB35+sy2HkRktFg9qgKyzk4vwF+qFarZ5/VpYSFhalWslsMQcHBxwcHExxKYuSHR7eYD5c6pqoX1eXSmQBdPn5UsNTWBzHwEBywg1Xf8mNjubadz9wbdmPoNMV/VD/hyIUr/wlyWztiFz5M9n/loxr8b/HpaShsDhVGmbw/PPPExUVVenjFUXh+eefJ6YeFr/ODg/n1CuzSdqzl5zwCJL27OXUK7PJDg83d2j1nqIopB0/YXCfrBcuLI2hqgdqezuaP/EojXv3LOoh02j0iWyx4gUYasL48eNRqVRs37693GNeffVVOnbsiKurK76+vtxzzz1V+ny4meJhBgUpKWhyckxyTVFadng4Fz9ewNHHniRmbVE7ajJ4EF79+5o5MiGqrkrJ7GeffUZ8fHylj9fpdHz22WckJCRUObC6Lnr1ujLrstfkh4v4T8zqtWjL+YCT9cKFpSmueuA1oD9OzYLxGtCfTh+8i++okbR59WW6fbcYGw93g+fWxJe35cuXk1OJBFKlUrF06VKSkpI4f/48KpWKO++80yQxFCezIL2zNaFkZ0x+8eezSoX3CKknKyxTlYYZKIrCtGnTcHJyqql4LEZ5HyLSM1izUg4fIWLFT0W/qNWleqtkvXBhqQxVPShm6+GBW8eOJO3ZW2afqb+8RUdHM2fOHPbt20dQUNBNj33vvff+i9HWlpdffpnOnTuTmpqKh4dHteKw9/FBZWWFotWSGx2DS6uW1bqeKM1QZwyKQvxfm3EtZ7EPIeqyKiWzDz74YJUfoH379ngamK1r6cob5yY9gzUnJyqaS598BoqCvZ8fLWc8Sfzmrfoxy7JeuKiv/CeOI+XwkVIJiKm/vCmKwvTp05kzZw6BRryPbd26laCgoColsmFhYWW2eXl50aRJE+x9fciNjpGKBjVAOmNEfVOlZPaHH36oqTgsjqEPF5WVlfQM1hBNVhbn330fbW4uVo6OtJ39Co7+/ri2a2vu0ISocYYWYKjMl7dp06axbNmycvcPGDCAXbt2AfDVV1+hKAqPPfZYlePbvn07b7zxBqtXr67SeePGjSuzbe7cucybNw8Hf39yo2NkmEENsHV3JwfpjBH1h0mqGTREJT9c0k6eojAtDVQq7Ly8zB1avaNotVyc/yl5sXGgUtHqhWdx9Pev+EQh6pGbDUUoz6JFi5g/f365+23+XSnvypUrvPXWWxw8eLDKcW3YsIH77ruPFStWGFwR8mbWrVtHSEhIqW1e/76HOvo3JYWiig7CdDRZ2WRHlp2oJ8O0hCWTZLYaij9c8q8ncfSx/6FoNFzfsxffUVV7Qxc3F758hb56QdD9U2nUrat5AxLCQjg7O+Ps7FzhcXv37iU5OZmuXUv/bU2cOJHJkyfzzTffGDxv5cqVPPnkk/z666+MGDGiyvGFhITQvpxVpoongeXGxaMrLJQlqk1AURTCFn1JYUoKqFS4deqIJiNDhmkJiyfJrAnYeXni0aUzqUePkbBlOz4jR6BSqcwdVr2QuGs3sev+AMCz3200nTDOvAEJUQ9NmjSJoUOHltoWEBDA4sWLGT58uMFzFi1axGuvvcaGDRvo16+fyWNyKL77otORF5+AY4Dcjamu+M1bSA4t6n0PmDJJVvkS9YYksybiPXwoqUePkX3tGllhV3BpGVLxScKg4sUosi6HkfdvKTin5s0ImfGUfEkQogY4OjoaXL3R09OTRo0aAfDEE08QERHBpk1Fy0fPmDEDa2trRo0aVeqcTZs2mSS5dWhasjxXtCSz1ZR19RrXvlsKgGuH9gTcPdG8AQlhQlWqM1uXVabId3WOr0ijbl2x+XcWb8I201yzISpZ/zAvLg7+XZs+6P6pWNnZmTk6IRoORVFK9dZ+/fXX+kS2eH9hYSFZWVmlfkzVS2vt6IBt48YA5MgksGrR5uZy8aNPUAoLsXFzleVqRb1jdDKbmJjItWvX9L8risI333zDs88+y59//mmS4CqrskW+jT2+MlRWVngPHQzA9d170ebmmvT6DYXB+ofA9V17zBCNEMKc9ONmZRJYtVxZ/C15sbEAtHz2aewaNzJzREKYltHJ7LRp0/j000/1v7/++us8+eSTbN68mfHjx7N06VJTxFeh4iLfS5YsqZHjq8J72BCgaCWw63v3m/z6DUFOhNQ/FEIUKa5aIuW5jJf4906u79wFQNMJ4/Do0tm8AQlRA4xOZv/55x8GDy7qidTpdHz99de8++67XLhwgdmzZ7NgwQJTxViuqhb5rm5R8GJhYWGcPXu21E9iYiL23t6433oLAAlbZahBVSmKgtZAryxI/UMhGqLintmc6BiUf4ccicrLiY7mytdFHTcurVsROPUeM0ckRM0wOplNT0+n8b/jmY4dO0ZKSgpTp04FYPDgwQZXdqmMadOmoVKpyv0ZOHCg/tiqFvmuTlHwksaNG0eHDh1K/Xz55ZcAeA8vWts66/Jlsq+FV+txGpqI5Sv+Wye8BKl/KETDVJzM6vLyKEhOMXM0lkWbn8/FDz9Gl5+PlZMTrV58DrW1zPkW9ZPRLdvf359z587Rr18/Nm7cSJs2bWj67+zT9PR07O3tjbpuTRX5rk5R8BvdrNB3ox7dsHFzpTA9g/it22jx+KPVfryGIGbtemLWrAPApX1b7Bo3JjcqWuofCtGAlaxgkBsdjZ1nYzNGYxmKq8GkHT+OJjMLgJYznsS+SRMzRyZEzTE6mZ0+fTovv/wy27dvZ+PGjbz33nv6fQcPHqRtW+OWGa2pIt/GFgU35GaFvtU2NjQZMpiYNeu4vnsPwdMekFn4FUj8eyfhS5cD4NK2De3nviavmRACG3d3rJwc0WbnkBMVrR/GJQwrrgZz4zLr9r4+ZoxKiJpn9DCDmTNnsnDhQnx8fFi4cCFPP/20fl9KSgqPPPKISQIsz6RJk7h69SonTpzQ/wAsXryY999/v9rHV0fxRDBtdg7J+w+Y9Nr1TcrhI1xeWDREwzEokHZzZtbbRLay5eBCQ0MZPHgwLi4uuLu706dPH3Q6XS1FKUTdoVKp/psEFiOTwCpiqBqMotUSs2a9mSISonZUawDNAw88wAMPPFBm+9dff12dy1ZKVYt8V+Z4U3Hw88O1Q3syzpwlfut2mgweZNLr1xcZ585z8aNPQKfDrkkT2s19DetK9MpbosqWgwsNDWXUqFF8/vnnbNiwAVtbW44dOyaLRYgGy6FpUzIvXiInSspzVSQnIsLwdqkGI+q5ao8Gz8vLIzY2ljwDs9DbtWtX3ctXyY2zXStKqmtydqzPiGFknDlL5vkL5ERGymz8G2SHh3Pu7XfRFRRg4+ZG+zdeq7e1D4vLwe3bt4+goKCbHvvyyy/z8MMPl/qS2LNnzyo/pqEJmF5eXjSRcXPCwjgESM9sZUg1GNGQGZ3MRkdH89hjj7Fly5Yy+xRFQaVSodVqqxWcJWvcqyfWLs5oMrOI37qD5o88ZO6Q6oTs8HAiV/5CyrFjoNWhtrOj3dw5OPj5mTu0GlGVcnA5OTkcOHCA3r1706NHD65cuUJwcDCzZs1i4sSqLT05bty4Mtvmzp3LvHnzqnQdIcyteFnbwtQ0NFnZWDs7mTmiuilh6zbyExLLbJdqMKIhMDqZvf/++7l69SqLFi0iJCQEW1tbU8Zl8dS2tjQZNJDYPzZwfdcugh+YirqBv0bZ4eGcenkWuvz8/zYqOlRWlrWq8rRp01i2bFm5+wcMGMCuXbuAqpWDS0lJQafTsWzZMjZs2EDnzp35448/mDJlCrt376Z3796VjvFmFTeEsCSOAU31/86NicGldSszRlM3ZV4O4+o33wHg1LwZDv5NpRqMaFCMTmaPHj3KypUrGTNmjCnjqVe8hw8j9o8NaDKzSA49hNcA06xZbqkif/qldCIL6AoKiVmznlbPP2OmqKqupsrHubi4AEXJcvfu3QGYMGECgwYNYt26dVVKZm9WcUMIS2Lv7Y3K2hpFoyEnKlqS2RsUZmRw8YOPUDQabNzdaffabGwbeZg7LCFqldFdYu3atavUhJaGzDHAH5e2bQCI37rNzNGYV05UNKlH/zG8z8ImJzg7O+Pp6Vnuj5ubG1C6HFzxPigqB2eop9bNzY0WLVrIZC8hSlBZWeHg5wvIuNkbKVotlz5eQP71JFCraf3y85LIigbJ6GR24cKFfPDBB+zfv9+U8dQ7Pv+uCJZx5iz/PDmDix8vIDs83LxB1bLMS5c5PXMOSjljqOvr5ARjysHNmDGDpUuXcuLECXQ6HX/88Qe7d+9mwoQJtRi5EHWLw7/luaSiQWlRq34j7cRJAIIfuA83uRsjGiijhxnceuut9OjRg/79+2Nra6u/RVpSYmLZwegNjb1/yfFeseTGxJJy+AidPninQYxjSjtxkvPvfYguLw/Vv7fflcJC/f76PDmhquXjAJ555hlycnK48847SUtLo2XLlqxatcqoigZC1BfFy9rmRksyWyzl6DGiVv0GQOPePfEbJ0P+RMNldDL7yCOP8Ntvv3HXXXfJBLCbiPtzY5lturw8ixsnaoykA6Fc+ngBikaDtbMz7V6fjdrOlpg16/Xlyhra5ITKlI+bOXMmM2fOrK2QhKjzihdOyEtIRFdYiPrfL8YNVV5CApc//RwAez8/Qp7+v3oxPEmn01FYWCiLxDRgarUaGxsb1OqqDRwwOpldu3Ytn376KU888YSxl2gQyhsPamnjRKsqfus2rnz1Deh02DZqRPs3XtMPJ6jvSbwQwrSKe2bR6ciNjcMpqH4OTaoMXUEBFz6YjyYrC7WdHW1nvoS1gTtAliYzM5PCwkLs7e3rRWIujKPVasnOzsbGxsbgHf/yGJ3Menl5VVg3UxSNB80JL7sqS30dJ6ooCjGr1xLx40oA7H19aP/GXOy9pVi/EMI4Dk3/q0OdGx3TIJPZ7PBwolevI+2f42iysgAIeep/9eKzRKPRoNPpTL4ap7BMjo6OpKeno9FosLauXJpqdDL7+uuvM3/+fPr3749zPV2C1BT8J44j5fCRMutl+4waYaaIakbRG+1a0k6cQpORARTVO2w3dw627u7mDU4IYdGs7O2x8XCnMDWNq998S/LBQ/hPHNdghihlh4dz6pXZpT5HVFZWOAYFmDEq0yksLJShiqIUW1tbCgsLaz6Z3bhxI5cvXyYwMJBu3brhfkPColKpWLVqlbGXrzecgoPp9ME7xKxZT+aly+TFxQGQuHM3rv+W7bJ0BhdDUKtp/tjDksgKYSIRadGsP7+VqPRYAtz8GNt2OEHu/uYOq1Zkh4dTmF70JbkwLY2kPXsb1ETa6NXrynSIKFptg5h7IURlGJ3MJiUl6VcYKiws5Pr16yYLqr5xCg7Wv+Fc+2EZsev+IGHLVjz79sG9U0czR1d94ctWlFkMAZ2O+E1bcW3b1jxBCVGPRKRFM2fHfPI1RX9nEekxHIk9xdtDXmwQCW306nVww6SghjKRFiD76jWD2+v73AshKsvoZHbnzp2mjKPBCLx3CimHj5AXG0fYoi/p/NknWDk4mDssoyg6HTHr/iDtn+MG98sbrRCmsf78Vn0iWyxfk8/6C9t4utdDZoqq9jTUibQA2txcClJTDe6rD+NlhTAFoxdNuFFhidqhonxWdnaE/N+TAOQnJBKx4iczR2Sc/OtJnHltHhHLfiz3GHmjFcI0otJjq7S9vinvvaS+v8coWi2XPv0MbXZ2mX31uUa3JVu5ciWtW7c2dxgNTrWS2QMHDjBq1ChcXFywt7fHxcWF0aNHExoaaqr46iW39u3wvX0UAHEbN5Fx7ryZI6qa63v2cfyZ58k4cxYAl/btUNvZlTpG3miFMJ0AN78qba9v/CeOQ21vX2pbQ3iPCV++gpRDRwBoMmwoXgP649QsGK8B/RvMeOGquBabzvwVx5gxfyfzVxzjWmx6jT3WwIEDmTNnTpntU6dO5eLFi/rfp02bho2NDc7Ozvqf4Er8v+3atQuVSlXqPGdnZzZv3mzKp1FvGD3MYNu2bdx+++20bt2al156CW9vbxISEvj9998ZOHAgGzduZOjQoaaMtV4Jun8qKUeOkZ+YyOWFX3Drgo+xuiEhrAuKy8HkREbi4OeHriCf1KP/AKC2taXZww/hPWIYORERDXoxBCFq0ti2w9kfeRSF/xbdsLO2Y2ybYWaMqvYUT6S9/PkXZF+5Cmo17d+aV6/fY+K3bid23R8ANL6tNyFPPo6qioXkLV1+oZb45LK90obEJGbx8U//UFBYtGx6eFwGoWfieOHeLjRtUnHFJZ/GTtjZWFUr3vJMnjyZFStWGHVuWlpapWf0N2RGv0KzZ89mzJgx/Pbbb6UKHL/++utMnDiRWbNmSTJ7E1YODoT83/84+/ob5MXGEfnTLzR76EFzh1XKjeVgStbLdQ5pQcvnnsHx32LmJSe5CSFMy8nGUZ/Iejp60MarJWPbDGsQk7+KOQUH0/qF5/jnyRmg0+lLANZHaSdPcfXrbwBwbtmSls/MaJCJ7GPvbiMlI7/ig8tRUKjlvWVHKnVsI1c7vpk1rNoJ7dKlS5kzZw7RNbz08rRp08jNzaVRo0b8+uuvqFQq5syZw8SJE3nkkUc4cOAAAQEBfPfdd/Tu3btGY6kLjP7rOH36NI8++qjBlToee+wxTp8+Xa3AGgL3WzrhPaKoZyX2jw1kXrxk5ohKM1QOBsAxMICOH7yrT2SFEDVrf+RRAGysbPh45Os83euhBpXIFnNo6oe9X9HQitQjR80cTc3IiY7mwgfzUbRabD09aTv7lTp5106Y37p16xgyZAiJiYl8++23PP/88zzwwAPMnz+ftLQ0hg0bxrRp08wdZq0wumfW3d2dK1euGNx35cqVMnVnhWHBD95P6tF/KEhO5uy8N7Hz8sIxKKhOFAQvb6awysoKtdz2EKLW7Iss6l3q6tcRBxv7Co6u3xp170rs+lhSjhyj+RNKvVr6tDAjg/NvvYc2Oxu1vT3tXpuJrYeHucMyCzsbK76ZNazSwwyWbjjH0fMJZbZ3b+vNg3e0q/D8mhxm8Ouvv7Jhwwb97y+99BKzZ8+u1Lmenp76fzdt2pSzZ8/qf+/bty933XUXAOPGjcPNzY3hw4fTsWNRyc8HHniAzz//nPT0dNzc3EzxVOosozOSu+++m5kzZ+Lq6spdd92Fvb09eXl5/P7778yaNYsHH6xbt8zrKmsnJ/zvGs/Vxd+izcklJyKSnIjIOlEQ3NbdnRwazlK8QtRF0elxRKQV3bLsG9jdzNGYX6Pu3Yhd/ycFyclkXwvHuXkzc4dkErrCQi689yF58fGgVtP6pefN3qFhbnY2VgT5uFbq2AdGt+XMlSTyCrT6bfa2Vtw/um2lr1FTJk2aZPSY2aSkpHLHzPr6+pb63cnJqdQ2JycnADIzMyWZLc8HH3xAcnIyDz74IA8++CDOzs5k/bte9D333MMHH3xgsiDru4zzF8tsM3dB8LyERLLCyva8N4QZxELUJcW9so42DnT2bW/maMzPpW0brJwc0WbnkHr0mMUns/qlwP85jiarqBey2fQHadStq5kjsyzN/Nz4cEY/1uwMIyI+gyAfVyYMCqGZX/1O4kQRo5NZBwcHVq5cyWuvvcaRI0eIi4vD19eX7t2706ZN/VimtbaUdzs/O8I8BcE1Wdmce+sdNFlZYG2Ne8cOFKalSZUCIWqZoij68bI9/TtjY2Vj5ojMT21tjUeXziTt3U/K4aMETLrL3CEZ7cZJtlA0jMu1g3xpMUYzPzdemFp7XwK0Wi15N8wr0d2wUp2oHUYns3v27KFLly60adOmTPKanZ3NsWPH6N+/f7UDbAgcAwNLVQoops3NQZufX6uD/3WFhVz44CNyo4pua7Z+7mk8+95Wa48vhPjPlZQIErKKlgq/LbCbmaOpOzy6dSNp736yLl+mIDXVYseVGppkq2i1xK79Q6rDWID333+f999/v9S2JUuWmCmahs3oZHbQoEGEhobSo0ePMvsuXLjAoEGD0Gq1Bs4UN/KfOI6Uw0fKvKnlJyRy+pXZtJn5Evbe3jUeh6IoXPn6G9JPFVWiCLp/qiSyQphR8RADN3tXOjSRVYWKeXTtDGo16HSkHvsH76FDzB2SUYoXnrlRQ1im19Lt2rWr3H2PPPKI/t9Lly416voDBw5EUZRy9xu67o3lwNq0aXPTa9QnRpfmutkLlJWVhaOjo7GXbnCKC4IXr+7i2b8vngOKerWzr13j5PMvk3rsnxqPI2b1WhK3/w2A97ChNJ04vsYfUwhhmE6n48C/QwxuC+iKuoHVGb0ZGxcXXNsW3RFMOWyZJbpi1v1BQUqKwX2WPMm2NlfhEqJYlXpm9+zZU+rbyLfffltmabW8vDw2btyoLw0hKsfQogONunUlbNGXaLKyOPfWuwRMmUTApLtqpHj29b37ifhxJQBut3Si+ROGawgLIWrH2euXSMsrWhjgtiCpYnAjj25dyTh7jrSTp9AVFKC2tTV3SJUW9evvRK78uegXlQpKdA5Z8iTba7HpvLxkI1rPMNSNs4hJd+bQkhA+fPR2mYh1g5UrV/L4448b3PfUU0/JJPoqqlIye+jQIRYuXAiASqXit99+K1MywtbWljZt2vDRRx+ZLsoGyqt/XxyDArnw/ofkxcYR9fMq0k6dwtbNndzYWBwDA6tVj7Z4qdqsS5fJSyiqz+cYGECbV16UOrJCmNn+iKIhBt5OnoQ0CuZabDqr//5vpvbEwTUzU3v8+PGsW7eObdu2VWoVx6oebyqNenQjYtmP6PLySD9zFo8unWvtsY2lKAqRK38m+rfVADi3akmzhx4gfvO2erEU+I9/H4KWB7C2KhpiqHbMRPFIZMVOT16bOtzM0dUtU6dOZerUqeYOo96oUsby0ksv8dJLLwHQrFkz1q1bxy233FIjgYkiTkGB3DL/Ay4vWEjK4SNknj2v35cTHmF0PVpDs2gBgh56AOt/a9MJIcyjUFvIwejjANwW1I3wuAxeXrhXX0MzPC6DQ2fj+HBGP5MmtMuXLycnJ6fGjr+ZiLRo1p/fSlR6LAFufoxtO/ymq5w5NG2KvY8PefHxpBw5WueTWUVRCF+6nNh1fwDg2q4tbV+bhbWjI67tKi7qbwku5R9F5Vx6rozKSsvl7GOAJLOi5hh9v/ratWuSyNYSaycn2sx8GcfgoDL7dHl5XPr0c9LPnEWbX7SGdXZ4OBc/XsDxZ57n4scLyA4P1x+vyckh/ew5Ln/+hcGlapN27a2x5yGEqJwT8efIKcwFoG9gD1b/HVaqGDxAXoGWNTvDTPaY0dHRzJkzp9Kzsat6vCFhYWGcPXuWv4/tZta2D9kXeYSI9Bj2RR5hzo75+sUiDFGpVHh0L6rwkHrkaJ2e6KLodFz95lt9IuvWqSPt5s7Buh7NLdHpFAqt0wzuUztm1W4wosGp9r3kffv2cenSpTK11gCefPLJ6l5e/Otm42RzwiM4M/t1VNbWODRtSm50NMq/lSRywiNIPhCKa/t25CckFq0ucxMyi1YI89v37xCDILem+Lv5EhF/weBxEfEZJnk8RVGYPn06c+bMIbASk4+qenx5xo0bB0CP/xtCUN+Wpfbla/JZf2EbT/d6qNzzG3XvStyfG8i/nkRORCROBr7wm0vxMK6ciEh0+fn6916Prp1p/cpLtVpysTb8deAaWkVnsIesdRPLndAmLIPRyWxCQgJDhgzh3LlzqFQq/bfikpOGJJk1rfLq0apsbFAKC1E0GnIiyu5XNBrST54qfY6VlT7hvfExhBDmk1uYx7HYor/X4olfQT6uhMeVTVxvtkzntGnTWLZsWbn7BwwYoJ/Q+9VXX6EoCo899lilYqzq8eVZt24dISEhfB32Mwl5yWX2R6XH3vR813ZtsXJ0RJuTQ8qRo3UmmS1vGJdrx/a0mfkKapv6tfhFbFIWS/fsQt0iu8w+W7Ut93a53QxRiYbE6GT2hRdewM3NjaioKAICAjh06BDe3t6sWLGC5cuXs3HjRlPGKTBcj1Ztb0/H999GbWNL5vnzhC9drl8SsSQrBwd8bx+FU/PmODVvhjY3l9Mz55S5lqXOohWivjgSc5ICbSHw30IJEweHsPdENLoSd9Ltba2YMCik3OssWrSI+fPnl7vf5t+E6sqVK7z11lscPHiwUvFV9fibCQkJoX379rTMaE5CZNlkNsDN76bnq21scO98C8n7Q0k9cpSAuydWOyZTMLQYAoCth0e9S2S1OoWPfzmAKvAkAI0cGtHGsxkxmQlFY5/bDLvp2GchTMHoZHb37t189tln+Pr6AkW3nQIDA5k1axY6nY4nn3ySLVu2mCxQ8V892pg16w3OfHX0b0raqTMk7Sk77rVRj+4E3V965uTNriWEMI/i5Wtbe7bAy6kxAK5OtvpE1tPdgQ7NG1e47ryzszPOzs4VPt7evXtJTk6ma9fSy4BOnDiRyZMn880331Tr+MoY23Y4h2NO6JN4ADtrO8a2GVbhuY26dyN5fyiZly5TkJaOrbv5S0CVN1yreGXF+mT97jDCbfZiZVOAGite6fcEzTwCzB2WaGCMTmbT0tLw8vJCrVbj6upKYmKifl+fPn2kRloNMVSPtqTyem8N9bhWdC0hRO3KyMvkZPw5APoG/ldb9sCpOACsrVQsfGEgzo6mq6k6adKkMiW1AgICWLx4McOHl52BXtXjKyPI3Z8nOj3B5/8sApWCk86LRzvdW6kePY+uXUqsBnYM7yGDjYrBlMobElbfhnFFJWSy8tgmrAKSALjvlnENPpFduXIlb775JhcvXjR3KA2K0dUMmjVrRlxc0Rts+/btWblypX7fn3/+SaNGjaofnaiyG1cT8xrQ36jSXUKI2ncw+h90ig61Sk3vgC767XtPxADQuXUTkyayAI6Ojvj7+5f6AfD09NS/jz/xxBOMGjWq0sdX1bXYdD774Sq6HBcA0q878On3YZVaPcrG1RWX1q0ASD1yzKjHNzXPPr3LbKtvw7i0Wh0f/LYDddOiyYntPNswurV5v0hEpEXzeej3vLT5bT4P/f6m1TCqa+DAgcyZM6fM9qlTp5ZKZKdNm4aNjY3+TomzszPBlfg83rVrFyqVSn+Oj48PkydPJioqSn9Mfn4+s2bNIigoCCcnJ4KCgli+fHml4h8zZgxNmzbF1dUVX19fHnroIZKTyw71sRRG98zefvvtbN26lUmTJjFnzhzGjh2Lv78/NjY2REZGSs+sGUmPqxCWqXiIQSfvNrjaFyV2SWm5nA8vWva07y1NayWOG8tcff3111U6vqqKS4/Z5LigdspA5ZCpLz32wtSuFZ7fqHs3Ms9fIPX4CXSFhWYfl5pytOj/UWVjg4OfH07BQfVuGNeqv88T77wPtVrBwcqJZ297CLXKtKtTFmgKSMhOqtSxsZkJLDz4g36oSkR6DIdjTjCj10P4uXhXeL63kye21jWzitzkyZNZsWKFUeempaVhbW1NXFwcd999N/fff79+4ubdd99Nbm4uO3bsoEWLFly/fp3U1NRKXfett96iVatWODg4kJqayv/+9z8ee+wxVq9ebVSc5mZ0Mvvee+/p/z1q1CgOHDjA2rVryc3NZdiwYfpv8UIIISqWlJ3C+etFdWNvKzHEYN/John9NtZqenXwMUtsNa24xFhxz6zaMRNQKl16rFH3rkQsX4EuL4+Ms+dwv9V8NdALUlK5vmsPAP53TSBwyiSzxVJTrsWms/rieqy8iiYbP9NnGu725VfWMEaBpoAZG18nNa/i3vlyr6Et5OP9lRvD7WHvxsLb36x2Qrt06VLmzJlDdLRpe4V9fX2ZMmUKM2fOBGDHjh1s27aNiIgImjRpAkCTJk30/67IjesEqNXqUj3K8+bNY/v27QwePJglS5aQnZ3NE088wcyZM3niiSfYvHkzHh4efPbZZ4wd+98dh40bN/LWW29x8eJF1Go1gwYN4vfff6/u06+QydYs7datG926dTPV5YQQokE5EFXUm2djZUN3//8+aPadLBpi0KV1Exzt69dM+GLFpceU3KJkVmWtAZv8m5YeK8khIACbRo0oTEnh4vxPce98a7WW+q6O2A0bUTQa1La2+I4aUeuPX9MKNTreW/snVl5Ft7uHNhtAF78OZo6q/ouOjubnn3+mR48eAGzbto1mzZrxwQcf8NNPP2Ftbc3QoUP56KOP8PT0rNQ1Z86cyaJFi8jKysLBwaFMKb9Dhw5x9913ExkZyYkTJ+jTpw9///03n376KT///DMLFizgoYceIjo6GkdHR7Zt28Zdd93F8uXLGTt2LDqdjv3795v8tTDEZMmsEEII40SkRbP23GYAXO2cuZ6dTJC7P4kpOVyMKLpt2O/W2hliYA4TB4dw8Gwc+f/2zALYu2bftPRYSTkREWjSi3rwNJmZJO3Za/RS39WhycklfnNRFZ8mQwZh42b+ygqmtmzbMVLdD6MCvOy8eahrzZRDs7W2ZeHtb1Z6mMGKk2s5HnemzPYuvh2Yesv4Cs+vyWEGv/76Kxs2bND//tJLLzF79uxKnevp6YlKpcLV1ZX+/fvz/vvvA5CUlMT58+cZPHgwYWFhZGVlcd9993H//fezadOmSl37vffe47333iMsLIylS5fSqlWrUvsDAwN55pmiIYvdu3enQ4cOdO3alX79+gHw4IMP8sILL3D58mVuueUWPvvsMx5++GHuvvtu/TWGDBlSqViqS5JZIYQwo4i0aGZv/1A/1i85J5U5O+bz9pAXOXayqCqJrbWa7u0qHvdnqZr5ufHRjH68+PkelAJbVLYFjB7S+Kalx0qKXr2uzCIwurw8Ytasr9X5AwnbtqPNzgGVCr+xd9ba49aWS1EpbI5Zh9qlELVixczBj2NjVXN3C2ytbSusNVzs3k5jOXf9MvmafP02O2s77uk0ttLXqCmTJk0yesxsUlIS1tZlUzVXV1dUKhUffvghjo6OODk58eabb3LbbbeRk5ODYxWWSg4JCWHMmDGMGDGCqKgofQ3q4tKrxZycnEptc3JyAiAzMxOAa9euMWKEee5GSDIrhBBmtP781lL1VeG/pVyvnWwOQLd23vV2iEGxZn5udGjuyZlcF6xsk8nUVX5mdXl1XWtzeW6dRkPsH0W9b4179cThhkTAkh24fIHlR/4kWReN2qUAgPtuuQt/17rzHIPc/Xl7yIusv7CNqPTYer9gQ5cuXQxuL7kia1UUFhaSkJBAenp6pYcp3Cg4OJhLly4ZdW51mXbqoRBCiCopr3zQteRowqLSgNqrYmBugT4u+klgkRUsZVtSefVba7Oua/L+UAqSim6JNx1ff0pwHbh8gU+PLCLF6ioqm6JEVlFUeFjVvTsFQe7+PN3rIT4aMZunez1U44msVqslLy+v1I9Op6vRxyw2fvx4mjZtyqxZs8jLyyM5OZl58+YxevRofY9peS5dusSaNWvIyMhAURQuXrzISy+9RPfu3Y1OZAGeeeYZvvvuO1avXk1BQQF5eXns2LHD6OtVhVHJbF5eHsOHD9eXhxBCCGEcK7XhG2RWhUWTn+xsrejetu4lDjUh2NdVPwksJiMOjVZTqfP8J45DbW9fZrtbx/Ymja88iqIQs3Y9AK7t2urr3tYHPx7dgMqq9BAOlUphxTFZsv7999/HwcGh1E9tJbNOTk5s27aNs2fP4unpSceOHQkICCgzicsQRVH45JNPCAwMxMXFhREjRtCxY0f++OOPasU0fPhwfv75Z95//328vLzw9/dn8eLF1bpmZRk1zMDe3p4jR46gvWGMkhBCiMrT6LSk5qaV2W5nbUdOZBAA3dt6Y2/XMEaEBfm4osspWoJXq+iIyYyvVO9ayaW+s8MjyEtIQJeXR/Tva/Hs3w8rO7sajTv95Cmyr10DwG9c/emVBUjTJIGBeVHpGsstsG8KN+vMe+SRR/T/Xrp0qVHXHzhwYIXDBdq0acO2bduqfO3WrVuzb9++mx4zb9485s2bV2rbjefY29uXiXHs2LGlSnXVFqOHGYwZM4Z169aZMBQhhGhYDkYdIz2/aPLErT7tCHL3p29QD57p8iRRESqgflcxuFGAjwvkOVP8+RiZVvmhBsWLxXT+/BPavTYLgLz4eCJX/lwToZZS3Cvr4N+URt0rXuTBkqh0Vga3u1k3ruVIhCif0V/3R4wYwUsvvURcXByjR4/G29sblUpV6pjRo0dXO0AhhKiPFEXhz4vbAWjj2YJZA2bo963aXlS83MHOiq4NZIgBgJ2NFb6NXEnOc0LlkE1Eegz9jLiOW4f2+IwcTvzmrcT+uRHP2/rU2K3/7GvhpJ04CYDf2DGo1PVnKsrec5cotElDdcN2RWvF/d3vMEtM9cXKlSt5/PHHDe576qmnqrWKamRkJO3atTO4r1OnThw4cMDoaxujJp9rMaOT2fvuuw+ANWvWsGbNmjL7VSqVDEMQQohynLt+mWupRYXn72g9tNS+fSeKeiR7tPPFzsZwz1h9FeTryvUcF9QO2USlxxh/nQfvJ+XoPxQkJRG26Etu+eSjGlnmNmZdUa+sjbs7TQb2N/n1zSUzN48vj/yAyl4HWivcaEq2NgM368bc3/0O+rRsY+4QLdrUqVOZOnVqjVw7MDCQrKysGrm2MWryuRYzOpm99u/4ICGEEFVX3Cvr69yEbn6d9NujEjIJjytaxrXvreatj2kOwb6uHDnjAsQTkWZ8Mmvt6EjIk49z7s13yImMIvq31QTeO8V0gQL515NI2lu0wpHvHaNR29ZM0X1zeOPPZWjt0wAYHzKBe3oMNm9AQtyE0clsUFCQKeMQQogGIzojjn9iTwNwe+vBqEvcmt53oiiBc7S3pkvryq2zXp8E+biiO1xU0SAlN42s/Gyc7W5eaqg8Hl274DVwANd37Sb69zU07tPLpCuCxW7YiKLVora3x2fkcJNd19zWHT1IpHICAF91a0lkRZ1XrcE9+fn5fPXVVzz88MMMHz6cy5cvA7Bq1SrOnz9vkgCFEKK+2XjxbwBcbJ0YENy71L69J4uGGPRs74NtAxtiABDk64Lyb0UDgMhqDDUAaPbwQ9i4uaFotVz+/MsyK4UZS5OdTcKWopnk3kMHY+PiUsEZliE+PYWfL/4CgLrQiXmjH6ngDCHMz+hk9tKlS7Rq1YqZM2cSHh7Ojh079Eua7d27l/fee89kQQohRH2RnpfBnvCDAAwL6Y9difXgI+IyiEooeh9tSFUMSvJt7IS1zglFW5TIV2XxBENsXF1o/vijAGRfucKJF17m+DPPc/HjBWSHhxt1zezwcE7Pfh1tbi4A7l06VyvGukJRFOZuXoxinY+iU/HorQ/g4eRc8YlCmJnRyezTTz9NYGAg4eHhbNmypVStsQEDBlRYw0wIIRqiLWF7KNRpsFZbMzJkQKl9e08W9UI6Odhwa6uGN8QAwMpKTYC3q34lsOqMmy3WuE8vXDt2BCDnWjg54REk7dnLqVdmVzmhzQ4P59Qrs8i59t95Fz/82OjEuC75Zt+fpFK0BHBbu94M6dCpgjOEqBuMTmb37t3LzJkzcXd3L1OSy9vbm7i4uGoHVxXjx49HpVKxffv2mx4XGhrK4MGDcXFxwd3dnT59+tTaih1CiIatQFPAlrDdAPQP6oG7g5t+n6Io+vGyvTv4YmNdf0o8VVWQj4t+JbDqDjOAouo61k6OZbbr8vKIXr220tcpSEnl8udfoMvLL3OdmDXrqx2nOZ2Lv8aO6C0A2OR6M+sO006WE6ImGT0BzN7entx/b7HcKCYmBnd3d2MvXWXLly8nJyenwuNCQ0MZNWoUn3/+ORs2bMDW1pZjx46VScaFEKIm7A4/RGZ+UcmcG8txhcdlEHM9G2iYVQxKCvZ1ZU9UcTIbi07RoVZVL7nPi483uD15fygR3t40GTIYB18fssPDiV69jpzISBwDA/EdNYLc2Fiu795L+pmzUE7nR05kZLXiM6d8TQEf7PoG1DqUQhte6Dcde1vTlzFrCFauXMmbb77JxYsXzR1Kg2L0u8OwYcN49913SU9P129TqVTk5+ezcOHCWlswITo6mjlz5rBkyZIKj3355Zd5+OGHeeCBB3B0dMTa2pqePXtKMitqVUV3Edq3b4+zs7P+x9HREZVKxdq1le9BEnWPTtGx4VLR/3ln3/b4u/nq912LTWf+imMAWFupcHOu2eVX67pAH1f9JLB8TT6J2dVfOtUxMNDgdkWrJfq31fzzxFOceOFlTr74Kkl79uqHIpyeOYewhV+Sfup0USJbzudFede3BB/9vZxcVRoAvdxG0qWFZT6X7PBwLn68oNpjoitj4MCBzJkzp8z2qVOnlkpkp02bho2NTan39OBKVNTYtWsXKpVKf46Pjw+TJ08mKirKlE+j3jC6Z/ajjz7itttuIyQkhGHDhqFSqXjzzTc5e/YsBQUFBhdSMDVFUZg+fTpz5swhsII3kpycHA4cOEDv3r3p0aMHV65cITg4mFmzZjFx4sQqPW5YWFiZbV5eXjRp0jDHuInKq8xdhLNnz5b6/fPPP+fNN99k1KhRNRmaqGH/xJ4hLjMRKN0rey02nZcX7iWvoGiWvUarMPOLfXw4ox/N/NwMXqu+C/Z1RZf7X3WAyLQYfJy9qnVN/4njSDl8BF1enn6b2tYW966dST9xCm1uLtlhV8o9361TR7z698WhqR9n33in9HXs7Wk6ofbXo6+OiLRo1p/fyvnrYSTnpgLgmNmSpyeONHNk/9Hm55OfkFCpY3NjYrn0yWfoCgoAyAmPIOXgIVo9/wwOTSu+02Hn7Y2VXc18iZw8eTIrVqww6ty0tDSsra2Ji4vj7rvv5v7772fXrl2mDbAeMDqZDQgI4OTJk3zyySfs2LGDFi1a6F/s559/nsaNjVu3edq0aSxbtqzc/QMGDND/R3711VcoisJjjz1W4XVTUlLQ6XQsW7aMDRs20LlzZ/744w+mTJnC7t276d27d4XXKDZu3Lgy2+bOncu8efMqfQ3R8BTfRdi3b1+V6jQXl7+zt7evwehETdvw7yIJQe7+dGjSWr999d9h+kS2WF6BljU7w3hhatdajbGuaOxmj5ONA5p8e9R2eUSmx9DD/9ZqXdMpOJhOH7xDzJr1+iEETSeMxSk4GG1eHskHDnJl8TdlxsMCOAT40+Gtefrfy7uOpYhIi2b2to8o0BXotykKTO83vM6M1dbm53Ps8acoTE01+hq6ggIuvP9RpY618fCg6+Ivqp3QLl26lDlz5hAdHV2t69zI19eXKVOmMHPmTKCo53bQoEH88ssvvPbaa0RHRzNkyBB+/PFHPv74Y7799lsKCwt56qmneOONN0waS11kdDIL4OHhwVtvvcVbb71lqnhYtGgR8+fPL3e/zb/LEV65coW33nqLgwcPVuq6Lv/WAJw2bRrdu3cHYMKECQwaNIh169ZVKZldt24dISEhpbZ5eVWv10DUb1W5i1DS33//zaVLl3jiiSeq9Hhy96BuuZISwbnrRXW472w9tNTQpoj4DIPnlLe9IVCpVAT5uhKW6wJ2eUSmVa88VzGn4GBaPf9Mme1W9vY0GTyQ1OMnSNqzt8x+5+bNK3UdS7Hyn42lElkoGj2xP2Y//Vt3MFNU4maio6P5+eef6dGjR6ntGzdu5NixY+Tm5tK3b1969uzJjBkziIqK4ujRo/Tt25eRI0dWKcexRNVKZqGoC/zMmTPExcXh5+dH+/btqzX5q3h8SEX27t1LcnIyXbuW7rmYOHEikydP5ptvvim13c3NjRYtWphkfGxISAjt27ev9nWEZaupuwglffnll4wcOZJmzZpV6Ty5e1C3FC9d28jBnT6B3Urt8/N00i9fW1KQj2utxFZXBfm6cinSGSv360Skm7aXqzwGhyJY4BCCilxMjAQDH4UXE+vOJDYrOzu6Lv6i0sMMwpf9SOrRf8ps9+jWleAH76vw/JocZvDrr7+yYcMG/e8vvfQSs2fPrtS5np6eqFQqXF1d6d+/P++//36p/e+88w4uLi64uLhw++2389dff/F///d/APTq1Yu2bdty+PBhSWbLo9FomD17Nl988UWpMYCOjo48+eSTvPPOO/pe1JowadIkhg4tPRs4ICCAxYsXM3y44WUFZ8yYwXvvvceUKVPo1KkTGzZsYPfu3bz55ps1Fqeov2rqLkKx2NhY1q9fz7p166ocm9w9qBsi0qL59fQGjsSeBKCXfxes1aVX9XJ3LfsBam9rxYRBIWW2NyRBPq4oF4ruqMVnXidfU1BqgYmacLOhCPWJNt8aDIxa0uXUrQUSrOzsKj2xLuj+qaSfOVfmi0jQ/feafXLepEmTjB4zm5SUhLV1+amar+9/E0mdnJxK/V68rXhBq/rM6GT2+eef55tvvuH1119nwoQJNGnShMTERFavXs3bb79NXl4en3/+uSljLcXR0RFHx7J1Az09PWnUqBEATzzxBBEREWzatAmAZ555hpycHO68807S0tJo2bIlq1atomfPnjUWp6i/auouQrFvvvmGgIAAoyZ+yd0D84tIi2bOjvnka/4bg7nj6j4GNe9NkLs/AAWFWg6cLKrJ7dPIEQd7a4J8XJkwKKTBTv4qVnISmIJCdEYcLRpVfqy5sSx9CEFFMvIy0dikl9muaK1oaWe5Y7QbyhcRYZjRyeyPP/7Iu+++y/PPP6/f1qhRI2bPno29vT1vv/12jSazhpRchQzg66+/LnPMzJkz9QOohagNxtxF0Gg0LFmyhKeffhq1um5MyBBVs/781lKJLEC+toD1F7bxdK+HANj1TzRpWUXHzHqoR4NPYEsK8nFByXNC0alQqRUi02JqJZmtz3SKjk/3/4DOKh9FB7oMT1Q2+ehynbFKCuG+Ry27Y6e2v4hotVrySvQEA7IIk5kY/SmpVqvL7fnp0KGD1G4V4l+Ojo74+/uX+oGydxFK9sCuX7+e5ORkHn74YbPELKovopyVq6LSiyYzKYrCut1FpaBubeUliewNnB1taezqiJJXdPejvNdTVN6Gizs4m3QeAF1sG261uQO/1FHc5n47Hz56u7TBKnr//fdxcHAo9SPJrHkY3TN7//338+233zJixIgy+5YsWcJ991U84FqIhqqiuwgTJ06scv1jUbcUaAsMbg9wK6p5eexCIlEJRWPZxg9o2ONjyxPk48rpHBfUjplESTJbLWHJ4aw8uQ4AbZonD3S7nbHS7ox2s1qvjzzyiP7fS5cuNer6AwcOLPM5UdH+t99+u8xxVZ2rYamMTmaDgoJYvXo17du3Z8yYMfoxs+vXryczM5MXXniBL7/8Eigqs/K///3PZEELIURdti/iCAlZSWW221nbMbbNMADW/9srG+TjQufWMjnPkCBfV05dKOqZDU+LQVEUuetnhJyCXD7evwQFHUqBHS10A7izXwtzhyWEyRidzL7wwgsAxMTEcP78+TL7S46llWRWCNFQxGcm8s3RlQA0dfEh2MOf6Ix4Atz8GNtmGEHu/lyLTefE5esAjBtgmpKB9VGQjwu640WTwDLzs0jPy8DdQW6FV4WiKHx9ZAXJuSkoChDZmeef6I1aLW3OnFauXMnjjz9ucN9TTz3FBx98UMsRWTajk1kZFyKEEKUVagv5NPRb8jT5OFjb80r/Jw0uw1o8VtbdxY4BXfxrO0yLEeTrii6nxLK26bGSzFbRjqv7OBhdVH9VE9uCxwb1x6exk5mjElOnTmXq1KnmDqPekGnSQghhIitPruVaahQAj3efajCRTU7PZc/xokUA7ritGTbWVmWOEUUCvF1Qa+xQNEU1myPSZNxsVUSmxfDDP78CoM3woL1jL0b2DjZvUAZYWVmh1WorPlA0GFqtFiuryr83SjIrhBAmcCTmJH9d3gnAkOZ9y6z0VWzDvmtotAq2NlaM6lO1ld0aGjsbK3w9nfW9s5EyCazS8jT5fHrgWwp1GpRCG6xiuvD05C51ckiLjY0NeXl5N53wJBoORVHIy8ur0sJb1V7OVgghGrqk7BS+PLwcKKpWMK3z3QaPy83XsCk0HIAh3QNwdarZFa3qgyBfVxLTnME1hUjpma1QRFo0689v5Xj8WbILilbnLLjWkf8b1Z0mHmUXGqoLVCoV7u7uJCcnY29vX6UeOVG/FNfudXd3r9IXL+mZFUKIatDotHwW+h3ZBTnYWtnwXO9Hyl12dceRSLJzC1GpYGz/ujmbfPz48ahUKrZv337T40JDQxk8eDAuLi64u7vTp0+fGplLEeTz30pg0RlxaHVyO7o8xavO7Ys8ok9kFZ2Kdk2bMrSHeZd0rYi1tTWNGzfG1la+4DVktra2NG7c+KZL+BoiPbNCCFENv575k4vJVwGY3mUK/m6+Bo/T6hTW7yma+NWjnQ9NvSpeCrm2LV++nJycnAqPCw0NZdSoUXz++eds2LABW1tbjh07ViO3sIN8XdHtL0pmC3Ua4rIS8Xc1/Bo3dIZWnVOpFRqHxNTJ4QU3UqlUkswKoxjdM/vnn39KRQMhRIMVkRbNGzs/Zd35LQB09u3AoGa9yz3+0Jk44pOLEsVxA+per2x0dDRz5sxhyZIlFR778ssv8/DDD/PAAw/g6OiItbU1PXv2rJlk1scFJdeZ4uGU7+5exOeh3xORFm3yx7J0YUmRBrdHZcTVciRC1C6jk9lx48bh7+/PK6+8YrDOrBBC1FcRadHM2f4RZxMv6bedS7x00wlKxeW4QgLcad+8cY3HWBWKojB9+nTmzJlDYODNb0fn5ORw4MABrKys6NGjB40bN6Zr166sXr26yo8bFhbG2bNnS/0kJiaWOsbX0xkbpzz970k5KeyLPMKcHfMlob1BarbhXvWCzLo5VlYIUzE6mb1y5QqPPvoov/76Kx06dKB3794sWbKEjIwMU8YnhBB1zrrzW8i/YbnafG0B6y9sM3j8hYgUzoenADC+lhZJmDZtGiqVqtyfgQMH6o/96quvUBSFxx57rMLrpqSkoNPpWLZsGV988QUJCQnMnj2be+65h9DQ0CrFOG7cODp06FDqp3jlyGJWahXOQZHc+JLla/LLfb0bop1XD5CvyiyzXdFawfW6dydACFMyesxscHAwb7zxBm+88QZ///03P/zwA8899xzPPvssEyZMYPr06QwaNMiUsQohRJ1Qske2pKj0WIPbi3tlPd0d6NPJr8biKmnRokXMnz+/3P3FZW+uXLnCW2+9Vek13F1cisavTps2je7duwMwYcIEBg0axLp16+jdu/yhFjdat24dISEhpbZ5eZWtzat2yDJ4fnmvd0MTnhrNt8d+AUCX44wu1wW1fRa6XGc0cc0JaVu3J38JUV0mmQA2ePBgBg8eTGxsLFOmTGHlypX89NNPBAUFMWPGDGbMmFHlmWlCCFEXXU6+Rnqe4TtQAW6lE9Vrsems2HyBw2fjAeh7ix/WVrVTRMbZ2Rln54onme3du5fk5GS6du1aavvEiROZPHky33zzTantbm5utGhhmt7lkJAQ2rdvX+FxXvZNiMhPKbP9xte7IcopzOWTA0so1BWiFNpQcKkrSoGDfr+9rRUTBoXc5ApCWD6TvKvu3r2badOm0bp1a86cOcNTTz3F1q1bueuuu5g7dy4PPPCAKR5GCCHMKqcgl89Cv8NQaXc7azvGthmm//1abDovL9yrT2QBNoWGcy02vRYirbxJkyZx9epVTpw4of8BWLx4Me+//77Bc2bMmMHSpUs5ceIEOp2OP/74g927dzNhwoQaiXFw0MCi2+Ul3Ph6N0SKovD14RXEZyWiKFBwtRNDb23NwC7+NPNzZWAXfz6c0Y9mfrIEsKjfjO4ujYiIYNmyZSxbtozw8HAGDhzIN998w4QJE7CzswNgyJAh9O7dm/vuu89kAQtRE67FprP67zAi4jMI8nFl4uAQ+QAQpSiKwpJjP5GYnYwKFY91u5ez1y8TlR5LgJsfY9sMI8jdX3/86r/DyCsoXRM1v0DLmp1hvDC1642XNxtHR0ccHctOEPL09KRRo0YAPPHEE0RERLBp0yYAnnnmGXJycrjzzjtJS0ujZcuWrFq1ip49e9ZIjN2bteKr1b2wDTmB2iEbWysb3hz8QqnXuyHadHknB6P/AUAT24KegZ34v7tvRa2u+2W4hDAlo5PZ5s2b4+fnx7Rp05g+fTrNmhlelrF9+/b06NHD6ACFqGnFPWjFiUd4XAaHzsbx4Yx+Zo5M1CW7ww+yP/IoAOPbjWRIi74MadG33OMj4g0PRShve11y47KiX3/9dZljZs6cycyZM2slnsZu9tjrPMi/1h67docp0BYSkZBGM4+AWnn8uuhS0lWWnyiqIKFNb0ygqgvP39NFElnRIBmdzG7YsIERI0agVt98pEKrVq3YuXOnsQ8jRI37eevFMj1oef/2oI281d5MUYm6JDYjnu/+nWDTunFz7m5/e4XnuJSzVG2Qj6tJY2sIwuMyyCvQoOR7oBTYorIt4IttmwlyDWyQd1Ay8rP4eN8SdIoOpcAOp8TuvD6jN/Z2MjdFNExGj5kdNWpUhYmsEHXd6StJHCoxprEkS+hBEzWvUFvIgtDvyNcW4GjjwNO9p2Olvvna8YUaLQnJ2WW2y2Qc46z+O+zfRRNUaFO9AVBc41i987JZ4zIHnaJjwf7vSM1PQ1FU6MK78NqD/Wns5lDxyULUU0Z/jZs+fXq5+9RqNa6urtx6661MmDChUjNqhahNiqKwZmcYyzedR6dTUDlkYu17FbVjFrqconI2QT4NezyeKLLy1DrC/y3O/0T3+/ByqnjBg3W7r5CYmgvALS09ycguIMjHlQmDZCy2MUp+sdSmemPtHYXaPpcrCZFAN/MFVosi0qJZf34rpxMukJ5fVE9WE9WKF8cMpWWAh5mjE8K8jE5mT58+TVRUFImJiXh7e+Pl5cX169dJSEigSZMmuLm5sWjRImbPns2OHTto1aqVKeMWwmhZuYUs+PkffY+sm2ce+UEHUVkVDTVQO2Zi5ZFI7+7d4IbhB6Jh+Sf2NH9d+huAoc370iugS4XnJKbm8Mu2ojq0/Ts35aX7GkayVZOCfFwJjytKaHWZjVA0NqisC7Hzum7myGpHRFo0s7d9RIHuv4U6FEXFyA5due0WKU8mhNHjBN58803c3d05dOgQcXFxnDp1iri4OA4ePIibmxsfffQRFy9exMXFhZdeesmUMQthtKsx6Tz/6W59Itu5jRdB3aL0iWwxlZWWo8lVW82oropIi+bz0O95afPbsqZ9FaTkpvHF4eUA+Lv68mDnuyt13rfrz1BQqMXBzorpd1ZcQ1VUbOLgEOxt/x3aoajRpjYBoNCp/OWD65OV/2wslcgCqFQKyXbnzBSREHWL0T2zL7/8Mm+88YZ+BZhiPXr0YN68ebzyyiucP3+eV199lWeeeabagQphjJIlt+xsrLgSk4ZGq6BSwcgh7sTbhnI56YrBc6PSY8GllgM2sYi0aObsmE++Jr/o9/QYjsSe4u0hLzb4skbliUiLZt35LRyLPU2eJh9rlRXP9n4YO2vDE7pK+udCIqGn4wC4d0QbGcdoIs383PhwRj8W/XaCS5FpaFO8sfaK4XrudaLT4/B38zV3iDXqYmIEGChScCkxsvaDEaIOMrpnNiwsDAcHw2/Ujo6OhIeHAxAUFER+fr6xDyOE0YpLbu0+Hk14XAYXI1PRaBUcXAroMjyGXRm/cCEprNzz68PqQuvPb9UnssVkTfvyFSf/+yOPklf8uqlUKAaXSSitUKNl8dpTAAT6uHBH3+Y1GWqD08zPjdem90KtAl2GJzaqonrmB6OPmzmymqUoCvnaPIP7dDkyH0UIqEYy27lzZ9544w3i40vPBI+Li+ONN97QL40YERGBn5/lJwXC8qz+O4x8qzRsmp/ErsN+bFocxzr4NKo2uzmXehqAxo4eTOk4pkyvW31ZXai8tetlTXvDDCX/Gp2mUsn/2l1XiE0qqmDwxPhOtbZsbUPi7mJHu+aNQVHjWFD0uXKoniezv57ZgM46t8x2RWtFS7u6s/iGEOZk9DCDr7/+mhEjRhAcHEzXrl31E8COHTtGo0aN2LJlCwCxsbE8+uijJgu4LimeXapfAajtcLl1W0coisLpmCvYtSs9sQtAARys7RnfbiSjWw7C1tqWrn4dWX9hW5nVnM7G1K2lR6sqwM2PiPSy4wr9Xev3bVljnS+np76i5D8xJYdV24smfQ3o7E/HEE+TxyaK9Onox5krySRHumPdouh9OD7rOj7OXuYOzeRCo46x+txfAGgzPFAK7FE7ZKHLdcYqKYT7Hq2ZFdeEsDRGJ7OdOnXi6tWrfP/99xw9epT4+HhatWrF1KlTeeihh/RDEF599VWTBVuXyFjEuisnr5Avfz9FpvMFrK3KViOwV9z4/PZZuNn/V7w+yN2fp3s9VJth1oqxbYdzJPZUmd7GrPwstDpthfVSG5IDkUdJzkk1uK+iISff/lE86cuah+5sVxPhiX/17ujLN+tOU5jqib3KBo1SyKGo44xtO9zcoZnU1ZQIFoYuA4qGE7RVRuDu7kRkfGZRmbfxUuZNiGJGJbN5eXnMmDGDhx9+mKeeesrUMVmEm41FrI9JkaW4FJnKRyuOkpgfjW3rBIPHeDg7lkpk67Mgd3/eHvJiUa9zWgy5mnwSs5M4mXCehYeWMqPnNElogYNR//D5wR+Aonk2JUfIVjTk5NiFBJn0VYs83R1oHejBxchUnDR+pFtFcCi6fiWzKblpvLfnSzRKIUqhDe5JfZk1ow+O9jbmDk2IOsmoZNbe3p5ffvmFqVOnmjoeiyFjEesWnU5h7a4wftx5GKumF7HzKL/+ZAvPhrWee8leZ51Ox5dHlrMn/BAHIo+iRsVTPR9s0AntkZiTfBb6HTpFh5dTYx7tei97Ig6VGXJyo2ux6fy24xKhp4vmDfg2duSOvs1qO/wGqU8nXy5GppIa5YE6OIKwlHCSclLwdGxk7tCqLV9TwId7vyI9PwNFp0J3tSuzHxkoiawQN2H0MIPBgwezc+dOBg4caMJwLEd5YxHd7Cy8lpMFOXD5Aj8e3UCaJgklz4GCQh027RNR/VvCprGDBxn5mRTqNPpz6svELmOp1Wqe7P4AKLAn4hD7Io8A8H89pzXI5an/iT3DJweWoFV0NHb0YO6g52ji1JhbfW8+VKC4UkZeiUU1ktPziErIlFu/taBPJz9+2HCO/CRPnJtZoVW0HI4+wehWg80dWrUoisJXh5dzNbWo5FZheHv+N3ygtCkhKmB0MvvUU0/xyCOPkJ2dzejRo/H29kalKl0Ir127+jt2rLyxiNfSokjLy8C9gdzGNpcDly/w6ZFFRZO7bAHbdH1jdrF15u4OtzO0eV9iMuMNTuxqyNRqNU/2eAAFhb0Rh4sSWpWK/+vxYINKaE/Gn+Pj/YvR6rQ0cnDXJ7KVsfrvsFKJLECBRseanWG8MFVmmNc0n8ZONPdz42psOs5aP9LVURyKPm7xyezqc39xIOoYAIVxwfQP6sXwnoFmjkqIus/oZHbkyJEAfPLJJ3zyySelEllFUVCpVGi19Xcp0FJjEdNjcbd34XTCRTLzs/gs9DvmDHi6Qd+6rWnLjv5RZtUuAAfFg4V3vIajTdG4xfo6sau61Go1T/V4EEVR2Bd5hH0Rh8kpyMHBxp7o9Lh6W52juALJpeRrXM9JRlEU3O1deX3Qs1WaDR8eZ7jKRUR8hqlCFRXo08mXq7HppEc3gsAoLly/QlpuOu4OltWLWdwmLyRdISknBQBtmhc+Bd14cuItZTqJhBBlGZ3M7ty505RxWKQbE6XNl3fx/T+rOJt4iVVn/uTeTuPMF1w9VaAtZNOF3aSqwg3uLyxU9ImsuDm1Ws1TPR9EQWF/5FH+iTuj31cfq3PcWIGk2CNd78HPxbvS19HpFLLzCg3uC/KROzK1pU8nP1ZsvkDu9cY4BarRoeNwzEmGh/Q3d2iVZqhNKgqoEkOY+Wh37O2M/ogWokEx+i9lwIABpoyjXhgRMoCLSVfYH3mUdee30LJxM7o3vcXcYdULhdpC/r56gN/P/EV6QYbBpR0B3Kwrd5tYFLFSW/F/PadxMemqvleoWH2rzmGoAgkUrSDVw//WSl/n560XSUoruyKTva0VEwaFVCdEUQUB3i4EeDsTlZCFi+JLuiqGQ9HHLSqZNdQmVSpo1SWDQPliJESlVftr36ZNmzh69ChRUVHMmTOHwMBA9uzZQ0hISINb+UulUvF4t6lEpMUQnRHHF4eW8f6wV/FxaWLu0CxKycUomrr64uvixa7wg6VqgGrTG6F2SUOl1um3KVor7u9+hzlCtmhWaqtye7PrS3WOvMI8TsSfNbivKs9x7/EYftl2EYAOzRvT2M2eyIR/634Okrqfta13Rz+iEi6RGdsImsZwNvESmflZuNhZxjKvl5IiDG7PUlIMbhdCGGZ0MpuQkMCYMWM4duwYwcHBXLt2jSeeeILAwEB++OEH7O3t+eqrr0wZq0Wwt7HnhdseY+a298kpzOXj/d/w9tCXyyyXKgwztBhFSdpUL6yut+bFcYNR7NP58egG0jXJuFk35v7ud9CnZRtzhF2nXYtNZ/XfYUTEZxDk48rEwWWTrkA3PyINVOfIKsgmIy8TV3vLrdJxKekqCw/+QFZBjsH9FS2IUOxyVCoLfvkHgGBfV157uKeUSzKzPh19+XX7JbITPHFoCjpFx8tb3qWtV0idH/OdXZBDcnaawbtMBZmOtR6PEJbM6GR2xowZZGVlceHCBYKDg7G1/S9ZGzp0KG+88YZJAqzLyksSmrr68L8e9/PpgW+JSI/h22M/82SPB2QgfyWsPbfF4K1gXYEdBZe74G3vy5xHehLk6wr4SvJagRtLSIXHZXDobBwfzuhXKqEtrzpHck4qL2x+i8e6T7W4ITManZbfz25k7fnNKErRMghWqqIyTsUqW6otOT2Xt78/TIFGh5uzLXOmSyJbFzRv6oZ3I0cSczMpXu4iOTeVfZFH6vSY7zxNPu/v/RKtqqDMPkVrBUktzBCVEJbL6GR28+bNLFu2jJCQkDJVC/z9/YmJKdvLU59UlCT0DujKxVZX+evS3+wOP0hMRjyF2sJ6O0u8ugo0Bfx97QCHoo8bPqDQllv8Qnj5gW64OEovd2UZKiGVV6AtU0Lqxuoc/q6+uNg6su3qPtLzM/lo39f0D+7JQ50n4WRb93uNYjLiWXjwB329zsaOHjzV40Fc7JyqXKotv1DL2z8cJiUjD2srFTMf7IF3o7r/GjQEKpWKPp382BAZSul12+rumO9CbSHz9y3mYtKVot/jglDZFKB2yEKX64wmrjkhbaUclxBVUa0xs9bWhk9PSkrCwaF+zyivTJJw3y0TOJt4iYi0aMJSwoH6OUu8OvIK89h2ZR9/XtxGWl75ZY28HJow78FeWFk1nDqoplBeqShD2w2VMRvcvC9fHFpKRHoMe8IPcTLuHP5uvmTlZ9epL2bF46wj02OxVlsRmR6L5t/FMvoG9eDhLpP1SXhVkhtFUfj8l+OERaUB8NRdt9C+uUwyrEv6dPLlr6Qsg/vq2phvrU7LZ6HfcyrhPACFMc3RxLQqdYxMJBSi6oxOZvv168fnn3/O6NGj9duKb6N///33DB5s2cWrK1KZJMFabUUTp8ZEpEWXOqau9hjUhuKkIyItGrVKzfXsFHI0ufr9zlofMklEZVV6Ypef0kkSWSME+bgSHle2rXp5VK5nMdjDn/eGvcrv5zay5txm0vMzSU/MBOrOF7PySm45WNvzePep9AnsZvS1V22/xJ4TRXeZxg1owdAeQdWKVZheqwAPbDRuaMkss6+y46Frg07R8dXhHzkccwIATXwgmpiWDO0RiEaj0w9Xk4mEQlSd0cnsBx98QN++fenQoQPjx49HpVKxZMkSzp49y+nTpzl48KAp46xzyksSrNSlx8UmZiUZPD/KwGSb+i4iLZo52z8iX1t2nFifgK60c+rBVyvD0dpmYO17tdRtt+seMj7RGBMHh3DobFyZuwiXI1NJSsvF073iOyjWVtZM6TiWKykRnIw/X2pfviaf9ee38nTv6SaNuypWnlxncJx1R+82RiWyxWPhz4Uncz216ItWt7beTLujfbVjFaanVqvo2qgPh/JXl1lIpX9QTzNFVZqiKPzwz6/siTgEgOa6H4WRbenf2Z8Zd9+KWi3zKYSoDqO7ujp06MCxY8fo1q0bS5cuxcrKijVr1uDv78+hQ4do1apVxRexYBMHh2BvW3aFr7DodHYdi9L/Xl7PQHpe5k1vq9c3V1Mi+HDf1wYT2U5N2qNEdGHh8mtotApKrguFV28h/+xtFF69BSXXRYrRG6mZnxsfzujHwC7+NPNzpUOLxqhVkJqZz2uLD5CeVTYJLE9aruH2ejT2FNEZcaYKuVI0Oi0HIo/x+o755ZbcSsg2/EXyZorHwu8+Hq1PZFUqmDS0ZZkvqqLuGHFrR/LP9UKT5IuXgxdqVdFH258Xt6JTdBWcXfN+Of0HW8J2A6BN8abwWgc6tPDk2SmdJZEVwgSqNWa2RYsW/Pjjj6aKxaIUJwlrdhZVM/DzdCYsOo2ElBwW/HIcJwcburfzKXeWeFpeBi9veYene02ng3drMz2LmqXRaTkU/Q+bL+3iYvLVco87FRFF7umiLwBNPBxIy8ynQPPfB5CMIaueZn5upSZ77fonmk9+OkZ0YhbzloTyzv9uq9TM/AA3vzKl0qBoZvZLm9/mjtZDmdh+NPbWdiaNv2TdYW8XLzzs3TgSc5KU3LQK460qQ2PhFQX+2h9O22AZK1tXtWvWGCc8yLp6C+kJtgS2SSbcaj+nEy6y5fJuRrUaVOsxFbfbM4kX9R0XugxPCq7cQoC3K7On9cDGWpY8F8IUZK28argxSUhOz+XlhXtJTM3l/WVHePPxPrRvXnqWeICrL40c3dl46W/S8jJ4a/dn3NVuNBPbjUatttwxoYYSjsMxJ0jN/W8Ne1srGwq0ZZcB1WQ7YW2l4q7BrZg0tCXRiVn6Lwkyhsz0BnbxJzu3kK/XnCIsOp23vj/EvEd7Y2dz8w9WQ1/MrNXW2FrZkFOYy/oLW9kfeZRpne+miVNj/ihZNcDIiWI3Dk25MZlu2bgZ3fw6sebcplK9/pUtuXWjc+HJhuMoZ4y8qBsi4zPIzS/6EpKeXUD6MWfs23ihcr3OylNrucWnLX6uPrUWj8Fx3AoURLXAw9mBeY/0xlmqsghhMtVKZn///XfWrFlDdHQ0eXlll3c8fPhwdS5vcRq7OfDm4314ZdFe0rMKeOu7g7z3VF+a+ZWdJd7Lvwufhn7L9exkfju7keNxZ2jk4EFC1vU6NUu8Mipa6KC1ZwtGtRyINteRz499WWpcm6K1wlvbkZnPD9Qv33jjlwRherff1oys3AJWbLrAmSvJfLD8CLOm9cD6JpPsbizfVVzaqrGDBz+dXs+OK/tIyklh/v7FqFVq/e3d8iaKlfwCVNzmGzm4E5keS0RaNJFpMRyOOWFwaEoTp8Y81+dRWjQqmpDVxa9DlUtulaQoCr9uv6QfWlDmucswlzpt9d9haHUlS3OpyLvSHudbD1CgLWDRoWW8NeRFrNS10xO65tzmsuO4VWDnG8Xrd4yjiZR2E8KkjE5m582bx5tvvsktt9xCu3btSi2a0JA19XJm3qO9mfXlfrLzNLz+TSgf/l8/fD2dSh0X0jiYD4fP4qsjP3I4+gRhKRFA0dKGdWWWeGUkZiWx8OBSgxNwmjh58nyfR2jeKAhFUZj55X7y43qVmdzVok2ArENuBpOGtCIrp5B1u69w5FwCb39/CCcHGyLjM8tdKcxQ+S6Ax7rdy6Bmvfn26M9cS4sqM04xX5PPe3u+oJ1XS+ys7cjXFhAaeUy/gEFEegz7Io9UOnYHGwd9InuzuCqjUKPl819PsOtYUdURlapoaEExGeZS9xnsOS+0xzmlM+mNDxGWEs76C1uZ0G5UjccSnhrF4egTBvc18i4kJMC9xmMQoqExOpn97rvvePXVV3n33XdNGU+9EOLvzmvTezJ3SShpmfm8+sU+Wgd5EJeUXSpJcLJ15IU+RUvfFhd3L5avyWft+S082/thMz2LIuX1noVG/cPeiMP6wt+GONjY07xREJciU/n+z7OcvZoMFE3uKikyoWxJHVHzVCoV0+9sT3ZuIdsOR3LsQqJ+X3krhd1My8bNeG/Yq/xvw6xSw0uKpeSmVTphdbCxJ8itKWl5GcRnXS+z31Qll9Kz8nnnh8OcD08BoEubJkwZ1oq/9ofLMBcLUl51mVau7bEKKORg1D/8dmYDnX070MwjoEZiUBSFzZd38ePJNaVWmStJnS9f2oWoCUYns5mZmQwZMsSUsdQrHUM8efn+bry7tGjloNDTRbO9b0wSVCoVWp3hN76DUcf41taRQc360NwjsNaXwzU0fOBA1DGAUj1vJW8pl+Rp34QPlh9h38mbFy6XW7jmo1KpeOruWzl+6TpJaaVvsRtaKawiarWa9l6tDCatXo6NCHL3J1+bz8WkaxSUM3xg7qDn8HRshEqlMjj20NjxsDeKiM/gre8OkZCSA8AdfZvxyJgOWFmpZbKXhSmvBJ2jnQ33d72H89fDSM/LYNGhpbw37FVsrUxb6i8zP4uvDv/I0dhTRRu0NigqLSp16XrZskytEDXD6GR2ypQpbN68WRLam+jVwZdWgR5cjEgttf3GJKG8WeI6RWFr2B62hu0hwM2PgcG9aeYRwN9X91d7Yk1FNFoNP55YU2b4QHHSqlKpuMW7Lf2CeuLt7MmbOz+jQPdfcqJWrDn4twOarKJE1svDgVG9gvh1x+VSHzgN7Rbu+PHjWbduHdu2bWPo0KEGj0lISOC5555jx44d5Ofn07ZtW95//30GDBhQIzFZqVU4OViTlFZ2nzETnwxNFLOztuPlfv/Tt9XPQ783mPC28myBl9N/iWR543SNbfPFNWQvRKSQlJaLVqegVqt4bGwHbu/b3KhrCvO7sbpMVm4h11Nz2XoogoFdA3ii+318sPdLotJjeWXru1irrEz2/nku8RKfH/xBX12jpUdLLu0NJF/JKzOkSpapFaJmGJ3MDhkyhFdeeYWkpCSGDRuGu7t7mWNKrg7WUOUXGO51LXlLzNCHv62VDb0DunEy/ixpeRlEpcfy48nVpa4RkR7DkZiTvD30pQon1hh6wy55nK+rNx28WpOSl8aF62FcTgmn0EDlAQAPBzc+GD4Ld/uiHtVrsenkn+uJ1jOs1Bu3kuuCk4MNk4a04o6+zbC1saJbO58GW6lg+fLl5OTkVHjck08+SWJiImfOnKFRo0YsWLCAO+64g4iICBo1alQjsQX7uBERV3a4h1qlQvdvwldZlUlAy0t4DfW4Vmc8bEnFNWRv7L17fFxHRt/WrNrXF+ZVcuJoVm4hz3y8k8TUXOavPMbCFwbSo+ktHI45SUxGPFC9uQkRadGsO7+FMwkXSc8v+ruxUqkZ4DeEnZvtyMvVADalhlQ1tC/uQtQmo5PZyZMnAxAeHs6yZcvK7FepVGi1hhO5hqS8sVwpGXnEXs/Cz8v5ph/+Wp2Wk/Hn2HXtIIeij6OglLpOvraA2ds/op1XCP5ufjhY27P+whZ9CazihPe5Po/gaudCRn4mGflZRKbFsunyzlITcA5G/VOp59S+SWt9IguwYvMF8jKcIKP0WFj/Js58OKMfLiVK0DTUSgXR0dHMmTOHffv2ERR08yVRw8LCmD59Ol5eXgA8/vjjvPjii1y+fJmePWtmRaPybtNeiUln3pJQnrunCx6u9pW+XkUJqKl7XCvjx7/Ol3l+AOfDUySZrWecHWx4cWo3Xv1iL0lpuSz67SS2IWUrdeRr8vntzEZe7Pt4pa99PvEyb+/+nEKdptT2/p63s3m9Bq1Og4OdFQ+ObseFiNQG+cVdiNpmdDJ77do1U8ZRbXXx9i2UnyRkZBcw4+NdPDi6LXf0bV7uh7+V2ooufh3p4teR5ze9aXClpQJtASfiz3Ei/pzBGPK1Bby/98tKxWtrZUO3prfQxrMFbvYufHloucH6nYUaHYfPxrP5YDgnLpWdoANgY60ulcg2VIqiMH36dObMmUNgYMW3GV955RWWLFnCpEmT8PT05IsvvqBFixZ06tSp0o8ZFhZWZpuXlxdNmjQxePyNt2l9GzuRkpnPhfAUjl+6zoyPd/LM5M50b2e6Wp2m6nGtSGZOAT9vvciR8wkG90sN2fqpbbNGTBnehp+2XGD/qViaukQbPO5wzAk+3PsVg5vfRmff9kRnxJW6s3Vn62FoFA0n4s5yIv4cl5MNf/ZtO38Ure4WGrvZM/eRXjTzc+P2vjX5DIUQxYxOZivqXapNdfn27Y1JQqCPCwFNXFi7K4zsPA1L1p/hwOk4npncuUz5rhsFu/sbTGaD3JrS1M2XmPQ4g2Nvb2SttkZRdGgNTNrydfEuVUFByXPix6MbSNck42bdmDuCh7JzXzo7jpwhrYKlUOvzxK5p06YZvCNRbMCAAezatQuAr776CkVReOyxxyp17dtuu40ff/wRPz8/rKysaNSoEWvXrsXBwaHS8Y0bN67Mtrlz5zJv3rxyz7mx11ynU9iw7yo/bDhHelYBb353iH63NkVBITohq9zyXXWFRqtj04Fwftpygaxcw0NmoH63U2PV1c6Bqpo0tBUnL1/n7NVkkhNsUJXzNn809hRHY0/hYutMTmGuUSXj1A5ZNPdz4/VHetLYrfJ/q0KI6qtSMvvTTz8xcuTIUolfZGQkfn5+WFv/d6nY2FiWLl3KrFmzTBdpOer67VswfGt9aI9AFv56gmMXEjl7NZmnPvqbIB8XNFql3CShvHGG/9drmv727Geh37E/8miZGG71accjXe/Bxc4Ze2s7Fh78weCbdMmSR9di0/n0+zDyCloCLckEvj4aUer4VoHudGntzdrdYaXGB9f38WGLFi1i/vz55e63sSmaLX3lyhXeeustDh48WKnr6nQ6Bg8eTP/+/UlOTsbV1ZWNGzcyevRodu/eza233lqp66xbt46QkNKvf3G7ryy1WsWY/i3oGOLJRyuOEZWQyd4T/31ZMqZ8V00pnthVfEu3bbAHG/ZfIzoxCyia5Na/c1MOnI5rUO3UGHW5c6CqrNQqnr+3C09/vIucmGY4uCeiqP8bHmBnZcvIlgM5HneWyPQYMguyyr2Wi60TnXzaEpWaSGRmZNn9Vo1576nKLQ0thDCtKiWz999/P6GhofTo0QMArVZLs2bNOHLkCF26dNEfFxUVxWuvvVbjyaw5bt9C1W/hGtLYzYG5j/Ri++FIvll3mrwCLWHRRbU5y0sSKjPOcFzbERyNPV0m4Z16y3iaOHvqt1U0ASc1I49Fv500OMbQSq1iRK8gRvQKpnnTovj6dPJtUBO7nJ2dcXZ2rvC4vXv3kpycTNeupb/MTJw4kcmTJ/PNN9+U2p6amsrVq1f5/fff9QnB2LFjadGiBVu2bKl0MhsSEkL79u0r92Qq0MzPjU+fG8DT83cSm5Rdal9egZbVf1/mxfu6meSxjHHjxK7wuAx2H//vlnLP9j48dGd7mno5Mz42vUG106qyhM6Bqmri4ciMu2/l/eVHyD3bE9dmUehsM3Czbsz9Xe6gT8s23NtpHFdSInh3zyKyCrLLXMPXuQmfjpqLWq3mjRVbUdQxZVYybGbVWRJZIcykSsmsoiiV2lYddf32LRh3C9cQlUrFsJ5BHD4bz8Gz8aX25RVoefv7Q0wf04GurZtgb1f0X2WqiTVB7v78r9MTpYYQjAgYwsGj2Xx+djeXItPKfQx/bxf+N7H0ZK+GOrGrIpMmTSpzmzYgIIDFixczfPjwMsc3btyYtm3b8sUXX/DJJ5/g7OzMX3/9xdmzZ8skxLXJzsYKWxvDS4HuPxVL0I5LDOsRhLuLnckf+8Ze15J3LVIz8ljy75fBGznZ2zDzwe7c0uq/Hmlpp+WzhLHdxrrtFj96d/Ql9DSkn2sHQCbw6akwfGf40szPjZDGwdzq087gHasWjYNJychn66EIjp/MQ2dXdiXDJA9JZIUwF6PHzNaUun77FkxzC7ek+BTDt/QSU3N5f9kRbK3VdGnThN4d/fBu5MCmAxEGP9iL6XJcKAjrRF58MAU+rugCXcC99LUvR6XxyXdh5Bf+N4Tgu6Nlb50Z0sxXxhhWlqOjI46OZddh9/T01Pe8PvHEE0RERLBp0yYA1q9fz0svvURISAh5eXkEBASwcOHCcscu1pbyKnNotArL/zrPT1su0KeTH6P7NMPBzoo1O6/ctJ3eLEktecyNva77TsXQ0t+D+JRs0jLLH7ft5eFQKpFtiOp654CpOgYqw8pAebm8Ai0f/3SMZyZ3pkVTd8a2Hc7h6JOlamZbq2xIOO/Dw79vRVfcd5NbdiXDoLbyviiEudS5ZLau374F097ChfKTBGcHG7JyCynQ6Dh4Jp6DZ0r33obHZXDgdCz3j2pLMz9XnBxsSM3I48Mfj5FfqC11zJ39mlNQoCU2KZvYpCzikw0n0FZqFbd18qN7ex+83O2Zt+Rgg17koCbceDfj66+/LvV7y5YtWbduXS1GVDmGKnPYWqvpEOLJqcvX0WgV9hyPYc/xGFQqKH6a4XEZhJ6J5cV7u9KsqRt2NlbEJWcz95vQUklq6JlYpo5sg5VaTUp6HsnpeZy4fL1Mr6tWq3AhIqXCeIPlS1ed7xwwdcfAzRSPn75RRFwmzy/Yg7uLHW2CPMiJ6AFeV/S9rnlxzTmVW5Tc2tpY0bmVF8cvJVJQ+N8EWnlfFMK8qpzMGlpStbaXWYX6c/sWDCcJ9rZWvPvkbTjYWXPwTBwHTsXp148vqVCj4/s/z970+oUaHWt2lr2dZ0iAtwsv3f/f+MeSlRhkjGHDdmNljpLtIS0zn22HI9h8MILElBxuHH1UUKjj3WU3nxVeUKjjhz8Nl5e7kZO9NRMHt6R5UzesrdS8/f0h+dJlQF3vHDB1x8DNlNdpYG2lQqNVSMvM/7fDwBkyS/e6OtpbM3VkGwZ3DcDZ0ZZrMvZaiDqlysnsiBEjSlUugKLVwEpu02g0N55mcvXp9u3NkgSAcQNCGDcghCc/3EFUQvmzbW9GrYIW/u74eTrj5+XE8YuJXLhhmV0o25slYwxFSeW1B3cXO+4e0ooJg1ry+HvbSShn6ExluDnb0tjVgUZu9sQkZhJn4C5C93Y+3D2klf53+dJVPfWpc6A85XUavP9UX/ILtRw5l8D6PVco1JQtWejdyJEx/Vrof5f3RSHqliols3Pnzq2pOEzCUm/fQuXeHJv7uRtMZvt3bspj4zqSk6fh67Wn+OdCYplj+nVuyotT/+tx7d3Rt8zSntKbJarLSq2iTVAjg8lst7be3DO8NfmFWn7ecpHTV5LKHNO/c1NeKlEZwdAStIbaqSQX1VOfOgfKU1GnQbtmjbmemluqEkYxqUUsRN1Wr5LZ+q68noW7BrfEzdkON2c7pt3ejnNXk8scM3FQy1LXquiNXQhjlddOHxjdVt++nMfZGExS7xos7bSusOTOgfJU9KWnvLYrX/KFqNvq3AQwUb7KfLDfeIy3uy3p4Xtxsu5h8Hp1vTcrMTGRL7/8kieffNLk5XpEzWiI7bQmSNuvfcXt8qdNZzh5IZJb2gRy76gOFvnlSdqPaFAUUWlnzpxRAOXMmTPmDqXSLDHmkswdv7kfv6osLd5ilhp3TarKa2Jpr19dj7eux1cZdfE51MWYRP2gNk8KLYQQQgghRPVJMiuEEEIIISyWJLNCCCGEEMJiyQSwKsjPL1o609B64nVVcayWFHNJ5o6/+HGL/+/rOktso2D+/+e6qCqvibRT06oP7bEuPgdLa6fCcqgU5ca1ekR51q9fb3AtcVH/rVu3jrFjx5o7jApJG23YpJ0KS2Ap7VRYDklmqyAtLY3du3cTEBCAnZ2ducMRtSA/P5+oqCgGDBiAu7u7ucOpkLTRhknaqbAEltZOheWQZFYIIYQQQlgsmQAmhBBCCCEsliSzQgghhBDCYkkyK4QQQgghLJYks0IIIYQQwmJJMiuEEEIIISyWJLNCCCGEEMJiSTJr4RRFYe7cufj5+eHk5ET//v05c+aMyY6vaVWN55dffqFfv364urqiUqnQaDS1GK2ojJpok3Wt3Rqjqs/h1VdfpWPHjri6uuLr68s999xDVFSUfv+8efOwsrLC2dlZ/3PPPffUxlOxaFV9D6mrba+qz0Pai6jPJJm1cPPnz+f7779ny5YtJCUlcdtttzFixAiysrJMcnxNq2o8Hh4ePPnkkyxYsKB2AxWVVhNtsq61W2NU9TmoVCqWLl1KUlIS58+fR6VSceedd5Y6pnfv3mRlZel/fv7559p4Khatqu8hdbXtGfNeKO1F1FuKsGjBwcHKggUL9L8XFhYqnp6eyvLly01yfE0zNp6dO3cqgFJYWFjTIYoqqok2WdfarTGq+xyOHz+uAEpKSoqiKIoyd+5c5bbbbquRWBuCyr6H1PW2V9nnIe1F1GfSM2vB0tPTCQ8Pp0ePHvpt1tbWdO7cmePHj1f7+JpW1+IR1VcTbbI+tBNTPIetW7cSFBSEh4eHftvx48fx8vIiKCiIe++9l2vXrpk89oasPrS9kqS9iPpKklkLlpGRAVBmjWsPDw/9vuocX9PqWjyi+mqiTdaHdlLd57B9+3beeOMNvv76a/22u+66i3PnzpGYmMiBAwdQqVQMHTrU7Le/65P60PaKSXsR9ZkksxbM1dUVgLS0tFLbU1NT9fuqc3xNq2vxiOqriTZZH9pJdZ7Dhg0buOuuu1ixYgUjR47Ub+/QoQNBQUGoVCqaNm3K999/T0xMDAcOHDB5/A1VfWh7xaS9iPpMklkL5ubmRnBwMEeOHNFv02g0nDhxgs6dO1f7+JpW1+IR1VcTbbI+tBNjn8PKlSuZOnUqq1atYvz48Td9DJVKhUqlQlEUk8Xd0NWHtlceaS+iXjH3oF1RPR9++KESEBCgnD59WsnJyVFmzZql+Pn5KZmZmSY5vqZVNR6NRqPk5uYqW7ZsUQAlKytLyc3NVbRabS1HLspTE22yrrVbY1T1OSxcuFBxd3dX9uzZY3D/qlWrlOvXryuKoijx8fHK/fffrwQFBSkZGRk19hzqg6q+h9TVtlfV5yHtRdRnksxaOJ1Op7z22muKt7e34uDgoPTr1085deqUfv/jjz+ujBw5stLH17aqxv/DDz8oQJmfnTt3miF6YUhNtMm61m6NUdXXBVCsra0VJyenUj/Fye2dd96peHp6Kg4ODoqfn58yZcoU5fLly7X+vCxNRe8hdf09s1hVn4e0F1GfqRRF7jEIIYQQQgjLJGNmhRBCCCGExZJkVgghhBBCWCxJZoUQQgghhMWSZFYIIYQQQlgsSWaFEEIIIYTFkmRWCCGEEEJYLElmhRBCCCGExZJkVgghhBBCWCxJZoUQQgghhMWSZLaOUalUqFQqVqxYUWbfihUr9PuLhYeHo1Kp2LBhQ22GWWWXLl1i3rx5pKWlmfS6wcHBvPjii9W+ztKlS/Wv7a233lrh8QMHDtQfv2jRomo/vqWRdlo10k6FEKLmSDJbBzk7O/PLL7+U2f7zzz/j7Oxcapuvry+hoaH07du3tsIzyqVLl3jjjTdMniSsXbuWp59+2mTX+/vvv/nxxx8rPO7LL78kNDTUZI9riaSdVp60U8swaNAg7rzzznL3Dx48mJYtW6LRaAD49ddf8fHxQVGUKp9bWSW/kCxYsOCmx86bN09/7F133VWlxxHCkkkyWwfdeeedbN26ldTUVP22lJQUtm3bxpgxY0oda2dnR69evXB3d6/lKGtOXl5epY/t3LkzgYGBJnvs7t2707FjxwqPa9euHb169TLZ41oiaafSTuubtm3bcuHCBYP7du7cyc6dO3nzzTextrYGYOPGjYwePRqVSlXlc6ti0KBBhIaGMmXKlJse98gjjxAaGkrnzp2r/BhCWDJJZv+1dOlSbG1ty/TInD17FpVKxfbt2/XbFi1aRMuWLbGzsyMkJIRPP/201Dnz5s3D09OT48eP06tXLxwdHencuTN79+6tVCy9e/fGz8+P1atX67etXr0aPz8/evfuXepYQ7dv//jjD7p27YqTkxMeHh707NmT3bt36/frdDref/99QkJCsLOzo1WrVixbtqzUdfft20e/fv1wdXXF1dWVW2+9ld9+++2mcb/33nuEhIRgb2+Pt7c3I0eOJD4+nl27dul7LJo1a4ZKpSI4OBj477bp4cOHGThwIA4ODnz00UcAvPrqq3Ts2BFnZ2f8/f2ZOnUq8fHxpR7zxtu306ZNo1u3bmzbto1OnTrh5ORE3759OXv2bEUvu0WQdirtVNSctm3bcu3aNQoKCsrse+2117jlllv0CaVOp2PTpk3cfvvtVT63qho1akSvXr3w8fG56XH+/v706tULV1dXox5HCEslyey/xo0bh0qlYu3ataW2r1q1Cm9vbwYNGgTAkiVLmDFjBmPGjOHPP//k7rvv5oUXXuD9998vdV5OTg4PPvggjz/+OKtXr8bOzo4JEyaQk5NTYSwqlYrJkyfz888/67f9/PPPlXojvHLlCnfddReDBw/mzz//ZOXKldxxxx2kpKToj5kxYwZvv/02jz32GBs3bmT8+PFMnz5dn2hkZGRwxx130Lx5c1avXs3vv//O/ffff9Nbr8uXL+fdd9/l+eefZ8uWLXz11VeEhISQnZ1Nly5dmD9/PgBr1qwhNDS0zOt8zz33cOedd/LXX39xxx13AJCYmMisWbPYuHEjCxYs4OrVqwwePBidTnfT1yAyMpKXXnqJ2bNn8/PPP5OYmMjkyZNRFKXC16+uk3Yq7VTUnLZt26LVagkLCyu1ffPmzezfv5933nlHPxb8yJEjpKWlMWzYsCqfK4QwMUXojRkzRhkxYkSpba1atVKeeuopRVEURavVKn5+fsq0adNKHfO///1PcXV1VXJzcxVFUZS5c+cqgLJjxw79McePH1cAZdOmTTeNAVAWLlyo/PPPP4qVlZUSHx+vxMXFKWq1Wjl+/LiycOFCpeR/27Vr1xRA+fPPPxVFUZTffvtNadSoUbnXv3z5sqJSqZSlS5eW2n7//fcr3bp1UxRFUY4cOaIASkZGxk1jLempp55SJkyYUO7+P//8UwGUa9euldr+ww8/KICyYMGCm15fo9Eo0dHRCqDs3r1bvz0oKEh54YUX9L8/+OCDipWVlXLp0iX9trVr1yqAcv78+XKvXxxHZmbmTeO4UfH/V22Sdirt1BLaqSUq/r9bs2ZNqe3dunVTbrvttlLbXnvtNWXw4MFGnVsVAwYMUCZOnFjj5whhyaRntoTJkyezY8cOkpOTAThx4gSXLl1i8uTJAERHRxMbG8vdd99d5ryMjAxOnz6t32Zra8vAgQP1v7dr105/jcro3LkzISEh/Prrr/z666+0atWqUrOXO3bsSHp6Og8++CBbt24lOzu71P4dO3agVqsZP348Go1G/zNkyBBOnDiBVqulRYsWODs7c++997J+/fpKTYa59dZb+euvv5g7dy6HDx9Gq9VW6nkWK75VV9KmTZvo06cPbm5uWFtb4+/vDxRN0rmZ4OBgWrZsqf+9qq99SSVfo6pO3Kgp0k6lnd6oLrZTS9S0aVNcXV1LjX1dv349R48e5b333it17MaNG0u1h6qcK4QwLUlmSxgzZgw2Njb6MYCrVq3C399fPwM7Li4OAG9v71LnFf9e8hapi4sLavV/L6+trS1QtUkjkydP5pdffuGXX37RJyoVad26NevXr+fq1auMHj0aT09P7r33Xq5fvw5AUlISWq0WNzc3bGxs9D/Tpk1Do9EQFxeHh4cH27Zto7CwkEmTJuHl5cXtt9/O1atXy33c6dOn8+677/Lrr7/Ss2dPvL29mTNnTqWThRtf0yNHjjBmzBj8/f358ccfCQ0N5eDBg0DFr+GNk4yMee2LlXyNbGxsqnx+TZB2Ku30RnWxnVqqNm3acPHiRQAUReH1119n1KhR9OvXT39MXFwcx48fL/PlpjLnnjlzhi5dutCyZUvGjBlDZmZmlWNUFKXUl5eqfikTor6RZLYEZ2dnbr/9dlatWgUUlV25++679eOcfH19gaIxciUlJCQARYP0TWnKlCn6D8eqTBy4/fbb2bt3L8nJyXz33Xds376dGTNm6GO0trbm0KFDHDlypMxPkyZNAOjVqxebN28mLS2NNWvWcOnSJe69995yH1OtVvPcc89x/vx5IiMjefHFF3nvvfdYsmRJpWK+cSzZ2rVr8fLyYtWqVYwZM6ZSkx9qwo2vT10g7VTa6Y3qYju1VCWrEvz222+cPn2ad955p9Qxf/31F82bN6d169ZVPveJJ57g7bff5vLly7Rp04YPP/ywyjHu3r271JeXIUOGVPkaQtQnVa8RUs9NmTKFyZMn8+eff3L16tVSH87+/v74+fnx22+/MWrUKP32X3/9FVdX10qVyqmKtm3b8thjjwFF3/irys3NjXvvvZfdu3fra00OHjwYrVZLenq6fuLCzTg4OHDnnXdy5syZSt8qCwgI4NVXX+WHH37g3LlzQNV7nXJzc7GxsSmVPKxcubJS55pSt27dav0xK0PaaWnSTutmO7VEbdu2Zf369Wi1WubNm8ekSZPKlLq6cYhBZc9NSEjg2rVrjB49GoCHH36Y8ePH89Zbb1Upxq5du5b60uLi4lKl84WobySZvcHo0aNxdHTk8ccfp1mzZvTo0UO/T61WM2/ePB5//HEaN27MsGHD2L17N1999RXvvvsu9vb2Jo/n66+/rtLxixcvJjQ0lJEjR+Ln58fly5f57bffeOCBB4Ci27tPPPEEU6ZM4eWXX6Zbt27k5eVx9uxZLl26xLfffsvGjRv5/vvvGTduHIGBgcTExLB48WIGDx5c7uM+/vjj+vIxbm5u7Ny5k8uXL/PBBx/oH7c4vilTpuDo6HjTpGrYsGEsWLCAZ599ljvvvJMDBw4YXG2qoZJ2Ku1U1Iy2bduSlpbGJ598wuXLl1m/fn2p/QUFBWzbtq1USbrKnhsdHa0fUw0QGBhIVFRUlWN0cXGRLzBClCDJ7A0cHBwYM2YMK1eu5NVXXy2z/9FHHyUvL4/PPvuMzz77DH9/fz7++GOee+45M0RbVqdOnfjjjz94/vnnSUlJwdfXl0cffZQ333xTf8wXX3xBq1atWLJkCa+//jqurq60a9eOhx9+GICQkBBUKhWzZs0iMTERLy8v7rjjDt59991yH7d3794sWbKExYsXk5eXR0hICEuWLGHcuHEABAUFMX/+fD7//HMWLlyIv78/4eHh5V5v9OjRfPDBByxcuJAlS5bQu3dvNmzYQKtWrUzyOlk6aafSTkXNaNu2LQCvv/46Dz30UKlJegB79uxBURQGDBhQ5XMVKbsmRI1QKfLXJQRLly7loYceIi0tDWdnZ6ysrG56vFarRVEUbGxsWLhwIf/3f/9XS5GKhkzaac3TarU4OjqiVqu5fPlyqZ5UgOeee45r166xbt26Kp8bHx9P165diYmJAeDixYuMGzeO8+fPlxvPwIED8fT05Pfff6/0czDmHCEsmUwAE6IEd3d3unbtWuFxQ4YMkVnjwmykndYcKysr8vPzyc3NLZOMQvnjZStzro+PD8HBwfz1118AfPfdd0yYMKHCmIqrF1TU96TT6Sp1nBD1jfTMCgEkJydz7do1ABwdHfU1P8tz8eJFfUmdoKAgvLy8ajxGIaSdWr5Tp07x4IMPkpWVRevWrVm5ciVubm7lHj9w4ED9Ms+ffvopzz77bLnHzps3jzfeeAOAiRMnSs+saDAkmRVCCCHqqJJfSAIDA/Vl6QyJjY0lNjYWKCpv17x581qJUQhzk2RWCCGEEEJYLBkzK4QQQgghLJYks0IIIYQQwmJJMiuEEEIIISyWJLNCCCGEEMJiSTIrhBBCCCEsliSzQgghhBDCYkkyK4QQQgghLJYks0IIIYQQwmJJMiuEEEIIISyWJLNCCCGEEMJiSTIrhBBCCCEsliSzQgghhBDCYkkyK4QQQgghLNb/AxGAiM8bvfIWAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 600x240 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 2], dpi=120)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    strain_strs = results_df[\"strain_str\"].dropna().unique()\n",
    "    strain_strs_latex = results_df[\"strain_str_latex\"].dropna().unique()\n",
    "    spec = fig.add_gridspec(1, len(strain_strs), wspace=0.35)\n",
    "    for i, (strain_str, strain_str_latex) in enumerate(zip(strain_strs, strain_strs_latex)):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        ax[i].set_title(strain_str_latex)\n",
    "        sub_df = results_df.loc[results_df[\"strain_str\"] == strain_str].dropna()\n",
    "        for sim_id in results_df[\"system_id\"].dropna().unique():\n",
    "            idx = np.arange(len(results_df.index))\n",
    "            sub_sub_df = sub_df.loc[sub_df[\"system_id\"] == sim_id]\n",
    "            name = sub_sub_df.index[0].split(\"-\")[2]\n",
    "            if strain_str == \"e_11 e_22 e_33\":\n",
    "                min_ind = np.where(sub_sub_df[\"scm_energy_per_atom\"] == sub_sub_df[\"scm_energy_per_atom\"].min())\n",
    "                ax[i].plot(\n",
    "                    sub_sub_df[\"vol\"] / sub_sub_df[\"vol\"].values[min_ind],\n",
    "                    sub_sub_df[\"scm_energy_per_atom\"],\n",
    "                    \"o-\",\n",
    "                    label=name,\n",
    "                )\n",
    "            else:\n",
    "                ax[i].plot(sub_sub_df[\"von_mises_strain\"], sub_sub_df[\"scm_energy_per_atom\"], \"o-\", label=name)\n",
    "        if strain_str == \"e_11 e_22 e_33\":\n",
    "            ax[i].set_xlabel(\"$V/V_0$ [-]\")\n",
    "        else:\n",
    "            ax[i].set_xlabel(\"von Mises strain [-]\")\n",
    "\n",
    "    ax[len(strain_strs) - 1].legend(loc=\"center left\", bbox_to_anchor=(1.0, 0.5))\n",
    "    ax[0].set_ylabel(\"Energy per atoms [eV]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"2-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "869e5889-2565-455e-abcf-eac84a03fd52",
   "metadata": {},
   "source": [
    "Finishing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b677b6ec-c9e5-4018-9b68-e221b27e6d7d",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5638c80c-e757-4a82-a6ce-7a2868baf751",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Vacancies `Single Poin` <a id='vacancies_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "aba0990c-0fdb-4f10-88c2-eeda5f36972e",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"3-Vacancies_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4354b96b-a27c-4f21-ba81-bf1faf8ee984",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "e4fb986f-ccec-4868-a962-e3dac092d652",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_3_SP\n",
    "\n",
    "print(settings_3_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0eab26e9-375f-41d0-aff9-9b8992d2d1d2",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "161d44d9-ef7e-4488-aad4-92e343a79c0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a68f6664-c20a-4249-8828-851b2ce85fe7",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c585240e-6a1a-421b-964c-9d57940f871b",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_3_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7c7af390-ece1-4f88-b078-783b8ab4543a",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "94d277cd-28be-4a1a-bf2a-a16c7857e261",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "aa2a3582-aa50-47f7-ba80-9245b8df17fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"3-vacancies\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "cb595c61-85c8-4b9e-9c1e-9017bca3a327",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "5350f87e-0e65-44f9-b244-e1aba88d4b23",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job_dir = os.path.join(RUN_DIR, \"jobs\", sim_name)\n",
    "    if os.path.exists(job_dir):\n",
    "        job = scm.AMSJob.load_external(job_dir, settings=settings_3_SP)\n",
    "        if not job.check():\n",
    "            print(f\"sim {sim_name} FAILE will rerun it\")\n",
    "        else:\n",
    "            continue\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_3_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8f49b5db-b879-4408-aff0-b837a84ac42a",
   "metadata": {},
   "source": [
    "### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "73771b2c-c15d-4705-9609-853eb8b2470f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10:33:40] JOB jobs STARTED\n",
      "[10:33:40] JOB jobs/SP-3.0-0-LiF_Fm-3m_-3.18_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.0-1-LiF_P6_3mc_-3.17_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.0-2-LiF_Pm-3m_-2.89_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.1-0-LiF_Fm-3m_-3.18_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.1-1-LiF_P6_3mc_-3.17_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.1-2-LiF_Pm-3m_-2.89_nv_1 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.10-1-LiF_P6_3mc_-3.17_nv_4 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.11-1-LiF_P6_3mc_-3.17_nv_4 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.12-1-LiF_P6_3mc_-3.17_nv_4 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.13-1-LiF_P6_3mc_-3.17_nv_4 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.14-1-LiF_P6_3mc_-3.17_nv_5 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.15-1-LiF_P6_3mc_-3.17_nv_5 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.16-1-LiF_P6_3mc_-3.17_nv_5 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.17-1-LiF_P6_3mc_-3.17_nv_5 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.2-0-LiF_Fm-3m_-3.18_nv_2 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.2-1-LiF_P6_3mc_-3.17_nv_2 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.2-2-LiF_Pm-3m_-2.89_nv_2 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.3-0-LiF_Fm-3m_-3.18_nv_2 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.3-1-LiF_P6_3mc_-3.17_nv_2 STARTED\n",
      "[10:33:40] JOB jobs/SP-3.3-2-LiF_Pm-3m_-2.89_nv_2 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.4-0-LiF_Fm-3m_-3.18_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.4-1-LiF_P6_3mc_-3.17_nv_2 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.4-2-LiF_Pm-3m_-2.89_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.5-0-LiF_Fm-3m_-3.18_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.5-1-LiF_P6_3mc_-3.17_nv_2 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.5-2-LiF_Pm-3m_-2.89_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.6-0-LiF_Fm-3m_-3.18_nv_4 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.6-1-LiF_P6_3mc_-3.17_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.6-2-LiF_Pm-3m_-2.89_nv_4 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.7-0-LiF_Fm-3m_-3.18_nv_4 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.7-1-LiF_P6_3mc_-3.17_nv_3 STARTED\n",
      "[10:33:41] JOB jobs/SP-3.7-2-LiF_Pm-3m_-2.89_nv_4 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.8-0-LiF_Fm-3m_-3.18_nv_5 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.8-1-LiF_P6_3mc_-3.17_nv_3 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.8-2-LiF_Pm-3m_-2.89_nv_5 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.9-0-LiF_Fm-3m_-3.18_nv_5 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.9-1-LiF_P6_3mc_-3.17_nv_3 STARTED\n",
      "[10:33:42] JOB jobs/SP-3.9-2-LiF_Pm-3m_-2.89_nv_5 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "221461b5-e587-4d6b-94b3-fea7024c369a",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "9c4b6e17-0ba7-40f4-93d2-297d28a80219",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"3-vacancies\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "4833283c-b9a4-430a-9fe1-4f2d8631e29f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vacancy_site</th>\n",
       "      <th>vacancy_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3.0-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.1-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963]]</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.2-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [0....</td>\n",
       "      <td>[F, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.3-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [2....</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.4-0-LiF_Fm-3m_-3.18_nv_3.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [0....</td>\n",
       "      <td>[F, Li, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                     vacancy_site  \\\n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif         [[1.6670525006, 1.1787841278, 2.88741963]]   \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006, 1.1787841278, 2.88741963], [0....   \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006, 1.1787841278, 2.88741963], [2....   \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif  [[1.6670525006, 1.1787841278, 2.88741963], [0....   \n",
       "\n",
       "                               vacancy_element  unitcell_repetition  \n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif            [Li]                    4  \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif             [F]                    4  \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif         [F, Li]                    4  \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif          [F, F]                    4  \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif     [F, Li, Li]                    4  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "18328189-e4ab-4f3c-b88b-64191992ff5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_3_SP)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "70999e98",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vacancy_site</th>\n",
       "      <th>vacancy_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3.0-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-250.853627</td>\n",
       "      <td>-4.733087</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.1-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963]]</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-250.588127</td>\n",
       "      <td>-4.728078</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.2-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [0....</td>\n",
       "      <td>[F, Li]</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-246.515831</td>\n",
       "      <td>-4.740689</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.3-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [2....</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-242.263743</td>\n",
       "      <td>-4.658918</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.4-0-LiF_Fm-3m_-3.18_nv_3.cif</th>\n",
       "      <td>[[1.6670525006, 1.1787841278, 2.88741963], [0....</td>\n",
       "      <td>[F, Li, Li]</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-238.689686</td>\n",
       "      <td>-4.680190</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                     vacancy_site  \\\n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif         [[1.6670525006, 1.1787841278, 2.88741963]]   \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006, 1.1787841278, 2.88741963], [0....   \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006, 1.1787841278, 2.88741963], [2....   \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif  [[1.6670525006, 1.1787841278, 2.88741963], [0....   \n",
       "\n",
       "                               vacancy_element  unitcell_repetition  \\\n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif            [Li]                  4.0   \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif             [F]                  4.0   \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif         [F, Li]                  4.0   \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif          [F, F]                  4.0   \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif     [F, Li, Li]                  4.0   \n",
       "\n",
       "                                scm_energy  scm_energy_per_atom  system_id  \n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif -250.853627            -4.733087        0.0  \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif -250.588127            -4.728078        0.0  \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif -246.515831            -4.740689        0.0  \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif -242.263743            -4.658918        0.0  \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif -238.689686            -4.680190        0.0  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "d19127ed-dac9-4e9d-9b0e-89e2f657f1e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x160 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 1.6], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    system_id = results_df[\"system_id\"].unique()\n",
    "    spec = fig.add_gridspec(1, len(system_id), wspace=0.35)\n",
    "    for i, sim_id in enumerate(results_df[\"system_id\"].unique()):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        idx = np.arange(len(results_df.index))\n",
    "        sub_df = results_df.loc[results_df[\"system_id\"] == sim_id]\n",
    "        xticks = np.arange(len(sub_df))\n",
    "        name = sub_df.index[0].split(\"-\")[2]\n",
    "        ax[i].bar(xticks, sub_df[\"scm_energy_per_atom\"], label=name)\n",
    "        ax[i].set_ylim([-4.8, -4.2])\n",
    "        ax[i].set_xticks(xticks)\n",
    "        xticks_labels = [\",\".join(sites) if sites == sites else \"\" for sites in sub_df[\"vacancy_element\"]]\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=90, ha=\"center\")\n",
    "        ax[i].set_title(name)\n",
    "    ax[0].set_ylabel(\"Energy per atoms [eV]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"3-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bd8c5dd8-4b60-40c8-a139-a3980eaedebc",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2053e1e-b476-4118-8c09-cbf4d382e211",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4738e113-5803-47a9-bbf2-84fb40ebb0db",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Substitution `Single Poin` <a id='substitution_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "d07eb665-553d-4cc1-956e-f88f1e639ffb",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"4-Substitution_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "96554a37-8305-468d-bec7-85e36ae7753e",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "9654bc2d-ce4c-43e1-a1c8-ec2444bb456f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_4_SP\n",
    "\n",
    "print(settings_4_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8a80f548-9b4b-4436-bab6-46a4ce7346a3",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "8434a3f1-6f90-4ae7-97e4-33a5bfc3ce06",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 8\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d7333205-92f6-407f-bbb7-b1d23ffd235e",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "307e7089-9042-486b-978c-ae99034129ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_4_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "349a23ce-0ae2-4f8a-9ca7-9763c0bd6c0d",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "99f95c01-81f5-4a26-8809-98f87cd5c07c",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "fb1cdfc9-c7af-4c2c-8db3-d3c930d018a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"4-substitutions\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "602e9283-766b-4ffd-9613-ef99bf86f00f",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b35621a1-c3bc-4805-950a-4081d98b66d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_4_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = f\"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\n",
    "export OMP_NUM_THREADS={CPUS_PER_CORE}\n",
    "\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dcd3b54f-2b56-4cb1-aafb-dab4fa7ca13f",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "76806d93-a1ba-4889-9380-a8a164c2902c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[18:57:04] JOB jobs STARTED\n",
      "[18:57:04] JOB jobs/SP-4.0-0-LiF_Fm-3m_-3.18_ns_1 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.0-1-LiF_P6_3mc_-3.17_ns_1 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.0-2-LiF_Pm-3m_-2.89_ns_1 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.1-0-LiF_Fm-3m_-3.18_ns_2 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.1-1-LiF_P6_3mc_-3.17_ns_2 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.1-2-LiF_Pm-3m_-2.89_ns_2 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.2-0-LiF_Fm-3m_-3.18_ns_3 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.2-1-LiF_P6_3mc_-3.17_ns_3 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.2-2-LiF_Pm-3m_-2.89_ns_3 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.3-0-LiF_Fm-3m_-3.18_ns_4 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.3-1-LiF_P6_3mc_-3.17_ns_4 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.3-2-LiF_Pm-3m_-2.89_ns_4 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.4-0-LiF_Fm-3m_-3.18_ns_5 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.4-1-LiF_P6_3mc_-3.17_ns_5 STARTED\n",
      "[18:57:04] JOB jobs/SP-4.4-2-LiF_Pm-3m_-2.89_ns_5 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e64b12f1-943f-48ba-9acd-507271e333c7",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fd8c770c-710b-4139-9f23-7af063e68448",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "1f7a9fc4-87e3-499d-af00-a5ab418ea555",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"4-substitutions\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "# data_path = os.path.join(ALLSIM_DIR, f'results_0-UnitCell_SP.json')\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "4bdf9d61-90c6-4529-b1f1-ade478c31499",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>substitution_site</th>\n",
       "      <th>substitution_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4.0-0-LiF_Fm-3m_-3.18_ns_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.1-0-LiF_Fm-3m_-3.18_ns_2.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.2-0-LiF_Fm-3m_-3.18_ns_3.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.3-0-LiF_Fm-3m_-3.18_ns_4.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.4-0-LiF_Fm-3m_-3.18_ns_5.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li, F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                substitution_site  \\\n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "\n",
       "                               substitution_element  unitcell_repetition  \n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                  [F]                    4  \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif               [F, F]                    4  \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif           [F, F, Li]                    4  \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif        [F, F, Li, F]                    4  \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif     [F, F, Li, F, F]                    4  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "cceae5d5-e0d8-473e-92da-ac73118825c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_4_SP).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "24949e76",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>substitution_site</th>\n",
       "      <th>substitution_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4.0-0-LiF_Fm-3m_-3.18_ns_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4</td>\n",
       "      <td>-246.959376</td>\n",
       "      <td>-4.573322</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.1-0-LiF_Fm-3m_-3.18_ns_2.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4</td>\n",
       "      <td>-233.970146</td>\n",
       "      <td>-4.332780</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.2-0-LiF_Fm-3m_-3.18_ns_3.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li]</td>\n",
       "      <td>4</td>\n",
       "      <td>-223.081439</td>\n",
       "      <td>-4.131138</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.3-0-LiF_Fm-3m_-3.18_ns_4.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li, F]</td>\n",
       "      <td>4</td>\n",
       "      <td>-212.037549</td>\n",
       "      <td>-3.926621</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.4-0-LiF_Fm-3m_-3.18_ns_5.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...</td>\n",
       "      <td>[F, F, Li, F, F]</td>\n",
       "      <td>4</td>\n",
       "      <td>-200.977492</td>\n",
       "      <td>-3.721805</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                substitution_site  \\\n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif  [[0.0, 0.0, 0.0], [0.8335262503, 2.3575682556,...   \n",
       "\n",
       "                               substitution_element  unitcell_repetition  \\\n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                  [F]                    4   \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif               [F, F]                    4   \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif           [F, F, Li]                    4   \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif        [F, F, Li, F]                    4   \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif     [F, F, Li, F, F]                    4   \n",
       "\n",
       "                                scm_energy  scm_energy_per_atom  system_id  \n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif -246.959376            -4.573322        0.0  \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif -233.970146            -4.332780        0.0  \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif -223.081439            -4.131138        0.0  \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif -212.037549            -3.926621        0.0  \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif -200.977492            -3.721805        0.0  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "8d285519-6243-425d-8377-0538a8718489",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAc0AAADiCAYAAAAsyJxKAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAlgUlEQVR4nO3dfVhUdfo/8PcghPKsgIoyA7qKZqmIkprac7KVkqYuuaZoEkZZuWa68rWyTDMrH/LSUinENFe7MHLRzcrHXxtbImiUraI5MiqiGAIqGA/n94cXs4zCcM7wOTNnhvfruuZyhpn5zA1zO/ecz/mc++gkSZJARERETXJzdABERETOgkWTiIhIJhZNIiIimVg0iYiIZGLRJCIikolFk4iISCYWTSIiIplYNImIiGRi0SQiIpKJRZOIiEgmFk2N2bRpE3r06OHoMEhlfJ/J0ZiDtmHRdJD77rsP8+bNu+XnEyZMwLFjx8y3J0+eDA8PD/j4+Jgv4eHhTY6/b98+6HQ6i+f5+Pjgq6++EvlrUBPs/T537NgRcXFxMJlM5sdcv34dycnJCAsLg7e3N8LCwrBhwwZZ8cfGxqJz587w8/NDSEgIpkyZgkuXLsl6LmmDFnLQlbBoOoG4uDhcuXLFfDEajbKfe/nyZYvn/vnPf1YvUGoWEe9zbm4uzp49i4kTJ5rvGzduHA4ePIjdu3fjypUrOHjwIAYOHChr3AULFuDEiRMoKyvD0aNHUVFRgcTERKW/GjkJtXLQlbBoasz69esRGhqq+utMnjwZcXFxSEpKQmBgIIKCgrB8+XKYTCbExMTA19cXvXr1QlZWluqxtERqvc8hISF48skncejQIQDA7t278c0332DTpk3o1q0bdDod2rdvL3tarm/fvmjTpo35tpubm8XWyfz58zF06FC89tprCAkJgZ+fH2bPno2SkhLExcXB398f4eHh+PLLLy3G3bFjBwYNGoS2bdsiMDAQY8eOFfDbkxL2ysG6LdEtW7YgIiICXl5eGDlyJC5fvoxXX30VISEhCAoKwuuvvy48FjWwaLZgGRkZePDBB3HhwgWkpKRg5syZmDRpEt577z1cvnwZDz/8MCZPnuzoMEmBM2fOYPPmzbjrrrsAAN988w26dOmCd955ByEhIdDr9ZgyZQqKi4tljzl37lz4+vqiXbt2yMjIuOXD7YcffkBgYCAKCgqwe/duLFu2DA8//DCmT5+OkpISvPjii5gyZQquXbtmjmns2LF4+eWXUVRUhLNnzyIpKUncH4Ec6uYcrLNjxw4cOnQIRqMRx44dw8CBA9GhQweYTCZkZmZi4cKFTvElnUXTCWzduhUBAQHmy8KFC2U/NygoyPy8O+64w+K+oUOHYuzYsWjVqhVGjRoFf39/DB8+HL1790arVq0wadIkHD9+HKWlpaJ/JWpAc9/ntm3bYsiQIejatat5n2VxcTF+/fVXXL9+HSdOnEB2djbOnDmjaOrs7bffRnl5OfLz8zFz5kxERERY3G8wGPDSSy/Bw8MD0dHRuPPOO9G/f38MGzYMbm5uiI+PR0lJCfLz8wEAK1aswNSpUzFu3DjcdtttaN26NR588EHZ8ZB61MjBOgsXLoSvry/at2+Pxx57DAAwffp0uLu7Y9CgQbj99tvx448/Cv191ODu6ACoaX/5y1+wceNGm55bXFwMd/eG3+aQkBCL297e3hY/8/b2BgCUl5fD39/fptcn+dR4n/38/KDT6bBkyRJ4eXnB29sbb775JoYMGYJr167By8tL9mt069YNsbGxiImJgclkgoeHBwBleQQAp06dQkxMjOLfkdSn1mcNgFtyoqG8qcsRLeOWJpELi4qKavDnOp0OkiQpHq+qqgpFRUXNmn0IDw/H8ePHbX4+kSOxaDpQTU0NKisrLS61tbWODosEc+T7PHr0aHTu3BnJycmorKzEpUuXMH/+fDz66KPmLcDGHD9+HNu2bUNZWRkkScKxY8fwyiuvIDo6GkFBQTbH9NJLL+Hjjz9Geno6/vjjD1RWVmL37t02j0dN42eNOCyaDrR48WK0adPG4sJEdj2OfJ+9vb3xzTff4JdffkFQUBB69+4NvV6PtLS0Jp8rSRKWLl0Kg8EAX19fxMTEoHfv3ti+fXuzYho+fDg2b96MxYsXIzg4GKGhoVizZk2zxiTr+Fkjjk6SMUfzwQcfyBrM3d0dzz33XLODIiIi0iJZRdPd3R3Dhg1rcrCDBw/iypUrQgIjIiLSHEkGHx8fOQ+TAgICZD2Omm/jxo2St7d3g5fZs2c7OjwSRM33+fTp042OPXjwYEG/gXzMaW3i+2JJ1pbm4cOHERkZ2WQBPnLkCPr27SuilhMREWmOrIVAkZGRWL9+PSoqKqw+To2COXz4cPTp0weRkZEYNmwYcnNzG3xcfn4+7r77bkRERCA6Ohq//PKL8FjIOTBnSCnmDMkmd5PUYDBI/v7+UmJiovTjjz+qt+17k5KSEvP1bdu2SX369Gnwcffff7+UmpoqSZIkff7559KAAQPsEB1pEXOGlGLOkFyyOwIZjUZ8/fXXWL9+Pe655x5069YNCQkJeOqppxAYGKhaUQ8ICDBfLy0thU6nu+UxFy5cQHZ2Nr7++msAwJgxYzB9+nScOHEC3bp1a3Ts2tpaGI1GeHh4WIzr6ekJT09Pcb8E2USSJJSXl6NTp05wc5N/dBRzpmWzJW/UzBmAeaN1inLGlkpbUlIirVy5Uurfv7/k6ekpjRs3Ttq1a5fQal7fxIkTpdDQUCk0NFT66aefbrk/OztbioiIsPhZdHS0tHv3bqvjmkwmCQAvGr+YTCbmDC+q541aOcO8cZ6LnJyxqfdsQEAApk+fjunTpyMrKwsTJkxAeno6ampqFI0zePBgcxPnm+Xm5kKv1wOAufFvWloa5syZg507d9oS9i18fX0BAEePHjVfB4CExf/P5jG3LnrM4vZfknfYPFZLH6+srAx6vd7ivXHFnAG0/b5oObaGxrs5bxydM0DDeePKOaP18ZrKGWtsbtielZWF1NRUbN26Fe3atcP8+fNtGkOJ+Ph4PPvss7h06ZLFlLBer0dhYSGqq6vh7u4OSZJQUFAAg8Fgdby6aZK6M9PX8fCU38T6ZvXHae5YHO+G+tNZrpgzgLbfFy3H1tB4dereK0fnTP1Y6ueN1v+OrjxeUzljjaI2ekVFRXj33XfRq1cvPPDAA7hy5QrS09Px22+/4dVXX1UylCyXL1/GuXPnzLczMjIQGBiIdu3aWTyuffv2iIqKMnfnT09PR2hoaJP7Gcj1MGdIKeYMKSF7S3PkyJHYtWsX7rjjDjz33HOYMGEC2rZtq2ZsKC0txbhx41BRUQE3NzcEBwcjMzPT/G0gISEBsbGxiI2NxZo1azB58mQsWrQIfn5+SE1NVTU20ibmDCnFnCElZBdNg8GA//znP42eakgNYWFhVk9KmpKSYr7eo0cPpzjrN6mLOUNKMWdICdlFc9WqVebrJSUl2LlzJ86ePYvZs2fj3LlzqK2tRWhoqCpBEhERaYHiU4NlZWWhe/fu+PDDD7FgwQIAwK+//ornn39eeHBERERaorhozpgxAykpKfjuu+/g7n5jQ3Xw4MFWpzeIiIhcgeKiefz4cYwaNQrA/5bnenl54fr160IDIyIi0hrFRdNgMODIkSMWP8vJyUGXLl2EBUVERKRFiovm3LlzMXLkSKxcuRJVVVVYu3Yt4uLikJycrEZ8REREmqG4I9CTTz4JPz8/rF69GmFhYfjiiy+wbNkyjBgxQo34iIiINMOmNnqPPvooHn30UdGxEBERaZqs6dkTJ07IGuzkyZPNCoaIiEjLZBVNuV2AoqOjmxUMERGRlsmanq2oqMATTzzR5OOqqqqaHRAREZFWySqa8+bNkzVY//79mxUMERGRlskqmq+//rracRAREWme4uM0iYiIWioWTSIiIplYNImIiGRi0SQiIpJJcdHcuXOnudmB0WjE448/jjFjxuDMmTPCgyMiItISxUVz5syZaN26NQDglVdegY+PDwIDA5GUlCQ8OCIiIi1R3Hv2/PnzCA0NRXV1Nb799lsUFBTA09MTnTp1UiM+IiIizVBcNNu0aYOioiLk5eWhZ8+e8PX1RVVVFbsBERGRy1NcNCdNmoTo6Ghcv34db7zxBgAgOzsbXbt2FR4cERGRligumu+88w4eeugheHh44L777gMAeHh44P333xcdGxERkabYdD7Nhx9+2OL2gAEDhARDRESkZYqL5oULF7Bo0SLk5OTgypUrFvfl5OQIC4yIiEhrFBfNuLg4eHh4YPz48fDy8lIjJiIiIk1SXDRzcnJQXFwMDw8PNeK5xfDhw3H+/Hm4ubnB19cXH3zwAfr163fL48LDw+Hp6Yk2bdoAAObOnYu4uDi7xEjawpwhWzBvSA7FRXPgwIE4efIkevbsqUY8t9i6dSsCAgIAAF988QUmT56MI0eONPjYLVu2IDIyUvFrlJWVNSNC0hrmDNmCeUNyKC6aGzZsQFxcHIYOHYoOHTpY3Pfiiy8KC6xOXRIDQGlpKXQ6nfDX0Ov1FrdHzMwQ/hpkP8wZsoW984Y545wUF83ly5fjhx9+QGVlpcU+TZ1Op0rRBG4cG7p3714AN3rfWnucJEm46667sHjxYgQHB8sa32Qywc/Pz3x7wht7mxcwORxzhmxhz7xhzjgnxUVz9erVOHz4sJDp2cGDByM/P7/B+3Jzc83fyjZs2AAASEtLw5w5cxpM5gMHDsBgMKCqqgrz5s1DfHy81aSvz8/Pz+IDkLSLOUO2YN6QKIqLZnBwMMLDw4W8eFZWlqLHx8fH49lnn8WlS5cQGBhocZ/BYABwo9HCjBkzEBERISRG0hbmDNmCeUOiKC6a8+bNw7Rp0/Dqq6+iffv2FveJ/gZ1+fJlXLt2zdwMPiMjA4GBgWjXrp3F465evYqqqirzPonNmzc3uOqNXB9zhmzBvHE+/3z/cYe8ruKiOXXqVADAp59+at5RLkkSdDodampqhAZXWlqKcePGoaKiAm5ubggODkZmZqb5dRMSEhAbG4s777wTY8aMQU1NDSRJQteuXc3TLNSyMGfIFswbkktx0Tx16pQacTQoLCwMP/74Y6P3p6SkmK/n5ubaIyTSOOYM2YJ5Q3IpLpphYWHm68XFxQgKChIaEBERkVa5KX3CtWvXMG3aNHh5eaFDhw7w8vLCs88+i6tXr6oRHxERkWYoLpovv/wyjh8/jt27d+PcuXPYs2cP8vPzMWvWLDXiIyIi0gzF07Pbt29HXl6eeVVZhw4dsHXrVvTu3Rsffvih8ACJiIi0QvGWpiRJcHOzfJqbmxskSRIWFBERkRYpLpojRozA2LFjkZ2djYsXL+LgwYOIi4vDyJEj1YiPiIhIMxQXzaVLl8JgMGDo0KHo2LEj7rnnHoSGhuL9999XIz4iIiLNULxP08fHB5988gk+/vhjXLx4EcHBwaqcDYCIiFyHozr4iKa4aPbu3Rt5eXnQ6XQWbfQiIyNx+PBhkbGRE3KV/xhERA1RPD1rNBob/HlBQUFzYyEiItI02VuaM2fOBABUVVWZr9f57bff0LVrV7GRERERaYzsollSUgIAqK2tNV8Hbhxu0qtXLyxdulR8dERERBoiu2impqYCAKKiovDCCy+oFhAREZFWKV4IVL9glpeXWzQ14BnJiYjIlSleCFRQUIDhw4fDy8sLAQEBaNu2rflCRETkyhQXzeeeew5t27bF999/Dx8fH+Tk5CA2NhZr1qxRIz4iIiLNUDw9m5WVhdOnT8PHxwc6nQ59+/ZFSkoKhg0bhoSEBDViJCIi0gTFW5qtWrWCp6cngBv7MH///Xf4+/vDZDIJD46IiEhLFG9pRkZGYs+ePYiJiTFvXXp5eaFXr15qxEdERKQZirc0P/74Y/Ts2RMAsGLFCgQGBqKqqgppaWnCgyMiItISxVuaer3efD0oKAjr1q0TGhARETke+0g3TPGWJhERUUuleEuTXAu/TRIRycctTSIiIpkUFc3q6mq88sorqKysVCseIiIizVJUNN3d3fHJJ5/gtttuUyueRqWmpkKn0yEjI6PB+/Pz83H33XcjIiIC0dHR+OWXX+wbIGkOc4aUYs5QUxRPz8bFxWHTpk1qxNIoo9GIdevWYdCgQY0+Ztq0aUhMTMTx48cxZ84cTJ48Wfb4ZWVlFhdyfswZUkrtnAHAnHEBioumyWTC1KlTERUVhVGjRuGJJ54wX9RQW1uLhIQErFy50tyJ6GYXLlxAdnY2nnrqKQDAmDFjYDKZcOLECVmvodfr4e/vb76Qc2POkFL2yBnAMm/IOSlePTtgwAAMGDBAjVgatHTpUgwZMgT9+/dv9DEmkwkhISFwd7/x6+h0OhgMBhQUFKBbt25NvobJZLI4rdmEN/Y2P3ByGOYMKWWPnKkboy5vmDPOSXHRfP3114W9+ODBg5Gfn9/gfbm5uSgtLUV6ejoOHDgg7DUb4ufnx3OBOgnmDCmllZwBmDeuwKbjNPfu3YvPPvsM58+fxz//+U9kZ2ejvLwc999/v6JxsrKyrN6fmZkJo9GI7t27AwDOnz+PxMREFBYWIikpyfw4vV6PwsJCVFdXw93dHZIkoaCgAAaDQfkvR5rGnCGlmDMkkuJ9mikpKZg4cSI6dOhg/mbm4eGB1157TXhwSUlJKCwshNFohNFoxKBBg7B27VqLRAaA9u3bIyoqChs3bgQApKenIzQ0VPaUCbkO5gwpxZwhJRQXzSVLluDrr7/GW2+9BTe3G0/v1asXfv31V+HBNSUhIQHbt28HAKxZswZr1qxBREQEFi9ejNTUVLvHQ9rHnCGlmDNUn+Lp2UuXLplPA6bT6cz/1l1X0759+yxup6SkmK/36NGjyWkYanmYM6SUs+YMW2Lah+Itzb59+yI9Pd3iZ9u3b0dUVJSwoIiIiLRI8Zbme++9h+HDh+Ozzz7DtWvX8Ne//hV79uzBrl271IiPiIhIMxQXzaioKPz888/YuHEjQkJCoNfr8d5776FTp05qxEc34RQMEZHj2HTISceOHTFr1izRsRAREWma4n2a1dXVWLBgASIiIuDt7Y2IiAi8+eabqKqqUiM+IiIizVC8pfnSSy/h+++/x1tvvYWwsDCcPn0aixcvRlFREVatWqVGjERERJqguGhu3boVeXl56NixIwBg4MCBGDp0KPr06cOiSURELk3x9KyPjw+8vb0tfubt7Q1fX19hQREREWmR4qL5f//3fxg/fjwOHTqEixcvIjs7GxMnTsS8efN4rjgiInJpiqdnExMTAQA7d+6ETqeDJEkAbjQ9TkxMhCRJ0Ol0qKmpERspEbkkHkZFzkRx0Tx16pQacRAREWme4qIZFhamRhxERESap3ifJhERUUvFoklERCSTTW30iIioebgAyjkp3tI8ePCgGnEQERFpnuItzZiYGHTu3BlTpkzBxIkTERwcrEZcLoPfJomIXIfiollYWIht27YhLS0N8+bNQ0xMDJ5++mk89thjcHPjLlIiV8cvgtSSKa5ynp6eGD9+PL766iucOHECAwcOxMsvv4zOnTtj9uzZOHnypBpxEhEROVyzNg1PnDiBY8eOoaioCL169cKlS5fQv39/LF68WFR8REREmqF4evb06dNIS0tDWloaqqqqMHnyZBw+fBhdunQBACQnJyMqKgp///vfhQdLRETkSIqLZs+ePTFy5EisWrUKMTEx0Ol0Fvf/6U9/wvjx44UFSEREpBWKi6bJZEJQUJDVx3z00Uc2B0RERKRViovm0aNHG/y5p6cnDAYDQkJCmh0UERGRFikumo8//jiuXLmCmpoatG7dGpWVlWjVqhVat26Na9euYciQIdi0aRP0er0a8RKRDXiYCJEYilfPLly4EPHx8bhw4QKuXbuGoqIiTJkyBUuWLMHZs2cREhKC559/Xnigqamp0Ol0yMjIaPD+8PBw9OjRA5GRkYiMjMSWLVuEx0DOhTlDtmDekDWKtzQXLVqEkydPwtPTEwAQHByMFStWoHv37khKSsJHH32E7t27Cw3SaDRi3bp1GDRokNXHbdmyBZGRkUJfm5wTc4Zswbyhpije0qyursaZM2csfnbu3DlUVVUBAHx9fVFdXS0mOgC1tbVISEjAypUrzYVatLKyMosLOTfmDNnCEXlz/fp1VV6H1KO4aCYkJGD48OFYunQpPv/8cyxduhQxMTF45plnAAA7duzA7bffLizApUuXYsiQIejfv3+Tj500aRJ69+6NqVOn4uLFi7JfQ6/Xw9/f33wh58acIVs4Im/efvvt5oRMDqB4enbBggXo0qULNm/ejHPnzqFTp06YO3cunn76aQDAiBEj8Nhjj8kaa/DgwcjPz2/wvtzcXJSWliI9PR0HDhxocqwDBw7AYDCgqqoK8+bNQ3x8PHbu3CkrDpPJBD8/P/PtCW/slfU8sj/mDNlCq3mj1hYtqUdR0ayursb48ePx6aefYurUqQ0+plWrVrLHy8rKsnp/ZmYmjEajeR/p+fPnkZiYiMLCQiQlJVk81mAwAAA8PDwwY8YMREREyI7Dz8/PIpFJu5gzZAvmDYmiaHrW3d0d3333Hdzd7XPu6qSkJBQWFsJoNMJoNGLQoEFYu3btLUl89epVXL582Xx78+bN6Nevn11iJG1hzpAtmDckl+LqN23aNCxZsgTJyclqxKNIQkICYmNjceedd2LMmDGoqamBJEno2rUrNmzY4OjwSIOcJWd4XKW2OEvekPp0kiRJSp7Qr18//PzzzwgICEDnzp0tzqGZk5MjPEA1lZWVwd/fH6WlpZwy0SAtvj+NxTTy5S+bNS6LpDjOlDekDUreH8VbmjNmzLA1LiIiIqemuGjGx8erEQcREZHm2XQS6tTUVDz00EPo06cPAGD//v3YunWr0MCIiIi0xqbes8uWLcOTTz6JgoICAEBISAjeffdd4cERERFpieKimZKSgp07dyIhIcF8Aupu3brh5MmTwoMjIiLSEsVF8+rVq+ZzZtYVzaqqKna2ICIil6d4IdCgQYOwevVqvPDCC+afffLJJxgyZIjQwIicCQ8ZIWoZFBfN5cuX48EHH8T69etx5coV3H333SgqKsK3336rRnxERESaobhodu3aFUePHsWOHTtgNBqh1+sxYsQIeHt7qxEfERGRZth0yEmbNm0wduxYzJo1C3FxcS2mYF6/fh3z588Xdg48jtcyaPnvqOXY1BjPWWj979jSxqtPcRu98vJyLFu2DIcOHUJ5ebnFfXv27BEanNqUtrYS3QqL49n39UWwJSZH/x2dNTZbx3OFvNHC37EljadqG72JEyfCZDJhzJgxLWYLk4iICLChaO7btw8FBQWa+QZHRERkL4qLpl6vR1VVlRqx2F3dzHRZWZmsx9c9Tu7jOV7zxqt7nMI9CKpSmjP1H6vF90XLsdk6nivkjRb+ji1pPCU5o3if5po1a/D5559jzpw56NChg8V9db1oncWZM2eg1+sdHQY1wWQyITQ01NFhAGDOOBPmDSklJ2cUF83658+0GEinQ01NjZKhHK62thbnzp2Dr6+vubsRaYckSSgvL0enTp0azTt7Y85oH/OGlFKSM4qLJhERUUsl7GtYSUmJqKGIiIg0SXbRbN++vcXtmJgYi9thYWFiIiIiItIo2UWzoqLC4vbBgwctbnOWl4iIXJ3sotnUzmvu3CYiIlenjaVlRERETkB2c4M//vgDH3zwgfl2ZWWlxW1XaXhARETUGNmHnNx3331NTsHu3btXSFBERERaJHt6dt++fdi7d6/ViytKTEw0X//yyy8dGIn6MjMzzddFLOzKyclp9hjOiDlju5aaM+Q82NygCVFRUeb/yPWv22L+/PmYP38+gBurj6Ojo5sV29WrV+Ht7d1of0WlTfVF/q43j/HII4/gX//6V7PGcxai/475+fno3r07fvrppwbvV9q+UmTeMGfE2L9/P+69915s3769wftjY2MVjaf1z5p7770X+/fvR9u2bS1mMCVJgk6nw++//65ovPHjx2Pz5s3o169fgzOiIr+MKW7Y3pI19/vF9u3bzYk8bdq0Zr+Rw4YNQ05ODgICAqDT6cwJV/ev0raG9X8/Ed+l6o9x/vz5Zo/njET8Hf/2t78hMzMTjz/++C336XQ6/Pbbb4rGE5k3zBkxNm7ciHvvvRfLli275b6zZ88qLppa/6z5xz/+AQA4fPhws+KqM2vWLADA8uXLhYxnDYtmEyoqKpCXlwdJklBZWWm+XkfJt3zRHzB1/xFqa2ubPRZgediQiEOIRI/nLETmDPC/KdBTp04JiU9k3jBnxFi3bh2AhteF2NLoXeufNSEhIQAabooTFxeHLVu2KBqvf//+AG5swd5s1qxZDf7cVpyebUJ4eHij/3mVfsvv27cvvvvuO0iShGHDhpmv1xF5jtIhQ4bg3//+t6LneHl5oWfPngCA//73v+brdZR+Ww0ICMADDzwAANizZ4/5ep1t27YpGs9ZiMyZphgMBhQUFAgbT2neMGfUZ8t7rPXPGmtE57To8bil2QSj0ShsrLy8PAQEBJiT19/fv1lTHNbYkiSi9x+tWLHCfL2hqUVXJTJnmiL6O6/SvGHOaJPWP2usEZ3Tosdj0bQjUVMbctgytSVyCgMA4uPjhY5HtxI9hal0POaMGI0t8gJsOwZe6581zjweiyY1afXq1ejYsSOeeOIJIeN9+eWX0Ov1iIqKEjKeq7N29nmt7l1hzihjbau6TZs2dozEPhpb5SpJEi5cuKB4vNGjRzc63qVLl2yKsTHcp6kB8fHx8Pf3x6xZs2AwGGQ/7+bl2nXqTqhaXV0tJL7k5GQcOXIExcXF+OGHH5o9Xnx8PLKzsxEaGopdu3YJiNC1ubm5mafW6jRnqs0eecOc0SatfNbs37/f6v1KZzDS0tKs3i9yBoNFUwMyMjKQn5+PrKwsRQsdTp8+bfV+0adrq62tbfKs5koUFxcjKChI2Hgkjz3zhjmjLc7yWaNlLJp2lJiYiLVr1wK4Md2ktYUOog9gzszMxIgRIwD876Blci3MGRLJGab1WTTtSHT3lMYkJycjKCgI06ZNg7e3t+L46k8HNmca0F6/b0tl61RbY2zJG+aMNonuMNQYWz9rrI2n9Wl9LgRyEDW/q3Tu3BlHjx7FI488ggMHDsh+nugDmEUfYE2WRo8ejfz8fMyYMUPI8Yu25A1zRptEdxhqjK2fNY1ZtGgRAHH5VLevs7i4WMh4AIumXYnuFNOY559/Xsg49dlyAHNL7e4iWmPT+qNGjRL6OqLzhjnjOKI7DDXG1pwRPa3fFJH7wTk9a0eiO8WIbspsjV6vh8lkUvQc0d1iWiqtN4BvDHNGm2zpkCP6s0b0tL7oBvDWsGg6MXvu/7HlP5roZeUtVf33tl+/fsjNzW3WeCNGjEBmZia6dOlyy30i2/wxZ7TJlvdF6/uaCwsLERIS0ugqX5Grezk968S0vv+HH3BiaL0BvEjMGTFEdxiy52eNLdP6ohvAW8Oi6cRqa2tRXl4OSZIsrtdRul+gqQOYRRG9rNzVVVRUWCzcqH/d0Q3gmTPaJLrDkOjPGmtE97LNysoSOh6nZ52Y6E4x9jqAWfSychJH6X5I5kzLIPqzxhrRZyWxZd+6NSya5DCiu8VQ84n+wBKNOeP6eGowcnq2HsBs72XlJI89GsAzZ8ga0dP6ohvAW8Oi6YJEd4qx9QDmYcOGIScnBwEBAUKWlZMY9d+POvXfFxGYMy2DrZ81hw8fFhrH8uXLhY5nDadnXZCtTZmJiJRoiZ81LJpOTHQDeHs2S7BlWTlpj70aJQDMmZZAdC9bNVZds2g6MdEHHNvzAGbRK9qo+WyZarNXowSAOeNI9moAv2rVKhw9ehR5eXlCetmqseqa+zRdhIjvPvY8gJl9RbXHlgbw9myUwJxxHHs1gBfd/1h0A3iARdOpie4UY88DmMlx7NUAXuuHr5B8ohvAi94VZM9V15yedWKiG8CLPoC5qWXl1dXVisYjMew1DW/LdCpzxvlooZet6Abw1nBL04kZjUah44mcwgDELysn8dT8zmzLdCpzpmUQvStI9HldrWHRJNWIPLMAiSNyWl90owTmjDaJbgBvz11Bolddc3qWmiS6WYLoZeWkjMhpfXv1JGXOOFZDq6PraGFXkDWiV11zS5OaZMuqSmts7RZDYoic1rfHdBjAnHE00auj7ZU3gPhV19zSJDPRzRKIiBxN9CpuFk0yE72izZ4dhkhbbJ3St2eHIdIeW/PGnquuWTTJrH6h7NevH3Jzc4WNp3aHIdIWW3uS2rPDEGmPrXljr/O6AtynSfWIbpZgzw5D5BiiGyXYs8MQaY+teWPPVdcsmmRWUVFh0Q6r/nVbvuWzw5Dry87ONl9/4403VN0Pzg5DrsNevWzVWHXNoklmopsl5OXlISAgwFwo/f39eW5EF6b2bAJnK1yHvXrZqrHqmkWTVGPPZeXkGKKn9K1hw3bXIbqXbWNEN4AHuBCIiJpBdP9jax2G7rjjDp4arAWwZRreniv1WTTJ7kR3GCLXYc9OMaRNWmgAbw2nZ8nuRHcYItfBKf2WQXQvW3uu1GfRJNXY67yNRORcrK2ybtOmjeLx7LlSn9OzpBo2NyBROKVP1thzWt9N2EhEVvC7GTXH6NGjodfrMWPGDEeHQhpUW1uLmpoa1NbWmi91t0XvB+f0LKnGnocjkGvglD5pHadnSTWiD0cg18cpfRJJjWl9bmmSakR3GKKWhd/nqbnUWKnPoklEmsEpfRJJjWl9Ts8SkWZwSp9sYa8G8ACLJhEROblnnnkG69atw/3333/LfWfPnsXx48eFvRaLJhERuSy9Xi+0ZzGP0yQiIpcl+uw4LJpEREQycfUsERE5NdEN4K3hPk0iInJqXbp0afQ+0auuWTSJiIhk4j5NIiIimVg0iYiIZGLRJCIikolFk4iISCYWTSIiIplYNImIiGRi0SQiIpKJRZOIiEim/w+Kh2ksWCbG8AAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 500x160 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 1.6], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    system_id = results_df[\"system_id\"].unique()\n",
    "    system_id = [id for id in system_id if id == id]\n",
    "    spec = fig.add_gridspec(1, len(system_id), wspace=0.35)\n",
    "    for i, sim_id in enumerate(system_id):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        idx = np.arange(len(results_df.index))\n",
    "        sub_df = results_df.loc[results_df[\"system_id\"] == sim_id]\n",
    "        xticks = np.arange(len(sub_df))\n",
    "        name = sub_df.index[0].split(\"-\")[2]\n",
    "        ax[i].bar(xticks, sub_df[\"scm_energy_per_atom\"], label=name)\n",
    "        ax[i].set_ylim([-4.8, -3])\n",
    "        ax[i].set_xticks(xticks)\n",
    "        xticks_labels = [\",\".join(sites) if sites == sites else \"\" for sites in sub_df[\"substitution_element\"]]\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=90, ha=\"center\")\n",
    "        ax[i].set_title(name)\n",
    "    ax[0].set_ylabel(\"Energy per atoms [eV]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"4-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ed79abbc-0dcb-4ffd-a36f-26c5fae732f6",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "fe6383ff-e0ea-4ea8-bdcd-5f8022af7c3d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "817f1b88-12df-46d6-8c0a-23523ea798d0",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Interstitial `Pre Optimization` <a id='interstitial_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "c1ecff11-dd33-41bf-bd6a-21c43ea424aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"5-Interstitial_PreO\"\n",
    "TASK = \"PreO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f1042c0c-24e4-4322-b9d9-ac7cf60fabb2",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Geopmetry Optimization*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "0fbe65d1-ca0e-4617-822e-deeb62d5592e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      ams: \t\n",
      "          task: \tGeometryOptimization\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tyes\n",
      "                               Convergence: \t\n",
      "                                           Energy: \t3.8087988488664447e-05\n",
      "                                           Gradients: \t0.3808798848866444\n",
      "                                           StressEnergyPerAtom: \t0.01904399424433222\n",
      "                                           Step: \t0.05\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t100\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZ\n",
      "                 Core: \tMedium\n",
      "           Dependency: \t\n",
      "                      Core: \t0.8\n",
      "           xc: \t\n",
      "              LDA: \tVWN\n",
      "           scf: \t\n",
      "               mixing: \t0.3\n",
      "           numericalquality: \tBasic\n",
      "           beckegrid: \t\n",
      "                     quality: \tBasic\n",
      "           CPVector: \t512\n",
      "           KGRPX: \t6\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_5_PreO\n",
    "\n",
    "print(settings_5_PreO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7e6fec17-0c07-4ee9-9171-e1d813ded87e",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "841c670f-7aed-40e5-b44c-09a594948154",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 1\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "72540b94-554c-4c24-b61e-8490f14cd9c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_5_PreO, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "00b67846-b1da-407a-b722-0834bb9eb8f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "775dabb6-fd99-4820-bc35-5e46513d70f5",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3c841614-8cce-4aa9-ada5-6a095a930355",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "9cff6df8-98e8-410f-a025-c8c99259e040",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"5-interstitial\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d0e32113-094b-4adb-9f4a-e1b41209f534",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "id": "6e45f7c8-a7b2-464a-8661-caaba2de15b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_5_PreO, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6ec02877-1075-4d5a-8683-18655de3d8c5",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "id": "c6fc202e-7a72-45d0-ada7-27038308418e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[26.11|20:46:50] JOB jobs STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.0-0-LiF_Fm-3m_-3.18_ni_0 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.0-1-LiF_P6_3mc_-3.17_ni_0 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.0-2-LiF_Pm-3m_-2.89_ni_0 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.1-0-LiF_Fm-3m_-3.18_ni_1 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.1-1-LiF_P6_3mc_-3.17_ni_1 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.1-2-LiF_Pm-3m_-2.89_ni_1 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.2-0-LiF_Fm-3m_-3.18_ni_2 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.2-1-LiF_P6_3mc_-3.17_ni_2 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.2-2-LiF_Pm-3m_-2.89_ni_2 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.3-0-LiF_Fm-3m_-3.18_ni_3 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.3-1-LiF_P6_3mc_-3.17_ni_3 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.3-2-LiF_Pm-3m_-2.89_ni_3 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.4-0-LiF_Fm-3m_-3.18_ni_4 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.4-1-LiF_P6_3mc_-3.17_ni_4 STARTED\n",
      "[26.11|20:46:50] JOB jobs/PO-5.4-2-LiF_Pm-3m_-2.89_ni_4 STARTED\n",
      "[26.11|20:48:59] JOB jobs/PO-5.0-0-LiF_Fm-3m_-3.18_ni_0 SUCCESSFUL\n",
      "[26.11|20:49:20] JOB jobs/PO-5.0-2-LiF_Pm-3m_-2.89_ni_0 SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2823cff0-e1f4-4f23-9989-81249714118a",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dd537421",
   "metadata": {},
   "source": [
    "save pre-optimized configurations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "e302be6b-ee73-499c-a3d3-eab3a725a822",
   "metadata": {},
   "outputs": [],
   "source": [
    "conf_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"5.1-interstitial_PO\")\n",
    "make_dir(conf_dir)\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_5_PreO)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    name = sim_name[3:]\n",
    "    mol = job.results.get_input_molecule()\n",
    "    # print(os.path.join(conf_dir, f'{name}.cif'))\n",
    "    mol = toASE(mol)\n",
    "    mol.write(os.path.join(conf_dir, f\"{name}.cif\"), format=\"cif\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e2cd7c26-7a94-4264-a170-e7c51c5e0847",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "a3e19c6d-026d-4fe5-90aa-783c30867a18",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"5-interstitial\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "dbed5699",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>interstitial_site</th>\n",
       "      <th>interstitial_element</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5.0-0-LiF_Fm-3m_-3.18_ni_0.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.1-0-LiF_Fm-3m_-3.18_ni_1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815]]</td>\n",
       "      <td>[Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.2-0-LiF_Fm-3m_-3.18_ni_2.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.3-0-LiF_Fm-3m_-3.18_ni_3.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.4-0-LiF_Fm-3m_-3.18_ni_4.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li, Li]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                unitcell_repetition  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                    4   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                    4   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif                    4   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif                    4   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif                    4   \n",
       "\n",
       "                                                                interstitial_site  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                                                 []   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif        [[0.8335262503, 0.5893920639, 1.443709815]]   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "\n",
       "                               interstitial_element  \n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   []  \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                 [Li]  \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif             [Li, Li]  \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif         [Li, Li, Li]  \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif     [Li, Li, Li, Li]  "
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "ef9369cc-9743-4d35-8594-3a9171e26372",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job_result = scm.AMSJob.load_external(job_dir, settings=settings_5_PreO).results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_PO_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_PO_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))\n",
    "results_df.to_json(os.path.join(conf_dir, \"info.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "81a3141e-0231-4821-9c16-f0c2455bc799",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>interstitial_site</th>\n",
       "      <th>interstitial_element</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5.0-0-LiF_Fm-3m_-3.18_ni_0.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "      <td>-122.520549</td>\n",
       "      <td>-5.105023</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.1-0-LiF_Fm-3m_-3.18_ni_1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>-119.788653</td>\n",
       "      <td>-4.791546</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.2-0-LiF_Fm-3m_-3.18_ni_2.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li]</td>\n",
       "      <td>-121.977475</td>\n",
       "      <td>-4.691441</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.3-0-LiF_Fm-3m_-3.18_ni_3.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li]</td>\n",
       "      <td>-121.142548</td>\n",
       "      <td>-4.486761</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.4-0-LiF_Fm-3m_-3.18_ni_4.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li, Li]</td>\n",
       "      <td>-121.414971</td>\n",
       "      <td>-4.336249</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                unitcell_repetition  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                    4   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                    4   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif                    4   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif                    4   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif                    4   \n",
       "\n",
       "                                                                interstitial_site  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                                                 []   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif        [[0.8335262503, 0.5893920639, 1.443709815]]   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "\n",
       "                               interstitial_element  scm_PO_energy  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   []    -122.520549   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                 [Li]    -119.788653   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif             [Li, Li]    -121.977475   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif         [Li, Li, Li]    -121.142548   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif     [Li, Li, Li, Li]    -121.414971   \n",
       "\n",
       "                                scm_PO_energy_per_atom  system_id  \n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif               -5.105023        0.0  \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif               -4.791546        0.0  \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif               -4.691441        0.0  \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif               -4.486761        0.0  \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif               -4.336249        0.0  "
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "909697d0-e936-44bf-9755-943116277563",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x160 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 1.6], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    system_id = results_df[\"system_id\"].unique()\n",
    "    system_id = [id for id in system_id if id == id]\n",
    "    spec = fig.add_gridspec(1, len(system_id), wspace=0.35)\n",
    "    for i, sim_id in enumerate(system_id):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        idx = np.arange(len(results_df.index))\n",
    "        sub_df = results_df.loc[results_df[\"system_id\"] == sim_id]\n",
    "        xticks = np.arange(len(sub_df))\n",
    "        name = sub_df.index[0].split(\"-\")[2]\n",
    "        ax[i].bar(xticks, sub_df[\"scm_PO_energy_per_atom\"], label=name)\n",
    "        ax[i].set_ylim([-5.5, -3.8])\n",
    "        ax[i].set_xticks(xticks)\n",
    "        xticks_labels = [\",\".join(sites) for sites in sub_df[\"interstitial_element\"]]\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=90, ha=\"center\")\n",
    "        ax[i].set_title(name)\n",
    "    ax[0].set_ylabel(\"Energy per atoms [eV]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"5-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "903d2df6-8106-456a-b58a-c33459c87e43",
   "metadata": {},
   "source": [
    "### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "id": "e04c1188-70d0-44d3-b818-ff1f1160b7c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "64414118-5377-4463-a280-25ebe369a579",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## Interstitial `Geopmetry Optimization` <a id='interstitial_go'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "8d9643ba-b833-415f-8361-b15b4fe631ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"5-Interstitial_GO\"\n",
    "TASK = \"GO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "beae9e0f-4546-4fa9-9cc1-fd6fd81bd119",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Geopmetry Optimization*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "6c86467a-be6f-4298-94da-5f6d83475473",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tGeometryOptimization\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "                               Convergence: \t\n",
      "                                           Energy: \t3.8087988488664447e-05\n",
      "                                           Gradients: \t0.3808798848866444\n",
      "                                           StressEnergyPerAtom: \t0.01904399424433222\n",
      "                                           Step: \t0.05\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t100\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "                 Core: \tMedium\n",
      "           Dependency: \t\n",
      "                      Core: \t0.8\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "              MetaGGA: \tpostscf TPSS\n",
      "           scf: \t\n",
      "               mixing: \t0.3\n",
      "           numericalquality: \tNormal\n",
      "           beckegrid: \t\n",
      "                     quality: \tNormal\n",
      "           CPVector: \t256\n",
      "           KGRPX: \t4\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_5_GO\n",
    "\n",
    "print(settings_5_GO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dc009feb-9b23-4e6c-9eb5-e1858a6ef5c2",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "8493707b-8588-4fc9-b036-b84513d3464f",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fc790e9d-a35d-4a8d-b320-3f94e9da308e",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "15892d71-a453-4ea4-b990-8c710edd5b35",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_5_GO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "edcc2db1-baad-4eeb-908a-9b58a78bb6f6",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "cef19003-ebd8-483e-9c42-4ca02bd8aa88",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "ebdceccd-a4bb-4a92-b77e-883c10cbf25a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"5.1-interstitial_PO\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2875c143-c517-4a5f-a4c7-1ccdac879c87",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f9391319-da52-4b78-b585-6568a77c1175",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_5_GO, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "de9d9ed8-8de1-4b4a-9c0c-6587c9776f6f",
   "metadata": {},
   "source": [
    "### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fc0210ec-2ce7-444d-ba7b-bd63d1b92bd1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[22:46:01] JOB jobs STARTED\n",
      "[22:46:01] JOB jobs/GO-5.0-0-LiF_Fm-3m_-3.18_ni_0 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.0-1-LiF_P6_3mc_-3.17_ni_0 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.0-2-LiF_Pm-3m_-2.89_ni_0 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.1-0-LiF_Fm-3m_-3.18_ni_1 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.1-1-LiF_P6_3mc_-3.17_ni_1 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.1-2-LiF_Pm-3m_-2.89_ni_1 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.2-0-LiF_Fm-3m_-3.18_ni_2 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.2-1-LiF_P6_3mc_-3.17_ni_2 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.2-2-LiF_Pm-3m_-2.89_ni_2 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.3-0-LiF_Fm-3m_-3.18_ni_3 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.3-1-LiF_P6_3mc_-3.17_ni_3 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.3-2-LiF_Pm-3m_-2.89_ni_3 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.4-0-LiF_Fm-3m_-3.18_ni_4 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.4-1-LiF_P6_3mc_-3.17_ni_4 STARTED\n",
      "[22:46:01] JOB jobs/GO-5.4-2-LiF_Pm-3m_-2.89_ni_4 STARTED\n",
      "[22:57:31] JOB jobs/GO-5.0-1-LiF_P6_3mc_-3.17_ni_0 SUCCESSFUL\n",
      "[22:57:47] JOB jobs/GO-5.0-2-LiF_Pm-3m_-2.89_ni_0 SUCCESSFUL\n",
      "[23:07:44] JOB jobs/GO-5.1-1-LiF_P6_3mc_-3.17_ni_1 SUCCESSFUL\n",
      "[23:09:10] JOB jobs/GO-5.0-0-LiF_Fm-3m_-3.18_ni_0 SUCCESSFUL\n",
      "[00:08:35] JOB jobs/GO-5.1-2-LiF_Pm-3m_-2.89_ni_1 SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bc5e14ad-be46-4cc0-949b-bcc98b0d5063",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f167c045-38c7-465f-b979-689bdf6c92c5",
   "metadata": {},
   "source": [
    "Post Processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1d62edfc-ee99-481a-a8ed-6363e138b39f",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "208b7b11-96d4-4146-9c4b-bb5129ccb9c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"5.1-interstitial_PO\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "# data_path = os.path.join(ALLSIM_DIR, f'results_0-UnitCell_SP.json')\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "7392f5a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>interstitial_site</th>\n",
       "      <th>interstitial_element</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5.0-0-LiF_Fm-3m_-3.18_ni_0.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "      <td>-122.520549</td>\n",
       "      <td>-5.105023</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.1-0-LiF_Fm-3m_-3.18_ni_1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>-119.788653</td>\n",
       "      <td>-4.791546</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.2-0-LiF_Fm-3m_-3.18_ni_2.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li]</td>\n",
       "      <td>-121.977475</td>\n",
       "      <td>-4.691441</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.3-0-LiF_Fm-3m_-3.18_ni_3.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li]</td>\n",
       "      <td>-121.142548</td>\n",
       "      <td>-4.486761</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.4-0-LiF_Fm-3m_-3.18_ni_4.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li, Li]</td>\n",
       "      <td>-121.414971</td>\n",
       "      <td>-4.336249</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                unitcell_repetition  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                    4   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                    4   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif                    4   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif                    4   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif                    4   \n",
       "\n",
       "                                                                interstitial_site  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                                                 []   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif        [[0.8335262503, 0.5893920639, 1.443709815]]   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "\n",
       "                               interstitial_element  scm_PO_energy  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   []    -122.520549   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                 [Li]    -119.788653   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif             [Li, Li]    -121.977475   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif         [Li, Li, Li]    -121.142548   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif     [Li, Li, Li, Li]    -121.414971   \n",
       "\n",
       "                                scm_PO_energy_per_atom  system_id  \n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif               -5.105023        0.0  \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif               -4.791546        0.0  \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif               -4.691441        0.0  \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif               -4.486761        0.0  \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif               -4.336249        0.0  "
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "aa6515ca-5dbf-41e9-8328-0b22d8a1fda5",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_5_GO)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "b29de702-c5f2-4841-8c4f-e6d24b7ed706",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>interstitial_site</th>\n",
       "      <th>interstitial_element</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5.0-0-LiF_Fm-3m_-3.18_ni_0.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "      <td>-122.520549</td>\n",
       "      <td>-5.105023</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-115.490746</td>\n",
       "      <td>-4.812114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.1-0-LiF_Fm-3m_-3.18_ni_1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>-119.788653</td>\n",
       "      <td>-4.791546</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-111.743702</td>\n",
       "      <td>-4.469748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.2-0-LiF_Fm-3m_-3.18_ni_2.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li]</td>\n",
       "      <td>-121.977475</td>\n",
       "      <td>-4.691441</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-108.530593</td>\n",
       "      <td>-4.174254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.3-0-LiF_Fm-3m_-3.18_ni_3.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li]</td>\n",
       "      <td>-121.142548</td>\n",
       "      <td>-4.486761</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-104.112090</td>\n",
       "      <td>-3.856003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.4-0-LiF_Fm-3m_-3.18_ni_4.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.8335262503, 0.5893920639, 1.443709815], [3...</td>\n",
       "      <td>[Li, Li, Li, Li]</td>\n",
       "      <td>-121.414971</td>\n",
       "      <td>-4.336249</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-99.352794</td>\n",
       "      <td>-3.548314</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                unitcell_repetition  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                    4   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                    4   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif                    4   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif                    4   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif                    4   \n",
       "\n",
       "                                                                interstitial_site  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                                                 []   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif        [[0.8335262503, 0.5893920639, 1.443709815]]   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif  [[0.8335262503, 0.5893920639, 1.443709815], [3...   \n",
       "\n",
       "                               interstitial_element  scm_PO_energy  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   []    -122.520549   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                 [Li]    -119.788653   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif             [Li, Li]    -121.977475   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif         [Li, Li, Li]    -121.142548   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif     [Li, Li, Li, Li]    -121.414971   \n",
       "\n",
       "                                scm_PO_energy_per_atom  system_id  scm_energy  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif               -5.105023        0.0 -115.490746   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif               -4.791546        0.0 -111.743702   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif               -4.691441        0.0 -108.530593   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif               -4.486761        0.0 -104.112090   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif               -4.336249        0.0  -99.352794   \n",
       "\n",
       "                                scm_energy_per_atom  \n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif            -4.812114  \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif            -4.469748  \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif            -4.174254  \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif            -3.856003  \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif            -3.548314  "
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "11646038-6ed3-44d0-9b6c-22ffb88a359a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x180 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 1.8], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    system_id = results_df[\"system_id\"].unique()\n",
    "    system_id = [id for id in system_id if id == id]\n",
    "    spec = fig.add_gridspec(1, len(system_id), wspace=0.35)\n",
    "    for i, sim_id in enumerate(system_id):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        idx = np.arange(len(results_df.index))\n",
    "        sub_df = results_df.loc[results_df[\"system_id\"] == sim_id]\n",
    "        xticks = np.arange(len(sub_df))\n",
    "        name = sub_df.index[0].split(\"-\")[2]\n",
    "        ax[i].bar(xticks, sub_df[\"scm_PO_energy_per_atom\"], label=\"PreO\")\n",
    "        ax[i].bar(xticks, sub_df[\"scm_energy_per_atom\"], label=\"GO\")\n",
    "        ax[i].set_ylim([-5.7, -3])\n",
    "        ax[i].set_xticks(xticks)\n",
    "        xticks_labels = [\",\".join(sites) for sites in sub_df[\"interstitial_element\"]]\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=90, ha=\"center\")\n",
    "        ax[i].set_title(name)\n",
    "    ax[len(system_id) - 1].legend(loc=\"center left\", bbox_to_anchor=(1.0, 0.5))\n",
    "    ax[0].set_ylabel(\"Energy per atoms [eV]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"5-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b444b36e-558d-4f27-afe1-a1e4929e5139",
   "metadata": {},
   "source": [
    "### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "2f83cbd0-f099-4d86-8960-e4061dbd106f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "be7c923e-a186-4a08-b6fe-127f8ece42c3",
   "metadata": {},
   "source": [
    "## Slab `PRE Optimization`  <a id='slab_po'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "0203c5ca-7567-4a42-815a-0644225b1dcf",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"6-Slabs_PreO\"\n",
    "TASK = \"PreO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "468756ca-c3b0-4811-b1c8-d333e37d5ae5",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Geopmetry Optimization*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "49bcde91-09ee-43a0-a706-b7f490e61185",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      ams: \t\n",
      "          task: \tGeometryOptimization\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "                               Convergence: \t\n",
      "                                           Energy: \t3.8087988488664447e-05\n",
      "                                           Gradients: \t0.03808798848866444\n",
      "                                           StressEnergyPerAtom: \t0.03808798848866444\n",
      "                                           Step: \t0.05\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t50\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tSZ\n",
      "                 Core: \tMedium\n",
      "           Dependency: \t\n",
      "                      Core: \t0.8\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "           SCF: \t\n",
      "               mixing: \t0.1\n",
      "               Iterations: \t180\n",
      "               Method: \tMultiSecant\n",
      "           numericalquality: \tGood\n",
      "           beckegrid: \t\n",
      "                     quality: \tNormal\n",
      "           Convergence: \t\n",
      "                       Degenerate: \tDefault\n",
      "           RadialDefaults: \t\n",
      "                          NR: \t10000\n",
      "           CPVector: \t512\n",
      "           KGRPX: \t6\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_6_PreO\n",
    "\n",
    "print(settings_6_PreO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "de3cdbdb-a049-40f2-9eeb-e3d11b2a01f4",
   "metadata": {},
   "source": [
    "### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a1d48165-fea2-459f-b01a-911e9d6b47a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 2\n",
    "CORES_PER_NODE = 128\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e09519db-1cf2-4f02-bd2b-66604b0265bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8fc9a209-16c8-4a6b-8d94-7a8917bebf39",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_6_PreO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "31c55d5d-4bf9-4bc9-8f19-f829162673ba",
   "metadata": {},
   "source": [
    "### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "313d2423-3ba0-48a2-9d25-f5704c1df105",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1a0c4d9d-5df7-4339-93b4-c1d27ce1a00c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"6-slabs\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7aacc55b-e8f3-4cfa-8510-b7f6170bc215",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "e7c83385-f3ec-44b3-b480-5f01741d3f50",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_6_PreO, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f66f6dfb-8bd9-44ea-abb3-2bbf9f53f45b",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "b1652642-fcbd-4fd6-85da-ef70a719c746",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02.12|00:10:26] JOB jobs STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.0-0-LiF_Fm-3m_-3.18_001x2 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.0-1-LiF_P6_3mc_-3.17_001x2 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.0-2-LiF_Pm-3m_-2.89_001x2 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.1-0-LiF_Fm-3m_-3.18_001x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.1-1-LiF_P6_3mc_-3.17_001x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.1-2-LiF_Pm-3m_-2.89_001x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.10-0-LiF_Fm-3m_-3.18_012x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.10-1-LiF_P6_3mc_-3.17_012x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.10-2-LiF_Pm-3m_-2.89_012x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.11-0-LiF_Fm-3m_-3.18_012x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.11-1-LiF_P6_3mc_-3.17_012x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.11-2-LiF_Pm-3m_-2.89_012x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.2-0-LiF_Fm-3m_-3.18_001x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.2-1-LiF_P6_3mc_-3.17_001x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.2-2-LiF_Pm-3m_-2.89_001x4 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.3-0-LiF_Fm-3m_-3.18_011x2 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.3-1-LiF_P6_3mc_-3.17_011x2 STARTED\n",
      "[02.12|00:10:26] Job PO-6.11-0-LiF_Fm-3m_-3.18_012x4 previously run as PO-6.10-0-LiF_Fm-3m_-3.18_012x3, using old results\n",
      "[02.12|00:10:26] JOB jobs/PO-6.3-2-LiF_Pm-3m_-2.89_011x2 STARTED\n",
      "[02.12|00:10:26] Waiting for job PO-6.10-0-LiF_Fm-3m_-3.18_012x3 to finish\n",
      "[02.12|00:10:26] Job PO-6.11-1-LiF_P6_3mc_-3.17_012x4 previously run as PO-6.10-1-LiF_P6_3mc_-3.17_012x3, using old results\n",
      "[02.12|00:10:26] JOB jobs/PO-6.4-0-LiF_Fm-3m_-3.18_011x3 STARTED\n",
      "[02.12|00:10:26] Waiting for job PO-6.10-1-LiF_P6_3mc_-3.17_012x3 to finish\n",
      "[02.12|00:10:26] Job PO-6.11-2-LiF_Pm-3m_-2.89_012x4 previously run as PO-6.10-2-LiF_Pm-3m_-2.89_012x3, using old results\n",
      "[02.12|00:10:26] JOB jobs/PO-6.4-1-LiF_P6_3mc_-3.17_011x3 STARTED\n",
      "[02.12|00:10:26] Waiting for job PO-6.10-2-LiF_Pm-3m_-2.89_012x3 to finish\n",
      "[02.12|00:10:26] JOB jobs/PO-6.4-2-LiF_Pm-3m_-2.89_011x3 STARTED\n",
      "[02.12|00:10:26] JOB jobs/PO-6.5-0-LiF_Fm-3m_-3.18_011x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.5-1-LiF_P6_3mc_-3.17_011x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.5-2-LiF_Pm-3m_-2.89_011x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.6-0-LiF_Fm-3m_-3.18_111x2 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.6-1-LiF_P6_3mc_-3.17_111x2 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.6-2-LiF_Pm-3m_-2.89_111x2 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.7-0-LiF_Fm-3m_-3.18_111x3 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.7-1-LiF_P6_3mc_-3.17_111x3 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.7-2-LiF_Pm-3m_-2.89_111x3 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.8-0-LiF_Fm-3m_-3.18_111x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.8-1-LiF_P6_3mc_-3.17_111x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.8-2-LiF_Pm-3m_-2.89_111x4 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.9-0-LiF_Fm-3m_-3.18_012x2 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.9-1-LiF_P6_3mc_-3.17_012x2 STARTED\n",
      "[02.12|00:10:27] JOB jobs/PO-6.9-2-LiF_Pm-3m_-2.89_012x2 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e11d7f41-5a01-4d97-a0bc-961908ace837",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f5512f12-fd31-493e-8a26-4f5bc1ceacd1",
   "metadata": {},
   "source": [
    "Save last configuration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a0b39a30-89ab-4747-bf26-934eb19c74a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-1-LiF_P6_3mc_-3.17_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.2-1-LiF_P6_3mc_-3.17_001x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-0-LiF_Fm-3m_-3.18_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.8-0-LiF_Fm-3m_-3.18_111x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-2-LiF_Pm-3m_-2.89_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.0-1-LiF_P6_3mc_-3.17_001x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.8-1-LiF_P6_3mc_-3.17_111x4 FAILED\n"
     ]
    }
   ],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "conf_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"6.1-slabs_PO\")\n",
    "make_dir(conf_dir)\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_6_PreO)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    name = sim_name[3:]\n",
    "    mol = job.results.get_input_molecule()\n",
    "    mol = toASE(mol)\n",
    "    mol.write(os.path.join(conf_dir, f\"{name}.cif\"), format=\"cif\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b8eafaa3-125e-4e03-8b0e-6535579ff015",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7dfe219f-49da-4fcd-9234-417d96754665",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"6-slabs\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8b7ec490",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>millers</th>\n",
       "      <th>slabs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0-0-LiF_Fm-3m_-3.18_001x2.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.1-0-LiF_Fm-3m_-3.18_001x3.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.2-0-LiF_Fm-3m_-3.18_001x4.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.3-0-LiF_Fm-3m_-3.18_011x2.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.4-0-LiF_Fm-3m_-3.18_011x3.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   millers  slabs\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif  [0, 0, 1]      2\n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif  [0, 0, 1]      3\n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif  [0, 0, 1]      4\n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif  [0, 1, 1]      2\n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif  [0, 1, 1]      3"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0aa26aef-c9fe-4036-8c2f-aa8e0e4fe7f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-1-LiF_P6_3mc_-3.17_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.2-1-LiF_P6_3mc_-3.17_001x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-0-LiF_Fm-3m_-3.18_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.8-0-LiF_Fm-3m_-3.18_111x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.9-2-LiF_Pm-3m_-2.89_012x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.0-1-LiF_P6_3mc_-3.17_001x2 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_PreO/jobs/PO-6.8-1-LiF_P6_3mc_-3.17_111x4 FAILED\n"
     ]
    }
   ],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_6_PreO)\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"millers_str\"] = str(results_df.loc[sim_name[3:] + \".cif\", \"millers\"])\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_PO_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_PO_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))\n",
    "results_df.to_json(os.path.join(conf_dir, \"info.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "1545c14b-8f1d-432f-9220-7bcd50aeea75",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>millers</th>\n",
       "      <th>slabs</th>\n",
       "      <th>millers_str</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0-0-LiF_Fm-3m_-3.18_001x2.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-52.869952</td>\n",
       "      <td>-3.304372</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.1-0-LiF_Fm-3m_-3.18_001x3.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-79.203034</td>\n",
       "      <td>-3.300126</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.2-0-LiF_Fm-3m_-3.18_001x4.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>4</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-108.190278</td>\n",
       "      <td>-3.380946</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.3-0-LiF_Fm-3m_-3.18_011x2.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-57.632127</td>\n",
       "      <td>-3.602008</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.4-0-LiF_Fm-3m_-3.18_011x3.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-87.752700</td>\n",
       "      <td>-3.656362</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   millers  slabs millers_str  scm_PO_energy  \\\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif  [0, 0, 1]      2   [0, 0, 1]     -52.869952   \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif  [0, 0, 1]      3   [0, 0, 1]     -79.203034   \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif  [0, 0, 1]      4   [0, 0, 1]    -108.190278   \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif  [0, 1, 1]      2   [0, 1, 1]     -57.632127   \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif  [0, 1, 1]      3   [0, 1, 1]     -87.752700   \n",
       "\n",
       "                                 scm_PO_energy_per_atom  system_id  \n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif               -3.304372        0.0  \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif               -3.300126        0.0  \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif               -3.380946        0.0  \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif               -3.602008        0.0  \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif               -3.656362        0.0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "40f5c707-c36e-4dbb-9a11-399d1e6c26bd",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1662169/3732062782.py:6: MatplotlibDeprecationWarning: The get_cmap function was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap(obj)`` instead.\n",
      "  Set2 = mp.cm.get_cmap(\"Set2\", 8)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x350 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 3.5], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    slabs_millers = results_df[\"millers_str\"].unique()\n",
    "    Set2 = mp.cm.get_cmap(\"Set2\", 8)\n",
    "    spec = fig.add_gridspec(\n",
    "        len(slabs_millers) // 2 + len(slabs_millers) % 2, min([len(slabs_millers), 2]), hspace=0.7, wspace=0.25\n",
    "    )\n",
    "    for i, millers in enumerate(slabs_millers):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        sub_df = results_df.loc[results_df[\"millers_str\"] == millers]\n",
    "        systems_id = results_df[\"system_id\"].unique()\n",
    "        xticks = 1 / sub_df[\"slabs\"].unique()\n",
    "        xticks_labels = [f\"$\\\\frac{{1}}{{{ns}}}$\" for ns in sub_df[\"slabs\"].unique()]\n",
    "        ax[i].set_xticks(xticks)\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=0, ha=\"center\")\n",
    "        ax[i].set_title(millers)\n",
    "        for sys_id in systems_id:\n",
    "            sub_df\n",
    "            idx = np.arange(len(results_df.index))\n",
    "            sub_sub_df = sub_df.loc[sub_df[\"system_id\"] == sys_id]\n",
    "            if not len(sub_sub_df.index) > 0:\n",
    "                # print(sub_sub_df.head())\n",
    "                continue\n",
    "            name = sub_sub_df.index[0].split(\"-\")[2]\n",
    "            # print(1/sub_sub_df['slabs'], sub_sub_df['scm_PO_energy_per_atom'])\n",
    "            ax[i].scatter(\n",
    "                1 / sub_sub_df[\"slabs\"],\n",
    "                sub_sub_df[\"scm_PO_energy_per_atom\"],\n",
    "                color=Set2.colors[int(sys_id)],\n",
    "                label=name,\n",
    "            )\n",
    "\n",
    "            # ax[i].set_xticks(idx)\n",
    "            # ax[i].set_xticklabels(results_df.index, rotation=25, ha='right')\n",
    "        # ax[i].set_ylim([-5.2, -3.8])\n",
    "    #     ax[i].set_ylim([-5, -3.8])\n",
    "    ax[3 - 2].legend(loc=\"center left\", bbox_to_anchor=(1.02, -0.3))\n",
    "    for i in range(0, len(ax), 2):\n",
    "        ax[i].set_ylabel(\"Energy per atoms [eV]\")\n",
    "    for i in range(len(ax) - 2, len(ax), 1):\n",
    "        ax[i].set_xlabel(r\"$1 /(\\mathrm{\\#\\, slabs})$ [-]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"6-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "136ef930-d7f5-444a-8a11-fe03f7b8cd0b",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b901b689-b0b9-482d-83b7-632a2af9e38c",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "815dbb77-7927-4ffa-8af5-b4677b9a61ce",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Slabs `Geopmetry Optimization`  <a id='slab_go'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "4827bde4-28a7-4368-978c-dcc8a2527d79",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"6-Slabs_GO\"\n",
    "TASK = \"GO\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "182f8138-3ba1-44a1-995d-6920fa7823b2",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Geopmetry Optimization*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "2eb381d8-4b88-4341-ad75-2fdd1749b86f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tGeometryOptimization\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "                               Convergence: \t\n",
      "                                           Energy: \t3.8087988488664447e-05\n",
      "                                           Gradients: \t0.3808798848866444\n",
      "                                           StressEnergyPerAtom: \t0.01904399424433222\n",
      "                                           Step: \t0.05\n",
      "                               PretendConverged: \tyes\n",
      "                               MaxIterations: \t15\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "                 Core: \tMedium\n",
      "           Dependency: \t\n",
      "                      Core: \t0.8\n",
      "           xc: \t\n",
      "              MetaGGA: \tTPSS\n",
      "           SCF: \t\n",
      "               mixing: \t0.05\n",
      "               Iterations: \t150\n",
      "               Method: \tMultiSecant\n",
      "           numericalquality: \tGood\n",
      "           beckegrid: \t\n",
      "                     quality: \tGood\n",
      "           Convergence: \t\n",
      "                       Degenerate: \tDefault\n",
      "           RadialDefaults: \t\n",
      "                          NR: \t10000\n",
      "           CPVector: \t512\n",
      "           KGRPX: \t6\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_6_GO\n",
    "\n",
    "print(settings_6_GO)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2c2b7ec2-1097-4595-acc6-beba8fae423d",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "894b77f5-0e55-4ed9-a9d5-064cf59eb3b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 64\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fc86a67d-2be9-4910-a463-1d1793b89c69",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)  # TODO: Replace it with GridRunner\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE * NODES\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "5edfd480-e5d8-4b2c-8173-9b6f36febcd6",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_6_GO, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4318c509-f1b8-4c70-bd42-3d917de6597d",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6204b8ec-a8d7-465b-9491-d798ec98deb5",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "91d8f6da-2ac0-432c-93fc-6e1f418fa498",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"6.1-slabs_PO\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    # apply 2D simmetry\n",
    "    system_ase.set_pbc([True, True, False])\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "043b928c-4c63-46a5-9a2a-1ea7afd25a3f",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "91811f99-60d3-4930-a273-7407bf76ad62",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_6_GO, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "761b5dbd-c038-4e1b-b395-7d8c182cc656",
   "metadata": {},
   "source": [
    "### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "e1724583-7c6a-4eeb-8bdd-6e197df70cef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10:47:38] JOB jobs STARTED\n",
      "[10:47:39] JOB jobs/GO-6.0-0-LiF_Fm-3m_-3.18_001x2 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.0-2-LiF_Pm-3m_-2.89_001x2 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.1-0-LiF_Fm-3m_-3.18_001x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.1-1-LiF_P6_3mc_-3.17_001x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.1-2-LiF_Pm-3m_-2.89_001x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.10-0-LiF_Fm-3m_-3.18_012x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.10-1-LiF_P6_3mc_-3.17_012x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.10-2-LiF_Pm-3m_-2.89_012x3 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.11-0-LiF_Fm-3m_-3.18_012x4 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.11-1-LiF_P6_3mc_-3.17_012x4 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.11-2-LiF_Pm-3m_-2.89_012x4 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.2-0-LiF_Fm-3m_-3.18_001x4 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.2-2-LiF_Pm-3m_-2.89_001x4 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.3-0-LiF_Fm-3m_-3.18_011x2 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.3-1-LiF_P6_3mc_-3.17_011x2 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.3-2-LiF_Pm-3m_-2.89_011x2 STARTED\n",
      "[10:47:39] JOB jobs/GO-6.4-0-LiF_Fm-3m_-3.18_011x3 STARTED\n",
      "[10:47:39] Job GO-6.11-0-LiF_Fm-3m_-3.18_012x4 previously run as GO-6.10-0-LiF_Fm-3m_-3.18_012x3, using old results\n",
      "[10:47:39] JOB jobs/GO-6.4-1-LiF_P6_3mc_-3.17_011x3 STARTED\n",
      "[10:47:39] Waiting for job GO-6.10-0-LiF_Fm-3m_-3.18_012x3 to finish\n",
      "[10:47:39] Job GO-6.11-1-LiF_P6_3mc_-3.17_012x4 previously run as GO-6.10-1-LiF_P6_3mc_-3.17_012x3, using old results\n",
      "[10:47:39] JOB jobs/GO-6.4-2-LiF_Pm-3m_-2.89_011x3 STARTED\n",
      "[10:47:39] Waiting for job GO-6.10-1-LiF_P6_3mc_-3.17_012x3 to finish\n",
      "[10:47:39] Job GO-6.11-2-LiF_Pm-3m_-2.89_012x4 previously run as GO-6.10-2-LiF_Pm-3m_-2.89_012x3, using old results\n",
      "[10:47:39] JOB jobs/GO-6.5-0-LiF_Fm-3m_-3.18_011x4 STARTED\n",
      "[10:47:39] Waiting for job GO-6.10-2-LiF_Pm-3m_-2.89_012x3 to finish\n",
      "[10:47:39] JOB jobs/GO-6.5-1-LiF_P6_3mc_-3.17_011x4 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.5-2-LiF_Pm-3m_-2.89_011x4 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.6-0-LiF_Fm-3m_-3.18_111x2 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.6-1-LiF_P6_3mc_-3.17_111x2 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.6-2-LiF_Pm-3m_-2.89_111x2 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.7-0-LiF_Fm-3m_-3.18_111x3 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.7-1-LiF_P6_3mc_-3.17_111x3 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.7-2-LiF_Pm-3m_-2.89_111x3 STARTED\n",
      "[10:47:40] JOB jobs/GO-6.8-2-LiF_Pm-3m_-2.89_111x4 STARTED\n",
      "[11:41:13] Job jobs/GO-6.3-0-LiF_Fm-3m_-3.18_011x2 reported warnings. Please check the the output\n",
      "[11:41:13] JOB jobs/GO-6.3-0-LiF_Fm-3m_-3.18_011x2 SUCCESSFUL\n",
      "[11:46:18] Job jobs/GO-6.3-2-LiF_Pm-3m_-2.89_011x2 reported warnings. Please check the the output\n",
      "[11:46:18] JOB jobs/GO-6.3-2-LiF_Pm-3m_-2.89_011x2 SUCCESSFUL\n",
      "[11:51:19] Job jobs/GO-6.4-0-LiF_Fm-3m_-3.18_011x3 reported warnings. Please check the the output\n",
      "[11:51:19] JOB jobs/GO-6.4-0-LiF_Fm-3m_-3.18_011x3 SUCCESSFUL\n",
      "[11:57:21] Job jobs/GO-6.4-2-LiF_Pm-3m_-2.89_011x3 reported warnings. Please check the the output\n",
      "[11:57:21] JOB jobs/GO-6.4-2-LiF_Pm-3m_-2.89_011x3 SUCCESSFUL\n",
      "[12:04:18] Job jobs/GO-6.5-0-LiF_Fm-3m_-3.18_011x4 reported warnings. Please check the the output\n",
      "[12:04:18] JOB jobs/GO-6.5-0-LiF_Fm-3m_-3.18_011x4 SUCCESSFUL\n",
      "[12:14:05] Job jobs/GO-6.0-2-LiF_Pm-3m_-2.89_001x2 reported warnings. Please check the the output\n",
      "[12:14:05] JOB jobs/GO-6.0-2-LiF_Pm-3m_-2.89_001x2 SUCCESSFUL\n",
      "[12:20:52] Job jobs/GO-6.5-2-LiF_Pm-3m_-2.89_011x4 reported warnings. Please check the the output\n",
      "[12:20:52] JOB jobs/GO-6.5-2-LiF_Pm-3m_-2.89_011x4 SUCCESSFUL\n",
      "[12:28:42] JOB jobs/GO-6.1-1-LiF_P6_3mc_-3.17_001x3 FAILED\n",
      "[12:29:38] JOB jobs/GO-6.4-1-LiF_P6_3mc_-3.17_011x3 FAILED\n",
      "[12:44:52] JOB jobs/GO-6.10-1-LiF_P6_3mc_-3.17_012x3 FAILED\n",
      "[12:49:02] Job jobs/GO-6.6-2-LiF_Pm-3m_-2.89_111x2 reported warnings. Please check the the output\n",
      "[12:49:02] JOB jobs/GO-6.6-2-LiF_Pm-3m_-2.89_111x2 SUCCESSFUL\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ecb08fd7-9f13-4ae8-bc4d-6e8da28ba35d",
   "metadata": {},
   "source": [
    "### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a4315ee2-bb6b-4f62-bbe6-9cac79fb9626",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "b49abc30-1c99-465b-a012-01859d286e67",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"6.1-slabs_PO\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "d94271a0-6f17-4696-8803-949b9ae4fe01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>millers</th>\n",
       "      <th>slabs</th>\n",
       "      <th>millers_str</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0-0-LiF_Fm-3m_-3.18_001x2.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-52.869952</td>\n",
       "      <td>-3.304372</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.1-0-LiF_Fm-3m_-3.18_001x3.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-79.203034</td>\n",
       "      <td>-3.300126</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.2-0-LiF_Fm-3m_-3.18_001x4.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>4</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-108.190278</td>\n",
       "      <td>-3.380946</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.3-0-LiF_Fm-3m_-3.18_011x2.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-57.632127</td>\n",
       "      <td>-3.602008</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.4-0-LiF_Fm-3m_-3.18_011x3.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-87.752700</td>\n",
       "      <td>-3.656362</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   millers  slabs millers_str  scm_PO_energy  \\\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif  [0, 0, 1]      2   [0, 0, 1]     -52.869952   \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif  [0, 0, 1]      3   [0, 0, 1]     -79.203034   \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif  [0, 0, 1]      4   [0, 0, 1]    -108.190278   \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif  [0, 1, 1]      2   [0, 1, 1]     -57.632127   \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif  [0, 1, 1]      3   [0, 1, 1]     -87.752700   \n",
       "\n",
       "                                 scm_PO_energy_per_atom  system_id  \n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif               -3.304372        0.0  \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif               -3.300126        0.0  \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif               -3.380946        0.0  \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif               -3.602008        0.0  \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif               -3.656362        0.0  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "0ef74f8c-8e63-4c5b-8b09-c9157da80268",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.1-1-LiF_P6_3mc_-3.17_001x3 FAILED\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.11-1-LiF_P6_3mc_-3.17_012x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.5-1-LiF_P6_3mc_-3.17_011x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.4-1-LiF_P6_3mc_-3.17_011x3 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.7-1-LiF_P6_3mc_-3.17_111x3 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.11-2-LiF_Pm-3m_-2.89_012x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.10-1-LiF_P6_3mc_-3.17_012x3 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.11-0-LiF_Fm-3m_-3.18_012x4 FAILED\n",
      "Simulation step02-sim/LiF/6-Slabs_GO/jobs/GO-6.7-0-LiF_Fm-3m_-3.18_111x3 FAILED\n"
     ]
    }
   ],
   "source": [
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_6_GO)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "a0ccaf03",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>millers</th>\n",
       "      <th>slabs</th>\n",
       "      <th>millers_str</th>\n",
       "      <th>scm_PO_energy</th>\n",
       "      <th>scm_PO_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0-0-LiF_Fm-3m_-3.18_001x2.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-52.869952</td>\n",
       "      <td>-3.304372</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-67.675824</td>\n",
       "      <td>-4.229739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.1-0-LiF_Fm-3m_-3.18_001x3.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-79.203034</td>\n",
       "      <td>-3.300126</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-100.444650</td>\n",
       "      <td>-4.185194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.2-0-LiF_Fm-3m_-3.18_001x4.cif</th>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>4</td>\n",
       "      <td>[0, 0, 1]</td>\n",
       "      <td>-108.190278</td>\n",
       "      <td>-3.380946</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-140.808647</td>\n",
       "      <td>-4.400270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.3-0-LiF_Fm-3m_-3.18_011x2.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-57.632127</td>\n",
       "      <td>-3.602008</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-75.675830</td>\n",
       "      <td>-4.729739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.4-0-LiF_Fm-3m_-3.18_011x3.cif</th>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 1, 1]</td>\n",
       "      <td>-87.752700</td>\n",
       "      <td>-3.656362</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-114.351394</td>\n",
       "      <td>-4.764641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   millers  slabs millers_str  scm_PO_energy  \\\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif  [0, 0, 1]      2   [0, 0, 1]     -52.869952   \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif  [0, 0, 1]      3   [0, 0, 1]     -79.203034   \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif  [0, 0, 1]      4   [0, 0, 1]    -108.190278   \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif  [0, 1, 1]      2   [0, 1, 1]     -57.632127   \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif  [0, 1, 1]      3   [0, 1, 1]     -87.752700   \n",
       "\n",
       "                                 scm_PO_energy_per_atom  system_id  \\\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif               -3.304372        0.0   \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif               -3.300126        0.0   \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif               -3.380946        0.0   \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif               -3.602008        0.0   \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif               -3.656362        0.0   \n",
       "\n",
       "                                 scm_energy  scm_energy_per_atom  \n",
       "6.0-0-LiF_Fm-3m_-3.18_001x2.cif  -67.675824            -4.229739  \n",
       "6.1-0-LiF_Fm-3m_-3.18_001x3.cif -100.444650            -4.185194  \n",
       "6.2-0-LiF_Fm-3m_-3.18_001x4.cif -140.808647            -4.400270  \n",
       "6.3-0-LiF_Fm-3m_-3.18_011x2.cif  -75.675830            -4.729739  \n",
       "6.4-0-LiF_Fm-3m_-3.18_011x3.cif -114.351394            -4.764641  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "a2d19f6b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1662169/3962507603.py:6: MatplotlibDeprecationWarning: The get_cmap function was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap(obj)`` instead.\n",
      "  Set2 = mp.cm.get_cmap(\"Set2\", 8)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkQAAAFkCAYAAADBpCTxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABJB0lEQVR4nO3de1xUdf4/8NcBEZA7gshVQEUkUQQU1DQrk3TV3Nzk65p5CTWybd3KLdRW3LRcM+2qliaZmWutaWa0mYlp6c+vJKRmeYNRbqLEZRABB+b8/uDL5AjKHDhnrq/n4zEPmZkPZ17HmQ+8+ZzP+RxBFEURRERERDbMztQBiIiIiEyNBRERERHZPBZEREREZPNYEBEREZHNY0FERERENo8FEREREdk8FkRERERk81gQERERkc3rZEijN99807CNdeqEJ598skOBiIiIiIxNMGSl6k6dOmH48OFtbuzYsWO4du2aLMFIOaGhobh8+TI8PT1x+fJlU8eRLCgoCGVlZfDx8UFhYaGp4xDpsG8RWS6DDpk5OzsjKyurzZuDg4PSeUkmb7/9dosf2P/5z38QGRkJZ2dn9O3bF5999pnk7XZ0Gz/99BPGjBmD7t27QxAE7Nu3r0WbwsJCrF+/XnI2ImNQom8Z0i/k2Ab7FtkygwqiQ4cOGbSxAwcOdCQLmdDRo0fx6KOPYvny5VCr1Vi2bBmmTp2K7Oxso26jc+fOePjhh7Fnz5727AaR2TGXfsG+RXRnBhVEMTEx+OCDD1BbW3vHdgMGDJAlFBnf+vXrMWbMGEyaNAkODg6YNGkSHnzwQaxbt86o2+jbty9mz56N+Pj49uwGkdkxl37BvkV0ZwafZbZkyRL4+/tj7ty5OHbsmJKZyARyc3MxePBgvccGDRqEnJwco26DyNqwXxBZBoMLIpVKhe3bt0OtVmPEiBGIjo7GG2+8gd9++03JfGQkarUanp6eeo95eXlBrVYbdRtE1ob9gsgyGFwQCYKApKQkbNu2DSUlJZg7dy62bNmCwMBATJ48GXv37lUyJynM3d0dlZWVeo9VVFTA3d3dqNsgsjbsF0SWoV0LM3p6euKpp55CdnY2srKykJ2djTFjxsidjYwoJiamxaHQ7OxsDBw40KjbILI27BdElqHdK1UfOXIEc+bM0RVC6enpcmUiE5g7dy4yMzOxc+dOaDQa7Ny5E1999RWeeOIJXZsZM2Zg5MiRim5DFEXU1dWhrq4OAKDRaFBXV4eGhoYO7yORKRirX7BvEXWMpIKotLQUr776KqKionDffffh2rVr2LFjB/Ly8vDiiy8qlZGMIDExEVu2bEFaWhrc3NyQlpaGjz76CIMGDdK1uXjx4h1/4MqxjYsXL8LZ2RnOzs4AgLFjx8LZ2RnLli3r8D4SmYKx+gX7FlHHGLRSNQCMHz8eX3/9Ne666y48/vjjmDp1Kry8vJTORwro06cPiouL4eHhYfBqtDU1Nejbty9Onz4NV1fXdr2uHNsAgB49eqC8vBx+fn44f/58u7dDJDf2LSLLZXBBNG/ePDz++OOIjY1VOhMRERGRURlcEN2soqICmZmZKCoqwt///ncUFxdDq9UiKChIiYxEREREipI8qfrIkSPo3bs31q1bh5deegkA8Msvv2DevHmyhyMiIiIyBskjRAkJCUhLS8PEiRPh5eWFiooKXL9+HT179kRJSYlSOYmIiIgUI7kgai6CAMDb2xvl5eUtviYiIiKyJJIPmYWEhOCnn37Se+z48eMICwuTLRQRERGRMXWS+g1paWkYP348FixYAI1Gg/feew+vvvoqVqxYoUQ+RWm1WhQXF8PNzQ2CIJg6DtEdiaKI6upqBAQEwM6u3WuqGgX7FlkSS+pbpJx2nWWWmZmJtWvXQqVSITg4GPPmzcO4ceOUyKeowsJCBAcHmzoGkSQFBQVmf0Yn+xZZIkvoW6ScdhVE1qKqqgqenp4oKCjghRbJpERRhKq4CqVl1+Hn0wWhAR4tRlbUajWCg4NRWVkJDw8PEyU1DPsWmRNRFCGqfgZK8wG/MAihd+n1L0vqW6Qcgw6ZnT9/Hr169Wqz3YULF9CzZ88OhzKW5g7h7u5u9T+0RVFEflEVSq7WwN/XBWGBLX/hkmmIooivDuXjrKocWhE4nX8NEaENGDM8rNX3yBLeN1vqW2TeRFGE9qsNwNlsQNsInLYHIuJhN2Z2i75kCX2LlGNQQRQbGwu1Wt1mu0GDBvFMMzN06y9cOwGICPW+7S9cMq78oirdewMAWhE4qypHZLg3woM8TZqNyNKJ+Sd/L4aApn/PZkOMTIQQ3t+04cisGFQQ1dbW4uGHH26znUaj6XAgkh9/4Zq3kqs1uvemmVZsepzvD1EHlVz4vRhqpm1sepwFEd3EoIJo8eLFBm0sLi6uQ2FIGfyFa978fV1gJ0DvPbITgABfF9OFIrIW/j0BO3v9osjOHgiwnOkdZBwGFURLlixROgcpiL9wzVtYoAciQr1bHNIMDeTkTqKOEsKiIUbE/37YzK5pDpEQGm3qaGRmJK9DRJaHv3DNmyAIGDM8DJHh3pz0TiQzQRBgN2Y2xMjEpsNk/j0hhEWzf1ELLIhsAH/hmj9BEBAe5MlDmEQKEAShaQI15wzRHbAgshH8hUtERHR7XKOciIiIbJ7kgigzMxPnz58HAKhUKjz00EOYNGkSCgsLZQ9HREREZAySC6JnnnkGTk5OAIAFCxbA1dUVXbt2RWpqquzhiIiIiIxB8hyiy5cvIygoCA0NDdi3bx8uXboER0dHBAQEKJGPiIiISHGSCyJnZ2eUlpbi5MmTiIyMhJubGzQaDVepJiIiIosluSB67LHHMGjQINTX12Pp0qUAgOzsbISHh8sejoiIiMgYJBdE//rXvzBq1Cg4ODhg5MiRAAAHBwe89tprcmcjIiIiMop2rUP0wAMP6N2Pj4+XJQwRERGRKUguiK5cuYKXX34Zx48fx7Vr1/SeO378uGzBiIiIiIxFckGUnJwMBwcHTJkyBV26dFEiExEZUUZGBmbNmoWdO3di4sSJLZ4/d+4cpk+fjrKyMnh4eOCDDz7AXXfdZfygREQKklwQHT9+HGVlZXBwcFAiDxEZkUqlwoYNG5CYmHjbNnPnzsWcOXMwY8YM/Oc//8GMGTNw7NgxI6YkIlKe5IUZExIScOHCBSWyEJERabVapKSk4K233oKjo2Orba5cuYLs7Gw8+uijAIBJkyahoKBAt1p9W9Rqtd6tvr5etvxERHKSPEL04YcfIjk5GXfffTf8/Pz0nnv66adlC0ZEylq9ejWGDRuGuLi427YpKCiAv78/OnVq+lEhCAJCQkJw6dIl9OrVq83XCA4O1ru/ZMkSpKendyg3EZESJBdEr7/+Oo4ePYq6ujq9OUSCILAgIjITQ4YMwblz51p9LicnB1VVVdixYwcOHjyoaI6CggK4u7vr7t9uJIqIyNQkF0Rr165Fbm4uIiMjlchDRDI4cuTIHZ/fs2cPVCoVevfuDaDpkjxz5sxBSUmJ3nUJg4ODUVJSgoaGBnTq1AmiKOLSpUsICQkxKIe7u7teQUREZK4kzyHy9fVFaGioAlGIyFhSU1NRUlIClUoFlUqFxMREvPfeey0u0tytWzfExsbio48+AgDs2LEDQUFBBh0uIyKyJJILosWLF2Pu3Lk4f/58iwmTRGQdUlJSsHv3bgDAu+++i3fffRcRERFYsWIFMjIyTJyOiEh+giiKopRvsLP7vYYSBAEAIIoiBEFAY2OjvOkUplar4eHhgaqqKg7rk9mzpM+rJWUl4ueVgHbMIcrPz1ciBxEREZHJSC6IevToofu6rKwMPj4+sgYiIiIiMjbJc4iuX7+OuXPnokuXLvDz80OXLl3wxBNPoKamRol8JBNRFHHytyJ8rvoJJ38rgsQjpURERFZNckH07LPP4uzZs/j2229RXFyM/fv349y5c3juueeUyEcyEEURm84cxrrTB5FZ8DPWnT6ITWcOsygiIiL6P5IPme3evRsnT56Et7c3AMDPzw+ffPIJoqOjsW7dOtkDUsedKi/Gj1cvoRFNBVAjRPx49RIG+4YiumugidMRERGZnuQRIlEU9c40A5rOPONog/nKqy7TFUPNGiEir7rMRImIiIjMi+SCaNy4cfjTn/6E7OxsXL16FceOHUNycjLGjx+vRD6SQbibD+wh6D1mDwHh7pwQT0REBLSjIFq9ejVCQkJw9913o3v37hgxYgSCgoLw2muvKZFPT0ZGBgRBwK5du1o8V1dXh4kTJyIiIgIDBgzAAw88YPAVua1dP+8AxPmG6IoiewiI8w1BP68AEycjIiIyD5LnELm6umLTpk14//33cfXqVfj6+uoWaFSSSqXChg0bkJiYeNs2c+bMwZgxYyAIAt5++22kpKTgwIEDbW771lW2HR0d0blzZ+QXVaHkag38fV0QFuhhlP1UgiAImNVnKAb7hiKvugzhbj7o5x1gsftDREQkN8kjRNHR0QCafsl269ZN90s1JiZG1mA302q1SElJwVtvvXXbq2U7OTlh7NixujyJiYlQqVQGbT84OBgeHh662yuvvIKvDuVj9/7zOHqiBLv3n8dXh/Itep6UIAiI7hqIh0IHILprIIshIiKim0geIbpdkXHp0qWOZrmt1atXY9iwYYiLizP4e9544w089NBDBrUtKCjQW669pKweX31/Edr/q3+0InBWVY7IcG+EB3lKiU5EREQWwOCC6JlnngEAaDQa3dfN8vLyEB4e3q4AQ4YMwblz51p9LicnB1VVVdixYwcOHjxo8DZffvllnD9/Ht9++61B7d3d3fUKopMXinTFUDOtCJRcrWFBREREZIUMLogqKioANB2+av4aaDrlPioqCqtXr25XgCNHjtzx+T179kClUqF3794AgMuXL2POnDkoKSlBampqi/arVq3CZ599hn379qFLly7tyuTv6wI7AXpFkZ0ABPi6tGt7REREZN4MLogyMjIAALGxsfjLX/6iWKBbpaam6hU+I0eOxPz58zFx4sQWbVevXo1t27Zh37598PT0bPdrhgV6ICLUG2dV5dCKTcVQRKg3QgM92r1NIiIiMl+S5xDdXAxVV1frTTS++bCTsaSkpGDChAmIjY3Fs88+i/DwcNx7770Ams4WO3r0qORtCoKAMcPDEBnubRVnmREREdGdSS6ILl26hJSUFHz//feor6/Xe66xsVG2YLdz62n0Gzdu1H0t51lggiAgPMiTc4bIKERRxKnyYi6LQERkIpILoieffBJeXl44fPgw7rnnHhw8eBDp6en4wx/+oEQ+IqvXfPHd5uvNNS+cOavPUBZFRERGIrkgOnLkCC5evAhXV1cIgoABAwZg48aNGD58OFJSUpTISGTVePFdIiLTk7wwo729vW5xRHd3d5SXl8PDwwMFBQWyhyOyBbz4LhGR6UkeIYqJicH+/fuRlJSkGxXq0qULoqKilMhHZPWaL757c1HEi+8SERmX5BGi999/H5GRkQCaVoPu2rUrNBoNNm/eLHs4IlvAi+8SEZme5BGi4OBg3dc+Pj7YsGGDrIGIbA0vvktEZHqSCyIikl/zxXc5iZqIyDQkHzIjIiIisjYsiIiIiMjmSSqIGhoasGDBAtTV1SmVh4iIiMjoJBVEnTp1wqZNm9C5c2el8hAREclKFEXkFVbih5wi5BVWynqZJ7IekidVJycnY+vWrZg2bZoSeYiIiGQjiiK+OpSPs6pyaEXATgAiQr0xZngYz+QkPZILooKCAmzcuBFr1qxBSEgI7Ox+H2T67LPPZA1HRETUEflFVbpiCAC0InBWVY7IcG9evJv0SC6I4uPjER8fr0QWIiIiWZVcrdEVQ820YtPjLIjoZpILoiVLliiRg4iISHb+vi6wE6BXFNkJQICvi+lCkVlq12n3WVlZmD17NsaPHw8AyM7ORlZWlqzBiIgsASfsmrewQA9EhHrD7v+mCzXPIQoN9DBtMDI7kkeINm7ciPT0dMyYMQOffPIJAMDBwQH/+Mc/cOjQIdkDEhGZK07YNX+CIGDM8DBEhnuj5GoN/H1dEBbowfeHWpA8QrRy5Urs3bsXy5Yt002ojoqKwi+//CJ7OCIic3a7Cbv5RVWmDUZ6BEFAeJAnhg0MRHiQJ4shapXkgui3335DVFQUAOg+VIIg8ANGRAaxpkNMd5qwS0SWRfIhswEDBmDHjh2YNGmS7rHdu3cjNjZW1mBEZH2s7RATJ+wSWQ/JBdGqVaswevRofPzxx7h+/Tr+/Oc/Y//+/fj666+VyEdEVsTa1oRpnrB7a4HHCbtElkdyQRQbG4tTp07ho48+gr+/P4KDg7Fq1SoEBAQokY+IrIi1rQnDCbtE1kNyQQQA3bt3x3PPPSd3FiIygYyMDMyaNQs7d+7ExIkTWzwfGhoKR0dHODs7AwDS0tKQnJzcrteyxkNMzRN2LbGgI6LfSS6IGhoa8Morr2DLli0oKipCYGAgHn30UaSlpcHBwUGJjESkEJVKhQ0bNiAxMfGO7bZv346YmBjJ21er1Xr3A3yceIiJiMyS5ILor3/9Kw4fPoxly5ahR48euHjxIlasWIHS0lK88847SmQkIgVotVqkpKTgrbfewrPPPqvIawQHB+vdX7JkCZYsWcJDTERkdiQXRJ988glOnjyJ7t27AwASEhJw9913o3///iyIiCzI6tWrMWzYMMTFxbXZ9rHHHoMoihg8eDBWrFgBX19fg16joKAA7u7uuvuOjo48xEREZklyQeTq6goXF/3j/S4uLnBzc5MtFBF1zJAhQ3Du3LlWn8vJyUFVVRV27NiBgwcPtrmtgwcPIiQkBBqNBosXL8b06dORmZlpUA53d3e9goiIyFxJLogWLVqEKVOmYOnSpQgJCcHFixfxz3/+E4sXL9abL8AfgkSmc+TIkTs+v2fPHqhUKvTu3RsAcPnyZcyZMwclJSVITU3VaxsSEgKg6RI98+fPR0REhDKhiYhMSBAlLhPbfLkOoOnsipu/vfm+IAhobGyUL6VC1Go1PDw8UFVVxQKOzJ6Sn9eRI0di/vz5Lc4yq6mpgUajgaenJ4Cmw2y7du1qc2SJfYssCT+vBLRjhCg/P1+JHERkRlJSUjBhwgT069cPkyZNQmNjI0RRRHh4OD788ENTxyMikp3kESJrwr8KyJJY0ufVkrIS8fNKQDsu7kpERERkbVgQERERkc1jQUREREQ2T3JBdOzYMSVyEBEREZmM5IIoKSkJ0dHRWL16Na5evapEJiIiIiKjklwQlZSUYOHChdi7dy969OiBP/7xj/jiiy+g1WqVyEdERESkOMkFkaOjI6ZMmYL//ve/OH/+PBISEvDss88iMDAQf//733HhwgUlchIREREppkOTqs+fP48zZ86gtLQUUVFR+O233xAXF4cVK1bIlY+IiIhIcZJXqr548SI2b96MzZs3Q6PRYMaMGcjNzUVYWBgAYOHChYiNjcULL7wge1giIiIiJUguiCIjIzF+/Hi88847SEpKgiAIes/37NkTU6ZMkS0gERERkdIkF0QFBQXw8fG5Y5v169e3OxARERGRsUkuiE6fPt3q446OjggJCYG/v3+HQxEREREZk+SC6KGHHsK1a9fQ2NgIJycn1NXVwd7eHk5OTrh+/TqGDRuGrVu3Ijg4WIm8REREkoiiiFPlxcirLkO4mw/6eQe0mO5BJPkss+XLl2P69Om4cuUKrl+/jtLSUsycORMrV65EUVER/P39MW/ePCWyEhERSSKKIjadOYx1pw8is+BnrDt9EJvOHIYoiqaORmZGECV+KoKCgnDhwgU4OjrqHqutrUXv3r1RWFiIiooK9O7dG2VlZbKHlZtarYaHhweqqqrg7u5u6jhEd2RJn1dLykrW7eRvRVh3+iAa8fuvOnsISI0ageiugQD4eaUmkkeIGhoaUFhYqPdYcXExNBoNAMDNzQ0NDQ3ypCMiMnOiKOLkb0X4XPUTTv5WxJEHM5NXXaZXDAFAI0TkVZv/H+1kXJLnEKWkpGD06NGYN28egoODUVBQgLVr12L27NkAgC+//BJ9+/aVPSgRkblpPhzz49VLaIQIewiI8w3BrD5DOUfFTIS7+cAeQosRonD3O58tTbZHckH00ksvISwsDNu2bUNxcTECAgKQlpaGWbNmAQDGjRuHP/zhD7IHNTZOwiNShjX1rVPlxbpiCGgaefjx6iUM9g3VHY4h0+rnHYA435AWRWs/rwBTRyMzI6kgamhowJQpU7BlyxY8/vjjrbaxt7eXJditMjIyMGvWLOzcuRMTJ07scLs74V99RMqwtr51p8MxLIjMgyAImNVnKAb7hlpFEU7KkTSHqFOnTvj+++/RqZPkgaUOUalU2LBhAxITE2Vpdyu1Wq13O156sdW/+k6VF7d7H8xBfX090tPTUV9fb+oo1ApbeH9uN6JiqX2r+XDMzXg4xvwIgoAIVx/kfLATEa4+LIaoVZInVc+dOxcrV65UIkurtFotUlJS8NZbb+md2dbedq0JDg6Gh4eH7rb1691WOQmvvr4eS5cutepfuJbMFt4fa5vg2nw4prko4uEY82UL/Ys6RvJQz+eff45Tp05hzZo1CAwMhJ3d7zXV8ePHZQ0HAKtXr8awYcMQFxcnS7vWFBQU6J1qee56Od4/d4ST8IhkZm0TXJsPx/Rz7YZ5S9LwztJXMDiwF0cgiCyQ5IJo/vz5sr34kCFDcO7cuVafy8nJQVVVFXbs2IGDBw/ecTunTp0yqN3tuLu76xVEsW5uyK0s4iQ8IplZ4wRXQRDQ190PP27agb5rNrEYIrJQkgui6dOny/biR44cuePze/bsgUqlQu/evQEAly9fxpw5c1BSUoLU1FRdu0OHDhnU7lbN64Wo1eoWz/3Jvx/CxC549uV0vLYwHXH+4aiurpa8j+akeT9b218yvbben+bHLWGdG/YtMjd3eo8sqW+RciSvVA00ncm1detWXLlyBSdOnMB3332H0tJSTJ48WYmMOiNHjsT8+fPbPHvM0HaFhYW85hpZnIKCAgQFBZk6xh2xb5ElsoS+RcqRPEK0fPlybN++HU8//TSee+45AIC/vz+ee+45xQuiW6WkpGDChAmYMGFCu74/ICAABQUFcHNz4zA3mT1RFFFdXY2AAPM/vMS+RZbEkvoWKUfyCFFYWBgOHTqEoKAgeHl5oaKiAlqtFj4+PigvL1cqJxEREZFiJI8Q1dTUwN/fHwB0f/lpNBrJp7oTERFZGlEU0dDQAK1Wa+ooZAB7e3uD106UvA5RYmIi1q5dq/fYpk2bMGzYMKmbIiIishg1NTVQq9W8gLkFqa+vR1VVlUFtJR8yy8vLw/333w9vb2+cPHkS8fHxKC0txb59+xAWFtauwEREROassbER169fh5ubm6mjkEQ1NTVwdHRsc6SoXWeZ1dbW4ssvv4RKpUJwcDDGjRsHFxeXdoclIiIyZ80rXHN6iOUx9L1rV0FERERkS1gQWS5D3zvJk6qrq6uxZs0a/Pjjjy0WU9u/f7/UzRERERGZnOSCaNq0aSgoKMCkSZN4mIyIiIisguRDZp6enrh06ZLetb+sTXl5OR544AGcOXMG165dM3WcDrO2/bFGtvQeWdu+Wtv+WBu53p+OHDITRRGnyouRV12GcDcf9PMOkG3B0tDQUOzatQsxMTG6x1JSUjB16lTce++9SE9PxzvvvIPAwEDd85mZmbddhFJqe0ug2CGz4OBgaDSa9qWyEG5ubvjmm2+MvvK2Uqxtf6yRLb1H1rav1rY/1sbU748oith05nCLCxrP6jNUsVXcN27cqHd/6tSpeP311w3+fqntrYXkdYieeuopJCcn45tvvsGJEyf0btbCwcEB3t7epo4hG2vbH2tkS++Rte2rte2PtTH1+3OqvFhXDAFAI0T8ePUSTpUXK/aaI0eOxK5du2Td5oEDB9CvXz+kpqaif//+iI6OxokTJzBjxgxER0cjISEBRUVFsr6msUkuiFJTU7F//34kJSUhJiZGdxs4cKAS+YiIiCxWXnWZrhhq1ggRedVlRsuwdetW3e/qmTNntrv9r7/+ipSUFJw4cQITJ07EfffdhxdeeEG3JqGljypJPmTG5cqJiIgME+7mA3sIekWRPQSEu/sYLYNch8x69eqFuLg4AEB8fDx69eqFyMhIAMDgwYOxc+dOOeKajOQRotupqKiQa1NERERWoZ93AOJ8Q2CPpvlCzXOI+nlZ3iRlJycn3df29vYt7lv6JU0MLoi6deumdz8pKUnvfo8ePeRJZCZGjRqFnJwcjBo1CqdOnTJ1nA6ztv2xRrb0Hlnbvlrb/lgbU74/giBgVp+hSI0agbHBdyE1aoSiE6qp/Qw+ZFZbW6t3/9ixY3r3rW3B63379pk6gqysbX+skS29R9a2r9a2P9bG1O+PIAiI7hqI6K6BbTduh6SkJDg4OOju3zxyQ4YzeB0id3d3qNVq3X1vb2+Ul5ff9nlLdrvK3VKLPmvbH2tkS++Rte2rte2PtZHr/eGlOyyXoe+dbHOIrIkoihBFEWvWrMH999+vu2+pjh49iqFDh2LIkCFYtGiRxe+PNTp06BASEhIwZMgQrFq1yqrfI/YvMiZb6lvUMQYfMrtx4wbefPNN3f26ujq9+9a2WKNGo0Fubq6pY8hi4MCB+OGHHwAA999/P9RqtVWvNG6JwsPDcfDgQXTu3Bn33nsvUlNT0aVLF1PHUgz7FxmLrfUtQ2RmZmLhwoUtHk9LS0NycrIJEpkHgwuixMREvVPqEhIS9O4nJibKm8zEtmzZgilTpuDVV181dZQOaz623NjYiICAAJv/YWCObl4W397eHnZ21j14y/5FxmJrfcsQY8eOxdixY00dw+wY/Mk4cOAAsrKy7nizFlqtFl9//XWLM+ks2ccff4y+ffvC09MTnTpJXn6KjOSbb75Bz549rXpSJPsXmYIt9C3qGJbKrfjss88wYcIEU8eQ1Z///Gf8+uuvKC4uxsmTJ00dh1pRWFiIV155Ba+99pqpoyiK/YuMzVb6FnUMC6JWnDlzBh988AEefPBB5ObmtrhQnqVpnmFvZ2cHNzc3/oVkhurr6zFjxgysW7cOrq6upo6jKPYvMiZb6lvUMQafdm+rRo0aZfI1LDrq008/xTvvvAOtVosRI0Zg2bJlpo5Et8jIyMCiRYsQEREBoOlaQoGByqxZYk7Yv0hpcvUtnnZvuQx971gQERERtYEFkeXiOkRERERmQBRFaPNOQPvDTmjzTsi6DlJoaGiLJSxSUlJ0Jzqlp6fD19dXd/X6mJgYFBcX33Z7N7fv27cvkpOTddcqFUUR6enpiIiIQHR0NO699947ZsvPz0dcXBxiYmLQr18/PPLII2Z93VOOEBEREbWhvSNEoihC+9UG4Gw2oG0E7OyBiHjYjZkty/XMQkNDsWvXLsTExLT6fHp6OiorKw2+2v3N7RsbG/HII48gLCwMr732Gt544w189913+Pe//43OnTvj8uXL6N69+223VV9fD61WC2dnZwDAX//6VwDAG2+8IWkfO4ojRERERCYm5p/8vRgCmv49m930uEJGjhyJXbt2dXg79vb2GDVqFM6cOQMAePXVV7FixQp07twZAO5YDAFNBUhzMdTY2IiamhpdEXjgwAH069cPqamp6N+/P6Kjo3HixAnMmDED0dHRSEhIQFFRkW5b//rXvxAdHY0BAwYgMTER169f7/D+3YoFERERkVJKLvxeDDXTNjY9biRbt27VHS6bOXOmwd9XW1uLXbt2IS4uDmq1GqWlpfj888+RkJCAhIQEbN++vc1t3LhxAzExMfDx8cG5c+ewdOlS3XO//vorUlJScOLECUycOBH33XcfXnjhBZw8eRLx8fG6Ua3Nmzdjx44d+P777/HTTz/hq6++UmQuFwsiIiIipfj3bDpMdjM7eyCgp9EiTJ06Fbm5ucjNzUVGRkab7bdu3YqBAwdi6NChiIqKwgsvvICGhgY0NDSgtrYWR48exfbt2/G3v/0NP/300x231blzZ+Tm5qK0tBSRkZF49913dc/16tULcXFxAID4+Hj06tULkZGRAIDBgwfj3LlzAIA9e/bgiSeegIeHBwDAy8sL9va3/J/KgAURERGRQoSwaCAi/vei6P/mEAmh0aYNdgdTp05FTk4OcnJy8Prrr8PZ2Rne3t5wdXXFo48+CqBp7tKwYcNw7Ngxg7bZuXNnzJw5E1u2bNE9dvOaXfb29i3uNzQ0yLRHhmFBREREpBBBEJomUE94CkLCOAgTnpJtQrWxTZkyBf/9738BAOXl5fjf//1f9O/f/7btL168qJvro9Vq8emnn96x/e1MmDAB69evR1VVFQCgsrISjY2NbXyXdLzoDhERkYIEQYAQ3h8Il14MGCIpKUl3kWEAiq2W/sorr2DmzJlYu3YtAOD555/H4MGDb9v+xIkTWLRoEYCmgig2NhZvvvmm5NedNm0aiouLMXToUHTq1AkuLi7Yt2+f7BdS5mn3REREbeDCjJaLp90TERERGYiHzIiIiGxIZmYmFi5c2OLxtLQ0JCcnS95efHx8iwnQd911F7Zu3drujG2Rex8AHjIjIiJqEw+ZWS4eMiMiIiIyEAsiIiIisnksiIiIiMjmsSAiIiIim8eCiIiISEGiKCKvsBI/5BQhr7AScp7LFBoaitzcXL3HUlJSkJWVBQBIT0+Hr6+v7uKuMTExKC4uvu32bm7ft29fJCcno6KiQra85oyn3RMRESlEFEV8dSgfZ1Xl0IqAnQBEhHpjzPAwxS7fsXHjRr37U6dO1V053hDN7RsbG/HII49g2bJleO2112ROaX44QkRERKSQ/KIqXTEEAFoROKsqR35RlWKvOXLkSOzatavD27G3t8eoUaNw5swZAE2jUYsXL8bQoUMRHByM9evXIyMjA0OGDEFoaCj+/e9/d/g1TYkFERERkUJKrtboiqFmWrHpcWPZunWr7nDZzJkzDf6+2tpa7Nq1C3FxcbrHampqcPjwYWRlZeFvf/sbioqKcOTIEXz66af4y1/+okR8o+EhMyIiIoX4+7rAToBeUWQnAAG+LkbLIPWQ2datW/Hdd98BAO655x688MILuueaV4Hu1asXnJyc8Kc//QlA02rV5eXlqKyshKenp2zZjYkFERERkULCAj0QEerdYg5RaKCHqaPd1p0KKCcnJ93X9vb2uvuCIEAQhBaX8LAkLIiIiIgUIggCxgwPQ2S4N0qu1sDf1wVhgR6KTaim9mNBREREpCBBEBAe5InwIE9Ftp+UlAQHBwfd/ZtHcchwvLgrERFRG3hxV8tl6Htn0AjRm2++adCLdurUCU8++aRBbYmIiIjMhUEjRJ06dcLw4cPb3NixY8dw7do1WYKRckJDQ3H58mV4enri8uXLpo4jWVBQEMrKyuDj44PCwkJTxyHSsfS+NXToUOTk5KCurg4ajQadOnFWRTNrGiHKzMzEwoULWzyelpamO4vMmhj83okGcHV1NaSZ6OnpaVA7Mq0ePXqIGzZs0HssNzdXfPDBB0U/Pz8RgPjNN99I3q4xt5GRkSEGBgZK3j6RklrrW6Ioip9++qnYp08f0cnJSYyMjBR37Nghabty9K3NmzeLQ4cOFb28vERvb2/xnnvuEb///vsW7bKyskQAokajkfwa1kyj0Yg1NTWmjkHtcP36dfHGjRtttjNoYcZDhw4ZVIUdOHDAoHZkfjp37oyHH34Ye/bssfhtEJmTo0eP4tFHH8Xy5cuhVquxbNkyTJ06FdnZ2QZvQ45+UV1djX/84x+4ePEiLl++jIkTJ+LBBx/kKKuB7O3tUVdXJ+t1yMg4bty4YdBop0HjoTExMfjggw+QnJwMZ2fn27YbMGCA4QnJrPTt2xd9+/a1im0QmZP169djzJgxmDRpEgBg0qRJ+Oijj7Bu3Tq8//77Bm1Djn4xb948vfvz589Heno6jh07hqCgoA5t2xYIggAPDw9UVFTAyckJ9vb2po5EBmguhgxZ5sDgS3csWbIE/v7+mDt3Lo4dO9ahgEREtiI3NxeDBw/We2zQoEHIyckxUaImR48exbVr1/iHrAT29vbw8vJC586dTR2FDNSlSxe4uBi2KrjBM+ZUKhX27t2LDz74ACNGjECvXr2QkpKCRx99FF27dm13WCIia6ZWq1tcysDLywtqtdo0gQAUFBQgOTkZL7zwAsLDw02WwxIJgoBOnTpxwrkVMniESBAEJCUlYdu2bSgpKcHcuXOxZcsWBAYGYvLkydi7d6+SOYmILJK7uzsqKyv1HquoqIC7u7tJ8pw/fx4jRozA5MmTsWzZMpNkIDJH7bravaenJ5566ilkZ2cjKysL2dnZGDNmjNzZiIgsXkxMTItpBtnZ2Rg4cKDRs5w4cQLDhw/HrFmzsHLlSqO/PpE5a1dBBABHjhzBnDlzdIVQenq6XJnIBERRRF1dHerq6gAAGo0GdXV1ehfqmzFjBkaOHGnybRBZkrlz5yIzMxM7d+6ERqPBzp078dVXX+GJJ57QtTFG3zp8+DBGjhyJ559/Hi+++GKH94vI2kgqiEpLS/Hqq68iKioK9913H65du4YdO3YgLy+PHczCXbx4Ec7OzrqzCMeOHQtnZ2e9IfWLFy/e8QeusbZBZEkSExOxZcsWpKWlwc3NDWlpafjoo48waNAgXRtj9K1FixahsrISixcvhqurq+728ssvd3gfiayBwdcyGz9+PL7++mvcddddePzxxzF16lR4eXkpnY8U0KdPHxQXF8PDw8PgNUhqamrQt29fnD59Gq6uru16XTm2AQA9evRAeXk5/Pz8cP78+XZvh0hult63hg8fjp9++gk3btxATU0NTy0nm2JwQTRv3jw8/vjjiI2NVToTERERkVG162r3FRUVyMzMRFFREf7+97+juLgYWq2Wi3sRERGRRZI8qfrIkSPo3bs31q1bh5deegkA8Msvv7RYBZWIiIjIUkgeIUpISEBaWhomTpwILy8vVFRU4Pr16+jZsydKSkqUyklERESkGMkFUXMRBADe3t4oLy9v8TURERGRJZG89nhISAh++uknvevfHD9+HGFhYbIGMwatVovi4mK4ubkZdOE3IlMSRRHV1dUICAiAnV27lxAzCvYtsiSW1LdIOZILorS0NIwfPx4LFiyARqPBe++9h1dffRUrVqxQIp+iiouLERwcbOoYRJIUFBSY/QkM7FtkiSyhb5Fy2nWWWWZmJtauXQuVSoXg4GDMmzcP48aNUyKfoqqqquDp6YmCggKTXVeIyFBqtRrBwcGorKyEh4eHqePcEfsWWRJL6luknHZdrnfs2LEYO3as3FmMrnko393dnT+0yaREUcSp8mLkVZch3M0H/bwDbnuoyRIOQbFvyU8UReQXVaHkag38fV0QFuhhEZ8FS8L/T9tmUEF0/vx59OrVq812Fy5cQM+ePTscisiWiKKITWcO48erl9AIEfYQEOcbgll9hvIHNAFo+ox8dSgfZ1Xl0IqAnQBEhHpjzPAwfkaIZGLQ7DFDV6e++do8RGSYU+XFumIIABoh4serl3CqvNjEychc5BdV6YohANCKwFlVOfKLqkwbjMiKGDRCVFtbi4cffrjNdhqNpsOBiGxNXnWZrhhq1ggRedVliO4aaKJUZE5KrtboiqFmWrHp8fAgT5NkIrI2BhVEixcvNmhjcXFxHQpDZIvC3XxgD0GvKLKHgHB3HxOmInPi7+sCOwF6RZGdAAT4upguFJGVMaggWrJkidI5iGxWP+8AxPmGtJhD1M8rwNTRyEyEBXogItS7xRyi0ECeEUUkl3adZUZE8hEEAbP6DMVg31CDzjIj2yMIAsYMD0NkuDfPMiNSCAsiIjMgCAKiuwZyzhDdliAICA/y5JwhIoVwjXIiIiKyeSyIiIiIyOZJLogyMzNx/vx5AIBKpcJDDz2ESZMmobCwUPZwRERERMYguSB65pln4OTkBABYsGABXF1d0bVrV6SmpsoejoiIiMgYJE+qvnz5MoKCgtDQ0IB9+/bh0qVLcHR0REAATxEmIiIiyyS5IHJ2dkZpaSlOnjyJyMhIuLm5QaPRcJVqIiIisliSC6LHHnsMgwYNQn19PZYuXQoAyM7ORnh4uOzhiIiIiIxBckH0r3/9C6NGjYKDgwNGjhwJAHBwcMBrr70mdzYiIiIio2jXwowPPPCA3v34+HhZwhARERGZguSC6MqVK3j55Zdx/PhxXLt2Te+548ePyxaMiIiIyFgkF0TJyclwcHDAlClT0KVLFyUyERERERmV5ILo+PHjKCsrg4ODgxJ5iIiIiIxO8sKMCQkJuHDhghJZiIiIiExC8gjRhx9+iOTkZNx9993w8/PTe+7pp5+WLRgRERGRsUguiF5//XUcPXoUdXV1enOIBEFQvCDKyMjArFmzsHPnTkycOFHvubq6OvzP//wPTp8+DWdnZ3Tr1g3r1q1Dr169FM1ERERElk9yQbR27Vrk5uYiMjJSiTy3pVKpsGHDBiQmJt62zZw5czBmzBgIgoC3334bKSkpOHDgQJvbVqvVevcdHR3h6OjY0chERERkISTPIfL19UVoaKgCUW5Pq9UiJSUFb7311m0LFScnJ4wdOxaCIAAAEhMToVKpDNp+cHAwPDw8dLdXXnlFruhERERkASSPEC1evBhz587Fiy++iG7duuk95+7uLluwm61evRrDhg1DXFycwd/zxhtv4KGHHjKobUFBgV52jg4RERHZFskF0eOPPw4A2LJli240RhRFCIKAxsZGyQGGDBmCc+fOtfpcTk4OqqqqsGPHDhw8eNDgbb788ss4f/48vv32W4Pau7u7K1bMERERkfmTXBDl5+fLGuDIkSN3fH7Pnj1QqVTo3bs3AODy5cuYM2cOSkpKkJqa2qL9qlWr8Nlnn2Hfvn1cOJKIiIgMInkOUY8ePXQ3FxcXvftKSE1NRUlJCVQqFVQqFRITE/Hee++1WgytXr0a27ZtwzfffANPT09F8hBZm4yMDAiCgF27dsnSjojIEkkuiK5fv465c+eiS5cu8PPzQ5cuXfDEE0+gpqZGiXxtSklJwe7du1FYWIhnn30WlZWVuPfeexETE4OEhASTZCKyFIacvSml3a3UarXerb6+viNxiYgUI7kgevbZZ3H27Fl8++23KC4uxv79+3Hu3Dk899xzSuRr4cCBA3prEG3cuBETJkxAUFAQRFHEhQsXkJubi9zcXBw9etQomYgskSFnb0pp1xqewUlElkLyHKLdu3fj5MmT8Pb2BgD4+fnhk08+QXR0NNatWyd7QCJShqFnb7bnLM9mPIOTiCyF5IJIFEXY2ekPLNnZ2UEURdlCEVHHyHX25qlTpySf5XkznsFJRJZCckE0btw4/OlPf8KKFSvQo0cPqFQqLFq0COPHj1ciHxG1g1xnbx46dEjSWZ5ERJZKECUO7Vy7dg1PP/00Pv74Y2g0GnTu3BlTpkzBG2+8ATc3N6VyKkKtVsPDwwNVVVX8K5bMnpKf15EjR2L+/PktrhHY3nbsW2RJ+HkloB2Tql1dXbFp0ybU1taipKQE169fx6ZNmyyuGCKi22s+e5OIyFZIHiGKjo7GyZMnWzweExOD3NxcuXIZBf8qIEtiSZ9XS8pKxM8rAe0YIbrdBVMvXbrU0SxEREREJmHwpOpnnnkGAKDRaHRfN8vLy0N4eLi8yYiIiIiMxOCCqKKiAkDTIm3NXwNNp9xHRUVh9erV8qcjIiIiMgKDC6KMjAwAQGxsLP7yl78oFoiIiIjI2CSvQ3RzMVRdXa23ICMnoxEREZElkjyp+tKlSxg9ejS6dOkCT09PeHl56W5ERERElkhyQfTkk0/Cy8sLhw8fhqurK44fP44JEybg3XffVSIfERERkeIkHzI7cuQILl68CFdXVwiCgAEDBmDjxo0YPnw4UlJSlMhIREREpCjJI0T29va6K1a7u7ujvLwcHh4eKCgokD0cERERkTFIHiGKiYnB/v37kZSUpBsV6tKlC6KiopTIR0RERKQ4ySNE77//PiIjIwEAb7zxBrp27QqNRoPNmzfLHs6URFFEXmElfsgpQl5hJSRe4YSIiIgsiOQRouDgYN3XPj4+2LBhg6yBzIEoivjqUD7OqsqhFQE7AYgI9caY4WEQBMHU8YiIiEhmkkeIbEF+UZWuGAIArQicVZUjv6jKtMGIiIhIESyIWlFytUZXDDXTik2PExERkfVhQdQKf18X2N1yZMxOAAJ8XUwTiIiIiBQlqSBqaGjAggULUFdXp1QesxAW6IGIUG9dUdQ8hyg00MO0wYiIiEgRgijx9KmuXbvi6tWrsLOz/MEltVoNDw8PVFVVtbgOmyiKyC+qQsnVGvj7uiAs0IMTqsmk7vR5NTeWlJWIn1cC2nHILDk5GVu3blUii1kRBAHhQZ4YNjAQ4UGeLIaIiIismOTT7gsKCrBx40asWbMGISEheiNFn332mazhiIiIiIxBckEUHx+P+Ph4JbIQERERmYTkgmjJkiVK5CAiIiIymXbNjM7KysLs2bMxfvx4AEB2djaysrJkDUZERERkLJILoo0bN2LatGnw8/PDwYMHAQAODg74xz/+IXs4IiIiImOQXBCtXLkSe/fuxbJly3QTqqOiovDLL7/IHo6IiIjIGCQXRL/99huioqIAQHcquiAIPC2diIiILJbkgmjAgAHYsWOH3mO7d+9GbGysbKGIiIiIjEnyWWarVq3C6NGj8fHHH+P69ev485//jP379+Prr79WIh/JhCtvExER3Z7kgig2NhanTp3CRx99BH9/fwQHB2PVqlUICAhQIh/JQBRFfHUoH2dV5dCKv1+bbczwMBZFREREaEdBBADdu3fHc889J3cWUkh+UZWuGAIArQicVZUjMtwb4UGeJs1GRERkDiTPIWpoaMBLL72EiIgIuLi4ICIiAv/85z+h0WiUyEcyKLlaoyuGmmnFpseJiIioHSNEf/3rX3H48GEsW7YMPXr0wMWLF7FixQqUlpbinXfeUSIjdZC/rwvsBOgVRXYCEODrYrpQREREZkRyQfTJJ5/g5MmT6N69OwAgISEBd999N/r378+CyEyFBXogItS7xRyi0EAPU0cjIiIyC5ILIldXV7i46I8suLi4wM3NTbZQJC9BEDBmeBgiw715lhkREVErJM8hWrRoEaZMmYIff/wRV69eRXZ2NqZNm4bFixdDrVbrbmReBEFAeJAnhg0MRHiQJ4shIiKimwiiKIptN/td8+U6gKZfsjd/e/N9QRDQ2NgoX0qFqNVqeHh4oKqqCu7u7qaOQ3RHlvR5taSsRPy8EtCOEaL8/HzdLS8vr9X7eXl5sgfNyMiAIAjYtWuXLO2IiIiImkmeQ9SjRw8lctyRSqXChg0bkJiYKEu7W916iM/R0RGOjo6ScxIREZFlkjxCZGxarRYpKSl466237likGNquNcHBwfDw8NDdXnnllY7GJiIiIgti9gXR6tWrMWzYMMTFxcnSrjUFBQWoqqrS3dLS0tobl8jitHWYOTQ0FH369EFMTAxiYmKwfft24wYkIjKCdl26Qy5DhgzBuXPnWn0uJycHVVVV2LFjBw4ePHjH7Zw6dcqgdrfj7u7OiXRkkww9zLx9+3bExMRI3j4PRxORpZA8QnTs2DHZXvzIkSMoKytr9RYcHIxDhw5BpVKhd+/eCA0Nxf/7f/8Pc+bMwbp16/S2Y2g7IvpdRw4zG4qHo4nIUkg+7d7b2xuBgYGYOXMmpk2bBl9fX6WytTBy5EjMnz8fEydOlKUdT7UkSyL353XVqlWorq7G0qVL79hnQkND4e7uDlEUMXjwYKxYsaLNft+ctaCgQC8rR4jIHPF3AQHtGCEqKSnBwoULsXfvXvTo0QN//OMf8cUXX0Cr1SqR745SUlKwe/duo78ukbkbMmQIfHx8Wr0VFBToDjMvXry4zW0dPHgQJ06cwPHjx+Hj44Pp06cbnKP5cHTzjcUQEZkrySNENysuLsaHH36ITZs2obq6GtOmTcPcuXPRs2dPOTMqhn8VkCWR8/O6bt06/POf/9QVKJcvX4a7uzuWLl2K1NTU235fSUkJIiIiUF1dbbSsRErj55WADp5ldv78eZw5cwalpaWIiorCb7/9hri4OKxYsUKufESkgNTUVJSUlEClUkGlUiExMRHvvfdei2KopqYGlZWVuvvbtm3DwIEDjZyWiEh5ks8yu3jxIjZv3ozNmzdDo9FgxowZyM3NRVhYGABg4cKFiI2NxQsvvCB7WCIyjpSUFEyYMAH9+vXDpEmT0NjYCFEUER4ejg8//NDU8YiIZCe5IIqMjMT48ePxzjvvICkpqcVFQnv27IkpU6bIFpDIFoiiCDH/JFByAfDvCSEs2qgX4D1w4IDe/Y0bN+q+zsnJMVoOIiJTkVwQFRQUwMfH545t1q9f3+5ARLZGFEVov9oAnM0GtI2AnT3EiHjYjZlt1KKIiMiWSS6ITp8+3erjjo6OCAkJgb+/f4dDEdkSMf/k78UQ0PTv2WyIkYkQwvubNhwRkY2QXBA99NBDuHbtGhobG+Hk5IS6ujrY29vDyckJ169fx7Bhw7B161YEBwcrkZfI+pRc+L0YaqZtbHqcBRERkVFIPsts+fLlmD59Oq5cuYLr16+jtLQUM2fOxMqVK1FUVAR/f3/MmzdPiaxE1sm/J2Bnr/+YnT0QYBnLVxARWQPJ6xAFBQXhwoULegus1dbWonfv3igsLERFRQV69+6NsrIy2cPKjWtPkDlobQ4RWplDZEmfV0vKSsTPKwHtOGTW0NCAwsJCvcUXi4uLodFoAABubm5oaGiQLyGRlRMEAXZjZkOMTDTZWWZERLZOckGUkpKC0aNHY968eQgODkZBQQHWrl2L2bNnAwC+/PJL9O3bV/agRNZO1TkEJV26wr+zC8JMHYbMjiiKOFVejLzqMoS7+aCfdwCLZiIZSS6IXnrpJYSFhWHbtm0oLi5GQEAA0tLSMGvWLADAuHHj8Ic//EH2oETWShRFfHUoH2dV5dCKgJ0ARIR6Y8zwMP7CIwBNn5FNZw7jx6uX0AgR9hAQ5xuCWX2G8jNCJBNJBVFDQwOmTJmCLVu24PHHH2+1jb29fauPE1Hr8ouqdMUQAGhF4KyqHJHh3ggP8jRpNjIPp8qLdcUQADRCxI9XL2GwbyiiuwaaOB2RdZB0llmnTp3w/fffo1MnyQNLFqm+vh7p6emor683dRRZWNv+WIuSqzW6YqiZVmx6nAgA8qrLdMVQs0aIyKs2/5NXiCyF5NPu586di5UrVyqRxezU19dj6dKlVlNAWNv+WAt/XxfY3XLUw04AAnxdTBOIzE64mw/sof8hsYeAcPc7XzWAiAwneajn888/x6lTp7BmzRoEBgbCzu73mur48eOyhiOyBWGBHogI9W4xhyg00MPU0chM9PMOQJxvSIs5RP28AkwdjchqSC6I5s+fr0AMItslCALGDA9DSHcnPPfCUqxasQR39fbnZFnSEQQBs/oMRT/Xbpi3JA3vLH0FgwN78TNCJCPJBdH06dOVyGESzWtSqtXqVp9vfvx2z1saa9sfa+PZRYuvd76HjW++iOrq6hbPN79vEtdSNYm2+ha1TyCc8eOmHQhc+nqrnxFqH0vqW6QcyStVA0BGRga2bt2KK1eu4MSJE/juu+9QWlqKyZMnK5FRMYWFhbzmGlmcgoICBAUFmTrGHbFvkSWyhL5FypFcEC1fvhzbt2/H008/jeeeew6VlZU4e/Yspk6dimPHjimVUxFarRbFxcVwc3Pj0DOZPVEUUV1djYCAAL25e+aIfYssiSX1LVKO5IIoLCwMhw4dQlBQELy8vFBRUQGtVgsfHx+Ul5crlZOIiIhIMZJL4ZqaGvj7+wOA7i8/jUajd7FXIiIiIksiuSBKTEzE2rVr9R7btGkThg0bJlsoIiIiImOSfMgsLy8P999/P7y9vXHy5EnEx8ejtLQU+/btQ1gYL0lJRERElqddZ5nV1tbiyy+/hEqlQnBwMMaNGwcXF66qS0RERJapXQURERERkTWRvDBjdXU11qxZgx9//LHFwmD79++XLRgRERGRsUguiKZNm4aCggJMmjTJag+TlZeX44EHHsCZM2dw7do1U8fpMGvbH2vE94jaws8IkbIkHzLz9PTEpUuX4O7urlQmk9NoNKiursbkyZOxb98+U8fpMGvbH2vE94jaws8IkbIkn3YfHBwMjUajRBaz4eDgAG9vb1PHkI217Y814ntEbeFnhEhZkg+ZPfXUU0hOTsbzzz8PPz8/vef69+8vWzAiIiIiY5FcEKWmpgJoOYFaEAQ0NjbKk4qIiIjIiCQXRFqtVokcRERERCYj22V9Kyoq5NqUWRg1ahRycnIwatQonDp1ytRxOsza9sca8T2itvAzQqQcg88y69atG65cuaK7n5SUhK+//lp3393dHWq1Wv6ERERERAoz+JBZbW2t3v1jx47p3bemBa8FQWj1cUvdR2vbH2vE94jaws8IkbIMPmR2u85o6POWRBRFiKKINWvW4P7779fdt1RHjx7F0KFDMWTIECxatMji98caHTp0CAkJCRgyZAhWrVrF94haYD8mUpbkSdW2QqPRIDc319QxZDFw4ED88MMPAID7778farXaqhfWtETh4eE4ePAgOnfujHvvvRepqano0qWLqWORGWE/JlKWwQXRjRs38Oabb+ru19XV6d23tsUat2zZgilTpuDVV181dZQOc3BwAAA0NjYiICCAv2jNUEBAgO5re3t72NnJdr4DWQn2YyJlGfxTNzExETt37tTdEhIS9O4nJiYqmdOotFotvv76ayQlJZk6imw+/vhj9O3bF56enujUiQOD5uqbb75Bz5494eTkZOooZIbYj4mUI/laZrbgP//5D+rr6zF16lSMGjXKaq4bpNVq8cgjjyA9PR3R0dGmjkO3KCwsxGOPPYbdu3fD1dXV1HHITLEfEymD4/KtOHPmDD744AM8+OCDyM3NxcaNG00dqUPq6+sBAHZ2dnBzc+Pogxmqr6/HjBkzsG7dOhZD1Cr2YyJlcYSoDdYwQvTpp5/inXfegVarxYgRI7Bs2TJTR6JbZGRkYNGiRYiIiAAAbN26FYGBgSZOReaE/ZhIWSyIiIiIyObxkBkRERHZPBZEREREZPNYEBEREZHNY0FERERENo8FEREREdk8FkRERERk81gQERERkc1jQUREREQ2jwURERER2TwWRERERGTzWBARERGRzWNBZCP+/e9/Y/LkyYpsOzQ0FLt27ZK97c1Gjx7d5kV2Q0ND4ezsjPvuu++2bQYPHgwnJyfExMRIzkDUGvatJuxbZOlYEFmgt99+G/Hx8XB0dMTEiRNbPF9fXw9PT0+UlZUBALRaLRYuXIgXX3yx1e3NmjULW7ZsAQD07dsX+fn5imVvr0WLFmHBggVtttu2bRv2799/2+f/93//F+vXr5czGlkR9q3bY98ia8eCyAIFBARg8eLFmD17dqvPZ2VlITo6Gj4+PgCAzMxMeHt7Izo6utX2x44dw6BBg1BVVYXffvsNYWFhimVvrxEjRqCyshI//PCDqaOQFWPfIrJdLIgs0MMPP4yJEyfqfijf6osvvsCECRN093fv3n3boe6amhoUFRWhT58+OHbsGOLi4lptt3r1aoSEhMDNzQ2hoaHYuHHjbdv17t0bbm5u6NmzJ95+++0WbX7++WfExsbC3d0dSUlJKC4ubvN1BEHAfffdh927d7f+n0IkA/YtItvFgsgK3fpDOzc3F5GRkXptPvvsM3h6esLPzw/V1dXw8vLC+PHjceDAAXh6emL58uW6tmfPnsXixYuxd+9eVFdX4+jRoxg8eHCrr92jRw/s378farUaGzduxIIFC1r85blx40Z8/PHHuHz5Mrp3745HH33UoNeJiopCbm5uR/97iNqNfYvIerEgsjK5ublwcnJCnz59dI9VVFTA3d1dr93DDz+MyspKPPPMM0hPT0dlZSWGDh2K//73v6isrMSiRYt0be3t7SGKIn7++WfU1tbCz88P/fv3b/X1J02ahODgYAiCgHvvvRdJSUk4cOCAXpvU1FRERkaiS5cuWLlyJbKyslBYWNjm67i7u6OiokKG/yUi6di3iKwbCyIrc+tfsADg5eUFtVrdavsDBw7gnnvugUajQU5ODhISElq06dmzJzZv3oy3334bfn5+GD169G3/mty6dStiY2Ph7e0NT09PZGZm6iagNuvRo4fuaz8/Pzg6OqKoqKjN11Gr1fDy8jLwf6Ipi6urK1xdXXHXXXcZ/H1ErWHf0s/CvkXWhgWRldm9e3eLH9oxMTH49ddfdfebz5Tx9PTEDz/8gHHjxsHb2xtqtRrdu3fH8OHDW2x38uTJyMrKQmlpKQYMGIBp06a1aHPp0iVMnz4dK1euxJUrV1BZWYmxY8dCFEW9dhcvXtR9feXKFdTX1yMwMLDN1zl9+rSkU3qnTp2Ka9eu4dq1a/j5558N/j6i1rBv/Y59i6wRCyIL1NDQgLq6OjQ0NECr1aKurg43btxASUkJ8vPzMWzYML3248ePR1ZWlu6+o6MjKisrsX37dowbNw6VlZX429/+hmXLlqGyshKHDh3S+/4zZ87gm2++QW1tLTp37gxXV1d06tSpRa5r165BFEV069YNdnZ2yMzMxN69e1u0e/fdd3HmzBnU1tbi+eefx4gRIxAUFNTm62RlZWHcuHEd/e8jui32LSLbxYLIAi1btgzOzs5Yvnw5vvjiCzg7O2P06NHYs2cPxowZA3t7e732Y8eORVlZGU6dOqX3+M6dO3V/8X7++eetrrsCADdu3MCLL74IPz8/dO3aFfv378cHH3zQol1UVBQWLVqE++67D127dsX27dtb/EUNNK3NMmXKFPj5+aGoqAhbt25t83UOHToEd3f3Vv/CJpIL+xaRDRPJaowbN0785JNPWn3u448/FidPnmzkRPIZPXq0uHfv3ju2iYiIEN3c3MRRo0bdts2QIUNEV1dXcfDgwXJHJCvGvsW+RdZPEMVbDkKTxVq5ciVSU1Ph5uZm6ihEVoV9i8j6sSAiIiIim8c5RERERGTzWBARERGRzWNBRERERDaPBRERERHZPBZEREREZPNYEBEREZHNY0FERERENo8FEREREdk8FkRERERk8/4/3dwnb+8KT1MAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 500x350 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 3.5], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    ax = {}\n",
    "    slabs_millers = results_df[\"millers_str\"].unique()\n",
    "    Set2 = mp.cm.get_cmap(\"Set2\", 8)\n",
    "    spec = fig.add_gridspec(\n",
    "        len(slabs_millers) // 2 + len(slabs_millers) % 2, min([len(slabs_millers), 2]), hspace=0.7, wspace=0.25\n",
    "    )\n",
    "    for i, millers in enumerate(slabs_millers):\n",
    "        ax[i] = fig.add_subplot(spec[i])\n",
    "        sub_df = results_df.loc[results_df[\"millers_str\"] == millers]\n",
    "        systems_id = results_df[\"system_id\"].unique()\n",
    "        xticks = 1 / sub_df[\"slabs\"].unique()\n",
    "        xticks_labels = [f\"$\\\\frac{{1}}{{{ns}}}$\" for ns in sub_df[\"slabs\"].unique()]\n",
    "        ax[i].set_xticks(xticks)\n",
    "        ax[i].set_xticklabels(xticks_labels, rotation=0, ha=\"center\")\n",
    "        ax[i].set_title(millers)\n",
    "        for sys_id in systems_id:\n",
    "            sub_df\n",
    "            idx = np.arange(len(results_df.index))\n",
    "            sub_sub_df = sub_df.loc[sub_df[\"system_id\"] == sys_id]\n",
    "            if not len(sub_sub_df.index) > 0:\n",
    "                # print(sub_sub_df.head())\n",
    "                continue\n",
    "            name = sub_sub_df.index[0].split(\"-\")[2]\n",
    "            # print(1/sub_sub_df['slabs'], sub_sub_df['scm_PO_energy_per_atom'])\n",
    "            ax[i].scatter(\n",
    "                1 / sub_sub_df[\"slabs\"],\n",
    "                sub_sub_df[\"scm_energy_per_atom\"],\n",
    "                color=Set2.colors[int(sys_id)],\n",
    "                label=name,\n",
    "            )\n",
    "\n",
    "            # ax[i].set_xticks(idx)\n",
    "            # ax[i].set_xticklabels(results_df.index, rotation=25, ha='right')\n",
    "        # ax[i].set_ylim([-5.2, -3.8])\n",
    "    #     ax[i].set_ylim([-5, -3.8])\n",
    "    ax[3 - 2].legend(loc=\"center left\", bbox_to_anchor=(1.02, -0.3))\n",
    "    for i in range(0, len(ax), 2):\n",
    "        ax[i].set_ylabel(\"Energy per atoms [eV]\")\n",
    "    for i in range(len(ax) - 2, len(ax), 1):\n",
    "        ax[i].set_xlabel(r\"$1 /(\\mathrm{\\#\\, slabs})$ [-]\")\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"6-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "27f388c7-cc50-43de-b102-7d07585d599a",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "624904bc-752c-4651-9042-827a7a4234d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f9f04ad6-9cb7-4752-a471-0aa57c05f62e",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Amorphous `ab initio MD` <a id='amorphous_aimd'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3be55930-3f8e-422a-9c74-3b2dadf40a66",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"7-Amorphous_MD\"\n",
    "TASK = \"MD\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a3a86d56-bbf8-47f6-9e79-8cc6769f2658",
   "metadata": {},
   "source": [
    "#### Setting `DFTB` *ab initio MD*"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c0309503",
   "metadata": {},
   "source": [
    "MD settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2f005185-510f-4aab-92c9-a2380570b291",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      ams: \t\n",
      "          task: \tMolecularDynamics\n",
      "          MolecularDynamics: \t\n",
      "                            NSteps: \t1000\n",
      "                            TimeStep: \t0.25\n",
      "                            Trajectory: \t\n",
      "                                       PrintFreq: \t5\n",
      "                                       SamplingFreq: \t50\n",
      "                            InitialVelocities: \t\n",
      "                                              Temperature: \t2500\n",
      "                            Thermostat: \t\n",
      "                                       Type: \tBerendsen\n",
      "                                       Temperature: \t2500\n",
      "                                       Tau: \t25\n",
      "      DFTB: \t\n",
      "           Model: \tNonSCC-GFN1-xTB\n",
      "\n"
     ]
    }
   ],
   "source": [
    "STEPS = 1000\n",
    "TIMESTEP = 0.25  # [fs]\n",
    "TEMP = 2500  # [K]\n",
    "\n",
    "from tools.misc.default_settings import settings_MD\n",
    "\n",
    "settings_MD.input.ams.MolecularDynamics.NSteps = STEPS\n",
    "settings_MD.input.ams.MolecularDynamics.TimeStep = TIMESTEP\n",
    "settings_MD.input.ams.MolecularDynamics.InitialVelocities.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Tau = int(100 * TIMESTEP)\n",
    "\n",
    "print(settings_MD)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a8b5625d-82aa-4ef6-9035-624715ce33a4",
   "metadata": {
    "tags": []
   },
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bbc2e006-675f-4f8e-b56c-c51d46a5e516",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 1\n",
    "CORES_PER_NODE = 128\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2ed95b37-5d84-4f57-8303-23bfb5dc4102",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a607dec7-584e-4c20-9b8a-b6f5e3a9d2ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_MD, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d86a2cf0-6a17-4468-aaf8-7e1c89a97c4f",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dfd507a9-1e51-40e5-9c66-144334eb4f28",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "5a1736f4-e6c7-4861-83e7-e05117692e2c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7-amorphous\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dc318467-a5f5-4ed3-aba5-d9aecf00b39e",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "668faa77-49fa-487f-a811-159fc86071d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_MD, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6febd41a-984f-4028-84fc-9920cf28c07f",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c55f2b61-2c2b-4dc8-846f-542c92164a20",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[12:36:49] JOB jobs STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-0-LiF_Fm-3m_-3.18_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-1-LiF_P6_3mc_-3.17_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-2-LiF_Pm-3m_-2.89_2x2x3 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bfdc5e64-347f-4378-8b0f-f0058403e399",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "72d5228b-f43d-4d27-935f-3f9767a631ec",
   "metadata": {},
   "source": [
    "Post Processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ded8d51f-f22b-4b21-84cc-68568276ba76",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "38303c61-f11f-4930-a072-5909f0cee2d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7-amorphous\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b79f3694-27b0-41ab-9742-5d9d0088aa3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif                    27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif                   27\n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif                    27"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0328c68b-f5b6-4649-a7ef-faef458a1743",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_df = pd.DataFrame(index=results_df.index, columns=[\"scm_step\", \"scm_Temp\", \"scm_energy_per_atom_t\"])\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    Nentry = job_result.read_rkf_section(\"History\")[\"nEntries\"]\n",
    "    energy = scm.Units.convert(\n",
    "        job_result.read_rkf_section(\"History\")[f\"Energy({Nentry})\"], \"au\", \"eV\"\n",
    "    )  # scm.Units.convert(job_result.get_energy(), 'au', 'eV')\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_step\"] = np.array(job_result.get_history_property(\"Step\"))\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_Temp\"] = np.array(job_result.read_rkf_section(\"MDHistory\")[\"Temperature(1)\"])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_energy_per_atom_t\"] = (\n",
    "        np.array(job_result.get_history_property(\"Energy\")) / Na\n",
    "    )\n",
    "\n",
    "results_df = pd.concat([results_df, temp_df], axis=1)\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "0b281bdd-8f82-4b33-ac92-938ab82b6672",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "      <th>scm_step</th>\n",
       "      <th>scm_Temp</th>\n",
       "      <th>scm_energy_per_atom_t</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "      <td>-4163.214306</td>\n",
       "      <td>-77.096561</td>\n",
       "      <td>0.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[2600.0, 2227.472837617847, 1746.6540792818357...</td>\n",
       "      <td>[-2.844426643130874, -2.842366090997381, -2.83...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "      <td>-8318.727662</td>\n",
       "      <td>-77.025256</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[2600.0, 2201.8829583416705, 2011.420067554859...</td>\n",
       "      <td>[-2.8400756492875705, -2.837850360580547, -2.8...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "      <td>-4162.027782</td>\n",
       "      <td>-77.074589</td>\n",
       "      <td>2.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[2599.9999999999995, 2307.5056676702957, 1673....</td>\n",
       "      <td>[-2.8410294274160868, -2.8393977166355455, -2....</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition   scm_energy  \\\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif                    27 -4163.214306   \n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif                   27 -8318.727662   \n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif                    27 -4162.027782   \n",
       "\n",
       "                                  scm_energy_per_atom  system_id  \\\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif            -77.096561        0.0   \n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif           -77.025256        1.0   \n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif            -77.074589        2.0   \n",
       "\n",
       "                                                                           scm_step  \\\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif  [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "\n",
       "                                                                           scm_Temp  \\\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif   [2600.0, 2227.472837617847, 1746.6540792818357...   \n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif  [2600.0, 2201.8829583416705, 2011.420067554859...   \n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif   [2599.9999999999995, 2307.5056676702957, 1673....   \n",
       "\n",
       "                                                              scm_energy_per_atom_t  \n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif   [-2.844426643130874, -2.842366090997381, -2.83...  \n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif  [-2.8400756492875705, -2.837850360580547, -2.8...  \n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif   [-2.8410294274160868, -2.8393977166355455, -2....  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "6cea222f-da3f-4dc5-a3b0-8249754dc534",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAADTCAYAAADJaMQcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACqx0lEQVR4nOydd3gb9f3HX6ctWbblvXf2srMHJCEQwiZsCJuWkhZaoKUtpeXX0pZS2lIolJZCwwqrbAgQIAlJyJ6Os+O995BsWXvc74+TZDt2EjuxcQL3eh4/caTT6SvJunvfZ7w/giiKIjIyMjIyMjIyMt8oiuFegIyMjIyMjIzMdxFZhMnIyMjIyMjIDAOyCJORkZGRkZGRGQZkESYjIyMjIyMjMwzIIkxGRkZGRkZGZhiQRZiMjIyMjIyMzDAgizAZGRkZGRkZmWFAFmEyMjIyMjIyMsOAargXMNz4/X7q6uoIDw9HEIThXo6MzHcaURSxWq0kJyejUMjXiANBPpbJyJw+9PdY9p0XYXV1daSlpQ33MmRkZLpRXV1NamrqcC/jjEI+lsnInH6c6Fj2nRdh4eHhgPRGRUREDPNqZGS+23R0dJCWlhb6Xsr0H/lYJiNz+tDfY9l3XoQFw/YRERHygUtG5jRBTqcNHPlYJiNz+nGiY5lcdCEjIyMjIyMjMwzIIkxGRkZGRkZGZhiQRZiMTB+IPh+WvfvwezzDvZRTorK+g20H6od7GTIyMqcx7QcO0nGkcLiX8Z1EFmEyMn1Q/9nnHPzt72n6at1wL+WU+Oc7Bfzp5R20WBzDvRQZGZnTEHtVFQf+7xEOPfJH/G73cC/nO4cswmRk+qB1+w4AnE1Nw7ySU6OlXRJfjW32YV6JjIzM6YYoipT99yXw+/E5HHI0bBj4zndHysgcjdfuwHr4CAA+x5kdQbLapXSq2eoc5pUMDc8880y/tlOpVNx9991DvBoZmTOLtm07aN+3P/R/y54CTJMmDuOKvnvIIkxG5ija9x9A9PkA8DvPXPHi8vhwe6TX0dZ+5r6O4/Gzn/2MuXPnnnC7nTt3yiJMZtjpdNn42+bnOTt9OuePOPHf7VDic7kof+kVAJIuu4T6Tz7DUrAXbrtlWNf1XUMWYTIyR2EpKAj9PpiRsM7SMqxFRSQsPA+FWj1o+z3m89m76jvaOr6dIkyv17Nu3Ynr9qKiogbtOZ1OJzfccAOHDh1Cr9cTHx/Pc889x4gRI3ptu3LlSh5++GH8fj9er5df/OIX3HbbbQAsWrSIhoYGFAoF4eHhPPPMM0yePBmA4uJibrvtNlpaWoiMjOSVV15h/Pjxg/YaZIaHTVU7OdxcTFNnCwtzzh5WP7y6jz/B1dRE+NgxZN1xG83rvsZWVo7b0o7GFDls6/quIdeEycgchWVPQeh3n2PwxEvZ8/+l7D//Zf9D/4ezoWHQ9nssOu1dnZ3fVhG2cePGfm23fv36QX3eu+66i8LCQvbu3cvixYu58847e20jiiI333wzr7zyCgUFBXz66acsXboUq9UKwDvvvMO+ffsoKCjgZz/7GbfffnvosUuXLuWuu+6iqKiIBx98sMd9MmcuO2oKAGh1mKmw1AzbOlzNLdS8+z4IAtk/+D476vfhH5UOQPvefcO2ru8isgiTkemGs6EBZ30DBAauDqYIc9RJVhGdxcUU/PQXtGzeMmj77gtrt0iYucM1pM81XOTl5fHKK6/gOEHEMjc3d9CeU6fTcfHFF4eiGLNmzaKioqLPbQVBwGKxANIYk5iYGLRaLQAmkym0XXt7e2h/TU1N7Nq1i5tvvhmAq6++murqakpKSvq1vo6Ojh4/Lte387M/0+hwdXKouTj0/911wyd2Kl5djt/tJuH88zigsfD3zS+wSlUN9MwEfJuwexysKtnAgcbTq/lAFmEyMt0w79kLQOQEKfXjcw5OOtJrd+C1WlGFG4lfeB4+u53Cv/6d0v+8gOj3D8pzHI21eyTsW1qYD/C73/2OpKQkli5dys6dO7/x53/66adZvHhxr9sFQeDtt9/mqquuIiMjg7PPPptXX30VjUYT2ubWW28lLS2N//u//+O1114DpNmPSUlJqFSq0H7S09Opqqrq13rS0tKIjIwM/fz5z38ehFcpc6rsrt2HX/QTHxYT+P/+EzxCwl5TQ/vBg4O2DldLKy0bN6M0GNAuXshzO6S/u6ok6e+yJT8fURQH7fmGmw5XJ//bv4J7PvkNy3a/xV82/huLo324lxVCFmEyMt0IpiJj5swCBi8S5mpuBkCXmMjIn9zNqJ/dj0Kno+HzL6Vi2CGgR03Yt7QwH6CiooK3336bjo4O5s2bx8SJE3n66adpbW09qf3Nnj2b2NjYPn+qq6t7bPvYY49RUlLSp9Dxer08+uijfPDBB1RWVvLVV19xyy230NLSEtpm+fLlVFdX8+ijj/Lggw+e1HqPprq6mvb29tDPQw89NCj7lemNtbiEsmUv9at2dHvNHgCun3A5Rk0YpeZK2hwW7NU1OBsbe20viiL1n31OwX0PcODXv6Xj0OE+9yuKIuYBlBs46uoACBszin/sfwuH18n5OXOZO/l8zOFK/JYOmor7fq4zCZ/fx+t7P+SeT37DB4c+x+ZxEKOPwuVz897BlX0+xul10djZ/I2uUxZhMjIB/F6v1K6tUBAzcyYweIX5roDfmDY+DoC4+XNJuugCACn9OQR0j4R1OjyhTslvG4IgcMEFF/DWW29RX1/P0qVLee2110hJSeG6665j1apVA9rf1q1baWlp6fMnLS0ttN0TTzzBBx98wOeff47BYOi1n4KCAurq6pg3bx4A06dPJzU1lT179vTa9rbbbmPdunW0traSlpZGfX09Xq8XkE6yVVVVpKen92v9wQHewZ9g+lNm8Kl89TXqP/mMhi+O/zdm9zjY13gElULFtJRJTE6SIu17dq+j4P4H2P3DH1P4939gr5HqxLx2B0VPPEXZC8sQA38HRS+/3GeE6pONZdz6+y/Zdbi3kOsLV5MkMkowU91RT05UBrdPvpZbcq/ClpMAwKcfL8Pr8/bvTThN+bLka1YcWYXb7+HsjBk8ccHD/OG8B1ApVHxVtol6a08PSLvbwYOrHuPelb+joH7wIo8nQhZhMjIBrIVF+BwOjCNyUEeZEJTKQRNhzkbpC69LSAjdpomJBsDd1jYoz3E0nY6e7tff1uL87phMJn784x+za9cu1q1bx65du7jooosG/XmefPJJ3nrrLVavXt2jtqs7QTF1+LAUVSgpKaG0tJTRo0djsVioC0QkAD766CNiYmKIjo4mPj6eKVOm8PrrrwPw/vvvk5qa2mf3pczw4bU76Aj4CTauWXvcFN6e+gN4/V4mJY5Fr9YxNXkSiCKOd1ZKIksUadmwkT0/vp/Cvz3J3gd+ScumzTi0Ap/Mi8RqUOAqKsW8O7/Xvr/aJUVnNxbU9mvdwQvCQn8LRk0YPzvrB6iVapQKJWdfcD0AutJ6Xsx/e0Dvx+lEu7ODdw58CsDD8+/l3ll3kG5KIS4shgtHzMcn+vnf/hWh7UVR5D87X6fe2oQoijyz7WVabENzXD4aWYTJyAQIpgWjJuchCAJKvR7R6x2U+ZFHR8IANDFSbYir5eTSZieieyQMvr3F+UezdetW7rrrrpD4euSRRwZ1/zU1NTzwwANYLBYWLFhAXl4eMwORU4A777yTFStWkJCQwAsvvMB1111Hbm4uV155Jc8++yzp6em0t7dzxRVXMHHiRHJzc3n22Wf59NNPQ8X5zz//PM8//zyjRo3i8ccf5+WXXx7U1yDTN6LPh9tsprOsHHP+HprWrjumi3z7vn2hKJWjpobOouI+twPYHuiKnJmSB0Be4jiy671EVbahNpmY+vy/SLnqChRaLS2bNuOsq6MuVs1bF8aQe/5iduZGAFD66vIeNaQtFgdltVJ9U0FR83GFoCiK7G88wq79Ukex1aji3ll3EBeoUQNIyJsKSgWpTR7WF29kS9XuPve1u24/v1/3FA3fcOquv7y172PsHgdz0qcxMWFMj/uuHHcherWOrdW7KW2rBOCL4vVsq8knQmtkWvIkOt02ntzyXzy+nsfQVruZdWVbcHsHb7yT7BMmIxMgWA9mmpwHgFKvw9vZic/pPGVfr6AI08XHh27TBkTYUEXCgt2RkUYN7Z3uIYmEiaLI4T89jkKjZswvfz7o++8vjY2NLF++nJdffpny8nKuvPJK3n//fc4777xBf67U1NTjnuyWLVsW+n3JkiUsWbKk1zYZGRns2LHjmPsYPXo0W7duPbWFygwIv8dDwf0/x1HT0zpCUKmYtuw/aI7ymjPnS2llbUI8rsYmGr9aS/joUb326/a62VN/EIWgYGrKJAB0CjXn7pO+j+Kl89AlJJB52y3EXXYR7z//JxrtrRzMjebeOXcyJXkCL7pstB38iOiqGlo2biZuvmT0uvNQVylDW4eTmqZO0hLCezy/KIpsrtrFx0dWUWmp4ZpmMwALpl5IXlJP7zmVQU/EmDF0HDxEcrOHZbvfYmzcCIxu8HRYMaSlUtNezz+2vojL6+LVPe/y4NyhMUFubXdQ12JjYk7sgB5X0lrB2vItaFVabsm9qtf94Voji8cs4n/7V/DG3g9ZMmkxy/e+j4DAvbO+x6iYLB5a8xdK2ipYXvA+3596Az6/j8+L1/H2gU9xeV3o1TpmpU0ZlNcpR8JkZABPRwedJaUowwyEjxoJgEKnAwanLswZqMN4cWN9qDZLEx1IR55kAfmJCBbmZyRKV9FDIcK8HR2Yd+6idfNWvDbboO+/P1x22WWkpaXx5ptvcvfdd1NXV8ebb745JAJM5tuLeVc+jpoa6RgwZjTRs2ZiyMxA9Hpp3rCpx7aiKGLeLYmwET++GwSBlo2b8fVhB7K38TAur4txcSOJ0BoBaPxyFWFtDpqiVBSkKwHwi36eP/Ihn2Y7qZqVyZ8W/YopyRMAuHzcBWzLlcRV+RtvhKLzOw5JdWCxJj0gRcOOZkPFdp7Z9hKVlhoSwmJJ8kg1gvOnXtDn+2DKk+xcZtuiwNzB6r/8ll13/pA99/6U1uJC/r75BVxe6XXurtvPoaZjRwBPFlEU+eNL2/n1vzdT3Wjt9+P8op8X8/8HwNXjLiLG0LdJ88WjzsWki+BAUyGPbXgWn9/HNeMvZlLiWHRqHT8/ayk6lZYvS77mnQOf8qvVj7O84H1cXhfnZs1havLgjXaSRZiMDGDOLwBRxDRpIoJSOigq9dKBbTBGF7kCNWGbK50cKpdElzrKBAoFrta2IWkJD6Yj0wNXxkMhwoLeZwCOmv7VpAw26enpbNu2jT179vDjH/94UN3xZb47NK3/GoDM229j0l8eY+xDvyTrDmm6QfO6r3ts66iuwd3Sgi4xgYgJ4zHlTsJnt9O6dVuv/Qa7ImemStMQvJ2dVL0l1VttmGIkv+EAftHPm/s+Ylu1lBJ7eP5PSIlIDO0jLiyG5LPn0hCtwtPYTOPqr3C6vOwtbkalFLjpAinltre4pwjrdNlYvvd9AO6cuoSnzn8YRYcdhU6HKiKiz/chmAlI2d/I7Z+0krKnBr/bDX4/qz96iVprA6Nisrk5EGV6fe8Hg378OlTeRmmNlGYtrbH0uY3X76Og/iA17fX4/NKF7frybZS2VZJkjOeSUecec/86lZZrx18KgM1tZ2LCGK4ed3Ho/pSIRH44XRrf9N7Bz6i01JAWmcwfzv05P5xxC2rl4E08kdORMt9p7FXV1H74Ec1fS3USpimTQ/cFRdip2lR4bTa8nZ10KnV4FSpKa9rJGxWPQqVCHRmJx2zGZ7ejCgs7pec5mmAkLD3p5CNhos9H1VtvEzNrJsYROb3ud9Z3iTB7dXWf6Zih5l//+lfod7PZzMqVK6mtreWXv/wldXV1+P1+UlNTv/F1yZw5eKxWzLt2I6jVxM6ZHbo9cuIENDHR2MrLsVVUEJaZCXSlIpszTNy14kG+NzkXCvZS/flq3mwI5+YLxxJr0uP1+9hdK5myTk+VIkzV77yH19pJ9MwZKEa6MHfUs2z3/1hTuhGNUs2v5t5DgjGOo1k87gKeytvIVWstVP3vHSoSx+Dx+skbFcesCYk88w7sK2nB5/OjVErxlTf3fYTV1cm0lFwWjZgnTerw+9HFxx1zZJIxOwtVuBGvtROlSsm+bA1t8TrO2WJGUVhJxMg0fjrnTiK14awu2UBJWwXbavKZnTZ1QO95267d+J1O6X04qtzjk41lKCKbURgtlDdkcg5pPe73+/38Y8sydtQWAKBWqEiNTKLJJl3g3j7l2hMKpQXZc1hTthG728G9s+5AoegZk5qTPpWq9hpWlWzkirGLuHjUeagUygG9xv4gR8JkBkxdcyc+/5lt5ueoq+Pwnx5nz0/up2ntelAoSLz4QuLPXRDaRjlI6cigR1i7SkpFBAtpAbTBDsnWwa8LC0bC0uKl5z0ZEWbO30PNu+9T+fqbfd7fPRJmrx6+MSwgFeSPHDmS5557jj/+8Y8AHD58mHvuuWdY1yVz+tO6eSui10v09GmojF0XQ4JSSdx8yWKkqVs0LNiluFbXQIerk2ftmxH1WpxHDrNryyHeXlOE1+/j3QOfYvM4GBWTTbTehKO+nvrPPkdQqci849ZQjdia0o2hmqQRMZl9rjE1IonkaTOpSlTjbW+n5oOPAZgxLhGjQcOIVBMOl5fiagsARS1lrCnbhFap4XuTrwO6yiK03WpTj0ZQKhn1s/tJv/EGpr3wHxQ3XMSBNBUeJSS1eLgv7yZiDFGolCqWTJJMit/a9/GALC06y8o4/MfHKPzbk+z83l1UvPpayCut2exge/0uNKN2o04pZUP7hzg8XccuURR5ac/b7KgtIFxrJMOUih+RcnM1NredacmTmJw04YRrUCmU/Hnhr3jq4keI1PUdFbxh4mJevOJvXD5m0ZAIMJBFmMwAyS9sYunjX/Hx16XDvZSTRvT7OfLnv9K2YydKg4GUq69k2n+fI2fpD1CouoLDgxUJczYGRJhaEkOl3URYqENykOvCvD4/DpeXML06VC8yEEPHIMEUo72quu/7u6cjq/ve5pvi/vvvZ9myZWzatCnkNj979uzjFsDLnDmsL9/KewdXDknqvvnrDQDEnTM/dJvP72PFkdXsSvGHthF9PnwOB+2HDuFVQE2ChnOzz8KnFNiXKp1OJ3aUsvHIIX69+nE+PPwFCkHBpaOl+sSKl5cjer0kXXwh+qSkHrVFt02+hhmpecdd55VjL2DjZCMiEHdwM0avnenjJNubvFFS9KyguBmf38d/d78FwDXjLyE2TLrY66tLuy/UY8YTeelitDHR3Jp3NTERsdQkaFCIkNzY1Rk4K20KOVEZNHQ2s6Zs03H22JPK5W+E1uHt6KD2g4/YvfQeip/+J/9ZvwJV5j4EAUS3FpuykT+uf5pOt1Rz+tHhL1lVsgG9SsfD8+/lbxf8hteu+gdPXPAwPz9rKT+edXu/16FQKE4oroZ6yLqcjpQZEIUVUsSmoKiJqxYMn2+RKIon/eVo27ETe1U1+tRUJv31sWOmAbtE2ClGwpqkK7xgJKyupROHy4teq+ryChvkSFhweHe4QU1UhBTRO5lIWNBd293aitdm6/VeObtHwo4h1L4pioqKuOKKK4CuA6fBYJBnJ34LKGmt4LkdryEikh2VxpRBLIx2NjbScegwqvBwoqbkAdBib+Of217mcLM0r/PmWC0xLRYse/dhtraB10dtooZrJi/mqnEXkZc4jjfbnie32MEk5yF2ZTdQYYG0yGTunnErOdEZtO8/QNv2HajCw0m7/loARsZksWjEPOIMMVwcqGF6Z00RTreXWy8e12ut2dEZpI3L41DhVsaXOVnk2UBE+BUA5I6M492viikoasaYXiPVMUUkccnorgaVkF/hUZEwj9fHofI29hY3s6eomdIaC3qtin/89BySYsP4w3k/p8j8Dr73vsSyZy8xM2cAoBAU3JR7JX9Y/w/ePfgZ8zJnYlDrj/t+W/btx7KnAHVkJJOfeQpXczMNX6yi8at1NK1dT+f8SEjWcs3Iq3nvIwv+rO2UUMHv1z7F/KzZvLX/Y5QKJQ+cdRdZUVKaUqVUkW5KId2U0v8P/jRBjoTJDIhGsx2Asrr2YZsvVtNk5frfrOT9tQPvyhFFkZr3PgAg9Zorj1uHpdQH0pGnWJhvq5NEmE0fztjMaEQRyuukaNhQ2VQE7SmMBg1atZIwvRqr3YPHOzDX/OMV3ouiiKO+HgQBpcGAq7kFr31wzG1PhvT0dPbu7TkCKj8/n6ysrGFakcxg4PX7+M/O1xGRjjfvHvhsUI89zeulKFjs2WehUKvZWbuXX3z5Jw43l5BojGNC/Gj2Z0j1RRvefZHNX0hF9cL4HK4ceyEgRYRioy6jOVJFhMNLTq2LeM8kHj//V+REZyD6fJS/9AoA6UuuQ2WULsgUgoI7py5h8dhFAFisLl77/DDvflXMwbK+o+M/mXkHVdOm4VYJjKpt4pHlD/Lx4VXU+Q+hTSmj2LuFdw58AsCd05b0iPR0RcK6RJjX5+cX/9zIw//ZwrtfFVNSbUEQBOxOL8++W4DfLxKtNzHpHKlw/egB3xMSRjMlaQJWVyd/+vqfNBzlRN8dURSpXC6ZEKddfy1KvR5DejrZd92J7bI5AEw5ZCfRNofrpiwkLSYO15EZpBhTqGyvZXnBewDcM+NWJiWOPebznEnIIuw0pnnDRvLvuRdn07H/qL9pmtqkk2x7pxuzdXgiDBsL6nC4vOQXDvx9ad+7j87iErTx8cTOPTt0u93pYev+OvKPNFFSbaHJbEfUSK3cpxoJa6uSoknhyUmMTDMBXXVhXZGwE6cjRVGkwlyN3X3s9TSt/5qGVau7ImF66eQRHYiGDdSwtUek66h0o8dswe90oo2Lw5AhjdRx1A5PhyTAQw89xGWXXcY///lPPB4PL7zwAtdffz2//vWvh21NMsemzWHpVx3RiiOrqGqvJSMyhYSwWErNleTXHxiUNYiiSFNAhDWOjeef217mb5v+g81tZ27GDP6y6Nf83zn3cdYVt+AXILKwnrhyyWfr4sV3hiKuPp+fIweV7DZKtUgXbulEsU2JwymJxaZ1X2MrK0efmkLCBYuOuZ49RV3HtHe/KupzG6M2DGvDZLZFSyJk2rYm3tj7AS8X/A9FShHKxHKcASuFsXEjezw2OLJIl9AlwrYdqKe0pp1Yk54bLxjD334ylzd+fyHxUXr2lbTw5XbJ0FSfmoImJgZnfYNU4N+NO6ZcR1J4PMWt5fziyz+xqmg9DV+uorOsvMd2rVu20Vlcgi4xkYRFC0O3t9jbeFl1CJtOQWqzhyXJUhNQWkI4eDVcGH8DY2Kl227OvYqzM2Yc8z0805DTkacxzes34KippW37DpIvu3S4lwN0RcJAEhLBk/s3yZ6A+Grqtpb+Uv2u1K6dctXiHvVfb68u4oP1JT22ndFRzrmcughzNDSiBpJGpBKfEglAaU07X5dvY0PRx5wLNNdWkun3oeyjPkEURQoaDvLO/k8pNVdydsYM7p11R6/tfA4HJc/8C9HnQ/XA7wAIN2gAiI7QUt1opa3DSXx07zmHfeFzOHpE6I4uvHfUS+JSn5yENiEe6+EjOKqrCR85PGnqG264gYiICP7973+TkZHBhx9+yFNPPcWll54e3x0ZCY/Pw1v7PubToq8I1xo5O3068zNnkRWV1qvEoK6jgfcPrkQQBJZOv5nq9jqe2/ka7x74lClJE06pXsfr8/LhFy+TXFeHxajk1frPQRDQqrTcOeUG5mfNCm17zqSF7JuyFevuAtQOP+r4WCLSM0L3HyhtxWp3Yx81l+Tp6dR9tIIra75i6ydZLFx8dqixJfOO23ocd46m+4Xl7iNNlNe1k5Uc2WObZrODsrp2zAkzWCg2kdLcwlXubNwTs6mscbL7oJmZozO5a9olvfYfTEd2j4St3FwBwC0XjeXcaV1diD++No/fvrCVlz85yNQx8cRHGTBNzqVpzVosBftIvLDLQiPBGMdfF/2GN/Z+yBcl69n8/iuEbbOCQkHqVVeQdsN1IAih9yH9piU9OiI/PPQFHsFPfkoic0vriNq7GS6aQ3qCVDDf2Ozmdxf+lBZ7W5+do33htTsoX/YSUdOm9Oh4Pd3olwhbsWLFiTcClEoll1zS+4OXOTkcgfZ/W1nF8C4kgM/np8XSJUjK69qZNjbhOI/oH9XvvEfNex8QN38uKVddiT4p8Zjb2hweCqukK9EWiwO/X0Sh6N+BuOPwEToOHERtMpFwXk8PmdrmTgDGZUUjCAIdNheuDkkQNTaYyei1twFgkYRM9vhsEuKlA2pZbTvNpTtpUkjPW19TytIVv2J6Sh5J4XFEaMOJ0IbjF318fHgVha1lod0dPoY5orWwCNEnpRud2zYDyRgNPSNhA6kLcwQGiwtKJaLP16vwPhgl0yUloU9JAoa/Q/Liiy/m4osvPvGGMsNCXUcDT299iXJLNUpBgdXVyefF6/i8eB1pkcnMy5jJnPSpxIXFSOalu97E4/dy6ajzGBGTSVZUGh8c/oIycxW76/YzLdBdeDL878AntK//mmSgfEQEM9OmMC5uJNNTckOF7N1JXriQwt0FAMRMndpDAG7eJ12QnJWbQuY582losaHZ9BWKt5dRXLEHj9mMKS+XqKnHdln3+0UKCqVI1fkz0lm9o4r3virmF7dM67HdzsPS93Ly+BQyF91M0d//wYhN5Uy54X7K4uzsWP81dSpjL8sFv8eDu61N8ggLl9KhVQ0d7C9tISJMw9m5yT22nzw6PrSOf727l0d+MAtTXl5AhBWQeGHPiJ5WpeF7U69naspEDm96PPiiqHnvA5q3byc6bzLOujrCcrKJPXtO6HHNtlbWlm1BEBXsVk3nLPXntG3fgaO+gfREyeOwutGKUqHstwADqHjpFZq+WoulYC8xs2YiKE7PxF+/RNiVV15Jenr6CbdramrCNkyu2d82RJ8vZPBpK68YsudxW9rxtLcTlnHiz7e13Ym/mzVFd6uFU8G8Ox+/y0XjqjU0rllL3NyzSb3mKgzpab223VfSHFqD1yditjqJiTx+IWiQmvelWrC4yy5kZfkGWu0Wbph4OVqVBkunlKb70dW5ZAZ8tT79jwU+38reAzWktnSSHGsc0Otq6mzB0WFB7XVhVeqZnRNPmF6NWqWgqsmM3lyFOkyKVEU6ocPVyVfH6DCKNURz9biLeGv/x7Q6zHS6bBi1PevZOg4dDv0uFGxHSFzcLRI2cBHmDBTlR0wYT/vefb0jYQERpk9JwpAmfVb2qmpKayx4fX5GZ/Q+kQ02JSUl/RpsXVpaSk5Ob58zmW8Gv+hnfflWXs5/B5fPTXpkCvfN/h4qhYqvK7axoWI71e11vLHvQ97Y9yEjY7JIDk/gcHMxcWExXDfxMgCUCiVXj7uIf+9YzrsHP2Vq8sSTioYdaDzCZ4dW8b1K6Xv/vaV/wJCcfNzHRE+fhjIsDJ/NRlQ3P0GfX2TrAem7MGdSMoIgMO2nS3mhsIm85v20bt0OCgWZd9x23LWW1bVj6XSRnhjO7ZeOZ0NBLZv21nLTRWNCx54ms513v5IuwmaOTyR2YhL1n67EWlhE5RtvkX3brYQb1FTUd2C2OokK78pUuFpaQRTRJcSH1rFySwUgiT6NuncU/nuXT2D3kSbyC5v4amc183MngiBg2bcf0ecLGVt3Z3xEFrZ6Nz6FwAfnRTF3VwcJ1bXUV0ulCpm33txDEL26cwU+0Ye3KR2NOoa4886l9YsvqFvxCenX3gRAdX07lW+8RcumLSReeD5JF12IQqM55nvZtms3javXAFKph7WwiIixY3ptZ6+pRfR6Ccs8pcvsU6JfIsxgMFBeXn7C7WSn6sHD2dQUimrYq6vxezynPL+wL4r/8Qzt+w8w9T//Qht3/BldwVRkSpyR2ubOQRNhwXqohAvOp3nd1zR/vYHmjZuY9Pifepl/7ins6Qjd1ObolwizlVdg3rkbn07DY77NtBdIB1+v38v3p94Qqm+LCteGHjNnaiaHPwfB4+KPL27niXvnEabv32fg9fv4v7VPoG/o4FrAoY8g0ijtOyMpgjJzOV6/l5yEHJRhHWhtdp5Z9H8caCvF4uyg3Wml3WXF4XEwNXkS52WfhVqpZkv1bvY3HqGyvZbx8T3fm/aDh4BAV6fNSra9DqNBihScVCQsILIix4+js6QUV1MzPocj1DXq6BYJUydJJ7CaA8X886mvEQR49IdzmDSi/1euJ8OUKVPo6Og44XbTp0+nbYhmdMr0jdfv41BTETtqCthZuxezUzpeXDjyHG7OvQpNwEzzhomXc92ESznUVMyWql1sr9lDcWs5xa3SOecHU29Ep+r6Xs7NmMEHhz6n3FzNrrp9TE2eSFlbFTtqCyg3V6FVajGo9Rg0eiK14czPnEW0wRR6vNXVyT+3v0JykxuDy49x1MhjCrAvt1VwqLyNzKQIspMjSbzlFjzFRSFXeYAjFW1YrC7p/hjpwkitUqK66Eq2r/Ax03KIpIsuOOGJPlhmMWV0PBFhGi6YlcGKDWV8uL6Ue67JxWJ18dvnt9BicTB1TDwzxyciCAI5P1rKvl8+RN1HKzCOGMGkkXFs3lvHnsLmHulFV8CHK2hPYXd6WLurGkGAC2dn9rkmo17NPdfm8scXt7Ps4/1MH7cQY042nSWlWItLiBgzutdjLHv2IHo8lBlS8PrPp+r79ZSu3sSM/Taa0iMYO0p6H7w+P6+u2sV2y05AINmfywP3nEWSykXrqlU0rVlL6vXXEaVwM2/fKmq2SxHAipdepW7FZ6QvuZ74BfN7CUFPh5WSZ/8NQFhONrbSMlo2buolwnwOB/t/9Wv8Hi/TX/4vKkP/yjQGm36JsNdff71fO1u+fPkpLUami+4F0aLXi6OmlrCszEF/Hlt5BaLXi7W4+IQirKlNEmGTRsRitjqpb7Vhd3ow6I4vTCrfeAtBoSB9yfW97hP9ftxtZgS1mpwfLSXjpiWULXuZlg0badm0ubcICxSuZo/0UlYqXRmOzTpxxGXnK/9BDewaoaIdF5OTJnCoqYgvS75mctJ4LB1OFAohFDkCUBoksRGlhZqmTv76+i5++72ZITfq47G34RBmRztRgUJ4ZUzXe5uTEkmFxwLAqJgsNNHVOGx2TG4VC3PmHne/GaZUSYRZanqIML/HQ2dRMYJSSdqS66l46RVyO4oJN0gRhJOxqQiJrORkDOlpWA8fwV5TG6r5Crrl72mBVz/ZxW0KNQZnB5EakXa3wFNv7eHZny/ot3A9GRwOB1dd1XtI79F4ArP2ZL4Z1pVt4bW9H4S8nQAyIlO4YdLiPufuKQQFExJGMyFhNN+begMHGo+wvaaA5PAE8pJ6WjVI0bCL2bXsXzStfJyfXpJGvXjs+YIfHfmSmyZdwcKcuQgIPL/zDcyOdhZ0hAMWomdM7/NxVrub597f18uYOi1hDHdXtTMhMFg6mIqck5vUY7sF09N5YMs0ajIm88SdVxxzfUF2H+kSYQBXzh/Bys3lrNlRxeVzs3ni9d3UNtsYnx3Dr26bHjoOhWVlknPPDyl+6hlK/vkvpt96H5uBFz7aT3JsGGMypeNj0Kg1aE+xPr8Gh8vL9HEJIfHYFzPGJXLWpGQ276tjzY4qpuXl0llSiqVgb58irOiL9dK/YelUVPq5YvaFZN1/IS9ufY2SjmrWffU3fnnWPfzr9RJKhY2o4kWyNJN49McXolZJryn27Dm0bNhExbIXubWiAL3bhjImluxbltC4ag0dhw5T8s9/Uf3+R2TceB2xc2aHxFjZ8//FY7YQNW0qWd+7nfy7f0LLlq1kff+OHoKtaf0GvFapHKR97z5iZs9iOOiXCFu8eDF1dXUkHydc++GHH3LllVcO2sK+63S3BgCwlZcPighrsbfh9/uJN8bid7vxWCyB/VecsHgxKMISog1kJUdysKyVynrrcUWQz+Wi5h2prTj1mqt6RfM8HR2IPh/a2BgEQUAdGUnyZZfQsmEjHYcLe2xb19JJQ6udmCQH9VFfo85Mosl8Yr8gt8eFeKAEvwCGhXP5+5RLSYtMZk3pJl7Y9Qb/3vEabqYRbYzoUV8WjPhkxeiIjdCTf6SJD78u5ZpzRx7rqUJsrNgOQIRNimaGp3TVueWkRLKuxQLA6NgctDEHpVl0bW09upbK69r51b828YPFE1k4Q0oXZ5qk8TsVlp6pwc7SMvxuN8ZRI0lYeC6lr75Ojq0GrU8SXSfTHRm8EPi60kVOTAJwBEd1DeEjRyD6/TjrG0Ch4NnVVXhRYA+PRddez7N3TOCJtY3sLW7hPx/u44EbBzbOZCA8/PDD/dpu6tShW4OMVNZgr6ggMncS7x38jHcPfgZIPlgzU/OYkZJHYvixXdq7o1IoyUsaT17S+GNuMyNyJMIhB0q/SGxRE6qp2UxPzWVC/Bj8oh+7x4HNbedQczEbK3ewbPf/2Fi5kwnxo9lRW0CkNpwRDU5c0CO12J2t++vx+UXGZkYzPjuGsrp2ymrbqW7s5NfPbeaac0eyZNFotgZF2MSe58eRaSZS4owUN3eybMVBOu1uaps7aWyzc8mcLJZc0BWZsTs9HKloQ6NWMj5bsq2JNek5Z0oaa3ZW8bN/fI3T7SM7JZL/+95MdJqep+74c+bTWVxC/acrMX26nAvm3MyXe5t4+Pkt/Pr2GUwZHd9lT5EQjyiKrNwsRRovnnNiC5fL5mazeV8dq7ZXct6ludS89wGWPQWk33Bdj+0aG9txHdiHCoH4OTPgoJnlKw/xn18t5P8WPcDft7zA3obD/Gb13+jsGIN2RC0qQcWDFywJCTCAlCsX07JhEy0bN6MHSgwpZH7/HuLPGk3cOfMx787n4POv4KqrpeiJpyiKfZOcm65HqVTQsmkzqnAjI+75EZroKMKys7CVldN+8BCmSdL5QhRFGlZ+Hnq+tl27T28RBrBo0SI2bdqEyWTqdd/HH3/MLbfcQmdn52Cu7TuNM1AUHZaVia28gs6yCuKPPY+0X4iiyCNrn6TD1cmTF/2WsPauE7K9orLPx5j3FFD6r+cY8+tfhdKR8dEGslMkEVZW135cERYMgYOUdtQl9iy6D5qUBv2yAMKys1BoNNjKyvA5naHxQcFUZEq6SLEIyqgmGttO/DdXfCQfjVfEGq3nhwvuCt1+XvZZ5NcfYFftXjRZB4i0L+jxuKBPmMLj5v7rJ/Pw81vYcbDhhCLM7nGws06aFxfRKYmw+OyutEBWcgSKUgsgRcKag675LT1tKjbvrcPu9LJ6R2VIhKVHSmaEVZaeVhDBerCIcWNRhYXRmDiK5NpDqA/uglmjBpyOdHl8WCqrUQJv7m7jQjWMB2xVVXx0+EtSfUZpqG9MHF4UTBubwJjMcTStqcddV8v9N8zgx0+sY/3uGmaMS2Ru3tCYKP7ud78bkv3K9B9RFDn86J/pLC6m4tqZfKwuR6VQ8eOZtzEnfdqJd3ASNH+5BmUgQnWJO50pF/22z+3Oyzmb+Zmz+O+uNylsKaWwRZr0sXTEYlwvP4naZDrmxe3GAuk7dvm8bM7Olf5+/X6RjzeUsnyl5OW1qaCOlnYnaQnhkp1CNwRBYMG0VF7//AifbCzrcd//Vhcye1JyqP50X0kLPr9I3qiYHrVZVy0YwVe7qnC6faTEhfH7H8w+ZmQ5847bsJWV03HoMAsq1qKbexkfbyznjy9u4+c3TSOuWyTsYFkrlQ1WEmMMocjb8RiXFR0qQ6nWxKLQ6bAWFePttIXGPPl8ft547mNm+j20x6Vz7x1zaXlhKwVFzby/tpibLxrLg2ffze9X/YfCjoNoRxYAsHDE2T3SxQDG7GyiZ06nbedurLPO573GeK7r8Ife1+hpU/lgoxWVsJc55n3EtzRS+vQzocdnL70LTbRUHhV79lnYyspp2bQlJMI6DhzEXlWNSxuG1mWjbVc+ot8/LMX7/X7GGTNmcOGFF/YqvP/kk0+46aabePnllwd9cd9lgp2RMYHolK0fNXknwuxsp8nWitPr4s29H+FqbgndZ6uo6PMxjau/wtXcQtv2HSGPsIRoA9nJ0sEjaDp6LBz1XSKs+/OFbgsIj6BfFoBCpcI4aiSiz4e1qKsTMFgzERUtHXwFpY9y64nfl/J9u6T9pvVMFwiCwA+n34xRbUQZ1QzRVT3u7+6YPy47GpVSQUmNBY/Xf9zn21FTgMfnITdhPBGd0lrD07peX5jJi6BxIboMhGsjurzC2nqKsOAMuMJKM06X5KeUGpGIUlBQ3V6Hz99lvNpxSKoHixgnpW6K4yUPIc/2zYiiSFSEVFPjb2qgae26UKv60YiiyOa9ddz/p89QOu1YlXp8Kg2HbFKatq2shDf3fcSKjf8DoFNvAmBiTiyGdClKt79gE+8WvcfSq6RIxr/f20tr+/CZuMoMLa2bt9BZLH1PDWt2Y1Dp+M38nwyZAPO73TR8/oX0H4UC58HCPm1kRJ8P8+58JsSO5IkL/4+LR5yHgIIFaeeQWiddUJom5/V54m3vdLGvpAWtRsm0MV0d4AqFwJXnjODJ++eRnhhOfat0PjxrUt9ZosvOzuaqc0Zw4wVj+OUt03jmgXO45tyR+EV4/sN9IdPZ/KNSkUHSEsK5fG4OI1Ij+cPSOZi61awejUKlYvQvH0ATHY1ldz4Xq2q4+aIxeH0if31tJyUHJCG4v9XP++skO56LZmf1q7tcEAQWzZRquVbtriVy4njw+2neuDG0zf9WF6ErPQjAuEsWIAgCd14+AYUAH64voclsp63DTfHmLLyN0kWlWqHiirEX9Pmco3/xADNefZGYyy4HQaC6sSvlXNvcSUlNOw1Jo4l68LdsHHsRjRpJdB0Kz+Iv+T5e/ewQ+YVNmGZL59DWLVvxe6XjaN2nKwHYGD6WZo0Jr8WCrez45xLR76d163bKXngRt2Vw6qFhACJs2bJlJCUlccUVV4TqKz777DOWLFnCiy++yLXXXjtoi5LpSgXFzJFCpLbyilN2ie4ePdlUtZOKskOh/7uamvF29u5sDR5cnQ2NNLZJ98dHGULeNScqzu8eCetLhAWFh6ZbJGzNjkpW1UtBWusRKSXp9fnZV9KCQgCNoWt2WbPvxCKso0R6DdGje3fHRGiNnJsg1U016XbR1Nm1xu4iTK1SkpMSicfrP6bwbN26nfb9B9gQSEVm6cYTHqgZr1J2HUAqLFLU0Wc1UdfciSa69+giURQpClhx+Pwih8ql+1RKFSkRSXj8Xuo7pQO36PdjDaRug8WnFeo42tTheBsbsB4+gsrrYZE5n+sL36f46WfZfdePKPjZL6l57wMc9fWIosiOQw387B9f8/jynfiapX3HZKczITuGJpUkuh01UhpU3Sq9nnqkq+AJOTGhDsn64kOsK9+CL7KKeZNT6HR4ePp/e4ZtwoLM0OH3eKh8TfJ+8igh3uzloZgLezWNDCbNGzbiae8gYsJ4oibnInq9WAr29tqu5r0POPSHP1H9v3fQqjSIdWOx71xI06EMLPl7AEJjio5my/56/H6RGeMS0Wl7J4yykiN56v75XDE/h5FpppBAORqDTs0dl41nyaLRzM1LISs5kuvPH0VclJ4Dpa1sKqhDFMWQP9jkPqJSdy6ewFM/PYf4qBMXjmuiohjxk7sBaFy9husXjuaHV01CBISAVc6/1tSw63AjGpUiFGHvD+dOS0OpENiytw7TuZLRasXLy7FXVbOvpJl3Vh9mpE2ysUkKWFBkJEVwwaxM3F4/L39ykCffzMfm8DIn9nx+OudOfj3/J0QHLuSORqFWo46IID1ROvZUNXQ14HydLx2H5k5O4ZxpGTz45++T+sgf2TLvNlalzqOwysx7a4v53Qtb+eXrB1FnZuO1Wmnftx9XczOt23fgFlRUJ42j3ChdPBat6bszXfT5aFq/gYL7fsaRx/9K/Wcr6Szq20j3ZOi3CFMoFLz99tv4/X6uv/56PvnkE6677rqQK7XM4OH3enE2NaHQatGnpKBLSsRns+Fqbj7xg49DZUCERWgkAbX7YM8/OltlRY//uy2WkMOys6GBlnYnGrWSSKOG9MRwlAqByvoOfL5jR4a6Oyu7WvoQYX2kI/cUNVOpkYpeG/L3A1I0yOHyMjI9Cqun68vo0NXh9/d+flEU8ftF3D4PqlrpNWRO7LsAN9Kfirc5Bb/gZUv17tDtCo0GFIrQ2KLRmdKV1pGK3l12XpuNI399goN/+BPllYfRq3X4zLFEOjyIwAFP1/tQ2CqlRPydJspq29HG9k5H1rfa6HR0FZPvK+n67DMC89EqA3Vh9qpqvJ2dGNLTUEdIKRGrw8O+CKmAvuKV5ey5516mtB4ABMKmTEUdGYGttJTK194g/+57+fPv3uCPL26npKadxBgDt0yX1hSVmcakkbFYlQZ8ai20tqPyipis0hVlhUuDXisJVEOadDCLskjrfvfgZ3x/8RhiI3XsKWqmsW3g5roypzeNq1bjbGigIVrFlhmBAvBP1w6a4D56P6IoUrfiUwCSL7+UqOnSd7ptx84e2/k9HuoD0bKGz7+kpbmdTzeXg6jgQHETlr37QBAw5eX1+bybAqnIuXnHroPWqJV8//IJPHn/fOKi+meTA6DTqPj+ZZKz/kufHKC8roPGNjvxUXpS4wdmg9MXprxcNDEx2CsqsVVWcclZWbzwi3OI8Dvwa7TMnTWCCTkx3H7peCLCjm3z0Gu/4VpmTkjE7fWzxx9L0mWX4He5OPSXJ/jH8u2k2psw+FwYR+T0aPK66cIxGHQqNu2t42BZK0kxYdx9VS6z06b2S6zHmfRoNUrqW2x4vD5EUWR9QISdM0U65giCwIzxSfzigct589FL+euP53LzRWPITIqgurGTVTbpeNa8cTMb//sOgihyxJTDQz+cT865kmCs3rgNt6fnWLf2g4fIv/teip96OjBvOIWR9/+EqOmDF+UdUAJUo9GwYsUKamtrueqqq3jhhRe48cYbB20xMhKuxibw+9ElSS3IYYHZd6fqFxY8abceySBOH4s3IIDUUSagd12YtbArFWhvaMTvF0mI1ksF9ColaQnhuL3+kNFpXzgbukXC+hRhvdORjW126nRxiEBnYSFtFnsoFTl5VDytdou0oU+FoHFwpLFnGtHvF7nvyfX8+rnNFDaVENsmCSHTiL5ruSydLnxtUq3awaauZgBBEFDqdPhdLkSfjzHp0hqPVJp7v442M/j9iG43Mw7YmJ06hfKSJrQ+H516Bftai0MnlKIWKS3gt5oorW3vlo7sEndFVdJrTImTIk17S7reu4xAcX5QVHevBwMpcmZzeiiOGQUKBdbCItxtbTTHZPBi+mXobv8h019exoRHf49r1ETw+4ktLSAxxsB91+fx3IPnka6WhKcuOYnckXGSN5DOhABEdXgxWaWDVas6grGZMSiVCjSxsfg1KiJsPjR+BWZHO5tqt/Knu8/ip0smk9BPp36ZMwOv3U712+8CsGmykawLLkYbH09ncQmWPQWntG+P18dP//E11/76M370l6/47fNbeObtPWx8fy32yip0iYlET5tKdOBk2LZzd8jSB6Bt+w48Zou0zs5O1i57P3SCjbM24LM7MI4cEbpo6Y7Z6uRAaQt6rZIp3VKRg8mcSUlMGhFLS7uTP7+6A5CiYKcyASCIoFAQN08aydb8tTSSKdJnA1HEmJTAvTdM4c93n81lc7MHvO8LZmYC8OX2SjJuvZmwnGxcNTVMLd/EXK10jDq6wD3SqOWG86UuSqVC4Be3TD1hR313FAqBtIRw/KLUpV5cbaG+xUZSTFhoDFx31CoFY7OiuX7haJ766XyuPW8kh4wZiEDdhs34d20BYMqt15CTauLi68/FrdIS09nEuyu6LsJ9LhdFTz6Ns6GBsOwsRj/4cyb/8x/ELzhnUD6n0Ovr74aTJ09mypQpzJ07l9bWVnQ6HX//+9+ZMmVK6EdmcAjWg+mTpBqmYOHoqYqw4mYpVOzrjEDVOIHwQOdeRCAkbyuvRBRFOgPDn9uPHAk91mexoPJ7e4TEswMjeMrqju3T1CMS1ldNWEAIBlNyIHVhupQaOsLj0Pg9/PvfX7DzsCTmJo+Oo9UhiaBwtyROt1QU9NhnfauN8roODpa1smvPNrQeEW9sBCpD31er5g4nfmsUChQcbi7pMc+u+xDvYCSssLJ3JMxt7hJm40sczNbn0FQuiV5HhJYWexuNnc04PE4q22vRKDSIjnDKai1oogNDvLulI4urpf1dODsTg05FaY0l9Ll0RcKCIqxnPZjd6UEUQRFpIvnySzFkpDP6wZ9Teu6NtGkiMXc4EZRKjOPH854+D4Cxngb+9cB8Fs7IQKVUhNLh+pRkRqaaMOhU1ArSVXpshx+TVYo+mjVhTMjpimKaI1UoRPh+6vkAfHT4S0wRSs6dlj6oB66+WLlyJSUlUq1LRUUFixcv5uqrr6amZnhd/L+t1H74MZ72DsqTNbSkhHPJuPNJvUbqkK/+37vHjYaJosjh8rZQrePRfLqpnJJqCy63j5qmTvYUNbN6RxXl738MQNJllyAolWhjYwjLycFrtdJxpOsCqn6lFAWLO2c+ALr8jagUcNGcTLLs0vfmWO71W/bV4xdh5vgktH0YmA4GgiBw15UTUSgEGlqlCHF/CuT7S9z8eQC0bNiI6PeHMhrdxxWdDLmj4oiL0lNe10FZo42Ss67EJajJ6ygmvVaa5xk9a2avx116djaXzc3mgRunMjJt4H6i6QldzvnBVOT8KaknPKaolApuvXgcD/34fBqMiai9bsJ8TjxpOcw8V/r8VRo1kXm5CMDBLzdRUmMBoP6Tz3C3tBAxbiy5T/5NssEYgsL9fndH3n///YP+5DJ9ExoHkxwQYdmBSNgJCgePh9fvo8nRhOhXIDoNlB0RuMihBDzsjfWQCjQXHea/r31AcWMt43P9jN+ygzTAJ4BSBJOns8fcQakurJry2vZQWLg7os/XowC8r3RqUHgcdNczSxyN2+vHbHURpleTM3syLatW4SktpswkhbRTk3Q4djiJ1EWQ5BvJEYrZ33wQuCK0z+51ag2H9jMKMGQfuw3bbHWBX0VyWCo1tiqK28pDg2+DnZk+p5O46GiiI7Q0mR20dTh7zM0MXnX7A++Vb8VmtDapiFYdFwO42dd4hKRwqT08JyaTfATKattRhRsRVCrcbW2hDp3iQCRsTGY0E7Jj2XGogf2lrcyemNQtElaDKIq9ImHWgFgL12uoXjCGslw9k/JmEN0qieq2gE1F/pFGqm0CreEJxFgbcRQVoZkopUkcAbd8fVISSqWCCdmxtLTooQMyHTpMnR14FdAZ5WZCtpR6qLM20hgOMc0w3hfNtORJ7Krbx4rC1dww8fJjvv+Dxc9+9jPWrJFcsn/xi19gNBoJCwvjRz/6EZ988smQP/93CXebmbqPP0EENucZWZQzlwitEeO5C6h+532shYW079uPKbfvsUIrNpax7OMDjEo38Zcfz0XVzXvPanfz9hqp5ubxe84mLkpPi8VB0e4jxL1Wi1OhwTWhKx0UM3M6ttJS2nbsJHL8OGwVlXQcPIQ6MoIR9/yQqvyDxHS0cE2alzmzM9nzv4AI62a42p1gV+TRY3wGm4zECC49K4sVG8tQKAQp4jxIGDIzMKSnYa+qpuPwEZwBe4ruFjgng1IhcP70dN5cVch/PzpAYWUbYxJnc3n9BkS3G31aKobU3p3QapWCu644sZ3QsQiKsIr6DjYEPp9zpvY+5xyLiSNiMV5zEXWvSA2EE2/q6S2YetZMinftINtWw1Nv5fOLK0bT8J40YSXze7cP6QVkv2XdbbfddsIfmcEhNA4mefAiYXUdDYj4ER1GLp6TDSLorB78AnwsFuETwFVTS4V6Nar0wxS2HSG+1Y0I1CRKdQMmsY2EHpEwqWCy7BiF6u42M6LXK119CQKu5pZeV8futjZE4OmDb/NV2aYeXmTRE6XuukyPJN5yR8bR4ZaibjF6E9mRmYheFfWOGtqdXdG4UOG84MMQSIEmjj32ASA4smhsrCS8DjR2XVF3L84XBCE0iufoaFgwEnYgR49fo8K+YzsjbNIVW1RKemi/hYFU5Lj4ESTGGLDaPbS0u9BERyN6vXg6rHh9fkprLKgEkfAj+eSmSGIvWBdm0kUQqQ2nzWGhraYCd2sb2vj4UB1Gp12qyTIYFLyU/zYri9exv7Gwl2HrF1ul9HPYZOmKMFhXI4qi9DcoCOgCczwnjYylLVKKCqTVOlD4RdqNSlTRbYwIpAT21B+gNULaxlFTww0TL0dA4LOitVicx46WDhYNDQ2kpqbi9XpZs2YN//nPf3j22WfZunXrkD/3d43qt9/B73JxOFtPR4yey0ZLhdoKtZrUq64IbPNun48tr2vnlU+l6G1RlYW3V/cscn5nTRE2h4c5k5IYnx1DfJSBcVkxTGiSIi17I0bwt3f24whE0YJmq1Vfb+GX/9zIrlclX8KE8xdSa3axQSfVRo5v3Eey1keC24xdoaUzume3NEBru4ND5a0YdCqmjBm8yNSxWHLBGLKSIzhvWtqgmhoLghCKhjV/vTE0Au9UI2EAC2dkIAhwuKINvwgzrruYhAukyHfs2Wed8v77Ii0wQ3L1jiosVhcjAh5sAyFlwdkotFp0iQm9DHqjpk4GQSDHUU9NfTuf/PFZfA4HhhmzQubUQ0W/RFgwxH8iSktLT2kxpzOfvfc2733/Fg4e2D/kzxV0ItcF0pGa6GhUERG4mpr67GDsDyUtgeHLzgi+f/kEpmcaUft9uAwGsuKzsEbqUPsgvDoKb2M6oxumoPWIhKWnYQiIwGh9RY9IWHa3Dsm+Ug/BVKQ+NQVNVBR+pxOfras422t34LPbsesU+JUCb+z7iPJA2Dwh2hDq9BstWEiJM3Lh7EzaAvVg0YYoEqPD8bVLwmNP/cHQfoORMGW4hXizFBWKHHlsby9LQJRMTpYiSQea+hJh0jZjMoLF+T3rwlxmSZS1RCkxXbQQAZFJVun7kJw5OrTfoE/R6NhsclJMABwsb+1hU1FZ34Hb6+c8sZKKfz9H2gGprmNfX3Vhu7cBXVEw6IqEYWzB7pFa97fV5BMTMmx10mx2sOtwA3qtkrzLJH+0th27EEURj9mC3+lEGxcXMtfNHRlHW8AaRN8kCSpLuBJdjDlksrin/gBtkVJw3V5dQ7ophbkZM3B5XXxwqMsYcajQ6/U0Njayfv16xowZQ3h4OIIgyG75g4zHaqXxq3X4FQJbJxpYmH02Jn1k6P6E889DHRVFx8FDPeaZguQ/98Qbu/H6/JyVm4xSIfDOmsJQs0tDq41PN5WjVAjcdnGXU37r1m00rlqNoFRiHjeT6sZO/vXuXkRRpFUfjV0bjsrcgvlIEb49O/AjUJU6iTe+PMJ+YzY+rR7bgQPUrZAiouWGJPKLepdHbN5XhyjCrAlJqFVDk4rsjlGv5pkHFnDv9X0bxp4KsYG6sNYtW7qmXwyCCIuL0odSp1PHxHPlOSPIWfoDxv/+t6RePTSG7cFImCUwXm7+5P5HwYJoTCbynnqCCY892mvUkToyEuPIEWh8bm5PaievvRivoODJpmQeX76Tt1YVsmp7JbuPNFLfMrjzsfslwvpb7zV9et/dZ98GmvN3kNRip+DLT4f8uY6OhEnF+ZnAsf28TsSeKklIJxgS0KiV3DRLKjht9RsZ472cWq/0XN9PPQtf9XgUhdIfmnHUSCaMk3L8UUIrmrAug1ejQUN8lJ4Om7tPE1BnwJ7CadLjN0lXLa6WrpRk0J6iU69ArVBhc9v5slKq5UiINqCNi0UbF4toMfP09ycxZXQ8rXZJ/MToTcRHGfBbpIPBroA5KnRFwjJHuolvk66Ww46RjhRFEUunC5VSwcTkkaiVaopay3F6pdepCKYjAz5EoUhYVU8RZmmU0neG6FjGXn8jrm6z7qJS0smITKHTbWNfo3RSGhmTxawJUpRp3a7qUHeou6U15A82yiIJNrGyjIgwDVUNVsxW6X0O1oVZDkjiM2J8dxEmiY5OTVfDws6aAiKMkkBq7XCyekclfhHmTU4lemQ22vh4nA0NOGpqu1KRyV2RgozEcJzxLtyqrrC8JUyDW2XB7GjH4XFyqLmEjmjp/XJUS6L/ugmXolQoWV26kQ7X0Jo533rrrUyfPp2bbropFJnftWsX2dkDL0CWOTbN675G9HgoTdPiDNeyeMyiHvcrNBqSL7sEkEbDdOeVTw5S1WBlRJqJB26cyk0XjsEvwt/f3I3d6eG1zw/j9fm5aHYmyYFIh7WomKInnwZRJPuuO/nJXecSblDz9Z4a/rJ8F/c/9TWHdVLq8Bb7LrSil5KwVP76aRmb99ahMehIukjyoqr7aAUA5YYUdh9p5Gg2FUh/+0NlLvxNoouPJ2LcWLzWzlCUW5swOCnPH1wxkWvPG8kDN01FoRAQlEpMeblDMt8YJFskrUYSTgoB5k0+uc9Hn5KMtlsTWHeip0lTNRK2rkSBSMeks3AYItm8t443vzzCP98p4JH/buOuP68J+boNBv2qCZPns4EYaZL+bevdGTeY+D0eXC0tKHQ61N2mE4RlZdK+dx+28nIiJ0hpOq/NRt0nn+Ht6ABBkIoGlUri58/r5QJd2iqlxsYnSbeb/HYagHZlGJ+sKWKmNorxneUk+9rJSUklbrf0RxY+aiS6xEQaAVOnj+3NG5k5suuklpUcSZPZQVlte69B2kHX/68sB0jwuhmBVJwflimtIWjJYDMo+OmcH/D0tpcotu1HYTSEGgDCx47F1byRjkNHiJvfVZQfY4giLkovRcJEgX0Nh/H4PHTafbR1uIiJ1BGuaEDnFunQG1GF9T0brdPhwesTiTXp0Ko0jInNZn+jFLHKTRzXKxI2Is2EUiFQXG3B6/OHalk6mhtQA8mpObgUGraYxrOgJR+QUgAT/GOobK9FFEXSIpII0xiYNVGDXquioLiZG3Kk1K6rtY2iDhUmj5WwFqn2wVFdTV5eOBsOtbK/pIV5k1NDkTCxVBJa3SNhnXY3CH7aqAAg0RhHQ2czbX5pfy0WB5X1UjTrwtmZOL0uomdMo/7TlbTt2IkqXLrq1Kd01cSIoohgsNIWoSQxIGzbVCbAzb6GgCWH30dG1jgUWguO+gZqP/6E+AXzuX7CZWyq3ImCoS3M/8tf/sLChQtRq9Wcc845AKjVav7+978P6fN+lxBFkYYvVwGwL0fHgqzZvdzOQZr9V7n8ddq2bUdceieCUsmuw418urkcrUbJz2+ailql4KoFI9l9pImDZa08+tIO9pe2YNCpuGGRFD12NjZx+NE/43e7SblyMYkXSoLvZzdO5ffLtrF5Xx2CAPGzZ8IXhSjbpOPWyKsvJbVEoKapk2vOHUl6XgzNn3wS6qCsMKbgKWnB7fGFHOr3l7ZwuKINo149qPVZw0nsvLl0HDqMGDApHYxIGEBKnJFbu0UqhxqFQiAt3khJTTuTRsT1qMcdLKKmTaXqzf8h+nyoIiK46MEfMNevIr+widZ2B23tTlo7nHh9/kGxEgnSLxEmz2cDTZz0pVS1D21ti7OhEfx+9MlJPYoBQzYVZRWAJNYOP/YXOg4c7LWPik3rOOe/L/V4fJu7CVQwZ6R0cAt2KrrDpDTC5Hl58GE+tvIKxuflErVZiliFjxoFWimqE9Hp49OaHVxnvSg0By47JZLtBxsoq2tn+rieI4mCkbA2o4DOLoSet7HNTnunC2u5VOOBKYKpyRO5bvylvLb3fdSZh4iLlg62EePGBOZIHiFu/tyQPUW03iSlRr0asEfhDGvjYFMx/g4pPZmebMBfI0VjapUx1DRZSY3vox09EMGLCjhRj48fzf7GQg40FgZEWM9ImFatJCslkpJqC+V17aFOH7fZghrIyRxPUZWZ3RFjmGkrJVIroI2LZaI4hs+KvgJgVGwOIPkFzc1LYdX2Sio6BSKRauSK6zWM6z4JQBTJNTjYgJSSlERYCgaHH12bDXVkBPqUritDq82NIrwNDy4p4pY6hdf2vs9h80EggqoGyWh1RJqJI7Zd/PqD9/nViIulz2rHzlAaOJgOB6jvbMIveGkN05EYGBVlEeKBGvY2HkajlK6AJydPJGqam9bNW6h46RUql7/O2FkzOfeSqzFqjz0keLA4//zze/x/2rShcW3/rtJx6BCOmlos4UrqE7U8dAy3c11CAmE52dhKy+g4fARfeg5P/08ySP3B4omheh6lQuBnS6bwk7+vY3+pdEy65tyRRBq1eDttHPrjn/C0txMzZzYZt94c2v+0sQn84IoJbD/QwC0XjWVUSjg7vn4Xn8OBLjmJKVeey2xRclZPS5DS0rFnn0Xz1xsIy8kmIyeFg2WtHCxrZfLoeJxuL/98pwCAJYtG95hjeCYTe9Ycyv/7IqLPh9JgQHmMi9EzgZxUEyU17SyYNvBUZH8Iy85CHRWFx2wmfcn1qMLCMCGZ1A4l/RJh8nw2MCZJUQHdSdZk9ZdQ7j6pp6AxZmcCAed8v5+ip56h48BBdEmJpF57NYgiXxSuI25zIZHNHbTu20tsbh4ANa1t+FUO8GqYlCn9AQc9uxYtyuWsibMYF6di14cvYauoZMJlEbjdFnxKNYb0NBrb7PgQiOwU8ft9vHdwJT+edTvAcZ3zO2ulyEtnuBq7XQQctNVV8c8XttDYZue6uAOkA0mp2QiCwEWjFvDWrjV4De2UOAqYRXJIEHQcltJ4bYFIWKwhCq1aicmopbM1DnVYGxsqt5Nok4z3IhNsKMqk2qgGbQxb9tVz3cLeIixYlB8cBzIhvqt+C7pqwvzOrnTrmIwoSqotFFaaGZkWhc/vQ9kpibQx2ZNYubUJr0JFyw0/Zv45OShUKsbFjUApKPCJfkbHdkUSz52Wxqrtlext9DAPcDQ1U1WvZ1GnJMIiJoyn48BBUtwtgIF9xdLnlhKeSHqzdHUbPn5cD8FtdXhQRktRyFmpU5iZNpnX9r7P7oZ96HXzcDgle4lzpsfz7oEXANiubWWywYC1sKtIunskrNwsRdxadeGAJMKMcWOBGvY3HEYRaN2enDSe+J/Po/3CRTSuXkPrlm20bNpMy5atTH3+X4N2Jd4XTU1NPPbYY+Tn5/eaY5ufnz9kz/tdouGL1QDsH6FnTsZ04sJijrlt7JzZ2ErLqF67kWds9Vg6XcyemMSimT1d2uOjDfzoyon8+/WtZGmczHKWUvbCBix79+OoqSF89ChG3v+TXvYAl8/N4fK5OaH/R02bQsvGzSReeAGCQoESQm7rAKnXXo21sIikSy5iKtL8xN1Hmpg8Op43vyykvsXGmIwoLjn725O+VkeEY5oyGfPOXegSBseHbLi45aKxTB4dz5yJvRsqBgNBEMj50VI6i4pIWLRwSJ6jL/ptUfFdJyEjCx9gtLlOuO2p4DzKIyyIPiUFQa3GXl1N+bKXaN28BXVkBON+93/okxI50FjIZ62tTB5lYN6eTgo/ei8kwjYE/L7CFTGhOWFBu4jYzBRGBYos1ZGRuFtbybTWUIFIjTYGv6Cgqd1Jh9pIlMdKlFvFxqodXD7mfNJNKeSkSiKstKYPEVZfhwrIm3AWkcZW2LWJwsK91PoTQOmhvVFKkY7IljoXVQolYs0EyNrMqsrVXDL+LExpaSjDDNgrq/B22roiYQYpAhUXpae9KR5tRjGbKnegEwtRRGXh1RlIDaTNGrTRNO+r47qFvd2ZzQG7hqhwKeKVE52BXqWjzFyFzW3v0R0ZZHRGNJ9uKudIhZlLz4aypnJ0Lj9urZIIYxSFlZKAGzk6BY1Jen90ah2TEsdxoPFID5focVnRJMWEUVkjpUTa65qIdYUT7W5HGx9P8uWX0nHgIIqaSmIiJ1PfaqOpzU58tIHRrYHi0lE9R6Z02J0oo6S0zMy0ycSHxZATnUFpWyXhcVYc1WHotSradAdxeCVxubelkAVT8mjdtCU0Kqp7JKysTRJhnWHxQD0eQcmYMaNxcIDqDulvNiU8kQSjFDE2TZqIadJEPD+w0vz1Bjzt7T284IaC66+/HrVazZIlSzAYZGPYwcbT0UHrlq34FHA4S8fvRx//RBUzexaVr71B3YbN1KcnMj4nlvtvmIwgCHjtDqpef4PO0jLcbWY0ZjP3BcpZqrr6YtAlJzPm179CqT32zMQgWd+7g8iJE0hYeF6f9xvSUpn6/L8AmFrbzvKVh8kvbKSoKoWPvy5BpVRw7/WTUfZjluKZRPyC+Zh37uoRLT8TiTRqjzmjc7CImTmdmJnfbG27LML6SXpSMge1AgaXD5/DETo5DzaOozzCgghKJWEZ6XSWlFL/2ecotFrG/t9v0Ccl4vf7WV4gtWV7p43Bu3cXioIjuC0WNCYT+6pLQYD0yK4vYTAdqY3tGi8RlpWJpWAv7V+vB6BGE0N5XTtNbXYsKkmEnW+ayDuOAv7vqye4dsIlXDDiHCLCNDS22emwuUNjMBqaqlE53Nh1Cq7IuxSvqYbStzfhNbciJHSiCDdjqJZEkilRWpfD5aWzORxDVApOUy3bqvO5aNQCIsaMxrx7D9aiolBNWHDeWHyUgeLqMK7OuJEd5nVUtteiHVlAfitMaZMO6s7oJBpr22lotZEY0zMcf3QkTKlQMjZuBPn1BzjUXExyN5+wIMEOycIqqaPrcPleYgAiwhBFMWRfMTqjpynhvbPuwOq29YgeCILAedPT+CQwWcDZ0so4v3TFHzfvbMJHS4LNWlzMpPMXsi6/ln0lzSyckUFSvSQMW1MjezxPk6sGweAmyZBCfOC5ZqdNobStEiGqAapzmDPFxJryL1EICsI1YbQ5LAgTZsMmyU1aUKnQxXfVxZRbpNRuZOY4fAX7aNRGMzEnHp9tXEiETU4az9GoI8JDRdpDTX5+Pi0tLaiHqDj4u07T2vWIXi8lGVpGZIwjM+r4aZpqn54WXRSxTjPnJ/tZetfskPlp9f/epv6zro5ZQaVCExuLPikRQ3o6how0DOnpGHOypfFh/UATHUXiBYtOvCGQlRxBdISW6sZO/vraLvwi3LRoNGkJvaPlZzoxc2Yz6oGfEj5m6OZ5ypw8347E9zdAQoSJjjDpAGJvGLzOiKMJRcKSeyv+YF0YCgWjf/lAyL9kfcU2Kiw1pIQncv/5P6E0Q4/CL1Lz5ZeIokhVu9Txk5vWFWYPibBuM74MmVJExRwYbluni+VgWSuNbQ4saungNFufzfzMWTi8TpYXvM+vVj1GUoYkUEoDTsMAq7ZJztbERhFriCYmRdp3uM2HKqUEZWwtRruUFgsO7w56hJkI2C+0S+nM8LFS0XlrQQE2t51wrTFUgxSc2WbwJvP7c36Ju3QSuAyE2X0YXCLa+DimTA446++r6/WeHl0TBjAhIZCSbCzsVRMGUuemyailodWOxeqiolKKNOqiYqhrsWG1e0iJMxJu6HnyCNMYSDT2LvhdMC2NTnVA1HdYGBdIRcbOm4vGZEIbH4/HbCE3XrpmemdNMau+yEdrsWPVK6jS9JzJ2IL0+LyELqPMWalSh7NLX4PRoEJMKMbj87Agaw4zU6X2+NIkFQRSPrrEhFAbt1/0U2auQkBg8sTJvJFyIZ8lzWN0RhS5iV0NAVOSJ/R6bd8kM2fO/Fbb5Awn3Qvy94/Qc9no84+5rdPl5bNNZfzmuc0cNkipx0tMlpAAc9Q3UP/Z5whKJeP/+Agzlr/E7HffYvqLzzPh0d+Tfdf3SbxgERFjx/RbgA0UQRCYMlrqEG9ss5OdHMlVC4bWD2q4EASBuHlnD2kpgMzJI4uwfqJUKrEapBN/4ymODzoeXTVhvfPeMXNmoYqIYORP7g610x6oaOB/+yXBc0ve1YRrjbhnS+m9ui++oLbJilslRY8mpUgizO/x4DGbURoMPboGwwIijMBA7DpdHAdKW2gy2zGrpUJaX3Mr98y8jT+c+wAZkSlUd9RTZVyFZtRuNpbtwe/302RrpahIEnKJWdLVlyoiAlGlwujwo46qRxlu6SbCpDRVozlg1GqQ6uGqLD3Hi7Ru2w6iSEwgCgaEuihbGs1U1VnwtSaT0X4534uTxpUYc3JCIezNfYkwa890JMCEeKkO7UBTYSji6bXbKTdX4/V5A6atUpTrUHkzzfWS6WlkfNIxo2DHIz7KwISRCdiUOlReNxFeO5rUNMIypBNY+CjJ42y0soP0xHDqW22sf38dANWJGg41luEXpffS7/dj00hRq1mpeV3PYYwlOyodl+jge7dFsKN+J2qFiqvHX8SkgJDa11FO5Hip46n7319TZwsOj5Ok8HhmjEnFGpPC6NwcdFoVY+NGolfrCNeEMSZ2eE9iy5cvZ+nSpfzmN7/hmWee6fEjc2p0HDiIs64ec7gSxciMHuI7SIvFwSufHuSOP67iPx/ux+n2kTRfMu9s27o95CVY+epriF4viRdfhGnSRNSRkUMyDuZETB0riRKFQuDe6/N6uPbLyHxTnFQ6ct26dbz55ps0NDTwySefsGvXLqxWKwsWLBjs9Z1WdIbpAAfNlZUMxenG53LhbmlBaTCgjozodX/UlMnMWN7V9Zhf2MQfP3sNdUoHuYljQ+mgibPOo3llPnFtHRz4chOCvhMQSI2QTqzBodndo2BAD1sLVVQ0nSoDB8taSUsIxxeIhAU7HsfEjeDxRQ/xZcnXvLV3BS5TM5s6PqJo5UZMukjCrVIqMDJFSlkIgoBTH4He2kaY049dp8Dg9KPQ6VAF6ncaAzPUMkzJFDkUVLXX4vf7CcvKRJeYiLOhgVhLFNEpXQInPkpPuNdGxpt/p/kDLeMME8lKPpe0jjpqgLCcbEbmxBBuUFNUZaG900WksSvqdXQ6EiDdlEy4Jozq9joqBCnqua10O5+s2s/8zFncM/M2RmdEsf1gA9srjqCxSQ0A+piY0HDvMQMQYQDnTkvHvNZAmE+KzCUumBe6zzhqJC2bNuOpKOOZB25j95FGqp7dBUB1vJYicwn/t+ZvfG/qDbh9bkSVE78tnOy4njUgs9KmUGau4qX8/+ET/Vw4cgGxhmgMKj0KQcHB5mJunrmQ9v0HevwtlJklUZcdlU6kUcuLD58fOmFpVRr+eO7PUSgUqJTDW93wj3/8g+3bt+N0OnvUhAmCwL333juMKzvz6REFG3N+rwLvt1cX8taqQnx+SWhNHRPPFfNzyB0Zx56dH+OoqaWzuATR66V16zZURiNp11/zjb+O7kwfl8jsiUlMHhVHTqppWNci891lwNJ/2bJl3HLLLSQkJLBhg2TEp1ar+e1vfzug/TidTq644gpGjRpFbm4u559/fg9n/szMTEaPHk1eXh55eXm8/fbbofuKi4uZM2cOo0aNYvr06Rw8eLBf950qDqMUNeps6B1RGQycgbogXVLSMbtYut++s7gcVVIFogjXjFkcum9ayiSOjJZEk33zKgSlD5MmGo1KCu33VQ8GgeJ/lXQijRwzkrSEcKx2D0VVZiyBSFhwjSDVT1086lweO+d3uMvHoXBFSlGw1jKiOqWDsS5RCvk7XV6a/FK0Kc6lIczhRwCEyC6x0hSIhCXFhJMSnoDb56HB1owgCMTMmQXAiCpXz0hYtIHc9mLUHicKazuXN25i3OpXMO/aDYAxJxulUsGodOl5So/q4rSECvO7RJhCUDA+0CX5ZuFK6UanG4Wg4OuKbZSbqxmbKUXv9tQexuCQvIc0UVEUBkVY5sCK0OdMTMKu7opKxs09O/R7qC6ssAilQmD62ATS7VLEtNk/E7/TQHFbBb9e/Rf+tX259KD23sOHZ6VJEUWf6Een0nJlwF7AoNEzMjoTl9dF+/QRjP7lA6RcuTj0uGBnZFaUFJnTa1U9WvjTTSkhgT+c/Pvf/6agoIDt27ezbt260M/atWuHe2lnNNbiElo2b8WrgPpx8Zyd3rNw+cP1Jbz+xREEQeDC2Zn8+5fn8sgPZpM3Kj7w3Z0NQMvmLZS/9AoAaddfizp8eOuvtGolv759BhfNOfZcWRmZoWbAIuyvf/0rq1at4tFHHw21pY8bN47Dhw+f4JG9ueuuuygsLGTv3r0sXryYO++8s8f9b7/9NgUFBRQUFHD99deHbl+6dCl33XUXRUVFPPjgg9x+++39uu94dHR09PhxuXp3QXojpOiUp7n3uIvBwBlyyk88wZYSe9t2ISj8+JpTKSn2hW7XqjQYz5qJWyUQ11BNuM1HTkxXEW1f9WAgzX0zpEn1WMZRo5iQI9VqeX0i9oAhY3cRFiQ11kSYfQS2vbN4cM59LMyZyziFVPukS5Rey/aDDbQrpNTerRkLmeCQ/Jvc+i7Tu8ZucyPTA47wwZRk8EA+otpFjKFLuMVF6pholcT7wZw5mNXhKOurQsPOjTlSCjY7RSpeLz9KhAUd6LtHwgDOypDWFxkhvQfjIjK4drxUYP763g8YmxXD6PQobKoGwgKWDxgjqKjvQKdRhsZs9BedVkV4ovSeuZIyegzaNWZnIahU2MrK8Xu9OGpr8ZjN6JISGT96Jq79ZzE5Yi5qpYomm/TZ6p29i6YTjXFkmaTbLx19HhG6rjVODKYkmwqJPWtOKDoJUBYQYdnRPa0FTjfi4uLIDBgBywwOPpeL4qeeBr+fnePDOHfiwh4Rz9XbK3npk4MoFAIP3T6de67J7VXcHhv47tZ/upLO4hJ0SYkkXtS3v5iMzHeNAYuw1tZWxo2T6kaCkRdBEAbsP6LT6bj44otDj5s1axYV/RjJ09TUxK5du7j5Zsm47+qrr6a6upqSkpLj3nci0tLSiIyMDP38+c9/7rWNECVFNxTtln6+yoHhqD92PdjReH1eWpTF0u+NGXy+taLH/MbZI2dRmKlFAYwvcfQUYS1BEda7SNw0ZTKCSkXUlMlMyO7q4jPFmlCFh+OxWHp0CoL0+UtDnAUU9mjumnYjOotUyB6MhK3Pr6FDJUV6Ih0wTieJjE5V18k+KMLiowxdsxEDIsw4IgdPpIGYDh/R7d7QY9yHDxDptdOgjeZLzWheylhM6i03ozQYiBg3FnWkJL6CIqz7sHGfX6Td5kajVqLX9kylzUydzCtXPskvF0ppLKXHzyWjzyNKF8n+xiMcaDrMfUsmoTRaMDik973Jq8LvFxmZFoXyJOpLchdIwm/kVZf2uF2h0RCWmYHf7cZeUUn7PsnkNnLSJKaOiQdRibs2m39c/Aizk2fiqcsiUtV3JO6OKddzyajzuPyoUTPBGp/gWKUgoihSHkhHZpqGxiRxsHj44YdZunQpJSUlvS6qZE6OyuWv46itoyFGxb7cKM7PmRu6b8u+Op59twCAn94wmRnj+r54NGRmoEtKDLm2Z95+65CNt5GROdMY8JkiNzeX999/v8dtK1as6Pd8yWPx9NNPs3jx4h633XrrrUycOJHvf//7NAd8raqrq0lKSkIVSJsJgkB6ejpVVVXHve9EVFdX097eHvp56KGHem2jjwsM1LYOzQw851EzI4/HhrJ8ULkQ7NEkGBKobLByODAEF2B01Ej2ZUtXpOPLnKR3SxcFPcI0R6UjATJuWsL0l5cRlpnBhJyu+xOiDV2pxcbe3aEjAjUVJTUW/F6vNHpJq0VtMtHe6WJPYRO2QOTF3dJCrCAJuWaxKwIVEmHRhtBsxGCHpCAINI2QRKHxUHXXY1atAWBvxEjcHh/JiZFkXHMlM19/hfF/fCS0XXDYeHk3EWa1ufH7RaLCtX1eRBg0epR6SST6HA50Ki3XTpAE0ut7P8QiNoDCT1in9Nhd1cH5kgOrBwuSefH5TFv2H9IWndvrvu4pyfZ90hB506QJTBoRh1IhsK+khQh1JJdlXoG3ZjRGQ99dZWPicrht8jXoVD0jfyOiM9GrdZS1VWHtNuOx2dZKp9tGgjGOMM3p7b31/e9/n9dee41Ro0YRFRVFVFQUJpOJqKiT+zy+61gK9lL/6UpEtZIvZ0cwJ2Na6G9gb1Ezf3t9N34Rll45kXOmHtuuontKMmL8OKJnzvhG1i8jcyYwYBH2xBNPsHTpUq6++mrsdjs33ngjd999N3/9619PehGPPfYYJSUlPaJPGzZsYN++feTn5xMbGxsayDtURERE9PjR9mEOGB0Vi10noHV58Nodfezl1HA2SC7n/YmErSqW6vESGcuFszIB+GJrRej+lZuraPCn0RQldSTGVVtD9x0rHQmSH5k6QhJL0RE6kmOl6FVCtAFdQkCE9ZGSHBHwqiqpsUgiz++XbA4Egc376vD5RVJGpIWeP9wnvX+1ThWiKNLp8GBzeIgK16JVK8mIlKIuVZaa0HOUpUsnAP8eKVrjbjPTtnMXPqWaQ0apriMrEPESlEoUqq7oVmJMGDqNktqmTpxu6Yq8r6L8oznarHVB1mxSwhOptNTwyp53AYgIjExddcgCDLwoP4igUPQZnQSpOB8CIuyAFAmLmDCBML2aMZnRuD0+Dpa1YrVLTQLhhoFFGpQKJRPiRyMihqYFeP0+/rv7LYAeLv+nK+Xl5aGfsrIyysrKQr/LDAxvp43iZyRj093T47BEqDg3+6zQ/f96fy9en58bLxjDpf1wmE+5YjHJV1zOyPt+cka7tsvIDDYDFmFTpkzh4MGDzJ49mzvvvJPc3Fzy8/PJzc09qQU88cQTfPDBB3z++ec9OprS06X6E7Vazf3338/GjRsBKW1YX1+PNxDaFkWRqqoq0tPTj3vfYJAUGR3yCnM1Db5XmOsYXYtH09TZQpm1FNGrYlLsBM6bno5KKbBpbx0dNjcdNjfvryvG15rEkSypGN69rWtsy/FE2NGMD6Qku0fCXI29RVhOKBLW3tVgEExF7paEVN70UaHnFzssALT4NFisrpBHWHy09DcQpY/EqAmjydaK3SMJoOJIN516BZ7qOhy1dTR+tRb8fjqyxuNSSpGfYMTraBQKgazkSPwioeHVfXmEHY1SJ90XHFukVCi5MfcKAKrb60AUMbg8+AQFToW0hlEnKcKOR9CmonXrNrzWTgwZ6SE3/qljpNRufmETnXZJER7tUdYfJiUEUpINRxBFkWW732JvwyFiDdHcNOnKwXgZQ0pGRkboJywsrMf/ZQZG2X+X4W5tRTVuBJszfKREJDIyRrrQaWi1Ud9iI9ak54bz+2cAqo4IJ+uO23rUOsrIyAxQhHm9XpKSkoiMjOTnP/85zz77LA8++CDJfRiL9ocnn3ySt956i9WrV2MymUK322w2LBZL6P9vvfUWkydLhpLx8fFMmTKF119/HYD333+f1NRURowYcdz7BoOUqKEVYe42MwgCmm7vRV+sLd8MgK81mREpsZjCtcyZmIzH62ftrire/aoIu9PLrOzxxM+bhygItG3bgc/hQBRFqSZMoejXGJkbzh/NpWdlsXBGelc6MhCx606cSU+kUSO561dKKURdQgLlde0crmgjIkxD7rSACGtpxt0qpU6tKgPl9R09ivJBSmFkhIrz63D7PFjdNiozpMhcy+YtNK6WUpHilDmhdRxLhEH3urCACOvDI+xoBIUChU6Hz+kM1dxNS57EmMAQbqNPCT4fmqgo9DoV2cmRx93fyaJLSkIVbsTvliJdkZO6jFinjpE+l91HmrA6pPuPlY48HqG6sIZDfHj4C9aWbcag1vPQvHuI0h/7fT1dsNvtLF26FIPBQEJCAgaDgR/+8IfYbIM77/VEnd3dWblyJVOmTCEvL48JEybw6quvhu5btGgRkyZNIi8vj7lz57Jnz57QfcfrDh9qXK1tNK/fgNJgYM+CDBAEzs06KxTB2lMklTNMHhUnR7VkZE6RARn7qFQqjEYjbrcbne7UTjQ1NTU88MADZGdnh/zFtFot27dvp7Gxkauvvhqfz4coimRnZ7N8+fLQY59//nluv/12HnvsMSIiInj55Zf7dd+pkhwVExJhjj7qok4Fr92B3+lEHWUKOZX3hc/vY13ZVukxTalkJUsdmxfOyWRDQS0rNpZhsbpQKARuvWg8KXEzObSxHvPuPbRu207UtKn4nU40MTE90nXHIj7awNKrpBO+7zjpSEEQyEk1kX+kiaZSqQZPl5jI/76WHMwvnpOFLkyPKiICb0cHjjrJ5sOqCqOiriM00zIowkAas3SwqYiq9hpMeul1to5OgCNWaj/8GJ/djiE9DcaPgXzJNyszube/WpDgsPFgh6TFeuJ0JIBSp8PvdOJ3uVDqdAiCwM25V/G7tX9niiELaEAfE81/frUQjWrgBfn9QRAEwkeNxLxbOlGbJnW502clRxAVrqW60RoapD7QdCRAgjGOuLAYmm2t/G//CpSCgp+fdRdpkUM7r22weOCBBygqKuKrr74iOzub8vJyfvOb3/Dzn/+c5557blCf66677uKiiy5CEASeffZZ7rzzTtavX99jG1EUufnmm1m/fj2TJk2ioqKCMWPGcNVVVxEeHs4777wTuvj88MMPuf3229m7d2/o8W+//TZ5eXkDXtvRjQharbbP8opj4Q407uizM9nYcQSloGBeZlcd196ACMsd2XfqXEZGpv8M2F3xN7/5DXfccQePPPIIaWlpIZsKkOqq+ktqamqPbr7uZGdn97gqPJrRo0ezdevWAd93qoRrw0IizFpbP6j7drdJkSHNCYqI8+sPYHa247dFovaYSImTLB4mZMeQGm+kpkkqqr5wdmbovrhz5mPevYfm9RswBFIz/UlFHk1XYX5vEQZScX7+kSastfXoAZcxiq/X16BWKbjkrKzQ83o7OvCYLYgKBTaljor69lDkprsICxXnW2q7fKiyU1FHNuNpl8RGwqLz6Qw8JjZS18OI9WiyU6S/z6BQCdpTHC8dCaDU6/BYpLowZeDiY1RsNk9d/AgUVlDKVjTRJqIjBj8C1h3jqFGSCFMoiAg420NgBMuYeL7aWc3mvZK4PZlImCAI5CaMZU3ZJgCWTr+ZCQljBmfx3wArVqxg//79RAcivAkJCbzzzjtMnDhxUEVYsLM7yKxZs3jiiSf63FYQhFBUv6Ojg5iYmJAg6h79b29vH7SoUlpazyL53/3udzzyyCP9fry7TfK6s2h8eHweZqTmEamTvjs+v8jeYlmEycgMFgMWYd/73vcA6coteNAQRRFBEPD5fMd76BmPQlDQadADVjrqBleEeczSge9EKcI1pdIJ0tuYRlZSRMgKQRAELpqdyX8/PoBGrexRqxE9cwZKvR7Lvv1EFxYBJyfCtDExCCoVzsYmRL+/16iRYHG+r0WKEm6uduHziyyckR6KNmljY7GVSoXSqkgTCAIV9R3EmSQhFRxDBIRsKqostaGRONFh0UTPmknjl6sQ1GrizplHvCGM2ROTmJhz/NeUkRiBQiFQXt+Bzy/2qzAfjirO7yaSE41xNHVIZsBq09B34AU7JI0jcnqMmwKYOjqBr3ZW0+kI1oSdnAXA3MwZfF25navHXcQ5WbNPbcHfMKIo9rgoBFAoFMe82Bss+ursBuk7+fbbb3PVVVcRFhaG2Wzmgw8+QNNtHuKtt97KunXSCKqVK1f2ePytt96KKIrMmDGDxx9/nLhjNG0cTXV1dY8L4oFEwQDcgWNRhd8CwLlZXen+0hoLnQ4P2cmRJ/zeyMjInJgBi7Dy8vKhWMcZgzPMCDQNek1Y8OpTE33sk3mLrY2ChoOoBQ2OtkSyp/es01k4I519JS1MH5dITKQ+dLtSqyVm9iya1q6j7iNpzuTRbvn9QVAq0cbH4ayrx202o42J6XH/iNQoEEW0VgsIAp8dlCJOi+flhLbpLv708bGoBQXVjVbcHsnwNCGmS4SlRkiTAyrba2mxS5HCGIOJ+AVTaVy1mvgF80Ou27++/cRt7xq1krR4I5UNVuqaO7u55R8/ghWMfh3tjwbgNlukfR/ncxssTLmTyLzjNiK7pSKD5I2OQyFAYGoM4fqTG3w8Nm4kr131j15i5kzg0ksv5ZprruHxxx8nIyODiooKfvOb33DZZZcNaD+zZ8+muLi4z/v27NnTI9IU7Oz+6quvem3r9Xp59NFH+eCDD5g3bx47d+7k8ssvZ//+/cQGvn/BMotXX32VBx98MCTENmzYQHp6Oh6Ph4cffpjbbrutl0g7FsEO75MlKMLqBBtR+iRyE7uirqEo2Cg5CiYjMxgMWIR91zuNvBGS8BHNbSfYcmD0Jx254shqRFEkXjmCDr8qVA8WxKBT8/D3Zvb52LgF82lauy5Uz3UykTCQiu2ddfU4Gxp6ibBYk44knQ+134PHGInVLTJjXGIPB+3u4k8bE0O6JpzSmnZqmzsRBKnAP3S/SkOSMZ46ayNHWqTasmh9FBHZY5j6/L/61VhwNFkpkVQ2SLVTx3LLP5qjbSq6E4pgnqCZYjAQFApSrri8z/vCDRpGpUeFZlcaTzISBpyRAgykRp97772Xs88+G4/Hg0ajYcmSJfz9738f0H76W84Q7Oxes2ZNj87uIAUFBdTV1TFvnjQHdPr06aSmprJnzx7OP//8Htvedttt/PCHP6S1tZWYmJhe3eGjRvWvC3EwCF4Q2vQKzsmcjVLRVaNaEKgHy5NFmIzMoHDS6ci+eOmll05pMWcCBn0kNp2CMKcTr93eY7zLqeA+QTqypLWCL0u+Rq1Uo2odCbhCheb9IXLCeDQxMaHh3ZrYkzuIdnVINhI5fnyP+wRBYLJGKgouR1rblefk9Nimu/jTxESTaYygtEaKmMVE6FCrejYlpJtSqLM2hryrYgLjk4KeZQMlJyWS9btrKK9rP4l0ZB+RsEC9jzrKdFLrGUymjEkIibCTsag40zEajbz00ku8+OKLNDc3Exc3dN17wc7uNWvW9Kjt6k7QMufw4cOMHTuWkpISSktLGT16NBaLBbvdHuos/+ijj4iJiSE6OhqbzYbH4wntt3t3+DeBK3BBaNMrWJDdlYp0ur0cKm9DrVKErGtkZGROjQFf8nYf7RMZGYndbue9995D/R0ZQ2HSR9IRJr1tg5mSDEXC+khref0+nt/5OiIi1467hJpqKXV3dCTseAgKBXHzu0aOnEokDLrc/Y8myy4VhpfpkxmZZup1sO5uRqqJjiYzqUtIxkf3FrQZkVJxvscn1Tp1nxt5MgSFa3G1hQ6bG71WhU5z/GsRRTAd2ZcIC6aRTwNX9qBfGJxaJOxMZeLEiYB0MRAfHx8SYCfTYXg8gp3dFouFBQsWkJeXx8yZXRHoO++8kxUrVpCQkMALL7zAddddR25uLldeeSXPPvss6enptLe3c8UVVzBx4kRyc3N59tln+fTTTxEEgcbGRhYsWMCkSZOYOHEiX3/9dY/u8KHGFqjpjElMJdHY9X09VNaG1+dnbGZ0r+HwMjIyJ8eAI2FPPfVUr9vWrVvHf//730FZ0OlOTFgkHWFKklq9OBubCRukgcHBk7m6j5P5p4VrqGyvJSMyhVkJZ/Gi4ysSYwwYdAM70cYvmE/tBx8BJy/CjCOkyJZl734ybul5nyiKhNdLRfflhiSWnjOiVySi+6gkbWwMWUldQjKhLxEW6JAMEqMfHBF2qLwNUTxxZyScIB0ZiIRpToNI2IhUE8mxYfj8Yq9ZmN8FjjV7tj9jywbC8Tq7AZYtWxb6fcmSJSxZsqTXNhkZGezYsaPPx5+oO3yo8ZjNCEByas8o9p4iSZzJqUgZmcFjUI7U55xzDldcccVg7Oq0JyE8iirj4Bu2dkXCeqYjGzqbeffgZwgILJ1+M1WNkgXFQFKRQQzp6SQsWojf5UZlNJ7UOiPGjUUVHk5ncTGulla0sV2RLkdNLXRYMKvDMSQmMmdi7/FLGlMkglKJ6PNJ6chuIqyvSFh6t6HRerUOvfrUbCAiwjTEmvS0BAaM96fDS6kPRsJ6i7BgGln9DdSEnQiFQuCJ+6T6o++SiebPfvYzADweT+j3IGVlZWRnn/4jl04XRJ8POu14lJCZ0PN9CxblTx4lu97LyAwWAxZhRxsB2u12Xn31VRITEwdtUaczSaZoDgyya74oilKXnUIRGkUTvP2/u97E4/Nw0cgFjIjJ5O18qTaqewRpIIy450entFZBqSR6+jSa1q6jbft2ki7p8kuyFBQAkDxrGn/6/lkh+4yjH6+JjcHV2IQ2JgadUUt0hJa2DhcJUb1FWJwhGr1ah8PjPOUoWJCclMiQCOuPu30wEuY/qjvS53Lhs9lRhRtRnCbp+O9iLZg5IIT9fn/od5AaDMaNG8eTTz45XEs74/C0dyCIYNMrGRvVNe7NbHVSXtdBuEEdms8qIyNz6gxYhJlMph5X2aIokpGRwYsvvjioCztdSTHFhGrCnIMkwnyOvt3yN1buYH/jEWL0UdwwUeqKKw+M3BnOA2H0rJk0rV1H67YdPUXYHsntO+Ps6cRE6Y/1cDJvvRlbeQXaQH3ZyLQoth9sIC0xvNe2giCQEZnCkZbSUFH+qZKVHMn2g9LopVOJhHWlIoe/Huy7THAqxpQpU/jJT34yzKs5s3G0Sm75dr2CtMiuSPbeYun2SSPjUCq+O1FWGZmh5pR9woxGIzEx351OmWhDZNfooobBEWF9pSJFUeTjI6sAuH3KtaE0XFmd1El4MunIwcKUNwmFTkf7gYN4OqyoI8Lxezy0HzgICgWRE3v7WHUn9uyziD37rND/l145iXOnpTEmo+/O0HSTJMKiBykSFnTOhwHWhB0VCQvV8Z0GqUgZeggwq9Xao27rVHyzvks01ErHdzEiDLWyK7q7t9u8SBkZmcFjwN2Rf/nLX8jIyAj9BAXYj3/840Ff3OmIQa2nI9B5NliRsL6MWotby6luryMuLIbpKbkA2J0eGlpthOlUxB8n0jTUKLVaoqZMBr8f8y5pZqP1SCF+l4vw0aN6ubmfiLgoPXMmHXs+4cTA6Jyc6MHxqMtOMYV+N/UnHak7QSTsGzBqlTkxVVVVLFq0CIPBgMlkIioqKvQj0z8a6iQRpj3qgrBAHlUkIzMkDFiEvf76633e/uabb57yYs4EBEFAoTRg0ykQ7Xa8Ntsp77NLhHUd+L4q2wxII0MUgvQxVdZbEUXITI4c9sLrmFlSS37r1u0AWAqkVKQpL3fQn2tGSh7/uOh3LMw+e1D2Fx+lJ0wnBYFPpTtSjoSdXtx9991ERUWxZcsWjEYj+fn5XH755Tz//PPDvbQzhvZGyWImPK7Lh6+13UmLxUFclJ7EmIFdYMnIyByffqcjV6xYAYDP5+OTTz7pEeovLS09pmHhtxGdwkBHmIIwpx9XUzOqrFM7MB3tlm/3ONhStQsBgXWr4YuPV+Ny+7C7vABknwaFsVHTpiCoVFgK9uJzOjHvGToRJggCyRGD1/ghCAKjM6PJP9JEUuyJP7tjmbW65UjYacXWrVuprKzEaDRKw8hzc1m2bBlz587lzjvvHO7lnRHYW5uJAWISu0YzFVZJFxuj0uW/cxmZwabfIuy+++4DwOl0cu+994ZuVygUJCQk8Mwzzwz+6k5TwlRGOowBr7CmJsKyMk9pf13Du6WD3ObKXbh8bnTOZCqrPYAntG24QcNZx0ndfVOowsKInDgBy54CmtdvwFZWhjLMQPjIEcO9tH7xk2vzqGmy9hipdCyUxzBrDUUwv4Hh3TInRqlUhoZVR0RE0NbWRmRkJNXV1cO8sjMDv+jHa5Eaf5KSs0K3FwdFWJr8dy4jM9j0W4QFC/Kvu+463nnnnSFb0JlAhDYiVJzvajz1urCj05FrA6nI9qpEwg0anv3FAvRaFVq1EsVp1JkUM3smlj0FVL7xFogipkkTe3R3ns7EmvTEmvpXV3esdKTnNBpZJCM5469du5YLLrggFP0yGAyMGzfuxA+WobGzBZ1duuAL65aOLKqyADAq3TQMq5KR+XYz4Jqw77oAA4g2RGDTS2+dp739lPfXPR1ZYa6m1FyJXmHEb4ll1oREoiN06LWq00qAAUTPnAGCgDfgHReZO/ipyNOBkEXF0d2RwQimXPh9WvDiiy8yZozUxPH0008TExODx+Ph1VdfHeaVnRmUm6sJc0gj0YJReZ9fpKTGjEKQJjLIyMgMLgO2qHC5XDz55JOsX7+elpaWHrVh+fn5g7q405U4o4litSSIvDb7Ke+veyRsTdkX0u/WDEDBWbnDn3o8FhqTifAxo7EePgJA1ORvqwg7RiTMbAFkEXa6kJbWVccUGxv7nRmlNliUt1WS6vQjKhWowqU0fU2jFYfLR2ZSBLrv4CgsGZmhZsCRsJ/97Ge8/vrrXHzxxRQWFnLbbbdht9tZvHjxUKzvtCQxMhq3RnrrfPZTE2GSW74ZFAr8YXo2Ve5EQKCpJIYwnYpJI07vlvBgl6QuMRHdt3RqgqBUotBo8DkcoYsO0efDbbEgqNUow3o7/cvInGnUNJSj8oMiIjzUfV0kF+XLyAwpAxZhH330EZ999hn33XcfKpWK++67jw8//JD169cPwfJOT5JN0biCkTD7qVlU+Ox2/C4X6shIttXtwe5xkKTNwO8yMHNCEmrVgD+ib5S4c+ZhHJFD8hWXD/dShhSFTgd+P6JHqpnxWDvB70cTFTXsdiEyMqeKKIo01UuDzvXdzLe7OiNNw7EsGZlvPQOOL9tsNjIzMwHQ6XQ4nU7Gjh3L7t27B3ttpy0xBhPugAhzW09NhHVPRa4s3wqAv0VKq/Q1APt0Q2Mykfv3vw73MoYcpV6Ht6MDn8OBQqPBWV8PyKlImW8HbQ4LQnsnALpuIqw4VJQv/53LyAwFAw6zjBw5kr17JU+oiRMn8tRTT/Hcc88RGxs76Is7XYnURXSJsM5TFGGB4m7BFM7h5hJ0Kh2VhXr0WiWTR8ef8lplBoej68Ka138NgOlbWgd3puH1evnFL36B86jmCZn+UW6uxuDwAV0XFk63l4qGDrQaJen9sHKRkZEZOAOOhD322GN0dkpXTH/+859ZsmQJVqv1O+VKrVNpcaul0UXeU6wJC3ZGmtVeRERSNDmY/Qqmj0tEoz4z7B6+C4S8wpxOfC4XzRs2gSAQf96CYV6ZDIBKpeKll17iL3/5y3Av5Yyk3FxFmFPqjFQHRFhpTTt+v8iIVBNK5eldFiEjc6YyoG+Wz+fD7XYzbdo0AKZNm0ZxcTENDQ3fqcJ8QRBAIzmt+09ZhEmRsCokmwdnk1SIf7xZijLfPN1d81u3bsNnt2PKnYQuXo5Wni5cf/31vPHGG8O9jDOScktNL3uK4mq5KF9GZqgZUCRMqVRyww03YLVah2o9Zww6tRGPElQuJ6IonnRxdjASViGa0alMlBdq0GqUTB0jn9xPJ7qnIxtXfwVA/HnnDueSZI6iurqaZcuW8dRTT5Geno5C0XWN+cEHHwzjyk5/ys1VzAiKsEAkTDZplZEZegacjpwzZw67d+9m6tSpQ7GeM4ZwdThutQK104/f6QydpAdKMBLWqVOQqs1hv0/BzPHx6DSyJ8/pRDAdaSuvoOPAQVRGIzGzZgzzqmS6M23atFCUXqb/dLg6abWbiXBJF5JBEVYojyuSkRlyBnymz8vL45JLLuHGG2/sdbXZfabkt50IbThutUCYU6oLO1kRFpwbadMr0LdIo0JOh9mQMj0Jfr71KyUz3bj581BoNMO5JJmj+N3vfjfcSzgjqTBLszWDIkwdFYXF6qKpzY7JqCUu6uSObTIyMidmwCJsx44djB07lj179rBnz57Q7YIgfKdEmEkfEfIK89ns0K2teyA4W1sA8Bh1NB/RohBgypiEEzxK5psmOLrI3SJ9XvEL5VTk6ci6det48803aWho4JNPPmHXrl1YrVYWLJAbKI5FaVslgDQ3UhDQmCLZV9gMSPVgsg+ejMzQMWARtm7duqFYxxlHUngcbk1AhB01zqa/iKKIq82MKEBqwljqC2BkmgmjXj2IK5UZDLpHOsNysjFmZw3jamT6YtmyZTzyyCPcfvvtoRm3arWa3/72t2zcuHGYV3f6UmauQu3xo3B7UZtMCEqlbNIqI/MNcVJ9x2azmTfeeIO//e1vANTV1VFTUzOoCzvdGRWbiUstvX1e28l5hflsdgSPF7tOQbg3A4BJI747fmtnEsFIGEDCwvOGcSUyx+Kvf/0rq1at4tFHHw2VSYwbN47Dhw8P88pOb8rMVV2dkYF6sKBJ60i5M1JGZkgZsAjbunUrI0eO5LnnnuMPf/gDAIcPH+aee+4Z9MWdzoyIT8EVGCnk6jy5blFzoyRc7QYlLVURAEyURdhpiSJQmK/QaIibN3eYVyPTF62trYwbNw4glEITBEFOpx0Hq6uTZlsrcT4tAJpoE6Iods2MTDMN4+pkZL79DFiE3X///SxbtoxNmzahUknZzNmzZ7Njx45BX9zpTJhOi1shHbiaW+pOah8HivMBUEWZOFLRgVIhMC7r5GrLZIYWXaJUpxd79lmojGHDvBqZvsjNzeX999/vcduKFSuYMmXKMK3o9KfMLM2LzBRMgFSU32x20OnwkBQThtEgN5/IyAwlA64JKyoq4oorrgC6rjYNBgMul2tQF3Ym4BUMgIXG5nomnsTjS8sPMBbQRCbgbvcxNjMavVa2pjgdiRg3jol/fpSwnOzhXorMMXjiiSdYtGgRb775Jna7nRtvvJG1a9fy5ZdfDvfSTlvK2iQRliRKFxaaqCiaLVKNa0KMYdjWJSPzXWHAkbD09PTQ7Mgg+fn5ZGV99wqVRZWUQrSYmwb82B01BbQ2SK3hboWUgpRTkacvgiAQMW4sSq12uJcicwymTJnCgQMHmD17NnfeeSe5ubnk5+eTmyvP9zwWwUhYjEdqBtJERdESEGFxJtmaQkZmqBlw2OWhhx7isssu4xe/+AUej4cXXniBv/3tbzz++ONDsb7TGpVGKlq1WS0Delybw8LzO19nsl0qhq22Sx+DXJQvI3NqJCYm8vOf/3y4l3HGUBawpzA6RTqQRhYFRVisLMJkZIacAYuwG264gYiICP7973+TkZHBhx9+yFNPPcWll146FOs7rdEbpDmPHnsnfr+/h3HtsfCLfv69fTlWt40MIRJwUGIBlVHBmMzooV3wtxiv14vP5xvuZcj0A6VSGaonHUy8Xi9//vOfee2116itrSUlJYWbb76Zhx56CLVatn05Gqurk2Z7GxFaI0KH1OGtjoqipV4SYTGRsggbDuRj2ZnDYBzLTurRF198MRdffPEpPfG3AZ1RSkeq3V7qO5tIiUg84WM+L1rHvsbDxBiiSMWPnQY6FFrGZEahVSuHesnfStrb21GpVENyYpcZfFwuFzabjcjIyEHd73333ceWLVt49NFHycjIoLKykscff5zGxkb+9a9/DepzfRsIpiJzojNwm4uBQDqyvRyQ05HDgXwsO7MYjGPZSX3SmzZtYvny5T2uNufNm3fSizhT0UWGA6D1iJSbq04owiotNbyx7yMEBH484zbsHz0JQKfSwHkj4oZ8vd9GvF4vKpWKsDC5Y/FMQavVYrPZQp/dYPHOO++wf/9+EhOl7+HMmTM5++yzmTRpkizC+iDolJ8VlY7HvBMATZSpWzpSd8zHygw+8rHszGMwjmUDLsx/7rnnuOSSS1CpVMybNw+1Ws3ixYt57rnnTmoBZzL6yEBHkVukPDB/7Vi4vW6e2fYyXr+XxWMXkd6hxN3Sgl0Xjk2pk+vBThKfzydfNZ6BqFSqQU+5GI3GXiewsLAwwsPDB/V5vi0EI2FZxmS8nZ2ojEYUGg0tFicg14R908jHsjOTUz2WDfgT/8tf/sKXX37JrFmzQrfdeuutXHfddfzoRz866YWciRijpBCk1iNy0HJ8EfZKwXtUt9eRHZXOdeMvpfK/LwGwJywbjUYljweRkTlFfvOb37BkyRJ+//vfk56eTmVlJX/4wx94+OGH6ejoCG0XERExjKs8fSgP2FOkKyIpA9RRJjxeH5ZOFwadCoNOrqOTkRlqBizCOjs7mTZtWo/bpkyZgu0kR/ecyYRHR+CE/2/vzuOiqvf/gb9mgWEbBkQWlUUJUDFAARU1wyzz97WuUuZ21VwizKKya4uZfeXRVbRUculqdf2m3nux6zXULKm8uKSUGaaIZhoYm7IIsi8Ds3x+fwxzZGSbgRnODPN+Ph7n8eAsc85nPgNvPvM57/P5wFahxh+VhWCMtTs697nCX5B28yzsxXZYMe45CJQqlJ05CwgEuOz0AIKH9IONmPLBCOmJuLg4AEBqaioEAgEYYwCAr7/+GnFxcdzfJyU9AzUtSfkyiRQO9UoAmnywu9XUC0ZIbzK4EbZs2TK89957WLt2LUQiEVQqFdavX48XXnjBFOUza87OjqiHEBKFGvXNDShvqIC7o+6I96V1Zfg4418AgOcj/wwvqQfunD4DVX0D6gf6o8bGiW5FEmIEubm5fBfBYmgHafXv54fa6zcAAI6D/biBWqkRRkjvMDgnLDU1FRs2bED//v0xYsQI9O/fH+vXr0dqairCw8O5xRrInCRoEtpApAZEqrZ5YUqVElt//D80KuSYPGQ8HvIbDQC4c+IkAOCqSyAAGqS1rxk8eDAyMzN1tsXGxuLUqVMAgISEBLi7u2PkyJHcUlTU8dRXhh5vrfz8/PRaCPBHpSYp39/VF1WXMgEALqNG3kvKp+EpSAuKZ6ZlcE/YihUrTFAMyyR1tEWTyBYO6iZImhnyqgoxxnskt39/1hHcrMyHt/MALAmfAwCQl5aiOusKRI6O+EnpDpFQgABvF37eAOk1u3fv1lmfP38+tm7dqvfrDT2ekM5oe8KGOA1AzbXfILCxgfOIYJSnaxpn1BNGOkPxzHgMboQtWrTIFOWwSHa2IjQLW6b7aMkL00rP/xlf/34CtiIbvDY+FhKxZiLcOyc03x6cxo5DY4EQg9wdIRYZ3CFJOrBh38/4+ddSk5x7zAhPvL1oTLdeO2nSJKxYsYKbd9UYTp8+jfj4eEycOBE//PADGGNITk5GUlISfvnlFzg4OODQoUMYNGiQ0a5JALlcjrlz5+LatWuwt7eHh4cHdu3ahYCAgDbHpqamYs2aNVCr1VAqlXjjjTfaxNA9e/Zg6dKlOHz4MPf7kZ2djUWLFqG8vBwymQx79+7FiBEjjPYetE9GehU3olCphCw0BCKJpNWURTQ8Bd9MGcsAimfmolv//TMzM/HZZ59h+/btOou1EQgEUNlo5hKUKBjyWhphv5VlY+fP/wQAxEXOh49sIACAqVQobWmENT2ouTU50J3GhLFGycnJXFf8kiVLun389evXERsbi6ysLMTExGDy5MlYtWoVrly5gsjISPq2aSJxcXG4ceMGLl++jBkzZiA2NrbNMYwxLFiwAHv37kVmZia+/vprLFu2DLW1tdwxeXl5+Pvf/67ztDmgyb2Ni4vD77//jrfeeguLFy/Wu2w1NTU6S1NTk+5+eS3KGyogs3OG8rccAJpbkQC44SlotHxiCIpn3WdwT9iaNWuwZcsWhIaGwsHBgdsuEAjwyiuvGLVwlkAtsQPqABdmjxuNVbhedhOb0j+BUq3EMyOewMODx3LHVl+5iubycjgOGYzbkn4AbmNgfyf+Ct8HdfebXW8zVvd9QEAAIiIiAACRkZEICAjAsGHDAABjxozB4cOHjVFci5CRkYHRo0eb/Dp2dnY6M4ZERUVh8+bN7R4rEAhQVVUFQNM4cnNzg6RlEni1Wo3Y2Fjs2LEDK1eu5F5z584dXLhwAcePHwcAzJw5E/Hx8cjJyWm3t+1+Pj4+Outr165FQkICt67tBfN39UV12mUAgMtIzSTn5dWUmG8uLCWWARTPesLgRtiuXbtw8eJFDB8+3BTlsTwSTbDyFMhwA3ex/swONCmb8LDfWMwa8YTOoaVpJwAAHo9ORka5ZkiPQdQTRnrAzu7ebSORSNRmXalU8lEsXkydOhWDBg3CkiVLsHDhQri7984sFNu2bcOMGTPabBcIBDhw4ACefvppODo6orKyEocOHYKtrSY1ISkpCRMmTOD+6WgVFhZiwIAB3MCdAoEAvr6+KCgo0KsRVlhYqDMWmrbRp+VgYw+RQIjRjv5oKDgJG5kMjoM1DyzQ5N2ET9YYzwy+HdmvXz8MGTKkxxeWy+WIiYlBUFAQwsLCMGXKFOTk5HD7s7OzMX78eAQFBWH06NH49ddfe7zPFAT2mmAlU2saU03KJgS7B2LZ6Pk6Y4Y1ld/F3XPnIRCL4R79MIpaGmHUE0aIcRQXF2P16tU4fvw4/Pz88NRTT+Grr76CWq026Dzjxo1D//79210KC3WfgE5MTEROTg42bNjQ5jxKpRLr1q3DoUOHkJ+fjxMnTmDhwoUoLy/H1atXkZKSgjVr1vToPbfH2dlZZ7m/ERbU3x/7nv4QIVWa7S4jwyAQCtGkUKGmvhmO9jawl9DI7YT0BoMbYZs3b8ZLL72EnJycNrkHhuosr6KznIju7jMFccstWSeFpsU+UOqJ1x9aBhuR7mjThQf+A6ZUwmPyJNg4S3G7rE5zvDs1wvqiqVOnwtvbm1tu3brFd5H6PIlEgnnz5uHbb79FTk4Oxo4di5UrV2LQoEF48803cfPmTb3Oc+7cOZSXl7e7tL7Vt3nzZhw6dAjffPONTmqGVmZmJoqKirh5dUePHg1vb29cunQJZ8+eRV5eHgIDAzF48GD89NNPiIuLw65du+Dj44Pi4mLuWz9jDAUFBfD19TVCLWnYim1Rlal7K/JutTYpn3rBiC6KZybEDHT69Gnm7e3NhEIhtwgEAiYUCg09lY6MjAzm5+fHGGOstLSUSaVSplAoGGOMqdVq5unpybKzs7u9ryPV1dUMACssLGTV1dXcIpfL9Sr30cRPWPr0p9m3iTvZuYJfWHVjTZtjGm7dZukxz7Afn5nL5OXlTKFUsemvf8mefvMoU6nUhlQTuY9cLtf7syLmo6PPTfv3WF1d3aPzf//992zx4sXM2dmZTZ48mS1dupTJZDK2YcOGHp1Xa8uWLSw8PJxVVFR0eExJSQlzcnJi165dY4wxlp2dzVxdXVl+fn6bY6Ojo9nhw4d11vfs2cMYY+zgwYMsIiKiyzIZUndqlYr9tGAxS5/+NGu6q3kPl7PvsCf/coQl/P1cl68nxkexzDL1NJYZ3Oe8dOlSLF26FHPnzm332193tc6r6CwnQiaTdWtfV7kUXSWzdsRWqrkNqWxoQJRP+4PU5id/DqjVGPDE/0Di5oaisjqo1QwDPJwgFLad5ogQYrj8/Hzs27cP+/btg0KhwOLFi5GZmcmlT6xevRrh4eFYtWpVj65z69YtrFy5Ev7+/njkkUcAaHrhzp8/D0AzkOX06dMxffp0fPrpp5g9ezaEQiHUajU++ugjvXq0PvnkEyxevBiJiYlwdnbGnj17elTm+9Xn5kFZUwMHP1/Y9nMFQPlghPDB4EbY3bt3kZCQ0O4cid2lzas4ceKE0c5pqK6SWTti56y5nahqaGx3f13OTdz94UeIHB0waOZTAEC3IkkbqampWL16dZvtb7/9NubMmcNDiSzPsGHD8Kc//Ql/+9vfMHXq1DYx6oEHHsC8efN6fB1vb29uXsr2tB7Ict68eXpd8/Tp0zrrQ4cOxblz57pdxq60HiVfi5uySEZjhJGeoXimP4MbYXPnzsXRo0fbfRqoO7R5FWlpaVzPWuucCLFYrJMT4ezs3K19XdEmsRrKXiaFAoC6saHd/fn/TAYADHoqBjZSKQC0SsqnJyOJxrRp03SGPSCGKywsRP/+nU8B9vHHH/dSaczb/flgAHC3iibvJsZB8Ux/BjfCbt++jTlz5iAiIgKenp46+w4dOmTQuZKSkvD5558jLS0NLi4u3HYPDw+Eh4fjX//6FxYvXoyUlBR4e3tztxS7u88UHF2cUQUAcnmbfVVZV1CVeRk2Li4Y+Kd7w1Voe8IGUU8YIUZz7dq1drdLJBL4+vpiwIABvVwi86SSy1Hz23XNVEXB94Yaosm7Cel9BjfCIiMjERkZ2eMLd5VX0VlORHf3mYLUVdMIEzbrNsIYY8j/h6YXzGf2MxC1Gu+kuKylJ4waYYQYzYwZM1BXVweVSgU7OzvI5XJurKGGhgZMmDABycnJbfI/rU3Ntd/AlEq4jAyDqFXaBeWEEdL7DG6ErV271igX7iqvorOciO7uMwVnN80tTNF9jbCGgkLUZWdD4t4fno8/prOvqLwlJ4xuRxJiNOvXr8fFixexceNG9O/fH2VlZXjnnXcwatQoPPXUU3j11Vfx0ksv4ejRo3wXlVeS/v1h5+UFr/+ZqrNdO0SFG+WEEdJrujUi36lTp7B//36UlJTgq6++woULF1BbW8v1aFkTJ1dNI0ysbAZjjEsGlhcVAwCkw4dDaHNvzLBmhQplVY2wl4jhItUv+Z8Q0rXExETcvHmTe6jG3d0d27ZtQ2BgIJYvX46PP/4YgYGBPJeSfw6+Poj45G862+TNStQ2KCB1sIWdLQ3USkhvMXiw1t27d2PhwoXw9PTEmTNnAAA2Njb43//9X6MXzhKI7O2hhgC2agUam+5NqSAvLQUA2Hnp5s0V360HY5qJu435hCkh1k6pVLYZRLKoqAgKhQIAIJVK++S0J8Zw71Yk9YIR0psMboR98MEHOH78ONatWwehUPPy4OBg/Pbbb0YvnCUQCARQimxhy5Soqbk3TIW8pKURdt/DC0VlNF1RXzd48GBkZmbqbIuNjcWpU6cAAAkJCXB3d8fIkSO5paioqMPztT5++PDhmDNnDiorKwFocg8TEhIQFBSEkJCQLnujc3NzERERgZEjR+LBBx/ErFmzuHNZutjYWDz++ONISkrCwYMHkZSUhKlTp+L5558HABw7dozmvO0APRlJOkLxzLS6NU5YcHAwAHA9OQKBwKp7dZQ2EtiqmlB9txpeHprbkx31hBVxY4RRPpgpbPnhU/xSdMUk544YGIKVE+K69drWY0cBwPz587F161a9X689XqVSYdasWVi3bh22bNmC7du3IysrC1evXoWtrS1KSko6Pc/AgQORnp4O+5Y5T1999VUkJCRg27ZtBr8nc/PXv/4VQ4YMweeff46ioiIMHDgQb7/9NpYuXQoAePLJJ/HEE090cRbrRE9Gmh9TxjKA4pm5MLgnLCwsDCkpKTrbjh49ivDw9keLtwbqlhyU2op782fKW3557Ly8dI6libut06RJk3DkyJEen0ckEuGxxx7DjRs3AACbNm3Cxo0bYWtrCwDwuu/37X4SiYQLWCqVCvX19dwXqNOnT+PBBx/E8uXLERoaipCQEGRlZWHx4sUICQnB2LFjcfv2be5c77//PkJCQhAWFoaoqCg0NLQ/Vl5vUCqVmD17NubPn4+0tDRcu3YNaWlpeO6557j3JxKJuNk0iK7yau1ArdQII12jeGY8ekekJ554AseOHcPmzZvx+OOPY//+/WhoaMCf//xnnDx5Et99950py2nebDV5FHWVmkYYU6nQdKcMArGYmxJE694YYdQTZgrd/WbX25KTk7lR0keNGqX3UCqNjY04cuQIxo0bh5qaGpSWluLLL7/EF198AQD4y1/+0uWI1M3NzRgzZgzy8/MRGhqq87Tg9evXsW/fPuzatQvvvvsuJk+ejPT0dAwbNgwvvfQStm7dik2bNmHfvn1ISUlBeno6ZDIZKisr9Z5lwhTEYjHS09OpkdVNNDyF+bGUWAZQPOsJvXvCzp49C0AzGOrVq1cxbtw4xMbGIiwsDBcvXkRYWFgXZ+i7BC0t8YbqWgBAc0UFmFIJO08PCIS6VVxcTlMWEU13fGZmJjIzM/UKWMnJyRg1ahTGjx+P4OBgrFq1CkqlEkqlEo2NjTh//jwOHDiA1157DZcvX+70XLa2tsjMzERpaSmGDRuGTz75hNsXEBCAiIgIAJoxAQMCAjBs2DAAwJgxY5CdnQ0A+Prrr/HCCy9AJpMBAFxdXSESibpVF8aybNkyfPDBB7yWwVJpG2Hu1Agj3UDxrPu69bXRy8sLr7/+urHLYrFELdMtyWs0DSwuKf++fLDGJiUqapogdbCB1MG2dwtJLFp7ORf29vZwcnLCggULAGgSaCdMmICMjAy9vhTZ2tpiyZIleP755/Hmm28CAOxaDSqsHei09bo5P1345Zdf4urVq/jwww8xaNAg7sEhALh48SKPJTN/1BNGehPFs3v0boQ1Nzdjx44dnQ6w+sorrxilUJbGxlFza7GpVpPvpU3Kl7R5MpJ6wYhxzZs3D99++y1efPFFVFRU4Oeff8Ybb7zR4fH5+flwd3eHg4MD1Go1Dh48iNDQUIOvO336dOzYsQMzZ86ETCZDVVUVpFIpr71hK1as4O3alk7bCKOBWgmfrDGe6d0IUyqVnc4NKRAIrLYRZit1hAqAoq6lEdZBTxhN3G09pk6dCptWg/S2/gZmTBs2bMCSJUuwc+dOAMBbb72FMWPGdHh8VlYW3nnnHQCAWq1GeHg4tm/fbvB1Fy5ciKKiIowfPx5isRiOjo5IS0uDQ0uvMB8WLVrE27UtWYNcgXq5EjInW9ja8HtLmZgnimemI2CddW214uzsjJqamq4PtDA1NTWQyWSorq6Gs7Nzt85xaXcyGr46hBuBE7B0819wY8uHKD+TjmFvvwm3qLHccQf+ewP/+vY6Fvy/YZgzZaix3oJVa2pqAgBek8KJ4Tr63Hr697hnzx4kJyfjzp07yMrKwvfff4/S0lLMnj3bKOU2Z92tu8LSWrz4wUk84C3D1tcmmax8pHMUyyxTT2OZwUNUkLbsnTW3F9WNmkda5cXanjAanoKQ3rJ+/Xp8+OGHmDt3LgoKCgAAAwYMwKZNm3gumXkrbolLHq789WISYq30vh2pZ4eZVXJ0dcZdAJBr8iq4gVo9PXSO0+aEDaDhKch9UlNTsXr16jbb33777S4f0W5PZGRkm6TTESNGIDk5udtl7Iqx34Ohdu/ejbNnz8Lb25vLIwkICMDNmzdNfm1LdqNAM8L4A94ynktC+gqKZ/rTuxFWW1trtIv2NQ4yTc+WoKkJivp6KGtqYCOTQWSv+6TR7TLKCSPtmzZtGqZNm2a08124cMFo59KXsd+Doerr6zFgwAAA92bzUCgUdHunCzfyKwAAw3z78VwS0ldQPNMf3Y40AnHL05E2qmbU3ioG0DYp/+atKtQ2NMPd1R4OdjZtzkEI6ZmoqCguoVfrs88+w4QJE3gqkflTqxmyC6sgEACBvi58F4cQq0PDSxuB2FGTSyFRN6OqUDNx6f3DUxz5XnNLZMoYv94tHCFWYuvWrXj00Uexd+9e1NXVYfz48SgtLUVaWhrfRTNbt+7UokGuhK+XlL4cEsIDaoQZgXawVola0W5PWFllI85k3oatWIhp4wfzUURC+jx/f39cu3YNx44dQ15eHnx8fPDkk0/C0ZFu/3fkRr4mH2yor2sXRxJCTIEaYUYgdrjXE8ZN3N2qJ+yr9D+gVjNMHusHmRPlpxBiKvb29njmmWf4LobF0CblD/WjRhghfKCcMCPQJuDbqhVovnMHwL2esAa5At/9lAcAmPGwPy/lI71r8ODByMzM1NkWGxuLU6dOAQASEhLg7u6OkSNHcktRUVGH52t9/PDhwzFnzhxUVlaa8i1YpNraWrz33nuYMWMGJk+erLOQ9ml7woKoJ4x0gOKZaVFPmBEIRCKobWwhUTRDfbcMwL0xwo6fL0CDXImxI7zg7SHls5iER7t379ZZb2/utM5oj1epVJg1axbWrVuHLVu2GLmUlm3hwoUoLCzEzJkz6RakHhqblCgoqYGdrQi+Xt0bqJpYJ4pnxkONMCNhEjsIFc0QVd2FQCyGbT9XqFRqHD2rSciPiX6A5xJah+sbN6EiwzSPM/cbHYlhqzqex6wzkyZNwooVKxATE9OjMohEIjz22GNITU0FoPmWumDBApw8eRKFhYV45513IJFI8Omnn6K4uBgbN27E3Llze3RNS3H69GkUFBR0e+YLa5NzqwpqBgT6uEIkFPBdHHIfU8YygOKZuaDbkUYisLs3JpjIrT8gEOCHrCKUVTYiwMcFI/zdeCwdMTfJyclc1/2SJUv0fl1jYyOOHDmCiIgIblt9fT1+/PFHnDp1Cq+99hpu376Nc+fO4eDBg3j55ZdNUXyz5OPjA4VCwXcxLAaXlE/5YKSHKJ51H/WEGYmNkyNU5Zqfb9SK8H8bT0CpUgMAnop+gBs8kphWd7/Z9TZDu++Tk5Px/fffAwCio6OxatUqbp929OaAgADY2dlxiemRkZGoqKhAVVUVXFxcjFZ2cxUfH485c+bgrbfegud9Q8SEhobyVCrz9XsB5YOZM0uJZQDFs56gRpiRSF2dUZWn+bnJyYWbj83d1R4TQgfyVzDSJ3QW5Ozs7LifRSIRty4QCCAQCNpM99FXLV++HABw8uRJne0CgQAqlYqPIpktxhg3Uj71hJHeRvHsHmqEGYl2rDAAmPZEJKaMnYTM38sQ8oAbRCK660uIqanVar6LYDHKq+SoqGmCu6s9+jnbdf0CQohJUCPMSLSj5gOa4SkGDXDG4AGUIGytpk6dChubeyOQt/52R3pXZWUlXF2pt6c1uhVJDEHxzHSoEWYkrXvCtMNTEOuUl5fX6f6EhASDztfZ8fdfq7y8XGe9r3Xdt8fDwwN3WsbnAzT/ML777jtu3c/PDzU1NXwUzWxd107aTbciSRconpkW3SczEnGrcYnsPD14LAkh1qWxsVFnPSMjQ2edMdabxbEI1BNGiHmgRpiRaHvCbGQybgR9QvSVmpqqM+K0djlw4ADfRTN7XT15TE8m61Kq1Mi5VQ2RUIAHvF34Lg7pgyie6Y9uRxqJ2EHT8Go9cTch+po2bRqmTZvGdzGIFcgrrkGzQoUAbxkkNiK+i0P6IIpn+qNGmJHYtCT+2nt781wS6yISidDU1ASJhCZGtyRKpdJon1lzczO2b9/Orcvlcp11GsBVV1ml5vZt8BAaQNqcUCyzTD2NZQJm5QkTNTU1kMlkqK6u7tF0J0ylQtmZs5CFhkLi1s+IJSRdqa6uhlgshlhM3yksgVKphFKphEwma7OvO3+PkyZN6vKWo3ay4b5M37pTKFU4m3kbkcO94Oxo24slJF2hWGZZjBHL6JM2EoFIBI9HJvFdDKskk8mgVCppQE4LIZFIjDrB9unTp412LmtgIxZhcqQv38Ug7aBYZlmMEsuYlauurmYAWHV1dafHyeVytnbtWiaXy3upZOaL6kKD6kHDmPWg798jnxobG9mMGTNYYGAgCw0NZY899hjLzs5u99hjx46xUaNGsbCwMDZixAi2d+/eNsd89tlnDAA7fPgwt83Pz48FBQWxsLAwFhYWxv797393WS5D6o5+dzWoHjSoHjT4iGXUCNOzoizhn0NvobrQoHrQMGY9WEKdNjY2smPHjjG1Ws0YY2zHjh0sOjq6zXFqtZq5urqyy5cvM8YYy83NZRKJhNXU1HDH5ObmsnHjxrGoqKg2jbBLly4ZVC5D6s4S6rk3UD1oUD1o8BHLaIgKQggxgJ2dHaZNm8bloUVFRXU4oKVAIEBVVRUATY6Im5sbl8SrVqsRGxuLHTt2GDUZu6amRmdpamoy2rkJIcZFjTBCCOmBbdu2YcaMGW22CwQCHDhwAE8//TT8/Pzw0EMPYd++fbC11STDJyUlYcKECYiIiGj3vM8++yxCQkLw3HPPoaysTO/y+Pj4QCaTccuGDRu698YIISZn9Yn5rOXh0K6mNdHup+lPqC60qB40jFkP2nMwHh/aHjduHLKzs9vdd+nSJfj4+HDriYmJyMnJwYkTJ9ocq1QqsW7dOhw6dAgPP/wwMjIyMH36dFy5cgUlJSVISUnBmTNn2r3OmTNn4OvrC4VCgTVr1mDRokVITU3ttNzaOrt27RqkUim3XSKRtPls6HdXg+pBg+pBg5dY1uMbnxausLCQAaCFFlrMaCksLOQ7NHRp06ZNLCIiglVWVra7PyMjgwUGBupsi4yMZMePH2c7d+5kXl5ezM/Pj/n5+TGJRMLc3d3Zzp0725ynqKiIOTk5dVkeimW00GJ+S1exzOrHCVOr1SgqKoJUKqXpTQjhGWMMtbW1GDhwIIRC882WSEpKQnJyMtLS0uDq2v78i6WlpQgICMDPP/+M4cOHIycnB2PGjEFmZiZ8fXWHiJg0aRJWrFiBmJgY1NfXQ6FQwMXFhbvWkSNHOuw106JYRoj50DeWWf3tSKFQCG8a5Z4Qs9HewIfm5NatW1i5ciX8/f3xyCOPANDc8jt//jwAIDY2FtOnT8f06dPx6aefYvbs2RAKhVCr1fjoo4/aNMDuV1paipkzZ0KlUoExBn9/f/zjH//oslwUywgxL/rEMqvvCSOEEEII4YP59vcTQgghhPRh1AgjhBBCCOEBNcIIIYQQQnhAjTBCCCGEEB5QI4wQQgghhAfUCNNDdnY2xo8fj6CgIIwePRq//vor30UyCblcjpiYGAQFBSEsLAxTpkxBTk4Ot7+zeujLdbRnzx4IBAIcOXKE22ZNddHU1IT4+HgEBgYiJCQECxYs4PZZUz30Bdb0mVA8a4timRnGsi6HYSbskUceYXv27GGMMXbw4EEWGRnJb4FMpLGxkR07doyp1WrGGGM7duxg0dHR3P7O6qGv1lFubi4bN24ci4qKYocPH+a2W1NdrFixgsXHx3O/F8XFxdw+a6qHvsCaPhOKZ7oolplnLKNGWBdKS0uZVCplCoWCMcaYWq1mnp6eLDs7m+eSmV5GRgbz8/NjjHVeD321jlQqFXv00UfZhQsXWHR0NBe4rKku6urqmFQqZdXV1W32WVM99AXW/plYczyjWGa+sYxuR3ahsLAQAwYMgFismVxAIBDA19cXBQUFPJfM9LZt24YZM2YA6Lwe+modJSUlYcKECYiIiNDZbk11cfPmTfTr1w+JiYmIjIzExIkTucmqrake+gJr/0ysOZ5RLDPfWGb10xaR9iUmJiInJ4f7JbU2V69eRUpKSpfz9fV1SqUS+fn5CA4OxsaNG3Hp0iVMmTLF4nNDiHWx5nhGsUzDXGMZNcK64OPjg+LiYiiVSojFYjDGUFBQ0OX8b5Zs8+bNOHToENLS0uDg4ACg83pwdnbuc3V09uxZ5OXlITAwEABQUlKCuLg4FBcXY+bMmVZTF76+vhAKhZg/fz4AYNSoURgyZAiuXLmC0NBQq6mHvsAaYxlA8YximYbZxjKj3NTs46Kjo3WS8iIiIvgtkAlt2bKFhYeHs4qKijb7OquHvl5HrfMotOvWUhdTpkxhx44dY4wx9scffzA3Nzd269Ytxph11UNfYG2fCcWztiiWmVcso0aYHq5fv86ioqJYYGAgi4iIYFlZWXwXySQKCwsZAObv78/CwsJYWFgYGzNmDLe/s3ro63V0f+Cyprq4efMmmzRpEnvwwQdZaGgo++KLL7h91lQPfYE1fSYUz9pHscy8YpmAMcaM06dGCCGEEEL0RU9HEkIIIYTwgBphhBBCCCE8oEYYIYQQQggPqBFGCCGEEMIDaoQRQgghhPCAGmGEEEIIITygRhghhBBCCA+oEUZ4k5iYiHnz5vXKtfLy8iAQCODk5IS1a9dy23/66ScEBwdDKpVi+/btHb7+7t27cHJygo2NDVasWNELJSaEWAqKZaS7aO5IYjJOTk7cz42NjRCLxbCxsQEATJw4Ed98802vl+nWrVtwcXHh1tesWYN58+bh3Xff7fR1bm5uqKurw+LFi01bQEKI2aFYRkyFesKIydTV1XHLxIkT8f7773PrfASt9uTm5iIkJITvYhBCzBjFMmIq1AgjvElISEBMTAy3LhAI8NFHHyE4OBiOjo5YuHAhKisrMWfOHDg7O2PUqFG4fv06d3xdXR3i4+Ph6+sLDw8PPPvss6iurtb7+l5eXsjNzcW8efPg5OSE33//Hf/9738RGhoKqVQKT09PLF++3JhvmRDSB1EsI91FjTBiVr788kukp6cjOzsbx48fR3R0NF5++WVUVFRg5MiRePPNN7ljly5dioqKCmRlZSE3NxcKhQLx8fF6X6ukpAS+vr74/PPPUVdXh6CgICxatAhvvPEGamtr8ccff2DhwoWmeJuEkD6OYhnRB+WEEbPy+uuvo1+/fgCA6OhoiEQiPPTQQwCAWbNmIS4uDgBQVlaGlJQUlJeXc3kR7733HkaMGIG9e/dCJBJ16/o2NjbIyclBWVkZ3N3dMX78+J6/KUKI1aFYRvRBPWHErHh6enI/Ozg4tFmvq6sDoHlCSK1WY8iQIXBxcYGLiwtGjx4NoVCIkpKSbl//8OHDuHr1KoYOHYpRo0bhP//5T/ffDCHEalEsI/qgnjBikXx8fCAUClFUVAQHBwejnTc8PBwpKSlQq9U4cuQIZs+ejejoaJ0ASgghxkKxzLpRTxixSF5eXoiJiUF8fDzKy8sBaPIiDh8+3O1zNjc345///CcqKyshFAq5WwNiMX1XIYSYBsUy60aNMGKx9u7dy3XdOzs7Y+LEifjll196dM79+/cjICAAUqkUL7/8Mvbv3w83NzcjlZgQQtqiWGa9BIwxxnchCDG1/Px8DB06FHZ2dnjttdd0RprWR0VFBfz9/aFQKPDiiy9i06ZNJiopIYR0jGJZ30KNMEIIIYQQHtDtSEIIIYQQHlAjjBBCCCGEB9QII4QQQgjhATXCCCGEEEJ4QI0wQgghhBAeUCOMEEIIIYQH1AgjhBBCCOEBNcIIIYQQQnhAjTBCCCGEEB5QI4wQQgghhAf/H06Dx7AzTbA0AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 690x200 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[6.9, 2], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    spec = fig.add_gridspec(1, 2, wspace=0.35)\n",
    "    ax1 = fig.add_subplot(spec[0])\n",
    "    ax2 = fig.add_subplot(spec[1])\n",
    "    idx = np.arange(len(results_df.index))\n",
    "    for sim in results_df.index:\n",
    "        name = sim.split(\"-\")[2]\n",
    "        ax1.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_Temp\"], label=name)\n",
    "        ax2.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_energy_per_atom_t\"], label=name)\n",
    "\n",
    "    ax1.set_ylabel(\"Temperature [K]\")\n",
    "    ax2.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax1.set_xlabel(\"Time [fs]\")\n",
    "    ax2.set_xlabel(\"Time [fs]\")\n",
    "    ax1.legend()\n",
    "    ax2.legend()\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"7-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "649ff3e0-15cf-46a7-9376-f930b5913793",
   "metadata": {},
   "source": [
    "#### Save frames "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "2e5f5604-c7f3-43bf-be6e-2f851395daee",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "conf_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7.1-amorphous_frames\")\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7-amorphous\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)\n",
    "\n",
    "make_dir(conf_dir)\n",
    "last_n_frames = 10\n",
    "nskip = 2\n",
    "frame_results_df = pd.DataFrame(columns=[\"step\", \"time\", \"unitcell_repetition\"])\n",
    "\n",
    "mols_ase = []\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    name = sim_name[3:]\n",
    "    steps = job_result.get_history_property(\"Step\")\n",
    "    n_frame = len(steps)\n",
    "    n_rep = results_df.loc[f\"{name}.cif\", \"unitcell_repetition\"]\n",
    "    for i_frame in range(n_frame - last_n_frames * nskip + 1, n_frame, nskip):\n",
    "        frame_name = f\"{name}_{steps[i_frame] * TIMESTEP:1.1f}ps\"\n",
    "        mol = job.results.get_history_molecule(i_frame + 1)\n",
    "        # print(os.path.join(conf_dir, f'{name}.cif'))\n",
    "        mol = toASE(mol)\n",
    "        mol.write(os.path.join(conf_dir, f\"{frame_name}.cif\"), format=\"cif\")\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"unitcell_repetition\"] = n_rep\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"step\"] = steps[i_frame]\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"time\"] = steps[i_frame] * TIMESTEP\n",
    "        mols_ase.append(mol.copy())\n",
    "\n",
    "frame_results_df.to_json(os.path.join(conf_dir, \"info.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f0135cbd-d926-4ca7-be35-abda4e7a0106",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif</th>\n",
       "      <td>1600</td>\n",
       "      <td>400.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif</th>\n",
       "      <td>1700</td>\n",
       "      <td>425.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif</th>\n",
       "      <td>1800</td>\n",
       "      <td>450.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif</th>\n",
       "      <td>1900</td>\n",
       "      <td>475.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif</th>\n",
       "      <td>2000</td>\n",
       "      <td>500.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          step   time unitcell_repetition\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif  1600  400.0                  27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif  1700  425.0                  27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif  1800  450.0                  27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif  1900  475.0                  27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif  2000  500.0                  27"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frame_results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "76135ef7-5de3-46c0-a763-88ba2802c687",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ed657c183b6a4b3abecbfdca4b7573ec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d57b7feac2a6409a8c3b13a1f2fc1050",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=29), VBox(children=(Dropdown(description='Show', options=('All', 'Li', 'F')…"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(mols_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ea754e65-dcb7-4be8-b50f-e8caff786dca",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "8679444c-a3dd-46e3-9ad1-8588a4415106",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4a10bc10-72b7-4c14-83af-6cd5e01e7d34",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Amorphous `Single Point` <a id='amorphous_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "629124c9-a4da-4ad3-9258-27f76b78685b",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"7.1-Amorphous_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9f93885d-116b-47e4-917b-3dfc567e456c",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "9c90c48f-7dce-406a-a8ee-112aae871766",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_7_SP\n",
    "\n",
    "print(settings_7_SP)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3aa7c2ea-b75e-4fcf-8453-e7de1cdfed4b",
   "metadata": {
    "tags": []
   },
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "4b86e5f6-3c8d-4825-898f-b03abe3d877f",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "80932a55-55c4-4129-8159-8d146a22c87d",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "3c81fbea-e4ea-4b55-964f-11a13e992f9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_7_SP, file)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3a46150f-8010-49c7-bc03-e451917ccd0a",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dfef15b5-9ac3-4d4c-b600-1a0f3bce8071",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "4a76bb51-530f-4198-85a9-eb4991d1ae7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7.1-amorphous_frames\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0fa48390-4e61-424c-b5ed-51c9aa3de842",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "e05dad83-07eb-450d-a2bd-21dfd6ad567a",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_7_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2021.104\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "91712856-8217-48d4-b2ab-505feeb12cc1",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "acc131bf-6f60-4718-8e16-20f3284f593c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:00:37] JOB jobs STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps STARTED\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:25] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:05:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:07:43] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:08:55] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:09:01] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:09:07] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:09:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:10:26] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:12:44] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps reported warnings. Please check the the output\n",
      "[23:12:44] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:44] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:12:50] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps reported warnings. Please check the the output\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:13:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:14:02] Job jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps reported warnings. Please check the the output\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:14:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:41:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps FAILED\n",
      "[23:42:00] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "75d23540-8641-41c5-911a-7fbdbd90c813",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f7fa34bd-5441-4277-b361-370a7b8382e9",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "7a8273e9-6952-4a1c-a81b-cf4c6a3b043a",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"7.1-amorphous_frames\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "0f9611d5-57a9-4fe4-ab45-636e930ca2c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif</th>\n",
       "      <td>1600</td>\n",
       "      <td>400</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif</th>\n",
       "      <td>1700</td>\n",
       "      <td>425</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif</th>\n",
       "      <td>1800</td>\n",
       "      <td>450</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif</th>\n",
       "      <td>1900</td>\n",
       "      <td>475</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif</th>\n",
       "      <td>2000</td>\n",
       "      <td>500</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          step  time  unitcell_repetition\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif  1600   400                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif  1700   425                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif  1800   450                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif  1900   475                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif  2000   500                   27"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "3c8e8774",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_320.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_340.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_360.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_380.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_420.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_440.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_460.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_480.0ps FAILED\n",
      "Simulation step02-sim/LiF/7.1-Amorphous_SP/jobs/SP-7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps FAILED\n"
     ]
    }
   ],
   "source": [
    "for job_dir in np.sort(glob.glob(os.path.join(dir_jobs, \"*\"))):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_7_SP)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_per_atom_SP\"] = energy_per_atom\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "306193a5",
   "metadata": {},
   "source": [
    "### Finishing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "661c7651-7dca-4642-b772-79e279cba83c",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a0d16106",
   "metadata": {},
   "source": [
    "## Bulk 300 K `ab initio MD` <a id='300_aimd'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "a666b966",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"8-Bulk_300K_MD\"\n",
    "TASK = \"MD\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8928fced",
   "metadata": {},
   "source": [
    "#### Setting `DFTB` *ab initio MD*"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88fed954",
   "metadata": {},
   "source": [
    "MD settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "8646e1db",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      ams: \t\n",
      "          task: \tMolecularDynamics\n",
      "          MolecularDynamics: \t\n",
      "                            NSteps: \t1000\n",
      "                            TimeStep: \t0.25\n",
      "                            Trajectory: \t\n",
      "                                       PrintFreq: \t5\n",
      "                                       SamplingFreq: \t50\n",
      "                            InitialVelocities: \t\n",
      "                                              Temperature: \t300\n",
      "                            Thermostat: \t\n",
      "                                       Type: \tBerendsen\n",
      "                                       Temperature: \t300\n",
      "                                       Tau: \t25\n",
      "      DFTB: \t\n",
      "           Model: \tNonSCC-GFN1-xTB\n",
      "\n"
     ]
    }
   ],
   "source": [
    "STEPS = 1000\n",
    "TIMESTEP = 0.25  # [fs]\n",
    "TEMP = 300  # [K]\n",
    "\n",
    "from tools.misc.default_settings import settings_MD\n",
    "\n",
    "settings_MD.input.ams.MolecularDynamics.NSteps = STEPS\n",
    "settings_MD.input.ams.MolecularDynamics.TimeStep = TIMESTEP\n",
    "settings_MD.input.ams.MolecularDynamics.InitialVelocities.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Tau = int(100 * TIMESTEP)\n",
    "\n",
    "print(settings_MD)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8ba89b3",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "be1b1da0",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 1\n",
    "CORES_PER_NODE = 128\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "325dea28",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "6882f358",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_MD, file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ef82ab5",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2604019c",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "9e5b07d4",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8-bulk_300K\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad5e93fc",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a991b43",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_MD, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66ff125e",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9661d7c0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[12:36:49] JOB jobs STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-0-LiF_Fm-3m_-3.18_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-1-LiF_P6_3mc_-3.17_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-2-LiF_Pm-3m_-2.89_2x2x3 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "584392ab",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84d23232",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "b16c7d58",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8-bulk_300K\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "7a9aa46c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                    12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                   12\n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                    12"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "28ecc24c",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_df = pd.DataFrame(index=results_df.index, columns=[\"scm_step\", \"scm_Temp\", \"scm_energy_per_atom_t\"])\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    # sim_name = sim_name.replace(\"10.0-\", \"8.0-\")\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    Nentry = job_result.read_rkf_section(\"History\")[\"nEntries\"]\n",
    "    energy = scm.Units.convert(\n",
    "        job_result.read_rkf_section(\"History\")[f\"Energy({Nentry})\"], \"au\", \"eV\"\n",
    "    )  # scm.Units.convert(job_result.get_energy(), 'au', 'eV')\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_step\"] = np.array(job_result.get_history_property(\"Step\"))\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_Temp\"] = np.array(job_result.read_rkf_section(\"MDHistory\")[\"Temperature(1)\"])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_energy_per_atom_t\"] = (\n",
    "        np.array(job_result.get_history_property(\"Energy\")) / Na\n",
    "    )\n",
    "\n",
    "results_df = pd.concat([results_df, temp_df], axis=1)\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "e4023c6c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "      <th>scm_step</th>\n",
       "      <th>scm_Temp</th>\n",
       "      <th>scm_energy_per_atom_t</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-1855.994228</td>\n",
       "      <td>-77.333093</td>\n",
       "      <td>0.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[300.0, 244.72130136787882, 155.4044352312756,...</td>\n",
       "      <td>[-2.8434003513590658, -2.843102065120439, -2.8...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-3706.687014</td>\n",
       "      <td>-77.222646</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[300.00000000000006, 246.25676191160235, 191.3...</td>\n",
       "      <td>[-2.8394662073566574, -2.8391695700790827, -2....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-1855.288353</td>\n",
       "      <td>-77.303681</td>\n",
       "      <td>2.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[300.0, 261.24462668919415, 225.95779944928805...</td>\n",
       "      <td>[-2.8405855542363496, -2.8403753879099565, -2....</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition   scm_energy  \\\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                    12 -1855.994228   \n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                   12 -3706.687014   \n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                    12 -1855.288353   \n",
       "\n",
       "                                  scm_energy_per_atom  system_id  \\\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif            -77.333093        0.0   \n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif           -77.222646        1.0   \n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif            -77.303681        2.0   \n",
       "\n",
       "                                                                           scm_step  \\\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "\n",
       "                                                                           scm_Temp  \\\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [300.0, 244.72130136787882, 155.4044352312756,...   \n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [300.00000000000006, 246.25676191160235, 191.3...   \n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [300.0, 261.24462668919415, 225.95779944928805...   \n",
       "\n",
       "                                                              scm_energy_per_atom_t  \n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [-2.8434003513590658, -2.843102065120439, -2.8...  \n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [-2.8394662073566574, -2.8391695700790827, -2....  \n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [-2.8405855542363496, -2.8403753879099565, -2....  "
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "2a717825",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 690x200 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[6.9, 2], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    spec = fig.add_gridspec(1, 2, wspace=0.35)\n",
    "    ax1 = fig.add_subplot(spec[0])\n",
    "    ax2 = fig.add_subplot(spec[1])\n",
    "    idx = np.arange(len(results_df.index))\n",
    "    for sim in results_df.index:\n",
    "        name = sim.split(\"-\")[2]\n",
    "        ax1.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_Temp\"], label=name)\n",
    "        ax2.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_energy_per_atom_t\"], label=name)\n",
    "\n",
    "    ax1.set_ylabel(\"Temperature [K]\")\n",
    "    ax2.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax1.set_xlabel(\"Time [fs]\")\n",
    "    ax2.set_xlabel(\"Time [fs]\")\n",
    "    ax1.legend()\n",
    "    ax2.legend()\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"8-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a5fda97",
   "metadata": {},
   "source": [
    "#### Save frames "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "bba95370",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "conf_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8.1-bulk_300K_frames\")\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8-bulk_300K\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)\n",
    "\n",
    "make_dir(conf_dir)\n",
    "last_n_frames = 10\n",
    "nskip = 2\n",
    "frame_results_df = pd.DataFrame(columns=[\"step\", \"time\", \"unitcell_repetition\"])\n",
    "\n",
    "mols_ase = []\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    # sim_name = sim_name.replace(\"10.0-\", \"8.0-\")\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    name = sim_name[3:]\n",
    "    steps = job_result.get_history_property(\"Step\")\n",
    "    n_frame = len(steps)\n",
    "    n_rep = results_df.loc[f\"{name}.cif\", \"unitcell_repetition\"]\n",
    "    for i_frame in range(n_frame - last_n_frames * nskip + 1, n_frame, nskip):\n",
    "        frame_name = f\"{name}_{steps[i_frame] * TIMESTEP:1.1f}ps\"\n",
    "        mol = job.results.get_history_molecule(i_frame + 1)\n",
    "        # print(os.path.join(conf_dir, f'{name}.cif'))\n",
    "        mol = toASE(mol)\n",
    "        mol.write(os.path.join(conf_dir, f\"{frame_name}.cif\"), format=\"cif\")\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"unitcell_repetition\"] = n_rep\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"step\"] = steps[i_frame]\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"time\"] = steps[i_frame] * TIMESTEP\n",
    "        mols_ase.append(mol.copy())\n",
    "\n",
    "frame_results_df.to_json(os.path.join(conf_dir, \"info.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "5a408932",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps.cif</th>\n",
       "      <td>100</td>\n",
       "      <td>25.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps.cif</th>\n",
       "      <td>200</td>\n",
       "      <td>50.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps.cif</th>\n",
       "      <td>300</td>\n",
       "      <td>75.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps.cif</th>\n",
       "      <td>400</td>\n",
       "      <td>100.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps.cif</th>\n",
       "      <td>500</td>\n",
       "      <td>125.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         step   time unitcell_repetition\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps.cif   100   25.0                  12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps.cif   200   50.0                  12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps.cif   300   75.0                  12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps.cif  400  100.0                  12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps.cif  500  125.0                  12"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frame_results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "891533c1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b56023f2673d4d4c904226d60db9cf3f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=29), VBox(children=(Dropdown(description='Show', options=('All', 'Li', 'F')…"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(mols_ase, viewer=\"ngl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b856aba",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5426811b",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1dac2df4",
   "metadata": {},
   "source": [
    "## Bulk 300 K `Single Point` <a id='300_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c9202d9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"8.1-Bulk_300K_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c757efd",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8626c95e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_7_SP\n",
    "\n",
    "print(settings_7_SP)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9c0afcc",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "624ef865",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8e1410f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4289d597",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_7_SP, file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f05d3f4",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62d9094c",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ef840613",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8.1-bulk_300K_frames\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23ef2454",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5300da83",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_7_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2021.104\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77d6ddd8",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d594b680",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:00:37] JOB jobs STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps STARTED\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:25] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:05:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:07:43] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:08:55] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:09:01] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:09:07] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:09:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:10:26] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:12:44] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps reported warnings. Please check the the output\n",
      "[23:12:44] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:44] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:12:50] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps reported warnings. Please check the the output\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:13:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:14:02] Job jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps reported warnings. Please check the the output\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:14:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:41:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps FAILED\n",
      "[23:42:00] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c3c0d4e",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bda00240",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba0dabd8",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"8.1-bulk_300K_frames\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bc027cc2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif</th>\n",
       "      <td>1600</td>\n",
       "      <td>400</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif</th>\n",
       "      <td>1700</td>\n",
       "      <td>425</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif</th>\n",
       "      <td>1800</td>\n",
       "      <td>450</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif</th>\n",
       "      <td>1900</td>\n",
       "      <td>475</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif</th>\n",
       "      <td>2000</td>\n",
       "      <td>500</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          step  time  unitcell_repetition\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif  1600   400                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif  1700   425                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif  1800   450                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif  1900   475                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif  2000   500                   27"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4c52e70e",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in np.sort(glob.glob(os.path.join(dir_jobs, \"*\"))):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_7_SP)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_per_atom_SP\"] = energy_per_atom\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd46993c",
   "metadata": {},
   "source": [
    "### Finishing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa72fdb0",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3b0a04db",
   "metadata": {},
   "source": [
    "## Bulk 500 K `ab initio MD` <a id='500_aimd'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "40ca5525",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"9-Bulk_500K_MD\"\n",
    "TASK = \"MD\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0241c13c",
   "metadata": {},
   "source": [
    "#### Setting `DFTB` *ab initio MD*"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71f1ba50",
   "metadata": {},
   "source": [
    "MD settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "7926d994",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      ams: \t\n",
      "          task: \tMolecularDynamics\n",
      "          MolecularDynamics: \t\n",
      "                            NSteps: \t1000\n",
      "                            TimeStep: \t0.25\n",
      "                            Trajectory: \t\n",
      "                                       PrintFreq: \t5\n",
      "                                       SamplingFreq: \t50\n",
      "                            InitialVelocities: \t\n",
      "                                              Temperature: \t300\n",
      "                            Thermostat: \t\n",
      "                                       Type: \tBerendsen\n",
      "                                       Temperature: \t300\n",
      "                                       Tau: \t25\n",
      "      DFTB: \t\n",
      "           Model: \tNonSCC-GFN1-xTB\n",
      "\n"
     ]
    }
   ],
   "source": [
    "STEPS = 1000\n",
    "TIMESTEP = 0.25  # [fs]\n",
    "TEMP = 300  # [K]\n",
    "\n",
    "from tools.misc.default_settings import settings_MD\n",
    "\n",
    "settings_MD.input.ams.MolecularDynamics.NSteps = STEPS\n",
    "settings_MD.input.ams.MolecularDynamics.TimeStep = TIMESTEP\n",
    "settings_MD.input.ams.MolecularDynamics.InitialVelocities.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Temperature = TEMP\n",
    "settings_MD.input.ams.MolecularDynamics.Thermostat.Tau = int(100 * TIMESTEP)\n",
    "\n",
    "print(settings_MD)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35f290cf",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "2383b775",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 1\n",
    "CORES_PER_NODE = 128\n",
    "CPUS_PER_CORE = 2\n",
    "MAX_JOBS = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa340fb4",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "eb138ca8",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_MD, file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8cec58f7",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c51d66c7",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "6aa57046",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9-bulk_500K\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ca33be9",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b338311d",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_MD, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2022.102\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c2344ea",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5970936c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[12:36:49] JOB jobs STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-0-LiF_Fm-3m_-3.18_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-1-LiF_P6_3mc_-3.17_2x2x3 STARTED\n",
      "[12:36:49] JOB jobs/MD-11.0-2-LiF_Pm-3m_-2.89_2x2x3 STARTED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "579103c3",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e70196d1",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "07e55764",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9-bulk_500K\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "d8cf716c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                    12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                   12\n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                    12"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "01d80d0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_df = pd.DataFrame(index=results_df.index, columns=[\"scm_step\", \"scm_Temp\", \"scm_energy_per_atom_t\"])\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    # sim_name = sim_name.replace(\"11.0-\", \"9.0-\")\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    Nentry = job_result.read_rkf_section(\"History\")[\"nEntries\"]\n",
    "    energy = scm.Units.convert(\n",
    "        job_result.read_rkf_section(\"History\")[f\"Energy({Nentry})\"], \"au\", \"eV\"\n",
    "    )  # scm.Units.convert(job_result.get_energy(), 'au', 'eV')\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"scm_energy_per_atom\"] = energy_per_atom\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"system_id\"] = int(sim_name.split(\"-\")[2])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_step\"] = np.array(job_result.get_history_property(\"Step\"))\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_Temp\"] = np.array(job_result.read_rkf_section(\"MDHistory\")[\"Temperature(1)\"])\n",
    "    temp_df.at[sim_name[3:] + \".cif\", \"scm_energy_per_atom_t\"] = (\n",
    "        np.array(job_result.get_history_property(\"Energy\")) / Na\n",
    "    )\n",
    "\n",
    "results_df = pd.concat([results_df, temp_df], axis=1)\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "3ea873f6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>scm_energy</th>\n",
       "      <th>scm_energy_per_atom</th>\n",
       "      <th>system_id</th>\n",
       "      <th>scm_step</th>\n",
       "      <th>scm_Temp</th>\n",
       "      <th>scm_energy_per_atom_t</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-1855.574178</td>\n",
       "      <td>-77.315591</td>\n",
       "      <td>0.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[499.9999999999999, 416.923323063444, 271.1582...</td>\n",
       "      <td>[-2.8434003513590658, -2.842952894535068, -2.8...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-3705.435048</td>\n",
       "      <td>-77.196564</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[499.9999999999999, 389.0072525130963, 325.165...</td>\n",
       "      <td>[-2.8394662073566574, -2.8388537012790294, -2....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "      <td>-1854.542496</td>\n",
       "      <td>-77.272604</td>\n",
       "      <td>2.0</td>\n",
       "      <td>[0, 50, 100, 150, 200, 250, 300, 350, 400, 450...</td>\n",
       "      <td>[500.0, 424.927725103416, 320.27012569595587, ...</td>\n",
       "      <td>[-2.8405855542363496, -2.8401801382610117, -2....</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  unitcell_repetition   scm_energy  \\\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                    12 -1855.574178   \n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                   12 -3705.435048   \n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                    12 -1854.542496   \n",
       "\n",
       "                                  scm_energy_per_atom  system_id  \\\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif            -77.315591        0.0   \n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif           -77.196564        1.0   \n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif            -77.272604        2.0   \n",
       "\n",
       "                                                                           scm_step  \\\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [0, 50, 100, 150, 200, 250, 300, 350, 400, 450...   \n",
       "\n",
       "                                                                           scm_Temp  \\\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [499.9999999999999, 416.923323063444, 271.1582...   \n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [499.9999999999999, 389.0072525130963, 325.165...   \n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [500.0, 424.927725103416, 320.27012569595587, ...   \n",
       "\n",
       "                                                              scm_energy_per_atom_t  \n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif   [-2.8434003513590658, -2.842952894535068, -2.8...  \n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif  [-2.8394662073566574, -2.8388537012790294, -2....  \n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif   [-2.8405855542363496, -2.8401801382610117, -2....  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "c02ce972",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 690x200 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[6.9, 2], dpi=100)\n",
    "\n",
    "with plt.style.context(\"./articles_seaborn.mplstyle\"):\n",
    "    spec = fig.add_gridspec(1, 2, wspace=0.35)\n",
    "    ax1 = fig.add_subplot(spec[0])\n",
    "    ax2 = fig.add_subplot(spec[1])\n",
    "    idx = np.arange(len(results_df.index))\n",
    "    for sim in results_df.index:\n",
    "        name = sim.split(\"-\")[2]\n",
    "        ax1.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_Temp\"], label=name)\n",
    "        ax2.plot(results_df.loc[sim, \"scm_step\"] * TIMESTEP, results_df.loc[sim, \"scm_energy_per_atom_t\"], label=name)\n",
    "\n",
    "    ax1.set_ylabel(\"Temperature [K]\")\n",
    "    ax2.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax1.set_xlabel(\"Time [fs]\")\n",
    "    ax2.set_xlabel(\"Time [fs]\")\n",
    "    ax1.legend()\n",
    "    ax2.legend()\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(os.path.join(\"plots\", f\"9-{CHEM_FORMULA}_{TASK}_sim_energy.pdf\"), dpi=300, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb2576f9",
   "metadata": {},
   "source": [
    "#### Save frames "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "afe98808",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "conf_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9.1-bulk_500K_frames\")\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9-bulk_500K\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)\n",
    "\n",
    "make_dir(conf_dir)\n",
    "last_n_frames = 10\n",
    "nskip = 2\n",
    "frame_results_df = pd.DataFrame(columns=[\"step\", \"time\", \"unitcell_repetition\"])\n",
    "\n",
    "mols_ase = []\n",
    "\n",
    "for job_dir in glob.glob(os.path.join(dir_jobs, \"*\")):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    # sim_name = sim_name.replace(\"11.0-\", \"9.0-\")\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_MD)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    name = sim_name[3:]\n",
    "    steps = job_result.get_history_property(\"Step\")\n",
    "    n_frame = len(steps)\n",
    "    n_rep = results_df.loc[f\"{name}.cif\", \"unitcell_repetition\"]\n",
    "    for i_frame in range(n_frame - last_n_frames * nskip + 1, n_frame, nskip):\n",
    "        frame_name = f\"{name}_{steps[i_frame] * TIMESTEP:1.1f}ps\"\n",
    "        mol = job.results.get_history_molecule(i_frame + 1)\n",
    "        # print(os.path.join(conf_dir, f'{name}.cif'))\n",
    "        mol = toASE(mol)\n",
    "        mol.write(os.path.join(conf_dir, f\"{frame_name}.cif\"), format=\"cif\")\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"unitcell_repetition\"] = n_rep\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"step\"] = steps[i_frame]\n",
    "        frame_results_df.loc[f\"{frame_name}.cif\", \"time\"] = steps[i_frame] * TIMESTEP\n",
    "        mols_ase.append(mol.copy())\n",
    "\n",
    "frame_results_df.to_json(os.path.join(conf_dir, \"info.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "ce837531",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps.cif</th>\n",
       "      <td>100</td>\n",
       "      <td>25.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps.cif</th>\n",
       "      <td>200</td>\n",
       "      <td>50.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps.cif</th>\n",
       "      <td>300</td>\n",
       "      <td>75.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps.cif</th>\n",
       "      <td>400</td>\n",
       "      <td>100.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps.cif</th>\n",
       "      <td>500</td>\n",
       "      <td>125.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         step   time unitcell_repetition\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps.cif   100   25.0                  12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps.cif   200   50.0                  12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps.cif   300   75.0                  12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps.cif  400  100.0                  12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps.cif  500  125.0                  12"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frame_results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "515e6a53",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e82677d5780249c4a40f0359fded44ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=29), VBox(children=(Dropdown(description='Show', options=('All', 'Li', 'F')…"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(mols_ase, viewer=\"ngl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d8b1eb0",
   "metadata": {},
   "source": [
    "#### Finishing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c518c9a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3a8d009f",
   "metadata": {},
   "source": [
    "## Bulk 500 K `Single Point` <a id='500_sp'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c80e9124",
   "metadata": {},
   "outputs": [],
   "source": [
    "SIMsNAME = \"9.1-Bulk_500K_SP\"\n",
    "TASK = \"SP\"\n",
    "RUN_DIR = os.path.join(MAIN_DIR_JNB2, CHEM_FORMULA, SIMsNAME)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d52f3076",
   "metadata": {},
   "source": [
    "### Setting `BAND` *Single Point*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "00c99fd1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input: \t\n",
      "      AMS: \t\n",
      "          task: \tSinglePoint\n",
      "          Properties: \t\n",
      "                     Gradients: \tyes\n",
      "                     StressTensor: \tno\n",
      "                     Hessian: \tno\n",
      "                     PESPointCharacter: \tno\n",
      "                     ElasticTensor: \tno\n",
      "          GeometryOptimization: \t\n",
      "                               OptimizeLattice: \tno\n",
      "      BAND: \t\n",
      "           basis: \t\n",
      "                 type: \tDZP\n",
      "           xc: \t\n",
      "              GGA: \tPBE\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tools.misc.default_settings import settings_7_SP\n",
    "\n",
    "print(settings_7_SP)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58f8177e",
   "metadata": {},
   "source": [
    "#### Set-up `RUNNER`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "de622944",
   "metadata": {},
   "outputs": [],
   "source": [
    "NODES = 4\n",
    "CORES_PER_NODE = 32\n",
    "CPUS_PER_CORE = 4\n",
    "MAX_JOBS = 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "703c22dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.init(folder=RUN_DIR)\n",
    "RUNNER = scm.GridRunner(grid=\"slurm\", sleepstep=5, parallel=True, maxjobs=MAX_JOBS)\n",
    "RUNNER.settings.workdir = \"-D\"\n",
    "RUNNER.settings.special.nodes = \"--nodes=\"\n",
    "RUNNER.settings.special.qos = \"--qos=\"\n",
    "RUNNER.settings.special.walltime = \"--time=\"\n",
    "RUNNER.settings.special.tasks_per_node = \"--ntasks-per-node=\"\n",
    "RUNNER.settings.special.cpus_per_task = \"--cpus-per-task=\"\n",
    "RUNNER.settings.special.partition = \"--partition=\"\n",
    "RUNNER.settings.special.account = \"--account=\"\n",
    "RUNNER.settings.special.job_name = \"--job-name=\"\n",
    "\n",
    "scm.config.log.file = 7  # debug\n",
    "scm.config.log.stdout = 1  # 1: important 3: normal 5: verbose 7: debug\n",
    "scm.config.job.pickle = False\n",
    "scm.config.job.runscript.nproc = CORES_PER_NODE\n",
    "scm.config.job.runscript.shebang = \"#!/bin/bash\"\n",
    "scm.config.default_jobrunner = RUNNER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c098f76c",
   "metadata": {},
   "outputs": [],
   "source": [
    "settings_pkl_file = os.path.join(RUN_DIR, \"settings.pkl\")\n",
    "with open(settings_pkl_file, \"wb\") as file:\n",
    "    pickle.dump(settings_7_SP, file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b94e50bb",
   "metadata": {},
   "source": [
    "#### Set-up JOBS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ab87d89",
   "metadata": {},
   "source": [
    "Load systems"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7c3f1a74",
   "metadata": {},
   "outputs": [],
   "source": [
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9.1-bulk_500K_frames\")\n",
    "systems = []\n",
    "\n",
    "for path in np.sort(glob.glob(os.path.join(cis_dir, \"*.cif\"))):\n",
    "    system_name = path.split(os.sep)[-1].split(\".cif\")[0]\n",
    "    system_mat = Structure.from_file(path)\n",
    "    system_ase = matget2ase.get_atoms(system_mat)\n",
    "    systems.append((system_name, fromASE(system_ase)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5be8a3e5",
   "metadata": {},
   "source": [
    "Define jobs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a56b142",
   "metadata": {},
   "outputs": [],
   "source": [
    "jobs = []\n",
    "for name, system_obj in systems:\n",
    "    sim_name = f\"{TASK}-{name}\"\n",
    "    job = scm.AMSJob(molecule=system_obj, settings=settings_7_SP, name=sim_name)\n",
    "    job.settings.runscript.pre = \"\"\"module use --append /work/pr1u1751/pr1u1751/shared/modules\n",
    "module load ams/2021.104\"\"\"\n",
    "    jobs.append(job)\n",
    "# Group all\n",
    "multi_jobs = scm.MultiJob(name=\"jobs\", children=jobs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18163732",
   "metadata": {},
   "source": [
    "#### Run All"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3792b705",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[23:00:37] JOB jobs STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps STARTED\n",
      "[23:00:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps STARTED\n",
      "[23:00:38] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps STARTED\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:03:55] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:04:01] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_150.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:05:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:05:25] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:05:37] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_25.0ps SUCCESSFUL\n",
      "[23:07:43] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:08:55] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_225.0ps SUCCESSFUL\n",
      "[23:09:01] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:09:07] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:09:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:10:26] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:12:44] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps reported warnings. Please check the the output\n",
      "[23:12:44] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:44] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_75.0ps SUCCESSFUL\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_125.0ps SUCCESSFUL\n",
      "[23:12:50] Job jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps reported warnings. Please check the the output\n",
      "[23:12:50] JOB jobs/SP-11.0-1-LiF_P6_3mc_-3.17_2x2x3_100.0ps SUCCESSFUL\n",
      "[23:13:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_50.0ps SUCCESSFUL\n",
      "[23:14:02] Job jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps reported warnings. Please check the the output\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:02] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_200.0ps SUCCESSFUL\n",
      "[23:14:07] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_175.0ps SUCCESSFUL\n",
      "[23:14:13] JOB jobs/SP-11.0-0-LiF_Fm-3m_-3.18_2x2x3_250.0ps SUCCESSFUL\n",
      "[23:41:56] JOB jobs/SP-11.0-2-LiF_Pm-3m_-2.89_2x2x3_150.0ps FAILED\n",
      "[23:42:00] JOB jobs FAILED\n"
     ]
    }
   ],
   "source": [
    "result = multi_jobs.run(\n",
    "    jobrunner=RUNNER,\n",
    "    nodes=NODES,\n",
    "    cores=CORES_PER_NODE * NODES,\n",
    "    tasks_per_node=CORES_PER_NODE,\n",
    "    cpus_per_task=CPUS_PER_CORE,\n",
    "    walltime=\"24:00:00\",\n",
    "    partition=\"standard\",\n",
    "    account=\"pr1u1751\",\n",
    "    qos=\"standard\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9313f03",
   "metadata": {},
   "source": [
    "#### Post-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df5120e6",
   "metadata": {},
   "source": [
    "Load Material Project data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "254250d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_jobs = os.path.join(RUN_DIR, \"jobs\")\n",
    "\n",
    "cis_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, \"9.1-bulk_500K_frames\")\n",
    "data_path = os.path.join(cis_dir, \"info.json\")\n",
    "results_df = pd.read_json(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "90b5c553",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>step</th>\n",
       "      <th>time</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif</th>\n",
       "      <td>1600</td>\n",
       "      <td>400</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif</th>\n",
       "      <td>1700</td>\n",
       "      <td>425</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif</th>\n",
       "      <td>1800</td>\n",
       "      <td>450</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif</th>\n",
       "      <td>1900</td>\n",
       "      <td>475</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif</th>\n",
       "      <td>2000</td>\n",
       "      <td>500</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          step  time  unitcell_repetition\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_400.0ps.cif  1600   400                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_425.0ps.cif  1700   425                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_450.0ps.cif  1800   450                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_475.0ps.cif  1900   475                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3_500.0ps.cif  2000   500                   27"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7e6a887a",
   "metadata": {},
   "outputs": [],
   "source": [
    "for job_dir in np.sort(glob.glob(os.path.join(dir_jobs, \"*\"))):\n",
    "    sim_name = job_dir.split(os.sep)[-1]\n",
    "    job = scm.AMSJob.load_external(job_dir, settings=settings_7_SP)\n",
    "    # check\n",
    "    if not job.check():\n",
    "        print(f\"Simulation {job_dir} FAILED\")\n",
    "        continue\n",
    "    job_result = job.results\n",
    "    energy = scm.Units.convert(job_result.get_energy(), \"au\", \"eV\")\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_SP\"] = energy\n",
    "    Na = len(job_result.get_molecule(\"Molecule\").atoms)\n",
    "    energy_per_atom = energy / Na\n",
    "    results_df.loc[sim_name[3:] + \".cif\", \"energy_per_atom_SP\"] = energy_per_atom\n",
    "\n",
    "results_df.to_json(os.path.join(RUN_DIR, \"results.json\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34fcd066",
   "metadata": {},
   "source": [
    "### Finishing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26ba3335",
   "metadata": {},
   "outputs": [],
   "source": [
    "scm.finish()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  },
  "toc-autonumbering": true,
  "toc-showcode": false,
  "toc-showmarkdowntxt": false,
  "toc-showtags": false,
  "vscode": {
   "interpreter": {
    "hash": "68c6dd4486945bf06834acdebaf1156968f51509e7c23272cbd0a64c0886d511"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}