NuMuCC_validation_trackstudies_august5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%matplotlib inline\n",
    "#%load_ext nb_black"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "import helpfunction as helper\n",
    "import uproot\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.patches as patches\n",
    "from matplotlib import colors\n",
    "from matplotlib.colors import LogNorm\n",
    "import scipy as scipy\n",
    "from scipy.stats import norm\n",
    "import time\n",
    "import os\n",
    "\n",
    "pd.set_option(\"display.max_columns\", 50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "gr = (1 + 5 ** 0.5) / 2\n",
    "rgb_c2 = np.array([44, 160, 44]) / 255\n",
    "\n",
    "# Rename input files to ./input/run1/X_run1.root, where X is 'nu', 'on', 'off', and 'dirt'\n",
    "\n",
    "outdir = \"./output/dataMC/\"\n",
    "inputdir = \"./input/run1/\"\n",
    "main_tree_name = \"NuCCanalyzer\"\n",
    "#dict_names = [\"nue\", \"nu\", \"on\", \"off\", \"dirt\"]\n",
    "dict_names = [\"nu\", \"on\", \"off\", \"dirt\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "fidvol_vtx = [24.8, 31.55, 32.47, 32, 26.9, 30]\n",
    "\n",
    "mass_mu = 0.106\n",
    "mass_p = 0.938272\n",
    "kin_min = 0.04\n",
    "kin_min_mu = kin_min + mass_mu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "muon_candidate_selection = \"track_score>0.90 &\\\n",
    "                            vtx_distance<4.0 &\\\n",
    "                            track_chi2_proton>60.0 &\\\n",
    "                            track_chi2_muon<30.0 &\\\n",
    "                            (track_chi2_proton/track_chi2_muon)>7.0 &\\\n",
    "                            generation==2 &\\\n",
    "                            track_length>20\"\n",
    "\n",
    "event_selection = \"nu_pdg==14 &\\\n",
    "                   nu_contained &\\\n",
    "                   daughters_start_contained &\\\n",
    "                   longest_muon_candidate_length==track_length &\\\n",
    "                   nu_score>0.4\"\n",
    "\n",
    "# daughters_start_contained &\\ "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['strk_mid_gpar_vx', 'strk_mid_gpar_vy', 'strk_mid_gpar_vz', 'strk_mid_gpar_px', 'strk_mid_gpar_py', 'strk_mid_gpar_pz']\n",
      "['strk_tk1_gpar_vx', 'strk_tk1_gpar_vy', 'strk_tk1_gpar_vz', 'strk_tk1_gpar_px', 'strk_tk1_gpar_py', 'strk_tk1_gpar_pz']\n",
      "['strk_tk2_gpar_vx', 'strk_tk2_gpar_vy', 'strk_tk2_gpar_vz', 'strk_tk2_gpar_px', 'strk_tk2_gpar_py', 'strk_tk2_gpar_pz']\n",
      "('nu:\\tAll events: ', 1045525, '')\n",
      "('\\tSelected events: ', 153957, '(0.00%)', '')\n",
      "('\\tDuplicated:', 0)\n",
      "['strk_mid_gpar_vx', 'strk_mid_gpar_vy', 'strk_mid_gpar_vz', 'strk_mid_gpar_px', 'strk_mid_gpar_py', 'strk_mid_gpar_pz']\n",
      "['strk_tk1_gpar_vx', 'strk_tk1_gpar_vy', 'strk_tk1_gpar_vz', 'strk_tk1_gpar_px', 'strk_tk1_gpar_py', 'strk_tk1_gpar_pz']\n",
      "['strk_tk2_gpar_vx', 'strk_tk2_gpar_vy', 'strk_tk2_gpar_vz', 'strk_tk2_gpar_px', 'strk_tk2_gpar_py', 'strk_tk2_gpar_pz']\n",
      "('on:\\tAll events: ', 27129, '')\n",
      "('\\tSelected events: ', 27129, '(100.00%)', '')\n",
      "('\\tDuplicated:', 0)\n",
      "['strk_mid_gpar_vx', 'strk_mid_gpar_vy', 'strk_mid_gpar_vz', 'strk_mid_gpar_px', 'strk_mid_gpar_py', 'strk_mid_gpar_pz']\n",
      "['strk_tk1_gpar_vx', 'strk_tk1_gpar_vy', 'strk_tk1_gpar_vz', 'strk_tk1_gpar_px', 'strk_tk1_gpar_py', 'strk_tk1_gpar_pz']\n",
      "['strk_tk2_gpar_vx', 'strk_tk2_gpar_vy', 'strk_tk2_gpar_vz', 'strk_tk2_gpar_px', 'strk_tk2_gpar_py', 'strk_tk2_gpar_pz']\n",
      "('off:\\tAll events: ', 4716, '')\n",
      "('\\tSelected events: ', 4716, '(100.00%)', '')\n",
      "('\\tDuplicated:', 0)\n",
      "['strk_mid_gpar_vx', 'strk_mid_gpar_vy', 'strk_mid_gpar_vz', 'strk_mid_gpar_px', 'strk_mid_gpar_py', 'strk_mid_gpar_pz']\n",
      "['strk_tk1_gpar_vx', 'strk_tk1_gpar_vy', 'strk_tk1_gpar_vz', 'strk_tk1_gpar_px', 'strk_tk1_gpar_py', 'strk_tk1_gpar_pz']\n",
      "['strk_tk2_gpar_vx', 'strk_tk2_gpar_vy', 'strk_tk2_gpar_vz', 'strk_tk2_gpar_px', 'strk_tk2_gpar_py', 'strk_tk2_gpar_pz']\n",
      "('dirt:\\tAll events: ', 207426, '')\n",
      "('\\tSelected events: ', 3789, '(0.00%)', '')\n",
      "('\\tDuplicated:', 103301)\n"
     ]
    }
   ],
   "source": [
    "d = {}\n",
    "tree_names = [\"\"]\n",
    "\n",
    "flist = os.listdir(inputdir)\n",
    "\n",
    "for sample in dict_names:\n",
    "    file_name = \"\"\n",
    "    for fn in flist:\n",
    "        if sample + \"_\" in fn:\n",
    "            file_name = fn\n",
    "    file = uproot.open(inputdir + file_name)[main_tree_name]\n",
    "    d[sample] = {}\n",
    "    d[sample][\"subruns\"] = file[\"subruns\"]\n",
    "    d[sample][\"events\"] = file[\"Event\"]\n",
    "    d[sample][\"daughters\"] = file[\"Daughters\"]\n",
    "\n",
    "    # Flatten vectors into fields\n",
    "    keys = list(map(lambda x: x.decode('utf-8'), d[sample][\"daughters\"].keys()))\n",
    "    use_strk = 'strk_mid_gpar' in keys\n",
    "    \n",
    "    if use_strk:\n",
    "        n = len(d[sample][\"daughters\"])\n",
    "\n",
    "        def expand_gpar(field_name):\n",
    "            keys.remove(field_name)\n",
    "            data_rect = np.zeros(shape=(n,6), dtype=np.float32)\n",
    "            data = d[sample][\"daughters\"].array(field_name)\n",
    "            for i in range(len(data)):\n",
    "                data_rect[i:i + 1, 0:len(data[i])] = data[i]\n",
    "                \n",
    "            suffix = ['vx', 'vy', 'vz', 'px', 'py', 'pz']\n",
    "            fields = ['%s_%s' % (field_name, s) for s in suffix]\n",
    "            print(fields)\n",
    "            return fields, data_rect.T\n",
    "\n",
    "        mid_gpar_fields, mid_gpar_data = expand_gpar('strk_mid_gpar')\n",
    "        tk1_gpar_fields, tk1_gpar_data = expand_gpar('strk_tk1_gpar')\n",
    "        tk2_gpar_fields, tk2_gpar_data = expand_gpar('strk_tk2_gpar')\n",
    "\n",
    "    all_events = d[sample][\"events\"].numentries\n",
    "    sel_events = sum(d[sample][\"events\"].array(\"nu_mu_cc_selected\"))\n",
    "    print(sample + \":\\tAll events: \", all_events, \"\")\n",
    "    print(\n",
    "        \"\\tSelected events: \",\n",
    "        sel_events,\n",
    "        \"({:.02f}%)\".format(sel_events / all_events * 100),\n",
    "        \"\",\n",
    "    )\n",
    "\n",
    "    d[sample][\"event_hash\"] = helper.eventHash(\n",
    "        d[sample][\"events\"].pandas.df([\"evt_time_sec\", \"evt_time_nsec\"])\n",
    "    )\n",
    "    d[sample][\"forbidden_hashes\"] = d[sample][\"event_hash\"][\n",
    "        d[sample][\"event_hash\"].duplicated()\n",
    "    ]\n",
    "    print(\"\\tDuplicated:\", len(d[sample][\"forbidden_hashes\"]))\n",
    "\n",
    "    d[sample][\"df_daughters\"] = d[sample][\"daughters\"].pandas.df(keys)\n",
    "\n",
    "    if use_strk:\n",
    "        for x, y in zip(*(mid_gpar_fields, mid_gpar_data)):\n",
    "            d[sample][\"df_daughters\"][x] = y\n",
    "        for x, y in zip(*(tk1_gpar_fields, tk1_gpar_data)):\n",
    "            d[sample][\"df_daughters\"][x] = y\n",
    "        for x, y in zip(*(tk2_gpar_fields, tk2_gpar_data)):\n",
    "            d[sample][\"df_daughters\"][x] = y\n",
    "\n",
    "    d[sample][\"df_daughters\"][\"hash\"] = helper.eventHash(\n",
    "        d[sample][\"df_daughters\"][[\"evt_time_sec\", \"evt_time_nsec\"]]\n",
    "    )\n",
    "    d[sample][\"df_daughters\"][\"duplicated\"] = d[sample][\"df_daughters\"][\"hash\"].isin(\n",
    "        d[sample][\"forbidden_hashes\"]\n",
    "    )\n",
    "    d[sample][\"df_daughters\"] = (\n",
    "        d[sample][\"df_daughters\"].query(\"~duplicated\").copy().reset_index(drop=True)\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### POT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./input/run1/pot/run_subrun_on.txt\n",
      "./input/run1/pot/run_subrun_off.txt\n"
     ]
    }
   ],
   "source": [
    "# Data:\n",
    "if True:\n",
    "    samples = [\"on\", \"off\"]\n",
    "    for sample in samples:\n",
    "        out_name = inputdir + \"pot/run_subrun_\" + sample + \".txt\"\n",
    "        print(out_name)\n",
    "        run_subrun = np.array(\n",
    "            d[sample][\"subruns\"].arrays([\"run\", \"subRun\"], outputtype=pd.DataFrame)\n",
    "        )\n",
    "        np.savetxt(out_name, run_subrun, fmt=\"%d\")\n",
    "\n",
    "pot_on = 1.467e20\n",
    "triggers_on = 35425849\n",
    "triggers_off = 31725245\n",
    "\n",
    "d[\"on\"][\"scale\"] = 1\n",
    "d[\"off\"][\"scale\"] = triggers_on / triggers_off\n",
    "title_str = \"MicroBooNE {0:.2g} POT, INTERNAL\".format(pot_on)\n",
    "title_mc_str = r\"$\\nu_\\mu$ CC Filter, MicroBooNE INTERNAL\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('nu', ':\\t', 1.3081447199159453e+21)\n",
      "('dirt', ':\\t', 6.417133149580343e+20)\n"
     ]
    }
   ],
   "source": [
    "# MC\n",
    "samples_mc = [\"nue\", \"nu\", \"dirt\"]\n",
    "for smc in samples_mc:\n",
    "    if smc in dict_names:\n",
    "        pot = sum(d[smc][\"subruns\"].array(\"pot\"))\n",
    "        pot_corr = pot * (\n",
    "            1 - len(d[smc][\"forbidden_hashes\"]) * 2 / d[smc][\"events\"].numentries\n",
    "        )\n",
    "        d[smc][\"scale\"] = pot_on / pot_corr\n",
    "        print(smc, \":\\t\", pot_corr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Add fields to daughters: MC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Fidvol: 10cm everywhere, 50 cm on z end\n",
    "- abs(pdgcode)=14 and charged current\n",
    "- min kin E of the muon, 40MeV\n",
    "- for pi0 channel: exactly one daughter is a pi0\n",
    "- also split up using the interaxtion type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Add the signal categories, only needed for the nu sample:\n",
    "cols = [\n",
    "    \"mc_nu_ccnc\",\n",
    "    \"mc_nu_lepton_energy\",\n",
    "    \"mc_nu_pdg\",\n",
    "    \"mc_nu_vx\",\n",
    "    \"mc_nu_vy\",\n",
    "    \"mc_nu_vz\",\n",
    "]\n",
    "\n",
    "df_nu_event = d[\"nu\"][\"events\"].pandas.df(cols)\n",
    "\n",
    "nu_tpc_mask = helper.inTPC_mask(df_nu_event, \"mc_nu_vx\", \"mc_nu_vy\", \"mc_nu_vz\")\n",
    "\n",
    "nu_fid_vol_mask = helper.inTPC_mask(\n",
    "    df_nu_event, \"mc_nu_vx\", \"mc_nu_vy\", \"mc_nu_vz\", fidvol=fidvol_vtx\n",
    ")\n",
    "\n",
    "sig_def_ccinc = \"abs(mc_nu_pdg)==14 & mc_nu_ccnc==0 & mc_nu_lepton_energy>@kin_min_mu\"\n",
    "sig_mask_ccinc = df_nu_event.eval(sig_def_ccinc) & nu_fid_vol_mask\n",
    "\n",
    "# Has the event one pi0\n",
    "pi0_index = d[\"nu\"][\"events\"].array(\"mc_nu_daughter_pdg\") == 111\n",
    "sig_mask_pi0 = sig_mask_ccinc & (pi0_index.sum() == 1)\n",
    "\n",
    "# Has x protons\n",
    "sig_masks_proton = []\n",
    "m1 = d[\"nu\"][\"events\"].array(\"mc_nu_daughter_pdg\") == 2212\n",
    "m2 = d[\"nu\"][\"events\"].array(\"mc_nu_daughter_energy\") > mass_p + kin_min\n",
    "proton_index = m1 & m2\n",
    "sig_masks_proton = [\n",
    "    sig_mask_ccinc & (proton_index.sum() == x_proton) for x_proton in range(4)\n",
    "]\n",
    "\n",
    "\n",
    "# Add MC info\n",
    "mapper = dict(zip(d[\"nu\"][\"event_hash\"], nu_tpc_mask))\n",
    "d[\"nu\"][\"df_daughters\"][\"mc_nu_in_tpc\"] = d[\"nu\"][\"df_daughters\"][\"hash\"].map(mapper)\n",
    "mapper = dict(zip(d[\"nu\"][\"event_hash\"], nu_fid_vol_mask))\n",
    "d[\"nu\"][\"df_daughters\"][\"mc_nu_fid_vol\"] = d[\"nu\"][\"df_daughters\"][\"hash\"].map(mapper)\n",
    "\n",
    "mapper = dict(zip(d[\"nu\"][\"event_hash\"], sig_mask_ccinc))\n",
    "d[\"nu\"][\"df_daughters\"][\"mc_sig_nuccinc\"] = d[\"nu\"][\"df_daughters\"][\"hash\"].map(mapper)\n",
    "mapper = dict(zip(d[\"nu\"][\"event_hash\"], sig_mask_pi0))\n",
    "d[\"nu\"][\"df_daughters\"][\"mc_sig_nuccpi0\"] = d[\"nu\"][\"df_daughters\"][\"hash\"].map(mapper)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nu\n",
      "dirt\n"
     ]
    }
   ],
   "source": [
    "mapper_list = [\n",
    "    \"mc_nu_energy\",\n",
    "    \"mc_nu_pdg\",\n",
    "    \"mc_nu_ccnc\",\n",
    "    \"mc_nu_lepton_theta\",\n",
    "    \"mc_nu_vtx_distance\",\n",
    "    \"clustered_hit_completeness\",\n",
    "    \"matched_hit_fraction\",\n",
    "]\n",
    "\n",
    "for smc in samples_mc:\n",
    "    if smc in dict_names:\n",
    "        print(smc)\n",
    "        for field in mapper_list:\n",
    "            mapper = dict(zip(d[smc][\"event_hash\"], d[smc][\"events\"].array(field)))\n",
    "            d[smc][\"df_daughters\"][field] = d[smc][\"df_daughters\"][\"hash\"].map(mapper)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Add fields to daughters: All samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nu\n",
      "on\n",
      "off\n",
      "dirt\n"
     ]
    }
   ],
   "source": [
    "mapper_list = [\n",
    "    \"nu_pdg\",\n",
    "    \"num_tracks\",\n",
    "    \"num_showers\",\n",
    "    \"obvious_cosmic_chi2\",\n",
    "    \"nu_flash_chi2\",\n",
    "    \"nu_score\",\n",
    "    \"nu_contained\",\n",
    "    \"daughters_start_contained\",\n",
    "    \"event_weight\",\n",
    "]\n",
    "\n",
    "for sample in dict_names:\n",
    "    print(sample)\n",
    "    #if sample in samples_mc:\n",
    "    #    maplist = mapper_list + ['v1xpos','v2xpos','v1ypos','v2ypos','v1zpos','v2zpos']\n",
    "    #else:\n",
    "    #    maplist = mapper_list\n",
    "\n",
    "    for field in mapper_list:\n",
    "        mapper = dict(zip(d[sample][\"event_hash\"], d[sample][\"events\"].array(field)))\n",
    "        d[sample][\"df_daughters\"][field] = d[sample][\"df_daughters\"][\"hash\"].map(mapper)\n",
    "\n",
    "        d[sample][\"df_daughters\"][[\"theta_beam\", \"phi_beam\"]] = helper.cosmic_angles(\n",
    "            *d[sample][\"df_daughters\"][\n",
    "                [\"track_dirx\", \"track_diry\", \"track_dirz\"]\n",
    "            ].T.values\n",
    "        )\n",
    "\n",
    "        end_mask = helper.inTPC_mask(\n",
    "            d[sample][\"df_daughters\"],\n",
    "            \"track_endx\",\n",
    "            \"track_endy\",\n",
    "            \"track_endz\",\n",
    "            [10] * 6,\n",
    "        )\n",
    "        d[sample][\"df_daughters\"][\"contained\"] = (\n",
    "            d[sample][\"df_daughters\"][\"start_contained\"] & end_mask\n",
    "        )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Selection queries\n",
    "query_select_wo_trackcut = event_selection\n",
    "query_select = query_select_wo_trackcut + \" & track_length>20.0\"\n",
    "query_pure = \"mc_nu_vtx_distance<5 & abs(mc_pdg)==13\"\n",
    "query_event = query_select + \" & \" + query_pure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "for sample in dict_names:\n",
    "    # Tag the muon candidate\n",
    "    d[sample][\"df_daughters\"][\"muon_candidate\"] = d[sample][\"df_daughters\"].eval(\n",
    "        muon_candidate_selection\n",
    "    )\n",
    "    d[sample][\"df_daughters\"].loc[\n",
    "        d[sample][\"df_daughters\"][\"muon_candidate\"], \"longest_muon_candidate_length\"\n",
    "    ] = (\n",
    "        d[sample][\"df_daughters\"]\n",
    "        .query(\"muon_candidate\")\n",
    "        .groupby([\"hash\"])[\"track_length\"]\n",
    "        .transform(max)\n",
    "    )\n",
    "    d[sample][\"df_daughters\"].loc[\n",
    "        ~d[sample][\"df_daughters\"][\"muon_candidate\"], \"longest_muon_candidate_length\"\n",
    "    ] = -1\n",
    "\n",
    "    d[sample][\"df_daughters\"][\"chi2_ratio\"] = d[sample][\"df_daughters\"].eval(\n",
    "        \"track_chi2_proton/track_chi2_muon\"\n",
    "    )\n",
    "    d[sample][\"df_daughters\"].loc[\n",
    "        d[sample][\"df_daughters\"][\"muon_candidate\"], \"highest_chi2_ratio\"\n",
    "    ] = (\n",
    "        d[sample][\"df_daughters\"]\n",
    "        .query(\"muon_candidate\")\n",
    "        .groupby([\"hash\"])[\"chi2_ratio\"]\n",
    "        .transform(max)\n",
    "    )\n",
    "    d[sample][\"df_daughters\"].loc[\n",
    "        ~d[sample][\"df_daughters\"][\"muon_candidate\"], \"highest_chi2_ratio\"\n",
    "    ] = -1\n",
    "\n",
    "    # Tag the event selection\n",
    "    d[sample][\"df_daughters\"][\"python_select\"] = d[sample][\"df_daughters\"].eval(\n",
    "        query_select\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    0.684469\n",
       "0.0    0.242996\n",
       "2.0    0.069538\n",
       "3.0    0.002733\n",
       "4.0    0.000263\n",
       "Name: muon_candidate, dtype: float64"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How much muon candidates per event?\n",
    "grouped = (\n",
    "    d[\"nu\"][\"df_daughters\"]\n",
    "    .query(\"mc_sig_nuccinc\")[[\"hash\", \"muon_candidate\"]]\n",
    "    .groupby(\"hash\")\n",
    ")\n",
    "grouped.sum()[\"muon_candidate\"].value_counts() / len(grouped)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Efficiency/Purity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Muon neutrino CC inc selection efficiency: (0.00+-0.16)%\n",
      "Muon neutrino CC pi0 selection efficiency: (0.00+-0.45)%\n",
      "Muon neutrino CC inc selection purity (basic): (61.96+-0.41)%\n",
      "Muon neutrino CC inc selection purity (strict): (53.85+-0.43)%\n"
     ]
    }
   ],
   "source": [
    "# Using python:\n",
    "num_sel = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\")[\"python_select\"].sum()\n",
    "num_denom = sum(sig_mask_ccinc)\n",
    "error = helper.effErr(num_sel, num_denom)\n",
    "print(\n",
    "    \"Muon neutrino CC inc selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        num_sel / num_denom * 100, error * 100\n",
    "    )\n",
    ")\n",
    "\n",
    "num_sel = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccpi0\")[\"python_select\"].sum()\n",
    "num_denom = sum(sig_mask_pi0)\n",
    "error = helper.effErr(num_sel, num_denom)\n",
    "print(\n",
    "    \"Muon neutrino CC pi0 selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        num_sel / num_denom * 100, error * 100\n",
    "    )\n",
    ")\n",
    "# purity\n",
    "purity_nom = (\n",
    "    d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_select).sum()\n",
    "    * d[\"nu\"][\"scale\"]\n",
    ")\n",
    "purity_nom_plus = (\n",
    "    d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_event).sum()\n",
    "    * d[\"nu\"][\"scale\"]\n",
    ")\n",
    "purity_denom = 0\n",
    "\n",
    "for s in [\"off\", \"nu\"]: #, \"dirt\"]:\n",
    "    purity_denom += d[s][\"df_daughters\"][\"python_select\"].sum() * d[s][\"scale\"]\n",
    "\n",
    "error = helper.effErr(purity_nom, purity_denom)\n",
    "purity = purity_nom / purity_denom\n",
    "print(\n",
    "    \"Muon neutrino CC inc selection purity (basic): ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        purity * 100, error * 100\n",
    "    )\n",
    ")\n",
    "error = helper.effErr(purity_nom_plus, purity_denom)\n",
    "purity_plus = purity_nom_plus / purity_denom\n",
    "print(\n",
    "    \"Muon neutrino CC inc selection purity (strict): ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        purity_plus * 100, error * 100\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Muon neutrino CC inc selection efficiency: (0.00+-0.18)%\n",
      "Muon neutrino CC pi0 selection efficiency: (0.00+-0.47)%\n",
      "Muon neutrino CC inc selection purity (basic): (48.27+-0.34)%\n"
     ]
    }
   ],
   "source": [
    "# Using C++ analyser:\n",
    "num_sel = d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\")[sig_mask_ccinc].sum()\n",
    "num_denom = sum(sig_mask_ccinc)\n",
    "error = helper.effErr(num_sel, num_denom)\n",
    "eff = num_sel / num_denom\n",
    "print(\n",
    "    \"Muon neutrino CC inc selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        num_sel / num_denom * 100, error * 100\n",
    "    )\n",
    ")\n",
    "\n",
    "num_sel = d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\")[sig_mask_pi0].sum()\n",
    "num_denom = sum(sig_mask_pi0)\n",
    "error = helper.effErr(num_sel, num_denom)\n",
    "print(\n",
    "    \"Muon neutrino CC pi0 selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        num_sel / num_denom * 100, error * 100\n",
    "    )\n",
    ")\n",
    "\n",
    "purity_nom = (\n",
    "    d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\")[sig_mask_ccinc].sum()\n",
    "    * d[\"nu\"][\"scale\"]\n",
    ")\n",
    "purity_denom = 0\n",
    "\n",
    "for s in [\"off\", \"nu\"]: #, \"dirt\"]:\n",
    "    purity_denom += d[s][\"events\"].array(\"nu_mu_cc_selected\").sum() * d[s][\"scale\"]\n",
    "\n",
    "error = helper.effErr(purity_nom, purity_denom)\n",
    "purity = purity_nom / purity_denom\n",
    "print(\n",
    "    \"Muon neutrino CC inc selection purity (basic): ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "        purity * 100, error * 100\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13      0.941912\n",
       "211     0.034230\n",
       "2212    0.020282\n",
       "2112    0.002573\n",
       "22      0.000699\n",
       "321     0.000185\n",
       "3222    0.000119\n",
       "Name: mc_pdg, dtype: float64"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# What is the quality of the selected signal event:\n",
    "d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc & python_select\")[\n",
    "    \"mc_pdg\"\n",
    "].abs().value_counts() / len(\n",
    "    d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc & python_select\")\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2IAAAD6CAYAAADDcvQhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi40LCBodHRwOi8vbWF0cGxvdGxpYi5vcmcv7US4rQAAIABJREFUeJzs3XmYFcXZ9/HvzY5RAZEQZCBAIAoMq6yKiguLGMUVUB9BH5W4oEYTFXfc8qhRcQkuGBcwRiAogoYIiKKYVxxAEQVUUEA2BVk1KOv9/tE1hzPDLGfOnDMMM7/PdZ1r+lRXd1fPcNNdXdVV5u6IiIiIiIhIyamwrwsgIiIiIiJS3qgiJiIiIiIiUsJUERMRERERESlhqoiJiIiIiIiUMFXERERERERESpgqYiIiIiIiIiVMFTEREREREZESpoqYiEgZYWYLzKx73PdlZnbSPixSnnKXU0REpDxSRUxEJM3M7Dwzm2NmP5rZGjP7t5l1S3R9rn0tM7OfQt7sz2EA7t7S3WcUsF1aK2XhGNvN7NBc6R+bmZtZo8LKmYLjZ/9uNprZv8ysQQr3vdbMfhGXdomZzcjn+Nmfvxawv5Nyfc93/7n2uTvXcc4v7Ni51n9rZi+Y2YFFOb+49Bnh91u1oHMSEZGCqSImIpJGZnYd8AjwZ6Au0BB4AuibyPp8nOruB8Z9VqfxFDCzSkXIvhQ4N27bVsABJViGU939QKAe8B3weHGPHacicE0ix4/7DEnF/uP3CXyT6zgvJXjs7N9NW6AdcFNRzy9Upo8BHDitCOcmIiK5qCImIlIAM1tqZieH5Rpm9n18S0Ih29YA7gKudPdX3f2/7r7D3V939+sLW59EWfNskTCzF4kqeK+HFpEbQvphZvaKma0L53l1rn3daGbzgf8WoSL0IjAw7vsgYHR+5TSzBmb2aijD+jxacXKUwcyahxaZTaGLY56VAXf/GRgPtIjbX4HbJrDvvwB/MrOaCf4uiird+wfA3b8FphBVyIp6/IHALOAFor+tiIgkSRUxEZGCrQbqA7j7ZmAR0CXBbbsC1YAJSa5PCXe/gJytKA+YWQXgdeATovM7EfiDmfWK2/Rc4BSgprvvTPBws4CDQ6WmIjAA+HteGcP6N4DlQKNQjjG5ssXKAFgo81Tgl8BVwEtmdnge+z4A6B/Kg5lVLmjbwtYHc4AZwJ8S/F0UVbr3D4CZZQAnA0uSOP5A4KXw6WVmddNRRhGR8kAVMRGRgsUqYkE1ogpBImoD3xdQiSlsfX5eC602m8zstSJum60jUMfd73L37e7+NfAMUcUp22PuvsLdfyrivrNbxXoQVVxX5ZOvE3AYcH1oDfzZ3d/PlSe+DF2AA4H7QpnfJqrInRuX/zUz2wRsDsf/S0gvbNtE9g1wO3CVmdXJ55zi/zabzOzSfPLlp7D9F6SwY79mZj8AK4C1wB1FOb5F7y3+Ghjn7nOBr4DzkiiniIgARen3LyJSHq0iVMTMrAPQBJiZ4LbrgUPNrFI+la3C1ufndHd/qwj58/Jr4LBQaclWkZzntiLJfb8IvAc0Jle3xFwaAMsLOff4MhwGrHD33XFpy8lZUT7d3d8KrW19gXfNrEUC2yayb9z9MzN7AxhKVMnMrVh/mwT2X5DCjp39uzkO+AdwKBD/9y/s+IOAqe7+ffj+j5A2vIjlFBER1CImIlKY1UCGmR1EVMG4EthuZtOzM+Q1slzwAbANOD3J9ankub6vAJa6e824z0Hu3qeAbRI7kPtyokE7+gCvFpB1BdCwkPfP4suwGmgQulVma0geLW7uvsvdXwV2Ad0S2DbhfRO1JF1KrkpaCqV1/+7+LtE7Xg8menwzqw70A44Loy5+C1wLtDGzNukop4hIWaeKmIhIwVYTtYJNAP7u7mPC968AzOwQYGNeG4Z3ym4HRpjZ6WZ2gJlVNrOTzeyBwtan+Dy+C+XOlgX8EAbDqG5mFc0s08w65rcDi4Y8fyHB410MnODu/y0gTxawBrjPzH5hZtXM7OgC8n8IbAVuCL+n7sCp7P1eGRbpC9QiatkpbNuE9+3uS4CxwNW516VCuvcfPAL0yKsSlc/xTyeq1LYgGuSjLdCcqAU1fnCWyuHvmP1RzxsRkXyoIiYiUrBVwOHAu+5+b0hrBcwPy62BT/Pb2N0fAq4DbgXWEbUCDQFeS2R9Cv0fcGt4d+hP7r4L+B3RDfVS4Hvgb0CNAvbRAPhPIgdz96/cfU4heXYRVXaaEg0mspJogI388m8P+U8O5X0CGOjun8dle93MfgS2APcCg9x9QWHbJrjveHcBv8gjPXtkyuxPsgOx5Lf/giR8bHdfR9Rt9PYEjz8IeN7dv3H3b7M/wF+B8+MqXJOBn+I+w4p4DiIi5Ya5J9XzRESk3DKz24Ev3X2MmT0GvOfu4/d1udLJzKoQjbDY2t137OvyiIiI7O9UERMRKSIz+yewG6hMNMDFF+5+w74tlYiIiOxP1HdbRKToDnH3E/d1IURERGT/pXfERESKwMyqEY10KCIiIpI0dU0UEREREREpYWoRExERERERKWGqiImIiIiIiJQwVcRERERERERKmCpiIiIiIiIiJUwVMRERERERkRKmipiIiIiIiEgJU0VMRERERESkhKkiJiIiIiIiUsJUERMRERERESlhlfZ1AUraoYce6o0aNdrXxZAyZO7cud+7e519XY59TbElqabY2kPxJamm+NpD8SWplmh8lbuKWKNGjZgzZ86+LoaUIWa2fF+XoTRQbEmqKbb2UHxJqim+9lB8SaolGl/qmigiIiIiIlLC0loRM7NlZvapmc0zszkh7RAzm2Zmi8PPWiHdzOwxM1tiZvPNrH3cfgaF/IvNbFBc+pFh/0vCtpbO8xEREREREUmFkmgRO97d27p7h/B9KDDd3ZsB08N3gJOBZuEzGHgSooobcAfQGegE3JFdeQt5Lo3brnf6T0dERERERKR49sU7Yn2B7mF5FDADuDGkj3Z3B2aZWU0zqxfyTnP3DQBmNg3obWYzgIPdfVZIHw2cDvy7JE5ix44drFy5kp9//rkkDielQLVq1cjIyKBy5cr7uij7DcVJ2aV42PcUX2WX4mvfU3yVXaUpvtJdEXNgqpk58LS7jwTquvuasP5boG5Yrg+siNt2ZUgrKH1lHul7MbPBRK1sNGzYsDjns+dgK1dy0EEH0ahRI9Qjsuxzd9avX8/KlStp3Ljxvi7OfkNxUjYpHkoHxVfZpPgqHRRfZVNpi690d03s5u7tibodXmlmx8avDK1fnuYy4O4j3b2Du3eoUyc1I7X+/PPP1K5dW8FZTpgZtWvX1pOxIlKclE2Kh9JB8VU2Kb5KB8VX2VTa4iutLWLuvir8XGtmE4je8frOzOq5+5rQ9XBtyL4KaBC3eUZIW8WerozZ6TNCekYe+ZPSaOi/8l237L5T8kxXcJYv+nsnR7+3sqmof9dk/o+Vwim+yqZUxZdiq3gUX2VTUf6u6b52pa1FzMx+YWYHZS8DPYHPgElA9siHg4CJYXkSMDCMntgF2By6ME4BeppZrTBIR09gSli3xcy6hNESB8btq1wYPnw4LVu2JDMzk3PPPTdWu3/77bdp3749mZmZDBo0iJ07d+a5/ahRo2jWrBnNmjVj1KhRAGzbto3evXuTmZnJE088Ecs7ePBgPvroo6TKOWfOHK6++moAZsyYwf/7f/8vtu7CCy9k/Pjxe22zevVqzj777CId58ILL6Rx48a0bduWtm3bMm/evKTKK2XLo48+SmZmJi1btuSRRx7Jse7xxx/niCOOoGXLltxwww15bv/mm29y+OGH07RpU+67775Y+vnnn0/r1q25+eabY2n33HMPr732WpHKN2PGDH73u98VaZviWLp0KZ07d6Zp06b079+f7du375Vn+/btXHTRRbRq1Yo2bdowY8aMEiuf7F8Kii+Ahx56CDPj+++/z3P7dF+HXnjhBYYMGVKkbYpj7ty5tGrViqZNm3L11VcTdfzJaePGjZxxxhm0bt2aTp068dlnn5VY+WT/kt993sUXX0ybNm1o3bo1Z599Nj/++GOe2//f//0fTZs25fDDD2fKlCkArFu3jm7dupGZmZnjetW3b19Wr15dpPINGzaMBx98MMmzK7r8rsfxvvnmG44//njatWtH69atmTx5comVLxnpbBGrC0wItc5KwD/c/U0zmw2MM7OLgeVAv5B/MtAHWAJsBS4CcPcNZnY3MDvkuyt74A7gCuAFoDrRIB0lMlBHXgqqMSejsFr2qlWreOyxx1i4cCHVq1enX79+jBkzhoEDBzJo0CCmT5/Ob3/7W26//XZGjRrFxRdfnGP7DRs2cOeddzJnzhzMjCOPPJLTTjuNmTNn0q1bN26++WaOPvporrjiCj755BN27dpF+/bt8ylNwTp06ECHDtGgmTNmzODAAw/kqKOOKnCbww47LM8KWmH+8pe/FLkCJyWnpOPks88+45lnniErK4sqVarQu3dvfve739G0aVPeeecdJk6cyCeffELVqlVZu3btXtvv2rWLK6+8kmnTppGRkUHHjh057bTT2LlzJ9WrV2f+/Pn06NGDzZs3s3XrVj788ENuvfXWlJ5jqt14441ce+21DBgwgMsuu4xnn32Wyy+/PEeeZ555BoBPP/2UtWvXcvLJJzN79mwqVNDUk6VZaYovgBUrVjB16tR8380uyetQSbn88st55pln6Ny5M3369OHNN9/k5JNPzpHnz3/+M23btmXChAl8/vnnXHnllUyfPn0flVgSVVru8y688EKGDx/OwQcfDMB1113HX//6V4YOHZpj+4ULFzJmzBgWLFjA6tWrOemkk/jyyy95+eWXueyyyzjzzDPp06cPp59+Oq+//jrt2rXjsMMOS+k5plJ+1+MWLVrkyHfPPffQr18/Lr/8chYuXEifPn1YtmzZvil0AtJ2VXX3r929Tfi0dPd7Q/p6dz/R3Zu5+0nZlSqPXOnuv3H3Vu4+J25fz7l70/B5Pi59jrtnhm2GeF6PnsqwnTt38tNPP7Fz5062bt3KYYcdxvr166lSpQq//e1vAejRowevvPLKXttOmTKFHj16cMghh1CrVi169OjBm2++SeXKldm6dSs7duyIPcm77bbbuPvuu/MtR6tWrdi0aRPuTu3atRk9ejQAAwcOZNq0abEn/suWLeOpp55i+PDhtG3blpkzZwLw3nvvcdRRR9GkSZNY5WvZsmVkZmYC0RPNM888k969e9OsWbN8Wy7y8+OPP8ae7rdu3Tr2+zjwwAO5/vrradmyJSeddBJZWVl0796dJk2aMGnSpCIdQ0qnRYsW0blzZw444AAqVarEcccdx6uvvgrAk08+ydChQ6latSoAv/zlL/faPisri6ZNm9KkSROqVKnCgAEDmDhxIpUrV+ann35i9+7d7Nixg4oVK3L77bdz55135luWZcuWccwxx9C+fXvat2+fo2V4y5YtnHLKKRx++OFcdtll7N69G4CXX36ZVq1akZmZyY033gjAU089xfXXXx/bNv6J/9///nc6depE27Zt+f3vf8+uXbtylMHdefvtt2MPKwYNGpRnC97ChQs54YQTYr+XmjVrMmfOnL3ySflWUHwBXHvttTzwwAP5dgNK5XUoKyuLrl270q5dO4466ii++OKL2LoVK1bQvXt3mjVrliNGH374YTIzM8nMzIy15g0dOpQRI0bE8sQ/8f/LX/5Cx44dad26NXfcccdeZVizZg1btmyhS5cumBkDBw4sNL6OOOIIli1bxnfffZfvuZUlZlbNzLLM7BMzW2Bmd4b0F8xsaZh3dp6ZtQ3p5XqO2bzu84BYJczd+emnn/KMsYkTJzJgwACqVq1K48aNadq0KVlZWbH42rZtGxUrVmTnzp088sgjBd5bvf7663Tu3Jl27dpx0kkn5fj3+sknn9C1a1eaNWsWe4jn7lx//fVkZmbSqlUrxo4dC8CAAQP417/2VGize0Xt2rWL66+/PhZfTz/99F5lyO96nJuZsWXLFgA2b95cqiuXUDLziEka1K9fnz/96U80bNiQevXqUaNGDXr27Mmhhx7Kzp07YzdN48ePZ8WKFXttv2rVKho02PNKXkZGBqtWraJHjx4sW7aMLl26cPXVVzNp0iTat29f4D/ko48+mv/85z8sWLCAJk2axCpYH3zwQY6Wr0aNGnHZZZdx7bXXMm/ePI455hgguni9//77vPHGG3s90ck2b948xo4dy6effsrYsWPzPCeAW265hdatW3Pttdeybds2AO6++25q1KjBp59+yvz582MXwP/+97+ccMIJLFiwgIMOOohbb72VadOmMWHCBG6//fZ8z1f2H5mZmcycOZP169ezdetWJk+eHPu38+WXXzJz5kw6d+7Mcccdx+zZs/faPr84ad68OXXq1KF9+/aceuqpLFmyhN27dxf4tP6Xv/wl06ZN46OPPmLs2LGx7roQXWAef/xxFi5cyFdffcWrr77K6tWrufHGG3n77beZN28es2fP5rXXXuOss85iwoQJsW3Hjh3LgAEDWLRoEWPHjuU///kP8+bNo2LFirz00ks5yrB+/Xpq1qxJpUqVcpxPbm3atGHSpEns3LmTpUuXMnfu3HxjTsqvguJr4sSJ1K9fnzZt2uS7fSqvQ0cccQQzZ87k448/5q677srRZTgrK4tXXnmF+fPn889//pM5c+Ywd+5cnn/+eT788ENmzZrFM888w8cff0z//v0ZN25cbNtx48bRv39/pk6dyuLFi8nKymLevHnMnTuX9957b6/zycjY8+p6QfGVXWHNyspi+fLlrFy5cq98ZdQ24AR3bwO0JZqOqEtYd32Yd7atu2e/W1Bu55jN7z4v20UXXcSvfvUrPv/8c6666qq9ts8vvs477zwmTpxIjx49uPnmm3niiSe44IILOOCAA/ItS7du3Zg1axYff/wxAwYM4IEHHoitmz9/Pm+//TYffPABd911F6tXr+bVV19l3rx5fPLJJ7z11ltcf/31rFmzJkd8bd++nenTp3PKKafw7LPPUqNGDWbPns3s2bN55plnWLp0aULnk9uwYcP4+9//TkZGBn369OHxxx9P4Le97+yLecQkBTZu3MjEiRNZunQpNWvW5JxzzuHvf/87//M//8OYMWNiFZGePXtSsWLFhPdbqVIl/vGPfwDRHBq9evVi4sSJXHfddXzzzTcMHDiQ0047Lcc2xxxzDO+99x6//vWvufzyyxk5ciSrVq2iVq1a/OIXvyj0mKeffjoVKlSgRYsW+T4VPPHEE6lRowYALVq0YPny5TkCEqK+0L/61a/Yvn07gwcP5v777+f222/nrbfeYsyYMbF8tWpF/1dnd6WBqFWvatWqVK5cmVatWpXqZmxJXPPmzbnxxhvp2bMnv/jFL2jbtm0sHnbu3MmGDRuYNWsWs2fPpl+/fnz99dcJv8Qb/z7MqaeeytNPP829997LJ598Qo8ePbj00ktz5N+xYwdDhgyJVZK+/PLL2LpOnTrRpEkTAM4991zef/99KleuTPfu3cke6fX888/nvffe4/TTT6dJkybMmjWLZs2a8fnnn3P00UczYsQI5s6dS8eOHQH46aef8mzlS8T//u//smjRIjp06MCvf/1rjjrqqCL9PyLlQ37xtXXrVv785z8zderUpPabzHVo8+bNDBo0iMWLF2Nm7NixI7auR48e1K5dG4AzzzyT999/HzPjjDPOiF2jzjzzTGbOnMnVV1/N2rVrWb16NevWraNWrVo0aNCARx99lKlTp9KuXTsg6mmxePFijj02x2DQCRk6dCjXXHMNbdu2pVWrVrRr167cxFfouZT9QlPl8CmoN9N+NcdsKhV0nwfw/PPPs2vXLq666irGjh3LRRddlNB+a9SoEWuV2rhxI/fddx8TJkzg0ksvZePGjfzxj3+ka9euObZZuXIl/fv3Z82aNWzfvj3HsO99+/alevXqVK9eneOPP56srCzef/99zj33XCpWrEjdunVjDztPPvlkrrnmGrZt28abb77JscceS/Xq1Zk6dSrz58+P9YravHkzixcvTmp4+ZdffpkLL7yQP/7xj3zwwQdccMEFfPbZZ6W2a33pLJUU6q233qJx48bUqVOHypUrc+aZZ8a6OnXt2pWZM2eSlZXFscceG+umGK9+/fo5nnCvXLmS+vVzTsP2xBNPMHDgQGbNmkWNGjUYO3YsDz300F77OvbYY5k5cyYzZ86M3TiOHz8+1uJVmOyuYUCeLzbnzpPdlJ5bvXr1MDOqVq3KRRddRFZWVoHHrVy5cuymu0KFCrFjVKhQId8BTmT/c/HFF8eeXteqVSsWDxkZGZx55pmYGZ06daJChQp7DSiQSJxMnDiRI488kh9//JGvvvqKcePGMX78eLZu3Zoj3/Dhw6lbty6ffPIJc+bMyTFIRu7KX2GVwQEDBjBu3DheeeUVzjjjDMwMd2fQoEHMmzePefPm8cUXXzBs2LAc29WuXZtNmzbF/n3ndT4Q3QgPHz6cefPmMXHiRDZt2pTn/yMiecXXV199xdKlS2nTpg2NGjVi5cqVtG/fnm+//TbHtqm8Dt12220cf/zxfPbZZ7z++us5hqYuanydc845jB8/nrFjx9K/f38gujbddNNNsfhasmTJXu9e169fP0fLVn7xdfDBB/P8888zb948Ro8ezbp162IPYsoDM6toZvOIRs2e5u4fhlX3hu6Hw80s+6Kf1jlmzWyOmc1Zt25dsc8r1Qq6z8tWsWJFBgwYkOcrKInE1913380tt9zCyy+/TLdu3Rg1atRe1w2Aq666iiFDhvDpp5/y9NNPJx1f1apVo3v37kyZMmWv+Hr88cdj8bV06dIcrX+Jng/As88+S79+0fATXbt25eeff853sKDSQBWx/VTDhg2ZNWsWW7duxd2ZPn06zZs3B4gNOrBt2zbuv/9+Lrvssr2279WrF1OnTmXjxo1s3LiRqVOn0qtXr9j6jRs38sYbbzBw4EC2bt1KhQoVMDN++umnvfbVoEEDvv/+exYvXkyTJk3o1q0bDz74YJ5PCg866CB++OGHVP0aclizJpon3N157bXXYu+Y9ejRI0ef/40bN6bl+CUtPBkcb2afm9kiM+tqZoeY2bTQZ35adleN8tzPPjsevvnmG1599VXOO+88IGqJfeedd4Com+L27ds59NBDc2zbsWNHFi9ezNKlS9m+fTtjxozJ8SR+x44dsb718f30d+3atddohJs3b6ZevXpUqFCBF198Mcf7W1lZWSxdupTdu3czduxYunXrRqdOnXj33Xf5/vvv2bVrFy+//DLHHXccAGeccQYTJ07k5ZdfZsCAAUDUajx+/PjY+W7YsIHly5fnKIOZcfzxx8eeOo4aNYq+ffvu9TvbunUr//3vfwGYNm0alSpV2uuF6LLMzK4N7698ZmYvh/daGpvZhyEexppZlZC3avi+JKxvFLefm0L6F2bWKy69d0hbYmZ598feT+QVX61atWLt2rUsW7aMZcuWkZGRwUcffcSvfvWrHNum8jq0efPm2E3ZCy+8kGPdtGnT2LBhAz/99BOvvfYaRx99NMcccwyvvfZa7N/6hAkTYg8P+/fvz5gxYxg/fjznnHNOrKzPPfdcbHS6VatW7TXAT7169Tj44IOZNWsW7s7o0aPzjK9NmzbF/n/429/+xrHHHht756c8cPdd7t6WaNqhTmaWCdwEHAF0BA4BbiyBcqR8jtlUyu8+z91ZsmQJEN3vTJo0iSOOOGKv7U877TTGjBnDtm3bWLp0KYsXL6ZTp06x9YsXL2blypV07969SPGVPbpptokTJ/Lzzz+zfv16ZsyYQceOHTnmmGMYO3Ysu3btYt26dbz33nuxY/fv35/nn3+emTNnxnol9erViyeffDLWkv3ll1/GrkHZCrsex//esge/WbRoET///DOl8e+bTRWx/VTnzp05++yzad++Pa1atWL37t0MHjwYiF4obt68Oa1bt+bUU0+NvRM1Z84cLrnkEgAOOeQQbrvtNjp27EjHjh25/fbbOeSQQ2L7v+uuu7jllluoUKECvXr1YubMmbRq1YoLLrgg3/JkPzE/5phjWLVqFd26ddsr36mnnsqECRNyDNaRKueffz6tWrWiVatWfP/997HR62699VY2btxIZmYmbdq0id18lwGPAm+6+xFAG2ARMBSY7u7NgOnhO5TjfvZnnXUWLVq04NRTT2XEiBHUrFkTiLrfff3112RmZjJgwABGjRqFmbF69Wr69OkDRC1Df/3rX+nVqxfNmzenX79+tGzZMrbvESNGMGjQIA444ABat27N1q1badWqFUceeWTsONmuuOIKRo0aRZs2bfj8889zdNvt2LEjQ4YMoXnz5jRu3JgzzjiDevXqcd9993H88cfTpk0bjjzyyNhNXa1atWjevDnLly+PXdxatGjBPffcQ8+ePWndujU9evSIPZyId//99/Pwww/TtGlT1q9fH3uqP2nSpNi7kWvXrqV9+/Y0b96c+++/nxdffDFVf45Sz8zqA1cDHdw9E6gIDADuB4a7e1NgI5DdHHIxsDGkDw/5MLMWYbuWRLHzRGgJqAiMIIrJFsC5Ie9+Kb/4yk+6rkM33HADN910E+3atdurR0OnTp0466yzaN26NWeddRYdOnSgffv2XHjhhXTq1InOnTtzySWXxLodtmzZkh9++IH69etTr149AHr27Ml5551H165dadWqFWeffXaeDxWfeOIJLrnkEpo2bcpvfvOb2IiJTz31FE899RQQ3RxmZmZy+OGH8+9//5tHH3000V93meLum4B3gN7uviYM2rYNeJ7oegQFzzGbX3rK5pjdl/K7z8vu/ZB9v7NmzZrY/93x/4+3bNmSfv360aJFC3r37s2IESNydIG95ZZbuPfee4GoS/yTTz5Jx44dueaaa/Yqy7BhwzjnnHM48sgj93pg2bp1a44//ni6dOnCbbfdxmGHHRabnqFNmzaccMIJPPDAA7EHMT179uTdd9/lpJNOokqVKgBccskltGjRIjb10u9///u94rig6/Htt98eG2TtoYce4plnnqFNmzace+65vPDCC6V6PjgrZwMN0qFDB89r9K+iTti2aNGiWAsqgCO0AAAgAElEQVSUlB95/d3NbK67dyjJcphZDWAe0CR+tFAz+wLoHjdh+gx3P9zMng7LL8fny/64++9D+tNEE6bPAN4JlTzM7Nz4fHnJK7YUJ2VbUf6+yUyKuY9iqz4wi+jhxhbgNeBx4CXgV+6+08y6AsPcvZeZTQnLH5hZJeBboA7hIYi7/1/Y7xRgWDjMMHfvFdJvis+XH8VX+ZOK+CpoiPR9FF91gB3uvsnMqgNTiR5ezA3XLSN6oPGzuw81s1OAIUTTG3UGHnP3TuEh4lwgu3fHR8CRYcqjLKKHKR8STY30uLsXOJmU4qv8SfTvm+yEzonGlwbrENk/NQbWAc+bWRuiC9I1QN0w2TlEN4R1w3Ja+tmb2WCiFrZ85woS2Z+4+yozexD4BviJ6EZxLrDJ3bMf0cbHQyyGQiVtM1A7pM+K23X8NrljrnMaTkWkNKoHjAotwxWAce7+hpm9HSppRvSQMfudiv16jlmRwqgiJrJ/qkT0JPAqd//QzB5lTzdEIBqdyszS2uTt7iOBkRA9UUznsURKQuia25foYccm4J/so265etAhZY27zwfa5ZF+Qj75Hbgyn3XPAc/lkT4HyCxeSUVKht4RE9k/rQRWxo02NZ6oYvZd6JJI+Jn9Nnm57WcvUkQnAUvdfZ277wBeBY4Gaoauh5AzHmIxFNbXANZT9JjbS2kfTEBERIpHFbFiKG/v15V3penv7e7fAivM7PCQdCKwEJgEZI98OAjInnZ+EjAwjJ7YBdgcujBOAXqaWa3QEtATmBLWbTGzLqHP/sC4fRW1rMlsJqVcGf67fgN0MbMDwr/97Nh6Bzg75MkdW9kxdzbwdniKPwkYEEZVbEw04E0WUVeqZmEUxipEA3pMSrawZfjvUK7p71o66O9QNpWmv6u6JiapWrVqrF+/ntq1a5fq0VgkNdyd9evXU61atX1dlHhXAS+Fm7mvifrOVwDGmdnFwHKgX8i7T/rZK07KplIaDykRuvqOJ3r5fyfwMVH3238BY8zsnpD2bNjkWeBFM1sCbCCqWOHuC8xsHFElbidwpbvvAjCzIUQPQSoCz7n7gmTKqvgqm8pyfO1PFF9lU2mLL1XEkpSRkcHKlSspjZMASnpUq1aNjIyMwjOWEHefB+Q1Is+JeeTdJ/3sFSdlV2mLh1Ry9zuIpnWI9zV7htSOz/szcE4++7kXuDeP9MlED0eKRfFVdpXl+NpfKL7KrtIUX6qIJaly5co0btx4XxdDpFRTnIikj+JLJH0UX1IS9I6YiIiIiIhICVNFTEREREREpISpIiYiIiIiIlLCVBETEREREREpYaqIiYiIiIiIlDBVxEREREREREqYKmIiIiIiIiIlTBUxERERERGREqaKmIiIiIiISAlTRUxERERE0s7MqplZlpl9YmYLzOzOkN7YzD40syVmNtbMqoT0quH7krC+Udy+bgrpX5hZr7j03iFtiZkNLelzFCkKVcREREREpCRsA05w9zZAW6C3mXUB7geGu3tTYCNwcch/MbAxpA8P+TCzFsAAoCXQG3jCzCqaWUVgBHAy0AI4N+QVKZVUERMRERGRtPPIj+Fr5fBx4ARgfEgfBZwelvuG74T1J5qZhfQx7r7N3ZcCS4BO4bPE3b929+3AmJBXpFRKe0UsPKH42MzeCN/V/CwiIiJSDoX7wnnAWmAa8BWwyd13hiwrgfphuT6wAiCs3wzUjk/PtU1+6XmVY7CZzTGzOevWrUvFqYkUWUm0iF0DLIr7ruZnERERkXLI3Xe5e1sgg6gF64h9VI6R7t7B3TvUqVNnXxRBJL0VMTPLAE4B/ha+G2p+FhERESnX3H0T8A7QFahpZpXCqgxgVVheBTQACOtrAOvj03Ntk1+6SKmU7haxR4AbgN3he23U/CwiIiJS7phZHTOrGZarAz2Iek29A5wdsg0CJoblSeE7Yf3b7u4hfUB4raUx0AzIAmYDzcJrMFWIelRNSv+ZiSQnbRUxM/sdsNbd56brGIlS87OURWa2zMw+NbN5ZjYnpB1iZtPMbHH4WSukm5k9Ft6nnG9m7eP2MyjkX2xmg+LSjwz7XxK2tZI/SxERKUPqAe+Y2XyiStM0d38DuBG4zsyWED2EfzbkfxaoHdKvA4YCuPsCYBywEHgTuDJ0edwJDAGmEFXwxoW8IqVSpcKzJO1o4DQz6wNUAw4GHiU0P4dgyav5eWWCzc8UkC5SXhzv7t/HfR8KTHf3+8IANkOJLnAnEz0xbAZ0Bp4EOpvZIcAdQAeikavmmtkkd98Y8lwKfAhMJnpH898lc1oiIlLWuPt8oF0e6V8TvXKSO/1n4Jx89nUvcG8e6ZOJrlkipV7aWsTc/SZ3z3D3RkRNw2+7+/mo+VkkneLftcz9DuboMHTwLKIHIvWAXkRPJDeEytc0onld6gEHu/usEIej4/YlIiIiIsWUzhax/NwIjDGze4CPydn8/GJoft5AVLHC3ReYWXbz805C8zOAmWU3P1cEnlPzs5QzDkw1MweedveRQF13XxPWfwvUDctFfdeyfljOnS4iIiIiKVAiFTF3nwHMCMtqfhZJjW7uvsrMfglMM7PP41e6u4dKWtqY2WBgMEDDhg3TeSgRERGRMqUk5hETkTRw91Xh51pgAtEDju9Ct0LCz7Uhe1GH+l0VlnOn5y6DBsIRERERSYIqYiL7ITP7hZkdlL0M9AQ+I+e7lrnfwRwYRk/sAmwOXRinAD3NrFYYYbEnMCWs22JmXcJoiQPj9iUiIiIixbQv3hETkeKrC0wII8pXAv7h7m+a2WxgnJldDCwH+oX8k4E+RBOibwUuAnD3DWZ2N9HgNwB3ufuGsHwF8AJQnWi0RI2YKCIiIpIiqoiJ7IfCu5Zt8khfD5yYR7oDV+azr+eA5/JInwNkFruwIiIiIrIXdU0UEREREREpYaqIiYiIiIiIlDBVxEREREREREpYQu+ImVlFosEBYvnd/Zt0FUqkPFF8iaSP4kskPRRbIsVXaEXMzK4C7gC+A3aHZAdap7FcIuWC4kskfRRfIumh2BJJjURaxK4BDg+jsYlIaim+RNInqfgys5rA34hGDXXgf4EvgLFAI2AZ0M/dN4Z59h4lmh5iK3Chu38U9jMIuDXs9h53HxXSj2TP1BCTgWvCyKYi+wtdu0RSIJF3xFYAm9NdEJFySvElkj7JxtejwJvufgTRNBGLgKHAdHdvBkwP3wFOBpqFz2DgSQAzO4SoxaAz0Am4I0yaTshzadx2vZMoo8i+pGuXSAok0iL2NTDDzP4FbMtOdPeH01YqkfJD8SWSPkWOLzOrARwLXBjybge2m1lfoHvINgqYAdwI9AVGhxatWWZW08zqhbzTsidIN7NpQG8zmwEc7O6zQvpo4HQ0YbrsX5K6dplZA2A00btlDox090fNbBjRw4l1IevN7j45bHMTcDGwC7ja3aeE9N5ED00qAn9z9/tCemNgDFAbmAtcEOJYpNRJpCL2TfhUCR8RSR3Fl0j6JBNfjYluBp83szZEN3LXAHXdfU3I8y3RjSRAfaLWgWwrQ1pB6SvzSN+LmQ0mamWjYcOGCRZfpEQke+3aCfzR3T8ys4OAueEhBcBwd38wPrOZtQAGAC2Bw4C3zOy3YfUIoAdRDM02s0nuvhC4P+xrjJk9RVSJezKpsxRJs0IrYu5+Z0kURKQ8UnyJpE+S8VUJaA9c5e4fmtmj7OmGmL1fN7O0v9Pl7iOBkQAdOnTQO2RSaiR77QoPM9aE5R/MbBH5PIgI+gJj3H0bsNTMlhB19QVY4u5fA5jZGKBv2N8JwHkhzyhgGKqISSmVb0XMzB5x9z+Y2etEzcc5uPtpaS2ZSBmm+BJJn2LG10pgpbt/GL6PJ6qIfWdm9dx9Teh6uDasXwU0iNs+I6StYk9Xxuz0GSE9I4/8IqVeKq9dZtYIaAd8CBwNDDGzgcAcolazjUSVtFlxm8W3IOduce5M1B1xk7vvzCN/7uOrxVn2uYJaxF4MPx8sII+IJEfxJZI+SceXu39rZivM7HB3/wI4EVgYPoOA+8LPiWGTSUQ3kGOIbgQ3h8raFODPcQN09ARucvcNZrbFzLoQ3YAOBB5P7jRFSlxKrl1mdiDwCvAHd99iZk8CdxNV7u4GHiIarTRt1OIspUG+FTF3nxt+vmtmVYAjiALkC730KFI8ii+R9ElBfF0FvBS2/Rq4iGiU4XFmdjGwHOgX8k4mGrp+CdHw9ReFY28ws7uB2SHfXdkDdwBXsGf4+n+jgTpkP5GKa5eZVSaqhL3k7q+G/X0Xt/4Z4I3wNb8WZ/JJXw/UNLNKoVVMLc5SqiUyofMpwFPAV4ABjc3s9+6uC4dIMSm+RNIn2fhy93lAhzxWnZhHXgeuzGc/zwHP5ZE+h2iOMpH9UrKxFebdexZYFD/CYna33/D1DOCzsDwJ+IeZPUw0WEczICscs1kYIXEV0YAe54X3N98BziYaOTG+9Vqk1Elk1MSHgOPdfQmAmf0G+Bd6gieSCoovkfRRfImkR7KxdTRwAfCpmc0LaTcD55pZW6LWtWXA7wHcfYGZjSPqGrwTuNLdd4VjDgGmEA1f/5y7Lwj7uxEYY2b3AB8TVfxESqVEKmI/ZAda8DXwQ5rKI1LeKL5E0kfxJZIeScWWu79P1JqV2+QCtrkXuDeP9Ml5bRdGUuyUO12kNCpo1MQzw+IcM5sMjCN6UnEOe/q8i0gSFF8i6aP4EkkPxZZIahXUInZq3PJ3wHFheR1QLW0lEikfUhJfZlaRaKjfVe7+u9BffgzREL5zgQvcfbuZVQVGA0cSvczc392XhX3cRDTh5S7ganefEtJ7A48Sdfv4m7vfl+S5ipQ0Xb9E0kOxJZJCBY2aeFFJFkSkPElhfF0DLAIODt/vB4a7+xgze4qogvVk+LnR3Zua2YCQr7+ZtSB6ybkl0YvQb5nZb8O+RgA9iOZhmW1mk9x9YYrKLZI2un6JpIdiSyS1KuzrAohIcswsAzgF+Fv4bsAJRBPQAowCTg/LfcN3wvoTQ/6+wBh33+buS4mG4O4UPkvc/eswJPGYkFdEREREUkAVMZH91yPADcDu8L02sCnMnQJRS1b9sFwfWAEQ1m8O+WPpubbJL11EREREUqDAipiZVTCzfgXlKWDbamaWZWafmNkCM7szpDc2sw/NbImZjQ0TAmJmVcP3JWF9o7h93RTSvzCzXnHpvUPaEjMbmkw5RfaVYsbX74C12ZNr7itmNtjM5pjZnHXr1u3LoojkUJz4EpH8KbZEUqfAipi77yZ64p6MbcAJ7t4GaAv0NrMu7HmHpSmwkejdFYh7hwUYHvKR6x2W3sATZlYxDFIwAjgZaEE0B0WLJMsqUuKKGV9HA6eZ2TKiboMnEA2sUdPMst/9zCCa6JLwswFAWF+DaNCOWHqubfJLz30OI929g7t3qFOnTpKnIpJ6xYwvEcmHYkskdRLpmviWmf3JzBqY2SHZn8I28siP4Wvl8HH0DotIvGTj6yZ3z3D3RkQPKt529/OBd4CzQ7ZBwMSwPCl8J6x/2909pA8ILdKNgWZAFtEwxM1CC3aVcIxJKTljkZKTVHyJSKEUWyIpkMiEzv3Dzyvj0hxoUtiGodVqLtCUqPXqKxJ8h8XM4t9hmRW32/htcr/D0jmfcgwGBgM0bNiwsGKLlKSk4ysfNwJjzOwe4GPg2ZD+LPCimS0BNhBVrHD3BWY2DlgI7ASudPddAGY2BJhCNHz9c+6+IMkyiewrqY4vEYkotkRSoNCKmLs3Tnbn4YaurZnVBCYARyS7r+Jw95HASIAOHTr4viiDSF6KE19x+5gBzAjLXxO1FufO8zPRhJt5bX8vcG8e6ZOBycUtn8i+kor4EpG9KbZEUqPQrolmdoCZ3WpmI8P3ZmGggIS5+yaiLlNdKeF3WERKs1TEl4jkTfElkh6KLZHUSOQdseeB7cBR4fsq4J7CNjKzOqElDDOrTjQx7CL0DotIvKTiS0QSovgSSQ/FlkgKJFIR+427PwDsAHD3rYAlsF094B0zm09UaZrm7m8QvcNyXXhXpTY532GpHdKvA4aG4y0Ast9heZPwDkt4zyz7HZZFwDi9wyL7oWTjS0QKp/gSSQ/FlkgKJDJYx/bQouUAZvYboqHpC+Tu84F2eaTrHRaRPZKKLxFJiOJLJD0UW1KmNBr6r31y3ERaxO4gaolqYGYvAdPR/BEiqaL4EkkfxZdIeiQVW2G4+3fMbKGZLTCza0L6IWY2zcwWh5+1QrqZ2WNmtsTM5ptZ+7h9DQr5F5vZoLj0I83s07DNY2EqJJFSKZFRE6eZ2UdAF6Jm52vc/fu0l0ykHFB8iaSP4kskPYoRWzuBP7r7R2Z2EDDXzKYBFwLT3f0+MxtK9HrKjcDJRGMDNCOaouhJoHOYs+wOoANRq9xcM5vk7htDnkuBD4l6TfUG/p2iUxdJqUS6JgIcB3Qj+sdemWgoehFJDcWXSPoovkTSo8ix5e5rgDVh+QczW0Q0N2xfoHvINopoSpYbQ/roMHjbLDOraWb1Qt5p7r4BIFTmepvZDOBgd58V0kcDp6OKmJRSiQxf/wRwGfAp8BnwezMbke6CiZQHii+R9FF8iaRHKmLLzBoRjSXwIVA3VNIAvgXqhuX6wIq4zVaGtILSV+aRLlIqJdIidgLQPDyNwMxGARqdUCQ1FF8i6aP4EkmPYsWWmR0IvAL8wd23xL/G5e5uZp7i8uZVhsHAYICGDRum+3AieUpksI4lQPy/0AYhTUSKT/Elkj6KL5H0SDq2zKwyUSXsJXd/NSR/F7ocEn6uDemrwr6zZYS0gtIz8kjfi7uPdPcO7t6hTp06iRRdJOUSqYgdBCwysxlm9g7RfF4Hm9kkM9MEyiLFo/gSSR/Fl0h6JBVbYQTDZ4FF7v5w3KpJQPbIh4OAiXHpA8PoiV2AzaEL4xSgp5nVCiMs9gSmhHVbzKxLONbAuH2JlDqJdE28Pe2lECm/FF8i6aP4EkmPZGPraOAC4FMzmxfSbgbuA8aZ2cXAcqBfWDcZ6EPU2rYVuAjA3TeY2d3A7JDvruyBO4ArgBeA6kSDdGigDim1Ehm+/t2SKIhIeaT4EkkfxZdIeiQbW+7+PtFw93k5MY/8DlyZz76eA57LI30OkJlM+URKWiJdE0VERERERCSFVBETEREREREpYUWqiIWXIlunqzAi5ZniSyR9FF8i6aHYEkleIhM6zzCzg83sEOAj4Bkze7iw7USkcIovkfRRfImkh2JLJDUSaRGr4e5bgDOB0e7eGTgpvcUSKTcUXyLpo/gSSQ/FlkgKJFIRqxQm1+sHvJHm8oiUN0nFl5lVM7MsM/vEzBaY2Z0hvbGZfWhmS8xsrJlVCelVw/clYX2juH3dFNK/MLNecem9Q9oSMxuaqhMWKUFJX7/MrKKZfWxmb4Tvii2RPXRvKJICiVTE7iSaOG+Ju882sybA4vQWS6TcSDa+tgEnuHsboC3QO0x2eT8w3N2bAhuBi0P+i4GNIX14yIeZtQAGAC2B3sAT4Qa0IjACOBloAZwb8orsT4pz/boGWBT3XbElsofuDUVSIJGK2Bp3b+3uVwC4+9eA+gGLpEZS8eWRH8PXyuHjwAnA+JA+Cjg9LPcN3wnrTzQzC+lj3H2buy8lmjSzU/gscfev3X07MCbkFdmfJBVfZpYBnAL8LXw3FFsi8XRvKJICiVTEHk8wTUSKLun4Ck/X5wFrgWnAV8Amd98ZsqwE6ofl+sAKgLB+M1A7Pj3XNvml5y7DYDObY2Zz1q1bl0ixRUpSsvH1CHADsDt8r00JxxYovqRU072hSApUym+FmXUFjgLqmNl1casOBiqmu2AiZVkq4svddwFtzawmMAE4IuUFLbwMI4GRAB06dPCSPr5IXooTX2b2O2Ctu881s+7pK2XhFF9S2ujeUCS18q2IAVWAA0Oeg+LStwBnp7NQIuVAyuLL3TeZ2TtAV6CmmVUKT+YzgFUh2yqgAbDSzCoBNYD1cenZ4rfJL12ktCtOfB0NnGZmfYBqRDeYj6LYEgHdG4qkVL4VMXd/F3jXzF5w9+UlWCaRMq+48WVmdYAdoRJWHehBNEjAO0QXwzHAIGBi2GRS+P5BWP+2u7uZTQL+EeZ/OQxoBmQBBjQzs8ZEN4kDgPOSPmGRElSc+HL3m4CbAEKL2J/c/Xwz+yeKLSnndG8okloFtYhlq2pmI4FG8fnd/YR0FUqkHEk2vuoBo8IIbBWAce7+hpktBMaY2T3Ax8CzIf+zwItmtgTYQHTzh7svMLNxwEJgJ3Bl6PKImQ0hGhWrIvCcuy9IxQmLlKBUXr9uRLElkk33hiIpkEhF7J/AU0SjR+1Kb3FEyp2k4svd5wPt8kj/mmhUttzpPwPn5LOve4F780ifDExOtEwipVCxrl/uPgOYEZYVWyJ76N5QJAUSqYjtdPcn014SkfJJ8SWSPoovkfRIKrbM7Dkge0CczJA2DLgUyB4a9ObwsAIzu4lorr5dwNXuPiWk9yZ6d7Mi8Dd3vy+kNybqPlwbmAtcEKaJECmVEhm+/nUzu8LM6pnZIdmftJdMpHxQfImkj+JLJD2Sja0XiCY4z224u7cNn+xKWDKTouc38bpIqZRIi9ig8PP6uDQHmhS0kZk1AEYDdUP+ke7+aAjUsUT9ipcB/dx9Y5gA81GgD7AVuNDdPwr7GgTcGnZ9j7uPCulHEgV1daJuHte4u4b4lf1JUvElIglRfImkR1Kx5e7vmVmjBI8RmxQdWBrew8zuHrwkdBfGzMYAfc1sEdHE69mD34wChgFqFZdSq9CKmLs3TnLfO4E/uvtHZnYQMNfMpgEXAtPd/T4zGwoMJXoJ+mSiUaWaAZ2JAqdzqLjdAXQgCvK5ZjbJ3TeGPJcCHxJVxHoD/06yvCIlrhjxJSKFUHyJpEcaYmuImQ0E5hDdO24kmuh8Vlye+MnPc0+K3pmCJ14XKZUK7ZpoZgeY2a1hdBzMrFmY8LJA7r4mu0XL3X8AFhEFRF+ipxSEn6eH5b7AaI/MIpqzpR7QC5jm7htCYE4Deod1B7v7rNAKNjpuXyL7hWTjS0QKp/gSSY8Ux9aTwG+AtsAa4KEUFbNAZjbYzOaY2Zx169YVvoFIGiTyjtjzwHaimdQhmvfknqIcJDRDtyNquarr7mvCqm+Jui5CVEnL/YSjfiHpK/NIz+v4CjYprYodXyKSL8WXSHqkLLbc/Tt33+Xuu4Fn2NP9ML9J0fNLX0+YeD1Xen7HHenuHdy9Q506dZIpukixJVIR+427PwDsAHD3rUQTUibEzA4EXgH+4O5b4teFlqy0v9OlYJNSrFjxJSIFUnyJpEfKYiv0cMp2BvBZWJ4EDDCzqmE0xOxJ0WcTJkU3sypEA3pMCveU7xBNrA45J14XKZUSGaxju5lVJ1SYzOw3wLZEdm5mlYkqYS+5+6sh+Tszq+fua0LwrQ3pBT356J4rfUZIz8gjv8j+JOn4EpFCKb5E0iOp2DKzl4nu6Q41s5VEYwB0N7O2YV/LgN9D0pOi5zfxukiplEhFbBjwJtDAzF4CjiYacKNAYRTEZ4FF7v5w3KpJRE8p7iPn04pJRC9rjiF66XJzqKxNAf5sZrVCvp7ATe6+wcy2mFkXoi6PA4HHEzgfkdJkGEnEl4gkZBiKL5F0GEYSseXu5+aRnG9lqaiTouc38bpIaZXIqIlTzWwu0IWo2fkad/8+gX0fDVwAfGpm80LazUQVsHFmdjGwHOgX1k0mGrp+CdHw9ReF428ws7uJmqIB7nL3DWH5CvYMX/9vNGKi7GeKEV8iUgjFl0h6KLZEUqPQipiZvQ78g6j/7X8T3bG7v0/+/YVPzCO/A1fms6/ngOfySJ8DZCZaJpHSJtn4EpHCKb5E0kOxJZIaiQzW8SBwDLDQzMab2dlmVi3N5RIpLxRfIumj+BJJD8WWSAok0jXxXeBdM6tINGP5pUStUwenuWwiZZ7iSyR9FF8i6aHYkv1Ro6H/2tdF2Esig3UQRsY5FegPtGfPhMwiUkyKL5H0UXyJpIdiS6T4EnlHbBzRCDRvAn8F3g2T7olIMSm+RNJH8SWSHootkdRIpEXsWeDc7LkbRCSlFF8i6aP4EkkPxZZICuQ7WIeZ3QDg7lOAM3Ot+3OayyVSpim+RNJH8SWSHootkdQqaNTEAXHLN+Va1zsNZREpT4oVX2bWwMzeMbOFZrbAzK4J6YeY2TQzWxx+1grpZmaPmdkSM5tvZu3j9jUo5F9sZoPi0o80s0/DNo+FSdpF9ge6fomkh2JLJIUKqohZPst5fReRoilufO0E/ujuLYgm1LzSzFoAQ4Hp7t4MmB6+A5wMNAufwcCTEFXcgDuAzkT9/e/IrryFPJfGbaeLrOwvdP0SSQ/FlkgKFVQR83yW8/ouIkVTrPhy9zXu/lFY/gFYBNQH+rJn5KpRwOlhuS8w2iOzgJpmVg/oBUxz9w3uvhGYBvQO6w5291lhsvXRcfsSKe10/RJJD8WWSAoVNFhHGzPbQvSEo3pYJnzXpH0ixZOy+DKzRkA74EOgrruvCau+BeqG5frAirjNVoa0gtJX5pGe+9iDiVrYaNiwYVGKLZJOun6JpIdiSySF8q2IuXvFkixIaVbQBHDL7julBEsiZUWq4svMDgReAf7g7lviX+NydzeztD6hdPeRwEiADh066GmolAq6fomkh2JLJLUK6pooIqWYmVUmqoS95O6vhuTvQrdCws+1IX0V0CBu84yQVlB6Rh7pIiIiIpICqoiJ7IfCCIbPAovc/eG4VZOA7JEPBwET49IHhtETuwCbQxfGKUBPM6sVBunoCUwJ67aYWZdwrIFx+5L/3969h8tV1Wcc/75c5F54IICIhCBSFKlgiAkVxAiKEVuDBblUkXhLUSmilRJ9LCLaNkhrvdA+iJCGKCDXaERAEAhRIJgLCRxAIMQgoQjhIhe5aODXP9Yazj6TmXOd2TNz5v08zzxnz76svWad9R+P/ioAABYYSURBVNsza++11zYzs2GRNEvSo5J6CvM84q91JTfEzDrTfsAxwIGSluXXIcBM4N2S7gPeld8DXAmsBFYA3wc+DRARTwBfAxbl12l5Hnmdc/I29wNXlfHBzMxsVJvNuqPwesRf60r9DdZhZm0qIn5F/aGCD6qxfgCfqZPWLGBWjfmLgT1HkE0zM7M+ImJBHmSqaCowOU+fB8wHTqYw4i+wUFJlxN/J5BF/ASRVRvydTx7xN8+vjPjrE4nWlnxFzMzMzMxaqdQRf83ahRtiZmZmZtYW8tWvpo/CK2m6pMWSFq9Zs6bZuzOryQ0xMzMzM2ul0kf8jYizI2JCREzYdtttG/IhzIbKDTEzMzMzayWP+GtdyQ0xMzOzTNJOkm6QdJekOyV9Ns/38NpmDSDpQuAWYHdJqyV9HI/4a13KoyaamZn1Wgv8U0QslbQFsCSPyDaNNLz2TEkzSMNrn0zf4bUnkYbOnlQYXnsC6X6XJZLmRcST9A6vfSvph+YU/GPRukREHF1nkUf8ta7jK2JmZmZZRDwcEUvz9DPA3aRR16aShtUm/z00T78yvHYeMrsyvPZ7yMNr58ZXZXjtHcjDa+cfmXMKaZmZWRdxQ8zMzKyG/Kyjt5CuXHl4bTMzayh3TTQzM6siaXPgMuDEiHi6eBtXRISkUobXBqYDjB07ttm7MzPreONm/KzVWRgSXxEzMzMrkLQhqRF2fkRcnmd7eG0zM2uopjXEJM2S9KiknsI8jzplZmZtK3+XnAvcHRHfLCzy8NpmZtZQzbwiNps0ElTRDNKoU7sB1+X30HfUqemkEaUojDo1CZgIfKXSeKN31KnKdtX7MjMzG6r9gGOAAyUty69D8PDaZmbWYE27RywiFuQbnYumApPz9HnAfNLwv6+MOgUslFQZdWoyedQpgDyE8BRJ88mjTuX5lVGn/GVmZmbDFhG/Aur1sPDw2mZm1jBl3yPWklGnJE2XtFjS4jVr1ozsE5iZmZmZmY1QywbryGcRmz7qVN6Xb3g2MzMzM7O2UXZDrPRRp8zMzMzMzNpN2Q0xjzplZmZmZmZdr5nD118I3ALsLmm1pI/jUafMGsKPhzAzMzPrbM0cNfHoOos86pTZyM0GzgTmFOZVHg8xU9KM/P5k+j4eYhLp0Q+TCo+HmEC6X3OJpHkR8SS9j4e4lXSiZAo+2WFmZmbWMC0brMPMhi8iFgBPVM2eSnosBPnvoYX5cyJZCFQeD/Ee8uMhcuOr8niIHciPh8gnSeYU0jIzMzOzBmjaFbFuMW7Gz+ouWzXzfSXmxKw1j4cwMzNrBEmrgGeAl4C1ETEh9964CBgHrAKOiIgnc5f5bwOHAM8B0yJiaU7nWODLOdmvR8R5mLUhN8TMRqGICElNfzyEpOnAdICxY8c2e3dmZjb6vTMiHiu8b2S3exsF+rsI0mncNdFs9Cj98RB+Rp+ZmTVZQ7rdl51ps8FwQ8xs9PDjIczMrJMFcI2kJbnHBTSu271Z23HXRLMOlB8PMRkYI2k1qRvGTODi/KiIB4Aj8upXkvrQryD1o/8opMdDSKo8HgLWfTzEbGAT0miJHjHRzMyabf+IeEjSdsC1kn5TXNjIbvfuWm/twA0xsw7kx0OYmdloExEP5b+PSpoLTCR3u4+Ih4fQ7X5y1fz5NfZ1NnA2wIQJE5p+T7VZLe6aaGZmZmYtJWkzSVtUpknd5XtoULf7Ej+K2aD5ipiZmZmZtdr2wNx0azIbABdExNWSFtG4bvdmbcUNMTMzMzNrqYhYCexVY/7jNKjbvVm7cddEMzMzMzOzkvmKmJmZmZmZtY3R9NDm/viKmJmZmZmZWcncEDMzMzMzMyuZuyY2Ub3Lqqtmvq/knJiZNV+3dCUxMzNrBF8RMzMzMzMzK5kbYmZmZmZmZiVz10QzMzMzMytdt3dp9xUxMzMzMzOzkrkhZmZmZmZmVjJ3TWyB/i7DekRFMzMzM7PRzw0xMzMzszbX7ffSWOdy3a3PXRPNzMzMzMxK5itibcbdFs3MzMzMRj83xMzMzMzMbNjc/XB4Or4hJmkK8G1gfeCciJjZ4iw1zXArua+k2XB1U3zZ4PjLtjEcW1aL46sxHF/N4frZeB3dEJO0PvDfwLuB1cAiSfMi4q7W5qy9DCdw3Hgzx5dZczi2upt/zDZXo+Or3v9rNP9Och0tT0c3xICJwIqIWAkg6UfAVMBfZiPkIOxrNB9w++H4GuUc5y3T0NjyvcWD4/reNfzdlbnOt79Ob4jtCDxYeL8amFS9kqTpwPT89llJ99RIawzwWMNz2BydktdRk0+d3u/2OzcyM21kwPgaZGxB59SFsrg8CnR63fLo2tiCxsTXAMeu0crxlfUTW+D4GlF8ObasEfHV6Q2xQYmIs4Gz+1tH0uKImFBSlkakU/LqfI5+g4ktcBlXc3n05fKozfE1PC6PXi6L+hxfQ+ey6KsR5dHpzxF7CNip8P61eZ6ZjZzjy6w5HFtmzeP4so7R6Q2xRcBuknaR9CrgKGBei/NkNlo4vsyaw7Fl1jyOL+sYHd01MSLWSjoe+DlpiNJZEXHnMJMb8PJ0G+mUvDqfHayL46sMLo++uqo8Ghxb0GXlNwguj15dVxaOr6ZyWfQ14vJQRDQiI2ZmZmZmZjZInd410czMzMzMrOO4IWZmZmZmZlayrmuISZoi6R5JKyTNqLF8I0kX5eW3ShrXgjzuJOkGSXdJulPSZ2usM1nSU5KW5dcpZeezkJdVku7I+VhcY7kkfSeX6e2Sxrcgj7sXymqZpKclnVi1TtuUaafqhPgq0yDKY5qkNYU694lW5LMMkmZJelRST53lLT9OtDvHVy/HVi/H1sg5tvpyfPVqenxFRNe8SDdt3g+8DngVsBzYo2qdTwNn5emjgItakM8dgPF5egvg3hr5nAxc0eoyzXlZBYzpZ/khwFWAgH2BW9ugHvwe2Lldy7QTX50SX21WHtOAM1ud15LK4wBgPNBTZ3lbHSfa7eX4GnJZOLZ6lzu2+i8/x9bQy8Px1bt8RPHVbVfEJgIrImJlRPwJ+BEwtWqdqcB5efpS4CBJKjGPRMTDEbE0Tz8D3E16UnynmgrMiWQhsJWkHVqYn4OA+yPigRbmYTTqiPgq0WDKo2tExALgiX5WabfjRLtxfPVybBU4tkbMsdWX46ug2fHVbQ2xHYEHC+9Xs24D55V1ImIt8BSwTSm5qyFf/n4LcGuNxX8tabmkqyS9qdSM9RXANZKWSJpeY/lgyr1MRwEX1lnWLmXaiTouvppssPX+sNyd4VJJO9VY3i3a7TjRbhxfvRxbQ+PY6p9jqy/H19CMKL66rSHWUSRtDlwGnBgRT1ctXkrqWrcX8F3gx2Xnr2D/iBgPvBf4jKQDWpiXfik93PH9wCU1FrdTmVp3+CkwLiLeDFxL7xlXMxsZx5ZZ8zi+GqTbGmIPAcVW+2vzvJrrSNoA2BJ4vJTcFUjakNQIOz8iLq9eHhFPR8SzefpKYENJY0rOZiUvD+W/jwJzSZe1iwZT7mV5L7A0Ih6pXtBOZdqhOia+SjJgeUTE4xHxYn57DrBPSXlrR+10nGhHjq9ejq2hcWz1z7HVl+NraEYUX93WEFsE7CZpl3xl5ChgXtU684Bj8/ThwPWR78YrS+53fC5wd0R8s846r670T5Y0kfS/bEWDcTNJW1SmgYOB6pFl5gEfySPL7As8FREPl5zViqOp0y2xXcq0g3VEfJVowPKo6kf+ftL9oN2qnY4T7cjx1cuxNTSOrf45tvpyfA3NiOJrg+blq/1ExFpJxwM/J40KMysi7pR0GrA4IuaRGkA/kLSCdHPeUS3I6n7AMcAdkpbleV8CxgJExFmkA8GnJK0FngeOatFBYXtgbm6/bABcEBFXSzqukNcrSaPKrACeAz7agnxWGorvBv6hMK+Yz3Yp047UQfFVikGWxwmS3g+sJZXHtJZluMkkXUgamXSMpNXAV4ANob2OE+3K8dXLsdWXY2tkHFt9Ob76anZ8yb8zzczMzMzMytVtXRPNzMzMzMxazg0xMzMzMzOzkrkhZmZmZmZmVjI3xMzMzMzMzErmhpiZmZmZmVnJ3BAzs7Yl6dmq99MknZmnj5P0kcL81zQxH1dK2iq/Pt3AdM+QdKekM6rmT5b0tsL72ZIOb9R+zYocZ6+8b3icFdOUdI6kPRqZvjVeu8TDYOV6fMUA6+wt6ZCy8jQS3Ravbog1iaRDJYWkN9RZ3tAvmnYxnM8laZyk5wvPTBvJ/jeRtEzSnySNGWl61r4i4qyImJPfTgMa/oWYH9C4XkQcEhF/ALYCGhm304E3R8RJVfMnA29bd3WzcjnOGiciPhERd5W1P2u8MuKhSfYmPeuqE3RVvLoh1jxHA4vz31pqftFUvpAalYlGpzcIw/0CvT8i9h7pziPi+ZzO/400LWtvkk6V9IV89moCcH5uhG9Std58Sd/Oy3okTSxuX1ivJ58UGCfpHklzgB5gJ0mrcsN+JrBrTusMSXMkHVpI43xJU6v2r7xuj6Q7JB2Z588DNgeWVObl+eOA44DP5f28PS86QNLNklYWzwJKOknSIkm3S/pqnbJ6tnCW8ReSJuZyWan0UM4+Z33z+yskTc7TR+e890g6vSrdf5W0XNJCSdv3/1+zTuM4e2X9fuNM0vpKZ9Mr+/9cjXXmS5qQp6dIWppj57o8bzNJsyT9WtJt1Z/RWm8I8fD6fKxdnv/Pu/ZTRydLulHST3K9mynpQ7ke3CFp17zebElnSVos6V5Jf1Mjf+vUIUmvAk4Djsx5PbJeXcvfA5dLulrSfZK+UUj7YEm35M9ziaTN8/yZku7KsfEfed4H8+dcLmlBjXw6Xisiwq8Gv3IlegTYC7i3zjo/Ap4HlgGXAPcAc4A7gZ2BcUBPYf0vAKfm6Q8Dv87bfg9YvyrtcTXSq7kN8BHgdmA58INCGp8nfTn2ACfWyP9M4DOF96fmPBY/1xnAW3P6GwOb5fzsWSO/PVXz1slXXu83wGzgXuB84F3ATcB9wMTC9quAMa2uC36NOJZeynWp8vodcGaxzuXp+cCEOmnMB76fpw+o1LXi9vl9T65j44CXgX2r61ONuHwH8OM8vSXwW2CDqv0fBlwLrA9snz/DDnnZs3XyXJ232aTjxHrAHsCKPP9g4GxAedkVwAE10gvgvXl6LnANsCHpGLUsz59WKdv8/grSGcjX5DxvC2wAXA8cWkj3b/P0N4Avt7rO+OU4a0WcAfsA1xbeb1VI8/Bi+eVYehDYJc/fOv/9N+DDle1J33Obtbp+dNurQfFwK/CBPL0xsGm9Oko6zv4hT28EPAR8NW/7WeBbhbp0da6DuwGrc9qTgSv6q0Ose3zvb72VOc42Bh4AdiLF5YJKfQROBk4BtiH93lRVvb8D2LE4r6p8HK/55StizTEV+EVELAeelbRPjXVm0HsV6CRSUP1PRLwpIh6ol7CkNwJHAvvlbV8CPlRj1VfSIx0A1tlG0puALwMHRsRepIAn5/ejwCRgX+CTkt5Slf5FwBGF90fkea98rog4KSIWAfOAr5N+qP0wInrqfb68/5r5yl4P/Cfwhvz6e2B/UiPwS/2lax3p+VyX9s5195RhpnMhQEQsAP5C0lYDrP9ARCwcKNGIuBHYTdK2pKvfl0XE2qrV9gcujIiXIuIR4EbSCYqh+nFEvBypq0TlytPB+XUbsJQUE7vV2PZPpC9wSF+QN0bEn/P0uAH2+1ZgfkSsyZ/tfNIP7Uq6lXsTlgwiLWtPjrNew42zlcDrJH1X0hTg6X72sS+wICJ+mz/fE4X9zFDqpj+f9EN47DA+g43MiOJB0hakRshcgIh4ISKeo/86uigiHo6IF4H7SSfLYN1j9MW5ft5HqnPVt78Mtg71t951EfFURLwA3EU6mb8vqbFzU97m2Dz/KeAF4FxJfwc8l9O4CZgt6ZOkxlY1x2u2wVA3sEE5Gvh+nr44v18ywDaD+kICDiK15BdJAtgEeHSA9OptsyVwSUQ8Bn0q1/7A3Ij4I4Cky4G3kyo1ed3bJG2ndKPqtsCTEfFgvnxc7TRgESlYTxjEZzywTr4AfhsRd+R83Uk6YISkwfygtO4VNd6vpW/37I0L038cQtpzSFecjyKdwGiWFwvTKvz994j43gDb/jnyaTvSVYgXASLiZUmV74H+ymMw6b6Ev1O6XdfGWUQ8KWkv4D2kLlRHAB8b4r4FHBYR9wxxO+t8xXr3cuH9y/Q9rtaKsaKadUjSpCGsV8xL5bgu0hWkdW63UeqGfBBwOHA86ST6cTmt95G6Ge4TEY9Xb9sAHR+vviLWYJK2Jl1Jqpx9vpjUL1f1twLW/UKq9+Ul4LzC2ZrdI+LUAdIb7DZDdQkp8I4kXQ2rZxtSd80tGNyPu/4M9mBl3eUZUv2qp9L/fH/gqYh4itQNanyePx7YZZj7mQ2cCBC1b+z9JekYsH4+o38AqZvwUPdTy8+BjxX66u8oabtBbFfLKmBvSetJ2gmYmOf/GniHpDGS1iedWLpxmPuwzuY4o3acKd3btl5EXEbq0TG+n/QWku5t2SVvu3VhP/9Y+b1QoyeKtZea9ScingFWK9/XKGkjSZsyvDpa7YP5GL0r8DpSt8CienWoOq9DrWsLgf0kvT6vv5mkv8wxsWVEXAl8jtTdHUm7RsStEXEKsIbUvbHI8Zq5IdZ4hwNX5svLRMRK4GHSFaWigSrUI8B2kraRtBFQuSnzOuDwSqWStLWknQfIU71tricF9TaV+Xn9XwKHStpU0mbAB/K8aheRzk4eTmqU1ftc3wP+hdSl6XQGVi9fZvXMBs5SjZumsxck3QacBXw8z7sM2DpfWT2e1L+7X/mM3k35Bt8z8rxHgLuB/62z2Vx673e8HvjniPj9ALv6KfAB9b0puVZ+rgEuAG7JV4UvZXBfVLXcRLr35i7gO6QuHUTEw6Quxzfkz7AkIn4yzH1YZ5uN46xenO0IzM/dlH4IfLGf9NaQRoa7XNJyek9kfo107+btuby+NkD+rbVmUz8ejgFOkHQ7cDPwaoZXR6v9jtRguQo4LncfLKpXh24A9sh5PbKf9WrKdXYacGH+TLeQuvxtAVyR5/2KNL4AwBnKAzzlz7+8KknHa1a5uc4aRNINpDMCxf6m25DujfpU1boXAG8mfbm8MSL2rFp+Aun+qIdI/VlXRcSpOYi+SGpI/5k0aMbCwnbjSDdu7lmYV3MbSceS7lF7CbgtIqbl9T9P72XacyLiW3U+7x3AYxHxzhqf6ypS/+apEXFYPpt+M/DFiLh+gPyuk6/q9STNzu8vrbFsFekm2sdq5du6h6T5pBt8Fzcp/U1J9Xx8vgJg1nUcZ2bNVfzN0+q8WOO4IWYtV6sh1oA0V+GGmNHcH4iS3gWcC/xXvZMVZt3AcWbWXG6IjU5uiFnL5ftRbgYejxE+Syx3D7iFNIDIX1UN9GFmZmZm1hbcEDMzMzMzMyuZB+swMzMzMzMrmRtiZmZmZmZmJXNDzMzMzMzMrGRuiJmZmZmZmZXMDTEzMzMzM7OSuSFmZmZmZmZWMjfEzMzMzMzMSuaGmJmZmZmZWcn+H58BrKlxOAQ+AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 873.738x252 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(ncols=4, nrows=1, figsize=(7.5 * gr, 3.5))\n",
    "\n",
    "sig_selected_mask = d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\") & sig_mask_ccinc\n",
    "distances = d[\"nu\"][\"events\"].array(\"mc_nu_vtx_distance\")[sig_selected_mask]\n",
    "\n",
    "ax[0].hist(\n",
    "    distances,\n",
    "    range=(0, 8),\n",
    "    bins=20,\n",
    "    label=\"{0:0.1f}% within 5cm\".format(sum(distances < 5) * 100 / len(distances)),\n",
    ")\n",
    "ax[0].set_xlabel(r\"$\\Delta$ true reco vtx [cm]\")\n",
    "ax[0].set_ylabel(\"Events per bin\")\n",
    "ax[0].legend()\n",
    "\n",
    "hit_fraq = d[\"nu\"][\"df_daughters\"].query(\n",
    "    \"python_select & mc_sig_nuccinc & abs(mc_pdg)==13\"\n",
    ")[\"mc_hit_fraction\"]\n",
    "\n",
    "ax[1].hist(\n",
    "    hit_fraq,\n",
    "    range=(0, 1),\n",
    "    bins=20,\n",
    "    label=\"{0:0.1f}% above 0.9\".format(sum(hit_fraq > 0.9) * 100 / len(hit_fraq)),\n",
    ")\n",
    "ax[1].set_xlabel(r\"Hit purity of the muon\")\n",
    "ax[1].set_ylabel(\"Events per bin\")\n",
    "ax[1].legend()\n",
    "\n",
    "slice_purity = d[\"nu\"][\"events\"].array(\"matched_hit_fraction\")[sig_selected_mask]\n",
    "slice_completeness = d[\"nu\"][\"events\"].array(\"clustered_hit_completeness\")[\n",
    "    sig_selected_mask\n",
    "]\n",
    "\n",
    "ax[2].hist(\n",
    "    slice_purity,\n",
    "    range=(0, 1),\n",
    "    bins=20,\n",
    "    label=\"{0:0.1f}% above 0.9\".format(\n",
    "        sum(slice_purity > 0.9) * 100 / len(slice_purity)\n",
    "    ),\n",
    ")\n",
    "ax[2].set_xlabel(r\"Hit purity of the slice\")\n",
    "ax[2].set_ylabel(\"Events per bin\")\n",
    "ax[2].legend()\n",
    "\n",
    "ax[3].hist(\n",
    "    slice_completeness,\n",
    "    range=(0, 1),\n",
    "    bins=20,\n",
    "    label=\"{0:0.1f}% above 0.8\".format(\n",
    "        sum(slice_completeness > 0.8) * 100 / len(slice_completeness)\n",
    "    ),\n",
    ")\n",
    "ax[3].set_xlabel(r\"Hit completeness of the slice\")\n",
    "ax[3].set_ylabel(\"Events per bin\")\n",
    "ax[3].legend()\n",
    "\n",
    "fig.suptitle(title_mc_str)\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(top=0.88)\n",
    "#fig.savefig(outdir + \"muon_quality.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "230.0\n",
      "30.0\n",
      "85.0\n",
      "-83.53\n",
      "1010.0\n",
      "30.1\n",
      "33028509.4\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.0547711073514345"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fidvol ratio vs MCC8 analysis:\n",
    "fv1 = 1\n",
    "for i in range(3):\n",
    "    fv1 *= (helper.upper[i] - fidvol_vtx[2 * i]) - (\n",
    "        helper.lower[i] + fidvol_vtx[2 * i + 1]\n",
    "    )\n",
    "    print(helper.upper[i] - fidvol_vtx[2 * i])\n",
    "    print(helper.lower[i] + fidvol_vtx[2 * i + 1])\n",
    "print(fv1)\n",
    "\n",
    "\n",
    "fv2 = (233 - 70) * (256.35 - 24) * (1036.8 - 25 - 100 - 85)\n",
    "fv1 / fv2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plots"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Muon PID plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "def mc_track_contained(x_s, y_s, z_s, p_x, p_y, p_z, length):\n",
    "    l_n = length / np.sqrt(p_x ** 2 + p_y ** 2 + p_z ** 2)\n",
    "    x_e = x_s + p_x * l_n\n",
    "    y_e = y_s + p_y * l_n\n",
    "    z_e = z_s + p_z * l_n\n",
    "    df_end = pd.DataFrame({\"x_e\": x_e, \"y_e\": y_e, \"z_e\": z_e})\n",
    "    return helper.inTPC_mask(\n",
    "        df_end, \"x_e\", \"y_e\", \"z_e\", fidvol=[10, 10, 10, 10, 10, 10]\n",
    "    )\n",
    "\n",
    "\n",
    "a = d[\"nu\"][\"daughters\"].pandas.df(\n",
    "    [\"mc_vx\", \"mc_vy\", \"mc_vz\", \"mc_px\", \"mc_py\", \"mc_pz\", \"mc_length\"]\n",
    ")\n",
    "nu_daughters_contained = mc_track_contained(*a.values.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "ename": "IOError",
     "evalue": "[Errno 2] No such file or directory: './output/dataMC/momentum_comparison.pdf'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIOError\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-135-49b6bdf2a42f>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     68\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtight_layout\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     69\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplots_adjust\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtop\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.9\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 70\u001b[0;31m \u001b[0mfig\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msavefig\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutdir\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"momentum_comparison.pdf\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/home/pzhelnin/.local/lib/python2.7/site-packages/matplotlib/figure.pyc\u001b[0m in \u001b[0;36msavefig\u001b[0;34m(self, fname, **kwargs)\u001b[0m\n\u001b[1;32m   2060\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_frameon\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframeon\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2061\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2062\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprint_figure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2063\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2064\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mframeon\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/pzhelnin/.local/lib/python2.7/site-packages/matplotlib/backend_bases.pyc\u001b[0m in \u001b[0;36mprint_figure\u001b[0;34m(self, filename, dpi, facecolor, edgecolor, orientation, format, **kwargs)\u001b[0m\n\u001b[1;32m   2261\u001b[0m                 \u001b[0morientation\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0morientation\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2262\u001b[0m                 \u001b[0mbbox_inches_restore\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0m_bbox_inches_restore\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2263\u001b[0;31m                 **kwargs)\n\u001b[0m\u001b[1;32m   2264\u001b[0m         \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2265\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mbbox_inches\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mrestore_bbox\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/pzhelnin/.local/lib/python2.7/site-packages/matplotlib/backends/backend_pdf.pyc\u001b[0m in \u001b[0;36mprint_pdf\u001b[0;34m(self, filename, **kwargs)\u001b[0m\n\u001b[1;32m   2576\u001b[0m             \u001b[0mfile\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfilename\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_file\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2577\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2578\u001b[0;31m             \u001b[0mfile\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mPdfFile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmetadata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"metadata\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2579\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2580\u001b[0m             \u001b[0mfile\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnewPage\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwidth\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mheight\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/pzhelnin/.local/lib/python2.7/site-packages/matplotlib/backends/backend_pdf.pyc\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, filename, metadata)\u001b[0m\n\u001b[1;32m    436\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moriginal_file_like\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    437\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtell_base\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 438\u001b[0;31m         \u001b[0mfh\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopened\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcbook\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_filehandle\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"wb\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreturn_opened\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    439\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mopened\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    440\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/pzhelnin/.local/lib/python2.7/site-packages/matplotlib/cbook/__init__.pyc\u001b[0m in \u001b[0;36mto_filehandle\u001b[0;34m(fname, flag, return_opened, encoding)\u001b[0m\n\u001b[1;32m    609\u001b[0m             \u001b[0mfh\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbz2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mBZ2File\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mflag\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    610\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 611\u001b[0;31m             \u001b[0mfh\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mio\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mflag\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mencoding\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mencoding\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    612\u001b[0m         \u001b[0mopened\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    613\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'seek'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIOError\u001b[0m: [Errno 2] No such file or directory: './output/dataMC/momentum_comparison.pdf'"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 757.24x525.6 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "neutrino = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_hit_fraction\") > 0.5)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"track_length\") > 5)\n",
    "    & d[\"nu\"][\"daughters\"].array(\"start_contained\")\n",
    ")\n",
    "\n",
    "muons = (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 13) & neutrino & nu_daughters_contained\n",
    "protons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 2212) & neutrino & nu_daughters_contained\n",
    ")\n",
    "pions = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 211) & neutrino & nu_daughters_contained\n",
    ")\n",
    "un_muons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 13) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "un_protons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 2212) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "un_pions = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 211) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "\n",
    "data8 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_pions],\n",
    "]\n",
    "\n",
    "data9 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_pions],\n",
    "]\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(ncols=3, nrows=2, figsize=(6.5 * gr, 7.3))\n",
    "ax = ax.flatten()\n",
    "for i, l in enumerate(\n",
    "    [\n",
    "        \"Contained Muons\",\n",
    "        \"Contained Protons\",\n",
    "        \"Contained Pions\",\n",
    "        \"Uncontained Muons\",\n",
    "        \"Uncontained Protons\",\n",
    "        \"Uncontained Pions\",\n",
    "    ]\n",
    "):\n",
    "    ax[i].hist2d(\n",
    "        data9[i],\n",
    "        data8[i],\n",
    "        bins=50,\n",
    "        range=((0, 3), (0, 3)),\n",
    "        cmap=\"coolwarm\",\n",
    "        norm=LogNorm(),\n",
    "    )\n",
    "    ax[i].set_xlabel(\"Track Momentum using MCS [GeV/c]\")\n",
    "    ax[i].set_ylabel(\"Track Momentum using Range [GeV/c]\")\n",
    "    ax[i].set_title(l)\n",
    "\n",
    "fig.suptitle(title_mc_str)\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(top=0.9)\n",
    "fig.savefig(outdir + \"momentum_comparison.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pseudo-exiting tracks, MCS momentum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "neutrino = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_hit_fraction\") > 0.5)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"track_length\") > 5)\n",
    "    & d[\"nu\"][\"daughters\"].array(\"start_contained\")\n",
    ")\n",
    "\n",
    "muons = (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 13) & neutrino & nu_daughters_contained\n",
    "protons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 2212) & neutrino & nu_daughters_contained\n",
    ")\n",
    "pions = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 211) & neutrino & nu_daughters_contained\n",
    ")\n",
    "un_muons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 13) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "un_protons = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 2212) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "un_pions = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 211) & neutrino & ~nu_daughters_contained\n",
    ")\n",
    "\n",
    "data8 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_range_mom_mu\")[un_pions],\n",
    "]\n",
    "\n",
    "data9 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_mcs_mom\")[un_pions],\n",
    "]\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(ncols=3, nrows=2, figsize=(6.5 * gr, 7.3))\n",
    "ax = ax.flatten()\n",
    "for i, l in enumerate(\n",
    "    [\n",
    "        \"Contained Muons\",\n",
    "        \"Contained Protons\",\n",
    "        \"Contained Pions\",\n",
    "        \"Uncontained Muons\",\n",
    "        \"Uncontained Protons\",\n",
    "        \"Uncontained Pions\",\n",
    "    ]\n",
    "):\n",
    "    ax[i].hist2d(\n",
    "        data9[i],\n",
    "        data8[i],\n",
    "        bins=50,\n",
    "        range=((0, 3), (0, 3)),\n",
    "        cmap=\"coolwarm\",\n",
    "        norm=LogNorm(),\n",
    "    )\n",
    "    ax[i].set_xlabel(\"Track Momentum using MCS [GeV/c]\")\n",
    "    ax[i].set_ylabel(\"Track Momentum using Range [GeV/c]\")\n",
    "    ax[i].set_title(l)\n",
    "\n",
    "fig.suptitle(title_mc_str)\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(top=0.9)\n",
    "fig.savefig(outdir + \"momentum_comparison.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "neutrino = (d[\"nu\"][\"daughters\"].array(\"mc_hit_fraction\") > 0.5) & (\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\") > 5\n",
    ")\n",
    "muons = (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 13) & neutrino\n",
    "protons = (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 2212) & neutrino\n",
    "pions = (abs(d[\"nu\"][\"daughters\"].array(\"mc_pdg\")) == 211) & neutrino\n",
    "\n",
    "neutrino = (d[\"nu\"][\"daughters\"].array(\"mc_hit_fraction\") > 0.5) & (\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\") > 5\n",
    ")\n",
    "electrons = (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 11) & neutrino\n",
    "\n",
    "step1_nu = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"track_length\") > 20)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"track_score\") > 0.85)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"vtx_distance\") < 4)\n",
    ")\n",
    "step1_nue = (\n",
    "    (d[\"nu\"][\"daughters\"].array(\"track_length\") > 20)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"track_score\") > 0.85)\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"vtx_distance\") < 4\n",
    "    & (d[\"nu\"][\"daughters\"].array(\"mc_pdg\") == 11))\n",
    ")\n",
    "\n",
    "\n",
    "data1 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_proton\")[muons & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_proton\")[protons & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_proton\")[pions & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_proton\")[electrons & step1_nue],\n",
    "]\n",
    "data2 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_muon\")[muons & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_muon\")[protons & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_muon\")[pions & step1_nu],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_chi2_muon\")[electrons & step1_nue],\n",
    "]\n",
    "data3 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_length\")[electrons],\n",
    "]\n",
    "data4 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_score\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_score\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_score\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"track_score\")[electrons],\n",
    "]\n",
    "\n",
    "data5 = [\n",
    "    data1[0] / data2[0],\n",
    "    data1[1] / data2[1],\n",
    "    data1[2] / data2[2],\n",
    "    data1[3] / data2[3],\n",
    "]\n",
    "\n",
    "data6 = [\n",
    "    d[\"nu\"][\"daughters\"].array(\"vtx_distance\")[muons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"vtx_distance\")[protons],\n",
    "    d[\"nu\"][\"daughters\"].array(\"vtx_distance\")[pions],\n",
    "    d[\"nu\"][\"daughters\"].array(\"vtx_distance\")[electrons],\n",
    "]\n",
    "\n",
    "labs = [\n",
    "    r\"$\\nu_\\mu$ muons\",\n",
    "    r\"$\\nu_\\mu$ protons\",\n",
    "    r\"$\\nu_\\mu$ charged pions\",\n",
    "    r\"$\\nu_e$ electrons\",\n",
    "]\n",
    "\n",
    "fig, ax = plt.subplots(ncols=3, nrows=1, figsize=(8.5 * gr, 3.5))\n",
    "ax[1].hist(\n",
    "    data1, label=labs, range=(0, 350), histtype=\"step\", bins=35, lw=2, density=True\n",
    ")\n",
    "ax[1].set_xlim(0, 350)\n",
    "ax[1].axvspan(60, 350, alpha=0.2, color=\"green\")\n",
    "ax[1].legend()\n",
    "ax[1].set_xlabel(r\"Proton $\\chi^2$\")\n",
    "ax[1].set_title(\"Proton $\\chi^2$\")\n",
    "ax[0].hist(\n",
    "    data2, label=labs, range=(0, 60), histtype=\"step\", bins=30, lw=2, density=True\n",
    ")\n",
    "ax[0].axvspan(0, 30, alpha=0.2, color=\"green\")\n",
    "ax[0].set_xlim(0, 60)\n",
    "ax[0].legend()\n",
    "ax[0].set_xlabel(r\"Muon $\\chi^2$\")\n",
    "ax[0].set_title(\"Muon $\\chi^2$\")\n",
    "ax[2].hist(\n",
    "    data5, label=labs, range=(0, 40), histtype=\"step\", bins=40, lw=2, density=True\n",
    ")\n",
    "ax[2].set_xlim(0, 40)\n",
    "ax[2].axvspan(7, 40, alpha=0.2, color=\"green\")\n",
    "ax[2].set_xlabel(r\"Proton $\\chi^2$ / Muon $\\chi^2$\")\n",
    "ax[2].set_title(r\"Proton $\\chi^2$ / Muon $\\chi^2$\")\n",
    "ax[2].set_yscale(\"log\")\n",
    "ax[2].set_ylim(1e-3, 1)\n",
    "ax[2].legend()\n",
    "\n",
    "ax[0].set_ylabel(\"Area Normalised\")\n",
    "ax[1].set_ylabel(\"Area Normalised\")\n",
    "ax[2].set_ylabel(\"Area Normalised\")\n",
    "fig.tight_layout()\n",
    "fig.show()\n",
    "fig.savefig(outdir + \"muon_candidate1.pdf\")\n",
    "\n",
    "fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(6 * gr, 4))\n",
    "ax[0].hist2d(\n",
    "    data1[0],\n",
    "    data2[0],\n",
    "    bins=50,\n",
    "    range=((0, 350), (0, 100)),\n",
    "    cmap=\"coolwarm\",\n",
    "    norm=LogNorm(),\n",
    ")\n",
    "ax[0].fill_between(\n",
    "    [60, 210, 350],\n",
    "    [60 / 7, 30, 30],\n",
    "    facecolor=(*rgb_c2, 0.33),\n",
    "    edgecolor=(rgb_c2),\n",
    "    lw=1.5,\n",
    ")\n",
    "ax[0].set_xlabel(r\"Proton $\\chi^2$\")\n",
    "ax[0].set_ylabel(r\"Muon $\\chi^2$\")\n",
    "ax[0].set_title(\"Muons\", loc=\"left\")\n",
    "ax[0].set_title(\"MicroBooNE INTERNAL\", loc=\"right\")\n",
    "\n",
    "ax[1].hist2d(\n",
    "    data1[1],\n",
    "    data2[1],\n",
    "    bins=50,\n",
    "    range=((0, 350), (0, 100)),\n",
    "    cmap=\"coolwarm\",\n",
    "    norm=LogNorm(),\n",
    ")\n",
    "ax[1].fill_between(\n",
    "    [60, 210, 350],\n",
    "    [60 / 7, 30, 30],\n",
    "    facecolor=(*rgb_c2, 0.33),\n",
    "    edgecolor=(rgb_c2),\n",
    "    lw=1.5,\n",
    ")\n",
    "ax[1].set_xlabel(r\"Proton $\\chi^2$\")\n",
    "ax[1].set_ylabel(r\"Muon $\\chi^2$\")\n",
    "ax[1].set_title(\"Protons\", loc=\"left\")\n",
    "ax[1].set_title(\"MicroBooNE INTERNAL\", loc=\"right\")\n",
    "fig.tight_layout()\n",
    "fig.show()\n",
    "fig.savefig(outdir + \"muon_candidate2.pdf\")\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(ncols=3, nrows=1, figsize=(8.5 * gr, 3.5))\n",
    "ax[0].hist(\n",
    "    data3, label=labs, range=(0, 200), histtype=\"step\", bins=40, lw=2, density=True\n",
    ")\n",
    "ax[0].legend()\n",
    "ax[0].set_xlabel(r\"Track Length [cm]\")\n",
    "ax[0].axvspan(20, 200, alpha=0.2, color=\"green\")\n",
    "ax[0].set_title(\"Track Length\")\n",
    "ax[0].set_xlim(0, 200)\n",
    "\n",
    "ax[1].hist(\n",
    "    data4, label=labs, range=(0.5, 1), histtype=\"step\", bins=25, lw=2, density=True\n",
    ")\n",
    "ax[1].axvspan(0.85, 1, alpha=0.2, color=\"green\")\n",
    "ax[1].set_yscale(\"log\")\n",
    "ax[1].set_xlim(0.5, 1)\n",
    "ax[1].legend(loc=2)\n",
    "ax[1].set_xlabel(r\"Track Score\")\n",
    "ax[1].set_title(\"Track Score\")\n",
    "ax[2].hist(\n",
    "    data6, label=labs, range=(0, 6), histtype=\"step\", bins=30, lw=2, density=True\n",
    ")\n",
    "ax[2].set_xlim(0, 6)\n",
    "ax[2].axvspan(0, 4, alpha=0.2, color=\"green\")\n",
    "ax[2].set_xlabel(r\"Vertex Distance [cm]\")\n",
    "ax[2].set_title(r\"Vertex Distance\")\n",
    "ax[2].set_yscale(\"log\")\n",
    "ax[2].legend()\n",
    "ax[0].set_ylabel(\"Area Normalised\")\n",
    "ax[1].set_ylabel(\"Area Normalised\")\n",
    "ax[2].set_ylabel(\"Area Normalised\")\n",
    "fig.tight_layout()\n",
    "fig.show()\n",
    "fig.savefig(outdir + \"muon_candidate3.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Efficiency plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_panel_eff(ax, field, x_label, N_bins, x_min, x_max, legend=True):\n",
    "\n",
    "    denom = d[\"nu\"][\"events\"].array(field)[sig_mask_ccinc]\n",
    "    sig_mask_sel = d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\") & sig_mask_ccinc\n",
    "    nom = d[\"nu\"][\"events\"].array(field)[sig_mask_sel]\n",
    "    data = [denom, nom]\n",
    "    edges, _, bins, _, _ = helper.histHelper(N_bins, x_min, x_max, data)\n",
    "    fraq = bins[1] / bins[0]\n",
    "    err = np.sqrt(fraq * (1 - fraq) / bins[0])\n",
    "    fraq = np.append(fraq, fraq[-1])\n",
    "    err = np.append(err, err[-1])\n",
    "\n",
    "    ax.fill_between(edges, fraq - err, fraq + err, alpha=0.3, step=\"post\")\n",
    "    ax.set_xlabel(x_label)\n",
    "    ax.set_ylabel(\"Efficiency\")\n",
    "    ax.set_ylim(0, 1)\n",
    "    ax.set_xlim(x_min, x_max)\n",
    "    ax.step(\n",
    "        edges,\n",
    "        fraq,\n",
    "        where=\"post\",\n",
    "        label=r\"Efficiency {0:0.1f}%, Purity: {1:0.1f}%\".format(\n",
    "            eff * 100, purity * 100\n",
    "        ),\n",
    "    )\n",
    "    if legend:\n",
    "        ax.legend(loc=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(5 * gr, 3.5))\n",
    "\n",
    "\n",
    "plot_panel_eff(ax[0], \"mc_nu_energy\", \"True neutrino energy [GeV]\", 23, kin_min_mu, 3.0)\n",
    "plot_panel_eff(\n",
    "    ax[1], \"mc_nu_lepton_energy\", \"True lepton energy [GeV]\", 19, kin_min_mu, 2.5\n",
    ")\n",
    "\n",
    "ax[0].set_title(title_mc_str, loc=\"right\")\n",
    "ax[1].set_title(title_mc_str, loc=\"right\")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"energy_eff.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(ncols=3, nrows=1, figsize=(6.6 * gr, 3.7))\n",
    "\n",
    "field_arr = [\"mc_nu_vx\", \"mc_nu_vy\", \"mc_nu_vz\"]\n",
    "lab_arr = [\"True Neutrino vx [cm]\", \"True Neutrino vy [cm]\", \"True Neutrino vz [cm]\"]\n",
    "x_min_arr = helper.lower\n",
    "x_max_arr = helper.upper\n",
    "\n",
    "for i, (field, lab, x_min, x_max) in enumerate(\n",
    "    zip(field_arr, lab_arr, x_min_arr, x_max_arr)\n",
    "):\n",
    "    N = int(round(x_max - x_min) / 10)\n",
    "    N = min(N, 30)\n",
    "    plot_panel_eff(ax[i], field, lab, N, x_min, x_max)\n",
    "\n",
    "\n",
    "fig.suptitle(title_mc_str)\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(top=0.88)\n",
    "fig.savefig(outdir + \"vtx_eff.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Efficiency of the angles:\n",
    "lepton_mom = d[\"nu\"][\"events\"].arrays(\n",
    "    [\"mc_nu_lepton_px\", \"mc_nu_lepton_py\", \"mc_nu_lepton_pz\"]\n",
    ")\n",
    "theta, phi = helper.cosmic_angles(*list(lepton_mom.values()), df_out=False)\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(5 * gr, 3.5))\n",
    "sig_mask_sel = d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\") & sig_mask_ccinc\n",
    "\n",
    "# left\n",
    "x_ticks = [0, np.pi / 4, np.pi / 2, 3 * np.pi / 4, np.pi]\n",
    "x_labels = [r\"$0$\", r\"$\\pi/4$\", r\"$\\pi/2$\", r\"$3\\pi/4$\", r\"$\\pi$\"]\n",
    "x_min = 0\n",
    "x_max = np.pi\n",
    "N_bins = 16\n",
    "data = [theta[sig_mask_ccinc], theta[sig_mask_sel]]\n",
    "edges, _, bins, _, _ = helper.histHelper(N_bins, x_min, x_max, data)\n",
    "fraq = bins[1] / bins[0]\n",
    "err = np.sqrt(fraq * (1 - fraq) / bins[0])\n",
    "fraq = np.append(fraq, fraq[-1])\n",
    "err = np.append(err, err[-1])\n",
    "\n",
    "ax[0].fill_between(edges, fraq - err, fraq + err, alpha=0.3, step=\"post\")\n",
    "ax[0].set_xlabel(\"True lepton theta\")\n",
    "ax[0].set_ylabel(\"Efficiency\")\n",
    "ax[0].set_ylim(0, 1)\n",
    "ax[0].set_xlim(x_min, x_max)\n",
    "ax[0].step(\n",
    "    edges,\n",
    "    fraq,\n",
    "    where=\"post\",\n",
    "    label=r\"Efficiency {0:0.1f}%, Purity: {1:0.1f}%\".format(eff * 100, purity * 100),\n",
    ")\n",
    "ax[0].legend(loc=0)\n",
    "ax[0].set_xticks(x_ticks)\n",
    "ax[0].set_xticklabels(x_labels)\n",
    "\n",
    "\n",
    "# right\n",
    "x_ticks = [-np.pi, -np.pi / 2, 0, np.pi / 2, np.pi]\n",
    "x_labels = [r\"$-\\pi$\", r\"$-\\pi/2$\", r\"$0$\", r\"$\\pi/2$\", r\"$\\pi$\"]\n",
    "x_min = -np.pi\n",
    "x_max = np.pi\n",
    "N = 16\n",
    "data = [phi[sig_mask_ccinc], phi[sig_mask_sel]]\n",
    "edges, _, bins, _, _ = helper.histHelper(N_bins, x_min, x_max, data)\n",
    "fraq = bins[1] / bins[0]\n",
    "err = np.sqrt(fraq * (1 - fraq) / bins[0])\n",
    "fraq = np.append(fraq, fraq[-1])\n",
    "err = np.append(err, err[-1])\n",
    "\n",
    "ax[1].fill_between(edges, fraq - err, fraq + err, alpha=0.3, step=\"post\")\n",
    "ax[1].set_xlabel(\"True lepton phi\")\n",
    "ax[1].set_ylabel(\"Efficiency\")\n",
    "ax[1].set_ylim(0, 1)\n",
    "ax[1].set_xlim(x_min, x_max)\n",
    "ax[1].step(\n",
    "    edges,\n",
    "    fraq,\n",
    "    where=\"post\",\n",
    "    label=r\"Efficiency {0:0.1f}%, Purity: {1:0.1f}%\".format(eff * 100, purity * 100),\n",
    ")\n",
    "ax[1].legend(loc=0)\n",
    "ax[1].set_xticks(x_ticks)\n",
    "ax[1].set_xticklabels(x_labels)\n",
    "\n",
    "ax[0].set_title(title_mc_str, loc=\"right\")\n",
    "ax[1].set_title(title_mc_str, loc=\"right\")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"angle_eff.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Efficiency in function of interaction type\n",
    "dict_modes = {0: \"QE\", 1: \"RES\", 2: \"DIS\", 10: \"MEC\", 3: \"?\"}\n",
    "types_sig = d[\"nu\"][\"events\"].array(\"mc_nu_interaction_type\")[sig_mask_ccinc]\n",
    "types_sel = d[\"nu\"][\"events\"].array(\"mc_nu_interaction_type\")[sig_mask_sel]\n",
    "\n",
    "classes, denom = np.unique(types_sig, return_counts=True)\n",
    "nom = np.unique(types_sel, return_counts=True)[1]\n",
    "error = helper.effErr(nom, denom)\n",
    "\n",
    "for i, c in enumerate(classes):\n",
    "    print(\n",
    "        \"Muon Neutrino CC {0}\\t efficiency: ({1:0.2f}+-{2:0.2f})%\".format(\n",
    "            dict_modes[c], nom[i] / denom[i] * 100, error[i] * 100\n",
    "        )\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot efficiency per type and per channel in true neutrino energy\n",
    "fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(5.5 * gr, 4), sharex=True)\n",
    "\n",
    "# left\n",
    "for k, v in dict_modes.items():\n",
    "    if v == \"?\":\n",
    "        continue\n",
    "    type_sig = (d[\"nu\"][\"events\"].array(\"mc_nu_interaction_type\") == k) & sig_mask_ccinc\n",
    "    type_sel = (d[\"nu\"][\"events\"].array(\"mc_nu_interaction_type\") == k) & sig_mask_sel\n",
    "\n",
    "    denom = d[\"nu\"][\"events\"].array(\"mc_nu_energy\")[type_sig]\n",
    "    nom = d[\"nu\"][\"events\"].array(\"mc_nu_energy\")[type_sel]\n",
    "\n",
    "    data = [denom, nom]\n",
    "    edges, _, bins, _, _ = helper.histHelper(10, kin_min_mu, 3, data)\n",
    "    fraq = bins[1] / bins[0]\n",
    "    err = np.sqrt(fraq * (1 - fraq) / bins[0])\n",
    "    fraq = np.append(fraq, fraq[-1])\n",
    "    err = np.append(err, err[-1])\n",
    "\n",
    "    ax[0].fill_between(edges, fraq - err, fraq + err, alpha=0.3, step=\"post\")\n",
    "    ax[0].step(\n",
    "        edges,\n",
    "        fraq,\n",
    "        where=\"post\",\n",
    "        label=\"{0:>4}: Efficiency {1:0.1f}%\".format(v, len(nom) / len(denom) * 100),\n",
    "    )\n",
    "\n",
    "ax[0].set_xlabel(\"True neutrino energy [GeV]\")\n",
    "ax[0].set_ylabel(\"Efficiency\")\n",
    "ax[0].set_ylim(0, 1)\n",
    "ax[0].set_xlim(kin_min_mu, 3)\n",
    "ax[0].legend(loc=\"upper left\")\n",
    "\n",
    "\n",
    "# right\n",
    "for x_proton in range(4):\n",
    "    denom = d[\"nu\"][\"events\"].array(\"mc_nu_energy\")[sig_masks_proton[x_proton]]\n",
    "    nom = d[\"nu\"][\"events\"].array(\"mc_nu_energy\")[\n",
    "        sig_masks_proton[x_proton] & d[\"nu\"][\"events\"].array(\"nu_mu_cc_selected\")\n",
    "    ]\n",
    "\n",
    "    data = [denom, nom]\n",
    "    edges, _, bins, _, _ = helper.histHelper(10, kin_min_mu, 3, data)\n",
    "    fraq = bins[1] / bins[0]\n",
    "    err = np.sqrt(fraq * (1 - fraq) / bins[0])\n",
    "    fraq = np.append(fraq, fraq[-1])\n",
    "    err = np.append(err, err[-1])\n",
    "\n",
    "    ax[1].fill_between(edges, fraq - err, fraq + err, alpha=0.3, step=\"post\")\n",
    "    ax[1].step(\n",
    "        edges,\n",
    "        fraq,\n",
    "        where=\"post\",\n",
    "        label=\"{0:>5}: Efficiency {1:0.1f}%\".format(\n",
    "            r\"$1\\mu$\" + str(x_proton) + r\"$p$\", len(nom) / len(denom) * 100\n",
    "        ),\n",
    "    )\n",
    "ax[1].set_xlabel(\"True neutrino energy [GeV]\")\n",
    "ax[1].set_ylabel(\"Efficiency\")\n",
    "ax[1].set_ylim(0, 1)\n",
    "ax[0].set_xlim(kin_min_mu, 3)\n",
    "ax[1].legend(loc=0)\n",
    "\n",
    "ax[0].set_title(title_mc_str, loc=\"right\")\n",
    "ax[1].set_title(title_mc_str, loc=\"right\")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"modes_eff.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data/MC plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_panel_data_mc_pdg(\n",
    "    ax,\n",
    "    field,\n",
    "    x_label,\n",
    "    N_bins,\n",
    "    x_min,\n",
    "    x_max,\n",
    "    query=\"python_select\",\n",
    "    legend=True,\n",
    "    y_max_scaler=1.2,\n",
    "):\n",
    "    data = [None] * 6\n",
    "    scales = [\n",
    "        d[\"nu\"][\"scale\"],\n",
    "        d[\"nu\"][\"scale\"],\n",
    "        d[\"on\"][\"scale\"],\n",
    "        d[\"off\"][\"scale\"],\n",
    "        d[\"dirt\"][\"scale\"],\n",
    "        d[\"nu\"][\"scale\"],\n",
    "        d[\"nu\"][\"scale\"],\n",
    "    ]\n",
    "\n",
    "    data[1] = d[\"nu\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[2] = d[\"on\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[3] = d[\"off\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[4] = d[\"dirt\"][\"df_daughters\"].query(query).eval(field)\n",
    "\n",
    "    data[0] = (\n",
    "        d[\"nu\"][\"df_daughters\"]\n",
    "        .query(\"abs(mc_pdg)==13 & mc_hit_fraction>0.5\")\n",
    "        .query(query)\n",
    "        .eval(field)\n",
    "    )\n",
    "    data[5-1] = (\n",
    "        d[\"nu\"][\"df_daughters\"]\n",
    "        .query(\"abs(mc_pdg)==2212 & mc_hit_fraction>0.5\")\n",
    "        .query(query)\n",
    "        .eval(field)\n",
    "    )\n",
    "    data[6-1] = (\n",
    "        d[\"nu\"][\"df_daughters\"]\n",
    "        .query(\"abs(mc_pdg)==211 & mc_hit_fraction>0.5\")\n",
    "        .query(query)\n",
    "        .eval(field)\n",
    "    )\n",
    "    edges, edges_mid, bins, err, max_val = helper.histHelper(\n",
    "        N_bins, x_min, x_max, data, weights=scales\n",
    "    )\n",
    "\n",
    "    widths = edges_mid - edges[:-1]\n",
    "    # Data/MC\n",
    "    ax[0].errorbar(\n",
    "        edges_mid,\n",
    "        bins[2],\n",
    "        xerr=widths,\n",
    "        yerr=err[2],\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"BNB On (data)\",\n",
    "    )\n",
    "    ax[0].bar(edges_mid, bins[3], lw=2, label=\"BNB Off (data)\", width=2 * widths)\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[1],\n",
    "        lw=2,\n",
    "        label=r\"$\\nu$ other (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3],\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid, bins[0], lw=2, label=r\"muons (MC)\", width=2 * widths, bottom=bins[3]\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[5-1],\n",
    "        lw=2,\n",
    "        label=r\"protons (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3] + bins[0],\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[6-1],\n",
    "        lw=2,\n",
    "        label=r\"charged pions (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3] + bins[0] + bins[5],\n",
    "    )\n",
    "    #ax[0].bar(\n",
    "    #    edges_mid,\n",
    "    #    bins[4],\n",
    "    #    lw=2,\n",
    "    #    label=r\"$\\nu$ dirt (MC)\",\n",
    "    #    width=2 * widths,\n",
    "    #    bottom=bins[1] + bins[3],\n",
    "    #)\n",
    "\n",
    "    y_err = np.sqrt(err[1] ** 2 + err[3] ** 2 + err[4] ** 2)\n",
    "    val = bins[3] + bins[1] + bins[4]\n",
    "    for m, v, e, w in zip(edges_mid, val, y_err, widths):\n",
    "        ax[0].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, v - e),\n",
    "                2 * w,\n",
    "                2 * e,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "        sc_err = e / v\n",
    "        ax[1].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, 1 - sc_err),\n",
    "                2 * w,\n",
    "                sc_err * 2,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "    ratio = sum(bins[2]) / (sum(bins[1]) + sum(bins[4]) + sum(bins[3]))\n",
    "    ax[0].set_ylabel(\"Tracks per bin\")\n",
    "    ax[0].set_title(title_str, loc=\"right\")\n",
    "    ax[0].set_ylim(0, y_max_scaler * max_val[2])\n",
    "    ax[0].set_xlim(x_min, x_max)\n",
    "    if legend:\n",
    "        ax[0].legend()\n",
    "\n",
    "    # Ratio plots\n",
    "    ax[1].set_ylim(0.50, 2)\n",
    "    ax[1].set_xlim(x_min, x_max)\n",
    "    ax[1].errorbar(\n",
    "        edges_mid,\n",
    "        bins[2] / val,\n",
    "        xerr=widths,\n",
    "        yerr=err[2] / val,\n",
    "        alpha=1.0,\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"Data error\",\n",
    "    )\n",
    "    ax[1].set_ylabel(r\"$\\frac{Beam\\ ON}{Beam\\ OFF + MC}$\")\n",
    "    ax[1].set_xlabel(x_label)\n",
    "    return ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data order:\n",
    "# 0: signal numuccinc\n",
    "# 1: all numu\n",
    "# 2: all on\n",
    "# 3: all off\n",
    "# 4: dirt\n",
    "# 5: numu out of fidvol\n",
    "\n",
    "\n",
    "def plot_panel_data_mc(\n",
    "    ax,\n",
    "    field,\n",
    "    x_label,\n",
    "    N_bins,\n",
    "    x_min,\n",
    "    x_max,\n",
    "    query=\"python_select\",\n",
    "    legend=True,\n",
    "    y_max_scaler=1.2,\n",
    "):\n",
    "    data = [None] * 6\n",
    "    scales = [\n",
    "        d[\"nu\"][\"scale\"],\n",
    "        d[\"nu\"][\"scale\"],\n",
    "        d[\"on\"][\"scale\"],\n",
    "        d[\"off\"][\"scale\"],\n",
    "        d[\"dirt\"][\"scale\"],\n",
    "        d[\"nu\"][\"scale\"],\n",
    "    ]\n",
    "\n",
    "    data[1] = d[\"nu\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[2] = d[\"on\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[3] = d[\"off\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[4] = d[\"dirt\"][\"df_daughters\"].query(query).eval(field)\n",
    "\n",
    "    data[0] = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").query(query).eval(field)\n",
    "    data[5] = d[\"nu\"][\"df_daughters\"].query(\"~mc_nu_fid_vol\").query(query).eval(field)\n",
    "    \n",
    "    weights = [None] * 6\n",
    "    weights[0] = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").query(query).eval('event_weight')\n",
    "    weights[1] = d[\"nu\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "    weights[2] = d[\"on\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "    weights[3] = d[\"off\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "    weights[4] = d[\"dirt\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "    weights[5] = d[\"nu\"][\"df_daughters\"].query(\"~mc_nu_fid_vol\").query(query).eval('event_weight')\n",
    "\n",
    "    for i in range(len(weights)):\n",
    "        weights[i][np.isinf(weights[i])] = 1.0\n",
    "    \n",
    "    edges, edges_mid, bins, err, max_val = helper.histHelper(\n",
    "        N_bins, x_min, x_max, data, weights=scales, w=weights,\n",
    "    )\n",
    "\n",
    "    widths = edges_mid - edges[:-1]\n",
    "    # Data/MC\n",
    "    ax[0].errorbar(\n",
    "        edges_mid,\n",
    "        bins[2],\n",
    "        xerr=widths,\n",
    "        yerr=err[2],\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"BNB On (data)\",\n",
    "    )\n",
    "    ax[0].bar(edges_mid, bins[3], lw=2, label=\"BNB Off (data)\", width=2 * widths)\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[1],\n",
    "        lw=2,\n",
    "        label=r\"$\\nu$ bkg (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3],\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[0],\n",
    "        lw=2,\n",
    "        label=r\"$\\nu$ CC (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3] + bins[5-1],\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[5-1],\n",
    "        lw=2,\n",
    "        label=r\"$\\nu$ out-FV (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[3],\n",
    "        color=\"#c15a00\",\n",
    "    )\n",
    "    ax[0].bar(\n",
    "        edges_mid,\n",
    "        bins[4],\n",
    "        lw=2,\n",
    "        label=r\"$\\nu$ dirt (MC)\",\n",
    "        width=2 * widths,\n",
    "        bottom=bins[1] + bins[3],\n",
    "    )\n",
    "\n",
    "    y_err = np.sqrt(err[1] ** 2 + err[3] ** 2 + err[4] ** 2)\n",
    "    val = bins[3] + bins[1] + bins[4]\n",
    "    for m, v, e, w in zip(edges_mid, val, y_err, widths):\n",
    "        ax[0].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, v - e),\n",
    "                2 * w,\n",
    "                2 * e,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "        sc_err = e / v\n",
    "        ax[1].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, 1 - sc_err),\n",
    "                2 * w,\n",
    "                sc_err * 2,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "    ratio = sum(bins[2]) / (sum(bins[1]) + sum(bins[4]) + sum(bins[3]))\n",
    "    ax[0].set_ylabel(\"Events per bin\")\n",
    "    ax[0].set_title(title_str, loc=\"right\")\n",
    "    ax[0].set_ylim(0, y_max_scaler * max_val[2])\n",
    "    ax[0].set_xlim(x_min, x_max)\n",
    "    if legend:\n",
    "        ax[0].legend()\n",
    "\n",
    "    # Ratio plots\n",
    "    ax[1].set_ylim(0.50, 2)\n",
    "    ax[1].set_xlim(x_min, x_max)\n",
    "    ax[1].errorbar(\n",
    "        edges_mid,\n",
    "        bins[2] / val,\n",
    "        xerr=widths,\n",
    "        yerr=err[2] / val,\n",
    "        alpha=1.0,\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"Data error\",\n",
    "    )\n",
    "    ax[1].set_ylabel(r\"$\\frac{Beam\\ ON}{Beam\\ OFF + MC}$\")\n",
    "    ax[1].set_xlabel(x_label)    \n",
    "    return ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "query = \"python_select\"\n",
    "\n",
    "w_nu = d[\"nu\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "w_nu[np.isinf(w_nu)] = 1\n",
    "n_nu = np.sum(w_nu) * d['nu']['scale']\n",
    "\n",
    "w_on = d[\"on\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "w_on[np.isinf(w_on)] = 1\n",
    "n_on = np.sum(w_on) * d['on']['scale']\n",
    "\n",
    "w_off = d[\"off\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "w_off[np.isinf(w_off)] = 1\n",
    "n_off = np.sum(w_off) * d['off']['scale']\n",
    "\n",
    "w_dirt = d[\"dirt\"][\"df_daughters\"].query(query).eval('event_weight')\n",
    "w_dirt[np.isinf(w_dirt)] = 1\n",
    "n_dirt = np.sum(w_dirt) * d['dirt']['scale']\n",
    "\n",
    "rscale = n_on / (n_nu + n_off + n_dirt)\n",
    "print('Data/MC ratio:', rscale)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# track plot\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=1, nrows=2, figsize=(3.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax,\n",
    "    \"track_length\",\n",
    "    \"Muon candidate length [cm]\",\n",
    "    N_bins=23,\n",
    "    x_min=5,\n",
    "    x_max=350,\n",
    "    query=query_select_wo_trackcut,\n",
    ")\n",
    "ax[0].axvline(x=20, color=\"k\", alpha=0.3)\n",
    "ax[1].axvline(x=20, color=\"k\", alpha=0.3)\n",
    "ax[0].text(\n",
    "    100, ax[0].get_ylim()[1] * 0.9, \"Data/MC ratio: {0:.2f}\".format(ratio), fontsize=11\n",
    ")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_tracklength.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Muon candidate angles\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=2, nrows=2, figsize=(6.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "# left\n",
    "plot_panel_data_mc(\n",
    "    ax.T[0],\n",
    "    \"phi_beam\",\n",
    "    \"Muon candidate phi\",\n",
    "    N_bins=20,\n",
    "    x_min=-np.pi,\n",
    "    x_max=np.pi,\n",
    "    legend=False,\n",
    ")\n",
    "x_ticks = [-np.pi, -np.pi / 2, 0, np.pi / 2, np.pi]\n",
    "x_labels = [r\"$-\\pi$\", r\"$-\\pi/2$\", r\"$0$\", r\"$\\pi/2$\", r\"$\\pi$\"]\n",
    "ax[1][0].set_xticks(x_ticks)\n",
    "ax[1][0].set_xticklabels(x_labels)\n",
    "ax[0][0].set_xticks(x_ticks)\n",
    "ax[0][0].set_xticklabels(x_labels)\n",
    "\n",
    "# right\n",
    "# left\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax.T[1], \"theta_beam\", \"Muon candidate theta\", N_bins=20, x_min=0, x_max=np.pi\n",
    ")\n",
    "x_ticks = [0, np.pi / 4, np.pi / 2, 3 * np.pi / 4, np.pi]\n",
    "x_labels = [r\"$0$\", r\"$\\pi/4$\", r\"$\\pi/2$\", r\"$3\\pi/4$\", r\"$\\pi$\"]\n",
    "ax[1][1].set_xticks(x_ticks)\n",
    "ax[1][1].set_xticklabels(x_labels)\n",
    "ax[0][1].set_xticks(x_ticks)\n",
    "ax[0][1].set_xticklabels(x_labels)\n",
    "ax[0][0].text(\n",
    "    -3,\n",
    "    ax[0][0].get_ylim()[1] * 0.9,\n",
    "    \"Data/MC ratio: {0:.3f}\".format(ratio),\n",
    "    fontsize=11,\n",
    ")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_phi_theta.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# track momentum (MCS)\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=1, nrows=2, figsize=(3.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax,\n",
    "    \"track_mcs_mom\",\n",
    "    \"Muon candidate momentum (MCS) [GeV]\",\n",
    "    N_bins=19,\n",
    "    x_min=0.1,\n",
    "    x_max=2.0,\n",
    ")\n",
    "ax[1].axvline(x=20, color=\"k\", alpha=0.3)\n",
    "ax[0].text(\n",
    "    0.2, ax[0].get_ylim()[1] * 0.9, \"Data/MC ratio: {0:.3f}\".format(ratio), fontsize=11\n",
    ")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_trackmom.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# VTX efficiency in function of the xyz coordinates\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=3, nrows=2, figsize=(8.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "\n",
    "field_arr = [\"vx\", \"vy\", \"vz\"]\n",
    "lab_arr = [\"Muon candidate vx [cm]\", \"Muon candidate vy [cm]\", \"Muon candidate vz [cm]\"]\n",
    "x_min_arr = helper.lower + np.array([10, 10, 10])\n",
    "x_max_arr = helper.upper - np.array([10, 10, 50])\n",
    "\n",
    "for i, (x_min, x_max, lab, field) in enumerate(\n",
    "    zip(x_min_arr, x_max_arr, lab_arr, field_arr)\n",
    "):\n",
    "    N = int(round(x_max - x_min) / 10)\n",
    "    N = min(N, 30)\n",
    "    sc = 1\n",
    "    if i == 0:\n",
    "        sc = 1.7\n",
    "    ratio = plot_panel_data_mc(\n",
    "        ax.T[i],\n",
    "        field,\n",
    "        lab,\n",
    "        N_bins=N,\n",
    "        x_min=x_min,\n",
    "        x_max=x_max,\n",
    "        y_max_scaler=1.2 * sc,\n",
    "        legend=(i == 0),\n",
    "    )\n",
    "\n",
    "\n",
    "ax[0][2].text(\n",
    "    200,\n",
    "    ax[0][2].get_ylim()[1] * 0.9,\n",
    "    \"Data/MC ratio: {0:.2f}\".format(ratio),\n",
    "    fontsize=11,\n",
    ")\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_reco_muon_vtx.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(\n",
    "    ncols=3,\n",
    "    nrows=2,\n",
    "    figsize=(7.5 * gr, 4),\n",
    "    sharex=\"col\",\n",
    "    gridspec_kw={\"height_ratios\": [3, 1]},\n",
    ")\n",
    "\n",
    "field_arr = [\"num_tracks\", \"num_showers\", \"num_tracks+num_showers\"]\n",
    "lab_arr = [\"Track multiplicity\", \"Shower multiplicity\", \"Particle multiplicity\"]\n",
    "x_min_arr = [0.5, -0.5, 0.5]\n",
    "x_max_arr = [6.5, 4.5, 9.5]\n",
    "bins_arr = [6, 5, 9]\n",
    "\n",
    "for i, (x_min, x_max, lab, field, N) in enumerate(\n",
    "    zip(x_min_arr, x_max_arr, lab_arr, field_arr, bins_arr)\n",
    "):\n",
    "    ratio = plot_panel_data_mc(\n",
    "        ax.T[i],\n",
    "        field,\n",
    "        lab,\n",
    "        N_bins=N,\n",
    "        x_min=x_min,\n",
    "        x_max=x_max,\n",
    "        y_max_scaler=1.05,\n",
    "        legend=(i == 1),\n",
    "    )\n",
    "\n",
    "ax[0][2].text(\n",
    "    4, ax[0][2].get_ylim()[1] * 0.9, \"Data/MC ratio: {0:.2f}\".format(ratio), fontsize=11\n",
    ")\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_num_pfp.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data MC for events cuts: flash score and topo score after the muon selection\n",
    "# track plot PID\n",
    "query_str = \"nu_pdg==14 &\\\n",
    "             daughters_start_contained &\\\n",
    "             nu_contained &\\\n",
    "             longest_muon_candidate_length==track_length\"\n",
    "\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=3, nrows=2, figsize=(8.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax.T[0],\n",
    "    \"nu_score\",\n",
    "    r\"Topological Neutrino Score\",\n",
    "    N_bins=33,\n",
    "    x_min=0,\n",
    "    x_max=1,\n",
    "    query=query_str,\n",
    "    y_max_scaler=1.05,\n",
    "    legend=True,\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax.T[1],\n",
    "    \"nu_flash_chi2\",\n",
    "    r\"Neutrino Flash $\\chi^2$\",\n",
    "    N_bins=30,\n",
    "    x_min=0,\n",
    "    x_max=30,\n",
    "    query=query_str,\n",
    "    y_max_scaler=1.05,\n",
    "    legend=False,\n",
    ")\n",
    "\n",
    "ax[0][1].text(\n",
    "    10,\n",
    "    ax[0][1].get_ylim()[1] * 0.9,\n",
    "    \"Data/MC ratio: {0:.2f}\".format(ratio),\n",
    "    fontsize=11,\n",
    ")\n",
    "\n",
    "ax[0][2].hist2d(\n",
    "    d[\"nu\"][\"df_daughters\"].query(query_str)[\"nu_score\"],\n",
    "    d[\"nu\"][\"df_daughters\"].query(query_str)[\"nu_flash_chi2\"],\n",
    "    cmap=\"coolwarm\",\n",
    "    range=[[0, 1], [0, 30]],\n",
    "    bins=[20, 20],\n",
    "    norm=LogNorm(),\n",
    ")\n",
    "ax[0][2].add_patch(patches.Rectangle((0, 10), 0.25, 40, alpha=0.3, color=\"k\", lw=0))\n",
    "ax[0][2].add_patch(patches.Rectangle((0, 0), 0.06, 10, alpha=0.3, color=\"k\", lw=0))\n",
    "ax[0][2].set_xlabel(\"Topological Neutrino Score\")\n",
    "ax[0][2].set_ylabel(r\"Neutrino Flash $\\chi^2$\")\n",
    "ax[0][2].set_title(r\"Cut shown for $\\nu_\\mu \\; CC$ signal events\")\n",
    "ax[1][2].set_visible(False)\n",
    "\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_optopo_cuts.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# RUN THIS CELL ONLY AT THE END AND RESET AFTER\n",
    "\n",
    "trackscorecut = np.linspace(0.75, 0.95, 21)\n",
    "\n",
    "eff = []\n",
    "eff_err = []\n",
    "pur = []\n",
    "pur_err = []\n",
    "nue_pass = []\n",
    "nue_pass_err = []\n",
    "\n",
    "muon_candidate_selection_alt = \"vtx_distance<4.0 &\\\n",
    "                                track_chi2_proton>60.0 &\\\n",
    "                                track_chi2_muon<30.0 &\\\n",
    "                                (track_chi2_proton/track_chi2_muon)>7.0 &\\\n",
    "                                generation==2 &\\\n",
    "                                track_length>5\"\n",
    "\n",
    "for tr_sc_cut in trackscorecut:\n",
    "    print(\"{:.2f}\".format(tr_sc_cut))\n",
    "    muon_sel_str = muon_candidate_selection_alt + \" & track_score>@tr_sc_cut\"\n",
    "\n",
    "    for sample in dict_names:\n",
    "        # Tag the muon candidate\n",
    "        d[sample][\"df_daughters\"][\"muon_candidate\"] = d[sample][\"df_daughters\"].eval(\n",
    "            muon_sel_str\n",
    "        )\n",
    "        d[sample][\"df_daughters\"].loc[\n",
    "            d[sample][\"df_daughters\"][\"muon_candidate\"], \"longest_muon_candidate_length\"\n",
    "        ] = (\n",
    "            d[sample][\"df_daughters\"]\n",
    "            .query(\"muon_candidate\")\n",
    "            .groupby([\"hash\"])[\"track_length\"]\n",
    "            .transform(max)\n",
    "        )\n",
    "        d[sample][\"df_daughters\"].loc[\n",
    "            ~d[sample][\"df_daughters\"][\"muon_candidate\"],\n",
    "            \"longest_muon_candidate_length\",\n",
    "        ] = -1\n",
    "\n",
    "        # Tag the event selection\n",
    "        d[sample][\"df_daughters\"][\"python_select\"] = d[sample][\"df_daughters\"].eval(\n",
    "            query_select\n",
    "        )\n",
    "\n",
    "    num_sel = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\")[\"python_select\"].sum()\n",
    "    num_denom = sum(sig_mask_ccinc)\n",
    "    error = helper.effErr(num_sel, num_denom)\n",
    "\n",
    "    eff.append(num_sel / num_denom * 100)\n",
    "    eff_err.append(error * 100)\n",
    "\n",
    "    print(\n",
    "        \"Muon neutrino CC inc selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "            num_sel / num_denom * 100, error * 100\n",
    "        )\n",
    "    )\n",
    "    # purity\n",
    "    purity_nom = (\n",
    "        d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_select).sum()\n",
    "        * d[\"nu\"][\"scale\"]\n",
    "    )\n",
    "    purity_nom_plus = (\n",
    "        d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_event).sum()\n",
    "        * d[\"nu\"][\"scale\"]\n",
    "    )\n",
    "    purity_denom = 0\n",
    "\n",
    "    for s in [\"off\", \"nu\", \"dirt\"]:\n",
    "        purity_denom += d[s][\"df_daughters\"][\"python_select\"].sum() * d[s][\"scale\"]\n",
    "\n",
    "    error = helper.effErr(purity_nom, purity_denom)\n",
    "    purity = purity_nom / purity_denom\n",
    "\n",
    "    pur.append(purity * 100)\n",
    "    pur_err.append(error * 100)\n",
    "    print(\n",
    "        \"Muon neutrino CC inc selection purity (basic): ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "            purity * 100, error * 100\n",
    "        )\n",
    "    )\n",
    "    # passing nues:\n",
    "    num_nue_pass = sum(\n",
    "        d[\"nue\"][\"df_daughters\"]\n",
    "        .query(\"1.5>mc_nu_energy>0.05 & mc_nu_ccnc==0 & mc_nu_pdg==12\")\n",
    "        .query(\"python_select\")[\"event_weight\"]\n",
    "    )\n",
    "    num_nue_pass_err = np.sqrt(num_nue_pass)\n",
    "    pot_full = 1.32e21\n",
    "    scale_nue_full = pot_full / pot_on\n",
    "    print(\n",
    "        \"Number of electron neutrinos [50MeV-1.5GeV] passing CC inclusive filter: {0:.2f}+/-{1:.2f} per {2:.3g}POT.\".format(\n",
    "            num_nue_pass * d[\"nue\"][\"scale\"] * scale_nue_full,\n",
    "            num_nue_pass_err * d[\"nue\"][\"scale\"] * scale_nue_full,\n",
    "            pot_full,\n",
    "        )\n",
    "    )\n",
    "\n",
    "    nue_pass.append(num_nue_pass * d[\"nue\"][\"scale\"] * scale_nue_full)\n",
    "    nue_pass_err.append(num_nue_pass_err * d[\"nue\"][\"scale\"] * scale_nue_full)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Raquel request:\n",
    "for string in [\"contained\", \"~contained\"]:\n",
    "    query_str = \"track_score>0.85 & vtx_distance<4 & track_length>20 &\" + string\n",
    "    # track plot PID after doing the 3 first cuts.\n",
    "    fig, ax = plt.subplots(\n",
    "        ncols=3, nrows=2, figsize=(8.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    "    )\n",
    "\n",
    "    ratio = plot_panel_data_mc_pdg(\n",
    "        ax.T[0],\n",
    "        \"track_chi2_muon\",\n",
    "        r\"Track PID: muon $\\chi^2$\",\n",
    "        N_bins=30,\n",
    "        x_min=0,\n",
    "        x_max=60,\n",
    "        query=query_str,\n",
    "        y_max_scaler=1.05,\n",
    "        legend=True,\n",
    "    )\n",
    "    ax[0][0].axvline(x=30, color=\"k\", alpha=0.3)\n",
    "    ax[1][0].axvline(x=30, color=\"k\", alpha=0.3)\n",
    "\n",
    "    ratio = plot_panel_data_mc_pdg(\n",
    "        ax.T[1],\n",
    "        \"track_chi2_proton\",\n",
    "        r\"Track PID: proton $\\chi^2$\",\n",
    "        N_bins=35,\n",
    "        x_min=0,\n",
    "        x_max=350,\n",
    "        query=query_str,\n",
    "        y_max_scaler=1.05,\n",
    "        legend=False,\n",
    "    )\n",
    "    ax[0][1].axvline(x=60, color=\"k\", alpha=0.3)\n",
    "    ax[1][1].axvline(x=60, color=\"k\", alpha=0.3)\n",
    "\n",
    "    ratio = plot_panel_data_mc_pdg(\n",
    "        ax.T[2],\n",
    "        \"(track_chi2_proton/track_chi2_muon)\",\n",
    "        r\"Track PID: proton $\\chi^2$ / muon $\\chi^2$\",\n",
    "        N_bins=40,\n",
    "        x_min=0,\n",
    "        x_max=40,\n",
    "        query=query_str,\n",
    "        y_max_scaler=1.05,\n",
    "        legend=False,\n",
    "    )\n",
    "    ax[0][2].axvline(x=7, color=\"k\", alpha=0.3)\n",
    "    ax[1][2].axvline(x=7, color=\"k\", alpha=0.3)\n",
    "\n",
    "    ax[0][1].text(\n",
    "        15,\n",
    "        ax[0][1].get_ylim()[1] * 0.9,\n",
    "        \"Data/MC ratio: {0:.2f}\".format(ratio),\n",
    "        fontsize=11,\n",
    "    )\n",
    "\n",
    "    fig.tight_layout()\n",
    "    fig.savefig(outdir + \"numu_muon_pid2_pdg\" + string + \".pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Raquel data/data\n",
    "fig, ax_all = plt.subplots(\n",
    "    ncols=3, nrows=2, figsize=(8.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "query = \"track_score>0.85 & vtx_distance<4 & track_length>20 & ~contained\"\n",
    "\n",
    "fields_arr = [\n",
    "    \"track_chi2_muon\",\n",
    "    \"track_chi2_proton\",\n",
    "    \"(track_chi2_proton/track_chi2_muon)\",\n",
    "]\n",
    "labels_arr = [\n",
    "    r\"Track PID: muon $\\chi^2$\",\n",
    "    r\"Track PID: proton $\\chi^2$\",\n",
    "    r\"Track PID: proton $\\chi^2$ / muon $\\chi^2$\",\n",
    "]\n",
    "N_bins_arr = [30, 35, 40]\n",
    "x_min_arr = [0, 0, 0]\n",
    "x_max_arr = [60, 350, 40]\n",
    "\n",
    "for i in range(3):\n",
    "    ax = ax_all.T[i]\n",
    "    x_min = x_min_arr[i]\n",
    "    x_max = x_max_arr[i]\n",
    "    N_bins = N_bins_arr[i]\n",
    "    field = fields_arr[i]\n",
    "    x_label = labels_arr[i]\n",
    "\n",
    "    scales = [1, d[\"off\"][\"events\"].numentries / d[\"on\"][\"events\"].numentries]\n",
    "    data[0] = d[\"on\"][\"df_daughters\"].query(query).eval(field)\n",
    "    data[1] = d[\"off\"][\"df_daughters\"].query(query).eval(field)\n",
    "    edges, edges_mid, bins, err, max_val = helper.histHelper(\n",
    "        N_bins, x_min, x_max, data, weights=scales\n",
    "    )\n",
    "    widths = edges_mid - edges[:-1]\n",
    "    # Data/MC\n",
    "    ax[0].errorbar(\n",
    "        edges_mid,\n",
    "        bins[0],\n",
    "        xerr=widths,\n",
    "        yerr=err[0],\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"BNB On (data)\",\n",
    "    )\n",
    "    ax[0].bar(edges_mid, bins[1], lw=2, label=\"BNB Off (data)\", width=2 * widths)\n",
    "    y_err = err[1]\n",
    "    val = bins[1]\n",
    "    for m, v, e, w in zip(edges_mid, val, y_err, widths):\n",
    "        ax[0].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, v - e),\n",
    "                2 * w,\n",
    "                2 * e,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "        sc_err = e / v\n",
    "        ax[1].add_patch(\n",
    "            patches.Rectangle(\n",
    "                (m - w, 1 - sc_err),\n",
    "                2 * w,\n",
    "                sc_err * 2,\n",
    "                hatch=\"\\\\\\\\\\\\\\\\\\\\\",\n",
    "                Fill=False,\n",
    "                linewidth=0,\n",
    "                alpha=0.4,\n",
    "            )\n",
    "        )\n",
    "    ax[0].set_ylabel(\"Tracks per bin\")\n",
    "    ax[0].set_title(title_str, loc=\"right\")\n",
    "    ax[0].set_ylim(0, 1.1 * max_val[0])\n",
    "    ax[0].set_xlim(x_min, x_max)\n",
    "\n",
    "    # Ratio plots\n",
    "    ax[1].set_ylim(0.50, 1.50)\n",
    "    ax[1].set_xlim(x_min, x_max)\n",
    "    ax[1].errorbar(\n",
    "        edges_mid,\n",
    "        bins[0] / val,\n",
    "        xerr=widths,\n",
    "        yerr=err[0] / val,\n",
    "        alpha=1.0,\n",
    "        color=\"k\",\n",
    "        fmt=\".\",\n",
    "        label=\"Data error\",\n",
    "    )\n",
    "    ax[1].set_ylabel(r\"$\\frac{Beam\\ ON}{Beam\\ OFF}$\")\n",
    "    ax[1].set_xlabel(x_label)\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_muon_piddata_pdg.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Data MC for events cuts: flash score and topo score after the muon selection\n",
    "# track plot PID\n",
    "query_str = \"nu_pdg==14 &\\\n",
    "             daughters_start_contained &\\\n",
    "             nu_contained &\\\n",
    "             longest_muon_candidate_length==track_length\"\n",
    "\n",
    "fig, ax = plt.subplots(\n",
    "    ncols=3, nrows=2, figsize=(8.5 * gr, 5), gridspec_kw={\"height_ratios\": [3, 1]}\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax.T[0],\n",
    "    \"nu_score\",\n",
    "    r\"Topological Neutrino Score\",\n",
    "    N_bins=33,\n",
    "    x_min=0,\n",
    "    x_max=1,\n",
    "    query=query_str,\n",
    "    y_max_scaler=1.05,\n",
    "    legend=True,\n",
    ")\n",
    "\n",
    "ratio = plot_panel_data_mc(\n",
    "    ax.T[1],\n",
    "    \"nu_flash_chi2\",\n",
    "    r\"Neutrino Flash $\\chi^2$\",\n",
    "    N_bins=30,\n",
    "    x_min=0,\n",
    "    x_max=30,\n",
    "    query=query_str,\n",
    "    y_max_scaler=1.05,\n",
    "    legend=False,\n",
    ")\n",
    "\n",
    "ax[0][1].text(\n",
    "    10,\n",
    "    ax[0][1].get_ylim()[1] * 0.9,\n",
    "    \"Data/MC ratio: {0:.2f}\".format(ratio),\n",
    "    fontsize=11,\n",
    ")\n",
    "\n",
    "ax[0][2].hist2d(\n",
    "    d[\"nu\"][\"df_daughters\"].query(query_str)[\"nu_score\"],\n",
    "    d[\"nu\"][\"df_daughters\"].query(query_str)[\"nu_flash_chi2\"],\n",
    "    cmap=\"coolwarm\",\n",
    "    range=[[0, 1], [0, 30]],\n",
    "    bins=[20, 20],\n",
    "    norm=LogNorm(),\n",
    ")\n",
    "ax[0][2].add_patch(patches.Rectangle((0, 10), 0.25, 40, alpha=0.3, color=\"k\", lw=0))\n",
    "ax[0][2].add_patch(patches.Rectangle((0, 0), 0.06, 10, alpha=0.3, color=\"k\", lw=0))\n",
    "ax[0][2].set_xlabel(\"Topological Neutrino Score\")\n",
    "ax[0][2].set_ylabel(r\"Neutrino Flash $\\chi^2$\")\n",
    "ax[0][2].set_title(r\"Cut shown for $\\nu_\\mu \\; CC$ signal events\")\n",
    "ax[1][2].set_visible(False)\n",
    "\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_optopo_cuts.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "reload(helper)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# RUN THIS CELL ONLY AT THE END AND RESET AFTER\n",
    "\n",
    "trackscorecut = np.linspace(0.75, 0.95, 21)\n",
    "\n",
    "eff = []\n",
    "eff_err = []\n",
    "pur = []\n",
    "pur_err = []\n",
    "nue_pass = []\n",
    "nue_pass_err = []\n",
    "\n",
    "muon_candidate_selection_alt = \"vtx_distance<4.0 &\\\n",
    "                                track_chi2_proton>60.0 &\\\n",
    "                                track_chi2_muon<30.0 &\\\n",
    "                                (track_chi2_proton/track_chi2_muon)>7.0 &\\\n",
    "                                generation==2 &\\\n",
    "                                track_length>5\"\n",
    "\n",
    "for tr_sc_cut in trackscorecut:\n",
    "    print(\"{:.2f}\".format(tr_sc_cut))\n",
    "    muon_sel_str = muon_candidate_selection_alt + \" & track_score>@tr_sc_cut\"\n",
    "\n",
    "    for sample in dict_names:\n",
    "        # Tag the muon candidate\n",
    "        d[sample][\"df_daughters\"][\"muon_candidate\"] = d[sample][\"df_daughters\"].eval(\n",
    "            muon_sel_str\n",
    "        )\n",
    "        d[sample][\"df_daughters\"].loc[\n",
    "            d[sample][\"df_daughters\"][\"muon_candidate\"], \"longest_muon_candidate_length\"\n",
    "        ] = (\n",
    "            d[sample][\"df_daughters\"]\n",
    "            .query(\"muon_candidate\")\n",
    "            .groupby([\"hash\"])[\"track_length\"]\n",
    "            .transform(max)\n",
    "        )\n",
    "        d[sample][\"df_daughters\"].loc[\n",
    "            ~d[sample][\"df_daughters\"][\"muon_candidate\"],\n",
    "            \"longest_muon_candidate_length\",\n",
    "        ] = -1\n",
    "\n",
    "        # Tag the event selection\n",
    "        d[sample][\"df_daughters\"][\"python_select\"] = d[sample][\"df_daughters\"].eval(\n",
    "            query_select\n",
    "        )\n",
    "\n",
    "    num_sel = d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\")[\"python_select\"].sum()\n",
    "    num_denom = sum(sig_mask_ccinc)\n",
    "    error = helper.effErr(num_sel, num_denom)\n",
    "\n",
    "    eff.append(num_sel / num_denom * 100)\n",
    "    eff_err.append(error * 100)\n",
    "\n",
    "    print(\n",
    "        \"Muon neutrino CC inc selection efficiency: ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "            num_sel / num_denom * 100, error * 100\n",
    "        )\n",
    "    )\n",
    "    # purity\n",
    "    purity_nom = (\n",
    "        d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_select).sum()\n",
    "        * d[\"nu\"][\"scale\"]\n",
    "    )\n",
    "    purity_nom_plus = (\n",
    "        d[\"nu\"][\"df_daughters\"].query(\"mc_sig_nuccinc\").eval(query_event).sum()\n",
    "        * d[\"nu\"][\"scale\"]\n",
    "    )\n",
    "    purity_denom = 0\n",
    "\n",
    "    for s in [\"off\", \"nu\", \"dirt\"]:\n",
    "        purity_denom += d[s][\"df_daughters\"][\"python_select\"].sum() * d[s][\"scale\"]\n",
    "\n",
    "    error = helper.effErr(purity_nom, purity_denom)\n",
    "    purity = purity_nom / purity_denom\n",
    "\n",
    "    pur.append(purity * 100)\n",
    "    pur_err.append(error * 100)\n",
    "    print(\n",
    "        \"Muon neutrino CC inc selection purity (basic): ({0:0.2f}+-{1:0.2f})%\".format(\n",
    "            purity * 100, error * 100\n",
    "        )\n",
    "    )\n",
    "    # passing nues:\n",
    "    num_nue_pass = sum(\n",
    "        d[\"nue\"][\"df_daughters\"]\n",
    "        .query(\"1.5>mc_nu_energy>0.05 & mc_nu_ccnc==0 & mc_nu_pdg==12\")\n",
    "        .query(\"python_select\")[\"event_weight\"]\n",
    "    )\n",
    "    num_nue_pass_err = np.sqrt(num_nue_pass)\n",
    "    pot_full = 1.32e21\n",
    "    scale_nue_full = pot_full / pot_on\n",
    "    print(\n",
    "        \"Number of electron neutrinos [50MeV-1.5GeV] passing CC inclusive filter: {0:.2f}+/-{1:.2f} per {2:.3g}POT.\".format(\n",
    "            num_nue_pass * d[\"nue\"][\"scale\"] * scale_nue_full,\n",
    "            num_nue_pass_err * d[\"nue\"][\"scale\"] * scale_nue_full,\n",
    "            pot_full,\n",
    "        )\n",
    "    )\n",
    "\n",
    "    nue_pass.append(num_nue_pass * d[\"nue\"][\"scale\"] * scale_nue_full)\n",
    "    nue_pass_err.append(num_nue_pass_err * d[\"nue\"][\"scale\"] * scale_nue_full)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(5 * gr, 4), sharex=True)\n",
    "ax[0].errorbar(trackscorecut, pur, yerr=pur_err, label=\"Purity\", fmt=\".\")\n",
    "ax[0].errorbar(trackscorecut, eff, yerr=eff_err, label=\"Efficiency\", fmt=\".\")\n",
    "ax[0].axvline(x=0.85, label=\"Suggested cut\", color=\"k\", alpha=0.33)\n",
    "ax[0].legend()\n",
    "ax[0].set_xlabel(\"Minimum Trackscore Cut\")\n",
    "ax[0].set_ylabel(r\"%\")\n",
    "ax[0].set_title(title_mc_str, loc=\"right\")\n",
    "\n",
    "ax[1].errorbar(\n",
    "    trackscorecut,\n",
    "    nue_pass,\n",
    "    yerr=nue_pass_err,\n",
    "    label=r\"$\\nu_e$ passing selection\",\n",
    "    fmt=\".\",\n",
    "    color=\"C2\",\n",
    ")\n",
    "ax[1].axvline(x=0.85, label=\"Suggested cut\", color=\"k\", alpha=0.33)\n",
    "ax[1].axhline(y=45, label=\"Unblinding requirement\", color=\"C3\", alpha=0.5)\n",
    "ax[1].legend(loc=\"lower left\")\n",
    "ax[1].set_xlabel(\"Minimum Trackscore Cut\")\n",
    "ax[1].set_ylabel(r\"# $\\nu_e$\")\n",
    "ax[1].set_title(title_mc_str, loc=\"right\")\n",
    "\n",
    "fig.tight_layout()\n",
    "fig.savefig(outdir + \"numu_eff_pur_track_score.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nu = d['nu']['df_daughters'].query('python_select & muon_candidate & contained')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "pmcs = df_nu.mcs_trackMom_MuFwd_rm10.where(df_nu.mcs_trackIsBestFwd_Mu_rm10==1, df_nu.mcs_trackMom_MuBwd_rm10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "prange = df_nu.track_range_mom_mu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "ax.text?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(ncols=1, nrows=1, figsize=(5, 4), sharex=True)\n",
    "\n",
    "ax.hist2d(\n",
    "    prange,\n",
    "    pmcs,\n",
    "    bins=(50,50),\n",
    "    range=((0, 3), (0, 3)),\n",
    "    cmap=\"coolwarm\",\n",
    "    norm=LogNorm(),\n",
    ")\n",
    "ax.set_xlabel(\"p range [GeV/c]\")\n",
    "ax.set_ylabel(\"p MCS [GeV/c]\")\n",
    "#plt.title(\"MicroBooNE In Progress\")\n",
    "ax.text(2.9, 0.1, 'MicroBooNE In Progress', verticalalignment='bottom', horizontalalignment='right', fontsize=12)\n",
    "#plt.text(x, y, \"MicroBooNE In Progress\")\n",
    "#plt.savefig(\"MCSvsRangeMomentum.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pseudo-exiting track MCS momentum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,4))\n",
    "\n",
    "xbins = np.linspace(0.4, 1.8, 8)\n",
    "#8\n",
    "# Extract a profile histogram\n",
    "def get_profile(sample):\n",
    "    #  Selected contained muon candidate tracks\n",
    "    #muon_candidate & python_select\n",
    "    df = d[sample]['df_daughters'].query(\"python_select&muon_candidate & contained & hitsU > 150 & hitsV > 150 & hitsY > 150 & track_mcs_ll < 300 & mcs_trackMomErr_MuFwd_rm10 <7.5 & (mcs_trackMom_MuFwd_rm10 - track_range_mom_mu > -0.2)\")\n",
    "    \n",
    "    # MCS momentum, whichever of forwards and backwards fit is  best\n",
    "    pmcs = df.mcs_trackMom_MuFwd_rm10.where(df.mcs_trackIsBestFwd_Mu_rm10==1, df.mcs_trackMom_MuBwd_rm10)\n",
    "    \n",
    "    # Range-based momentum\n",
    "    prange = df.track_range_mom_mu\n",
    "    \n",
    "    flag = True\n",
    "    \n",
    "    x = xbins[:-1] + np.diff(xbins) / 2\n",
    "    y = np.empty(shape=len(xbins)-1)\n",
    "    ye = np.empty_like(y)\n",
    "    res = np.empty_like(y)\n",
    "    rese = np.empty_like(y)\\\n",
    "\n",
    "    # Loop over momentum bins\n",
    "    for i in range(len(xbins)-1):\n",
    "        # Select tracks in this bin\n",
    "        selection = np.where((prange >= xbins[i]) & (prange < xbins[i+1]) & (pmcs>0))\n",
    "        mcs = pmcs.values[selection]\n",
    "        rng = prange.values[selection]\n",
    "        \n",
    "        # Fractional bias mean\n",
    "        delta = (1.0 / mcs - 1.0 / rng) / (1.0 / rng)\n",
    "        #if(xbins[i] == 1.4):\n",
    "            #for j in range(len(specificxbins)-1):\n",
    "             #   selection = np.where((prange >= specificxbins[j]) & (prange < specificxbins[j+1]) & (pmcs>0))\n",
    "              #  mcs = pmcs.values[selection]\n",
    "              #  rng = prange.values[selection]\n",
    "        \n",
    "              #  if(np.mean((1.0 / mcs - 1.0 / rng) / (1.0 / rng))>0.14): \n",
    "               #     flag = False\n",
    "                #    continue \n",
    "               # beta = (1.0 / mcs - 1.0 / rng) / (1.0 / rng)\n",
    "               \n",
    "        if(flag==False):\n",
    "            flag = True\n",
    "            y[i] = np.mean(beta)\n",
    "            ye[i] = np.mean(beta) * np.sqrt(len(mcs)-1) / (len(mcs)-1)  \n",
    "\n",
    "            res[i] = np.std(beta)\n",
    "            rese[i] = np.std(beta) * np.sqrt(len(mcs)-1) / (len(mcs)-1)\n",
    "            continue\n",
    "                \n",
    "        y[i] = np.mean(delta)\n",
    "        ye[i] = np.mean(delta) * np.sqrt(len(mcs)) / len(mcs)  # sqrt(N) errors?\n",
    "\n",
    "        # Fractional bias width\n",
    "        res[i] = np.std(delta)\n",
    "        rese[i] = np.std(delta) * np.sqrt(len(mcs)) / len(mcs)\n",
    "\n",
    "    return x, y, ye, res, rese\n",
    "\n",
    "profiles = [\n",
    "    ('BNB MC, 2020',   'red',   get_profile('nu')),\n",
    "    ('BNB Data, 2020', 'black', get_profile('on')),\n",
    "]\n",
    "\n",
    "for name, color, (x, y, ye, res, rese) in profiles:\n",
    "    ax1.errorbar(x, y, ye, np.diff(xbins)/2, 'o', color=color, label=name)\n",
    "    ax2.errorbar(x, res, rese, np.diff(xbins)/2, 'o', color=color, label=name)\n",
    "\n",
    "    \n",
    "ax1.set_xlabel(r'$p_{range}$ (GeV/c)')\n",
    "ax1.set_xlim(xbins[0], xbins[-1])\n",
    "ax1.set_ylim(-0.1, 0.2)\n",
    "ax1.set_ylabel(r'$p^{-1}_\\mathrm{MCS}$ fractional bias')\n",
    "ax1.axhline(0, ls='dotted', c='black')\n",
    "ax1.legend(loc='upper left')\n",
    "#ax1.set_title(\"MicroBooNE In Progress\")\n",
    "ax1.text(1.75, 0.19, 'MicroBooNE In Progress', verticalalignment='top', horizontalalignment='right', fontsize=12)\n",
    "\n",
    "ax2.set_xlabel(r'$p_{range}$ (GeV/c)')\n",
    "ax2.set_xlim(xbins[0], xbins[-1])\n",
    "ax2.set_ylim(0, 0.3)\n",
    "ax2.set_ylabel(r'$p^{-1}_\\mathrm{MCS}$ fractional resolution') #it was bias previously but this makes more sense\n",
    "ax2.legend(loc='upper left')\n",
    "#ax2.set_title(\"MicroBooNE In Progress\")\n",
    "ax2.text(1.75, 0.29, 'MicroBooNE In Progress', verticalalignment='top', horizontalalignment='right', fontsize=12)\n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "#plt.savefig(\"FractionalBiasResoVsPRange.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.4\n",
      "1.45\n",
      "1.5\n",
      "1.55\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1) = plt.subplots(1, figsize=(10,4))\n",
    "xbins = np.linspace(1.4, 1.6, 5)\n",
    "df = d[\"on\"]['df_daughters'].query('python_select & muon_candidate ')\n",
    "prange = df.track_range_mom_mu\n",
    "pmcs = df.mcs_trackMom_MuFwd_rm10.where(df.mcs_trackIsBestFwd_Mu_rm10==1, df.mcs_trackMom_MuBwd_rm10)\n",
    "\n",
    "y = np.empty(shape=len(xbins)-1)\n",
    "\n",
    "for i in range(len(xbins)-1):\n",
    "    selection = np.where((prange >= xbins[i]) & (prange < xbins[i+1]) & (pmcs>0))\n",
    "    mcs = pmcs.values[selection]\n",
    "    rng = prange.values[selection]\n",
    "    print(xbins[i])\n",
    "    if(np.mean((1.0 / mcs - 1.0 / rng) / (1.0 / rng)) > 0.14): \n",
    "        print(pmcs.values[selection])\n",
    "        print(rng)\n",
    "    delta = (1.0 / mcs - 1.0 / rng) / (1.0 / rng) \n",
    "        \n",
    "    y[i] = np.mean(delta)\n",
    "ax1.hist(\n",
    "       y,\n",
    ")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split tracks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 3, figsize=(15,8))\n",
    "\n",
    "xbins = np.hstack((np.linspace(0, 1.5, 12), [2.5]))\n",
    "\n",
    "#& track_mcs_ll < 300 & vz > 100 & nu_score > 0.5 \n",
    "#& track_diry < 0.4 & track_diry > -0.4\n",
    "#& strk_mid_gpar_vz > 100\n",
    "# Selected muon candidate tracks in overlay and data\n",
    "#contained was removed and the figures look MUCH better\n",
    "#python_select & muon_candidate query('')\n",
    "\n",
    "df = d['nu']['df_daughters'].query('python_select & muon_candidate & contained')\n",
    "pmcs = df.mcs_trackMom_MuFwd_rm10.where(df.mcs_trackIsBestFwd_Mu_rm10==1, df.mcs_trackMom_MuBwd_rm10)\n",
    "g1 = df.strk_tk1_gpar_vx\n",
    "\n",
    "dfd = d['on']['df_daughters'].query('python_select & muon_candidate ')\n",
    "pmcsd = dfd.mcs_trackMom_MuFwd_rm10.where(df.mcs_trackIsBestFwd_Mu_rm10==1, df.mcs_trackMom_MuBwd_rm10)\n",
    "g1d = df.strk_tk1_gpar_vx\n",
    "\n",
    "# Loop over track fit parameters\n",
    "for i, var in enumerate(('v', 'p')):\n",
    "    for j, dim in enumerate(('x',  'y', 'z')):\n",
    "        \n",
    "        axis = ax[i,j]\n",
    "        v1 = 'strk_tk1_gpar_%s%s' % (var, dim)\n",
    "        v2 = 'strk_tk2_gpar_%s%s' % (var, dim)\n",
    "\n",
    "        x = xbins[:-1] + np.diff(xbins) / 2\n",
    "        y1 = np.zeros(shape=len(xbins)-1)\n",
    "        y1e = np.zeros_like(y1)\n",
    "        y2 = np.zeros(shape=len(xbins)-1)\n",
    "        y2e = np.zeros_like(y2)\n",
    "     \n",
    "        for k in range(len(xbins)-1):\n",
    "            \n",
    "            selection = (pmcs >= xbins[k]) & (pmcs < xbins[k+1])& (g1 != 0)\n",
    "           \n",
    "            selectiond = (pmcsd >= xbins[k]) & (pmcsd < xbins[k+1])& (g1d != 0)\n",
    "\n",
    "            # Difference in this parameter\n",
    "            delta = df[v1][selection] - df[v2][selection]\n",
    "            \n",
    "            # The shape is not Gaussian and has a long tail, so the variance is large.\n",
    "            # To control that, using the full width at half maximum; should confirm\n",
    "            # what was done in MCC8 studies.\n",
    "            tbins=np.linspace(-1, 1, 1001)\n",
    "            hist, _ = np.histogram(delta, bins=tbins)\n",
    "            br = tbins[np.where(hist>np.max(hist)/2)]\n",
    "            fwhm = np.max(br)-np.min(br)\n",
    "            \n",
    "            y1[k] = fwhm / 2.355 / np.sqrt(2)\n",
    "            y1e[k] = y1[k] * np.sqrt(np.sum(selection)) / np.sum(selection)\n",
    "\n",
    "            delta = dfd[v1][selectiond] - dfd[v2][selectiond]\n",
    "\n",
    "            hist, _ = np.histogram(delta, bins=tbins)\n",
    "            br = tbins[np.where(hist>np.max(hist)/2)]\n",
    "            fwhm = np.max(br)-np.min(br)\n",
    "            y2[k] = fwhm / 2.355 / np.sqrt(2)\n",
    "            \n",
    "            y2e[k] = y2[k] * np.sqrt(np.sum(selectiond)) / np.sum(selectiond)\n",
    "            \n",
    "            \n",
    "        \n",
    "        \n",
    "        axis.errorbar(x, y1, y1e, np.diff(xbins)/2, 'o', color='red', label='mc')\n",
    "        \n",
    "        axis.errorbar(x, y2, y2e, np.diff(xbins)/2, 'o', color='black', label='data')\n",
    "        \n",
    "        axis.set_xlabel(r'$p_{MCS}$ (GeV/c)')\n",
    "        axis.set_ylabel(r'$\\sigma(%s%s)$' % (var, dim))\n",
    "        axis.set_xlim(xbins[0], xbins[-1])\n",
    "        axis.set_ylim(0, 0.025)\n",
    "        \n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "#plt.text(x, y, \"MicroBooNE In Progress\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 3, figsize=(15,8))\n",
    "\n",
    "xbins = np.hstack((np.linspace(0, 1.5, 12), [2.5]))\n",
    "\n",
    "#& track_mcs_ll < 300 & vz > 100 & nu_score > 0.5 \n",
    "#& track_diry < 0.4 & track_diry > -0.4\n",
    "#& strk_mid_gpar_vz > 100\n",
    "#& strk_mid_gpar_vz > 150 & (-0.7 < track_diry < 0.7)\n",
    "# Selected muon candidate tracks in overlay and data\n",
    "\n",
    "df = d['nu']['df_daughters'].query('python_select & muon_candidate & hitsU > 20 & hitsV > 20 & hitsY > 20 & strk_mid_gpar_vz > 100 & num_tracks>=2')\n",
    "pmcs = df.mcs_trackMom_MuFwd_rm10.where(df.mcs_trackIsBestFwd_Mu_rm10==1, df.mcs_trackMom_MuBwd_rm10)\n",
    "g1 = df.strk_tk1_gpar_vx\n",
    "\n",
    "dfd = d['on']['df_daughters'].query('python_select & muon_candidate & hitsU > 20 & hitsV > 20 & hitsY > 20 & strk_mid_gpar_vz > 100 & num_tracks>=2')\n",
    "pmcsd = dfd.mcs_trackMom_MuFwd_rm10.where(dfd.mcs_trackIsBestFwd_Mu_rm10==1, dfd.mcs_trackMom_MuBwd_rm10)\n",
    "g1d = df.strk_tk1_gpar_vx\n",
    "\n",
    "# Loop over track fit parameters\n",
    "for i, var in enumerate(('v', 'p')):\n",
    "    for j, dim in enumerate(('x',  'y', 'z')):\n",
    "        \n",
    "        axis = ax[i,j]\n",
    "        v1 = 'strk_tk1_gpar_%s%s' % (var, dim)\n",
    "        v2 = 'strk_tk2_gpar_%s%s' % (var, dim)\n",
    "\n",
    "        x = xbins[:-1] + np.diff(xbins) / 2\n",
    "        y1 = np.zeros(shape=len(xbins)-1)\n",
    "        y1e = np.zeros_like(y1)\n",
    "        y2 = np.zeros(shape=len(xbins)-1)\n",
    "        y2e = np.zeros_like(y2)\n",
    "     \n",
    "        for k in range(len(xbins)-1):\n",
    "            \n",
    "            selection = (pmcs >= xbins[k]) & (pmcs < xbins[k+1]) & (g1 != 0)\n",
    "            selectiond = (pmcsd >= xbins[k]) & (pmcsd < xbins[k+1]) & (g1d != 0)\n",
    "\n",
    "            # Difference in this parameter\n",
    "            delta = df[v1][selection] - df[v2][selection]\n",
    "            \n",
    "            # The shape is not Gaussian and has a long tail, so the variance is large.\n",
    "            # To control that, using the full width at half maximum; should confirm\n",
    "            # what was done in MCC8 studies.\n",
    "            tbins=np.linspace(-1, 1, 1001)\n",
    "            hist, _ = np.histogram(delta, bins=tbins)\n",
    "            br = tbins[np.where(hist>np.max(hist)/2)]\n",
    "            fwhm = np.max(br)-np.min(br)\n",
    "            \n",
    "            y1[k] = fwhm / 2.355 / np.sqrt(2)\n",
    "            y1e[k] = y1[k] * np.sqrt(np.sum(selection)) / np.sum(selection)\n",
    "\n",
    "            delta = dfd[v1][selectiond] - dfd[v2][selectiond]\n",
    "\n",
    "            hist, _ = np.histogram(delta, bins=tbins)\n",
    "            br = tbins[np.where(hist>np.max(hist)/2)]\n",
    "            fwhm = np.max(br)-np.min(br)\n",
    "            y2[k] = fwhm / 2.355 / np.sqrt(2)\n",
    "            \n",
    "            y2e[k] = y2[k] * np.sqrt(np.sum(selectiond)) / np.sum(selectiond)\n",
    "            \n",
    "            \n",
    "        \n",
    "        \n",
    "        axis.errorbar(x, y1, y1e, np.diff(xbins)/2, 'o', color='red', label='mc')\n",
    "        \n",
    "        axis.errorbar(x, y2, y2e, np.diff(xbins)/2, 'o', color='black', label='data')\n",
    "        \n",
    "        axis.set_xlabel(r'$p_{MCS}$ (GeV/c)')\n",
    "        axis.set_ylabel(r'$\\sigma(%s%s)$' % (var, dim))\n",
    "        axis.set_xlim(xbins[0], xbins[-1])\n",
    "        axis.set_ylim(0, 0.025)\n",
    "        \n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "#plt.text(x, y, \"MicroBooNE In Progress\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split vertex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.optimize import curve_fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "import lmfit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 3, figsize=(15,4))\n",
    "\n",
    "xbins = np.linspace(-1, 1, 16)\n",
    "\n",
    "# Extract event-level variables of interest\n",
    "v = ['v%i%spos' % (i,j) for i in [1,2] for j in 'xyz']\n",
    "ds = d['on']['events'].pandas.df(v + ['num_tracks','nu_mu_cc_selected','nu_score'])\n",
    "ds = ds[(ds.num_tracks>=4)&(ds.nu_mu_cc_selected)&(ds.v1xpos!=0)&(ds.v2xpos!=0)&(ds.nu_score>0.5)]\n",
    "\n",
    "# Loop over x, y, z\n",
    "for i, dim in enumerate('xyz'):\n",
    "    v1, v2 = 'v1%spos' % dim, 'v2%spos' % dim\n",
    "\n",
    "    # Histogram the difference / sqrt(2)\n",
    "    delta = (ds[v1]-ds[v2])  / np.sqrt(2)\n",
    "    y, _ = np.histogram(delta, bins=xbins)\n",
    "    ye = np.sqrt(y)\n",
    "    x = xbins[:-1] + np.diff(xbins)/2\n",
    "    xe = np.diff(xbins)/2\n",
    "\n",
    "    ax[i].errorbar(x, y, ye, xe, fmt='o', color='black')\n",
    "    ax[i].set_xlabel(r'$(%s_2-%s_1)/\\sqrt{2}$ (cm)' % (dim, dim))\n",
    "    ax[i].set_ylabel('Events/bin')\n",
    "\n",
    "    ax[i].set_xlim(-1,1)\n",
    "\n",
    "    # Fit a Gaussian\n",
    "    par = lmfit.Parameters()\n",
    "    par.add('center', value=0)\n",
    "    par.add('amplitude', value=np.max(y))\n",
    "    par.add('sigma', value=0.2)\n",
    "\n",
    "    def residual(pars, x, data=None, err=None):\n",
    "        a, mu, std = pars['amplitude'], pars['center'], pars['sigma']\n",
    "        resids = data - a * scipy.stats.norm.pdf(x, mu, std)\n",
    "        return np.sqrt(np.square(resids) / np.square(err))\n",
    "\n",
    "    fit = lmfit.minimize(residual, par, args=(x[4:-5],), kws={'data': y[4:-5], 'err': ye[4:-5]})\n",
    "    \n",
    "    xbins2 = np.linspace(x[4],x[-5],1001)\n",
    "    x2 = xbins2[:-1] + np.diff(xbins2)/2\n",
    "    a, mu, std = fit.params['amplitude'], fit.params['center'], fit.params['sigma']\n",
    "    ax[i].plot(x2, a*scipy.stats.norm.pdf(x2, mu, std), c='r', zorder=-100)\n",
    "    \n",
    "    err_mu, err_amp, err_std = np.sqrt(np.diag(fit.covar))\n",
    "    \n",
    "    textstr = '\\n'.join((\n",
    "    r'$\\mathrm{N}=%1.1f\\pm%1.1f$' % (fit.params['amplitude'], err_amp),\n",
    "    r'$\\mu=%1.2f\\pm%1.2f$' % (fit.params['center'], err_mu),\n",
    "    r'$\\sigma=%1.2f\\pm%1.2f$' % (fit.params['sigma'], err_std),\n",
    "    r'$\\chi^2/\\mathrm{dof}=%1.2f$' % (fit.redchi)))\n",
    "   \n",
    "    props = dict(boxstyle='round',facecolor = 'white', alpha=0.5)\n",
    "\n",
    "    ax[i].text(0.05, 0.95, textstr, transform=ax[i].transAxes, fontsize=10,\n",
    "    verticalalignment='top', bbox=props)\n",
    "    # The parameters should go in a text box on the plot\n",
    "    #print(dim, pars)\n",
    "    \n",
    "    ymax = ax[i].get_ylim()[1]\n",
    "    ax[i].set_ylim(0, ymax*1.07)\n",
    "    ax[i].text(0.95, ymax*1.04, 'MicroBooNE In Progress', verticalalignment='top', horizontalalignment='right', fontsize=12)\n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "#plt.savefig(\"BNBsplitvertex.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['event',\n",
       " 'run',\n",
       " 'subrun',\n",
       " 'event_weight',\n",
       " 'evt_time_sec',\n",
       " 'evt_time_nsec',\n",
       " 'numpfp',\n",
       " 'hitsU',\n",
       " 'hitsV',\n",
       " 'hitsY',\n",
       " 'total_hits',\n",
       " 'caloU',\n",
       " 'caloV',\n",
       " 'caloY',\n",
       " 'hitsSps',\n",
       " 'num_primary_daughters',\n",
       " 'num_daughters',\n",
       " 'num_showers',\n",
       " 'num_tracks',\n",
       " 'daughters_stored',\n",
       " 'nu_vx',\n",
       " 'nu_vy',\n",
       " 'nu_vz',\n",
       " 'nu_contained',\n",
       " 'daughters_start_contained',\n",
       " 'nu_pdg',\n",
       " 'nu_score',\n",
       " 'nu_flash_chi2',\n",
       " 'obvious_cosmic_chi2',\n",
       " 'nu_mu_cc_selected',\n",
       " 'v1pos',\n",
       " 'v2pos',\n",
       " 'v1xpos',\n",
       " 'v2xpos',\n",
       " 'v1ypos',\n",
       " 'v2ypos',\n",
       " 'v1zpos',\n",
       " 'v2zpos']"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d['on']['events'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "contained & hitsU > 150 & hitsV > 150 & hitsY > 150 & track_mcs_ll < 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "2\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 3, figsize=(15,4))\n",
    "\n",
    "xbins = np.linspace(-1, 1, 16)\n",
    "\n",
    "# Extract event-level variables of interest\n",
    "v = ['v%i%spos' % (i,j) for i in [1,2] for j in 'xyz']\n",
    "ds = d['nu']['events'].pandas.df(v + ['num_tracks','nu_mu_cc_selected','nu_score','nu_contained','hitsU','hitsV','hitsY', 'total_hits'])\n",
    "ds = ds.query('num_tracks>=4 & nu_mu_cc_selected & v1xpos!=0 & v2xpos != 0 & nu_score>0.5')\n",
    "#& v1xpos!=0 & v2xpos != 0 & nu_score>0.5 & nu_contained & hitsU>150 & hitsV>150 &hitsY>150\n",
    "\n",
    "# Loop over x, y, z\n",
    "for i, dim in enumerate('xyz'):\n",
    "    print(i)\n",
    "    v1, v2 = 'v1%spos' % dim, 'v2%spos' % dim\n",
    "\n",
    "    # Histogram the difference / sqrt(2)\n",
    "    delta = (ds[v1]-ds[v2])  / np.sqrt(2)\n",
    "    y, _ = np.histogram(delta, bins=xbins)\n",
    "    ye = np.sqrt(y)\n",
    "    x = xbins[:-1] + np.diff(xbins)/2\n",
    "    xe = np.diff(xbins)/2\n",
    "\n",
    "    ax[i].errorbar(x, y, ye, xe, fmt='o', color='black')\n",
    "    ax[i].set_xlabel(r'$(%s_2-%s_1)/\\sqrt{2}$ (cm)' % (dim, dim))\n",
    "    ax[i].set_ylabel('Events/bin')\n",
    "    #ax[i].set_title(\"MicroBooNE In Progress\")\n",
    "\n",
    "    # Fit a Gaussian\n",
    "    par = lmfit.Parameters()\n",
    "    par.add('center', value=0)\n",
    "    par.add('amplitude', value=np.max(y))\n",
    "    par.add('sigma', value=0.2)\n",
    "\n",
    "    def residual(pars, x, data=None, err=None):\n",
    "        a, mu, std = pars['amplitude'], pars['center'], pars['sigma']\n",
    "        resids = data - a * scipy.stats.norm.pdf(x, mu, std)\n",
    "        return np.sqrt(np.square(resids) / np.square(err))\n",
    "\n",
    "    fit = lmfit.minimize(residual, par, args=(x[4:-5],), kws={'data': y[4:-5], 'err': ye[4:-5]})\n",
    "    \n",
    "    xbins2 = np.linspace(x[4],x[-5],1001)\n",
    "    x2 = xbins2[:-1] + np.diff(xbins2)/2\n",
    "    a, mu, std = fit.params['amplitude'], fit.params['center'], fit.params['sigma']\n",
    "    ax[i].plot(x2, a*scipy.stats.norm.pdf(x2, mu, std), c='r', zorder=-100)\n",
    "    \n",
    "    err_mu, err_amp, err_std = np.sqrt(np.diag(fit.covar))\n",
    "    textstr = '\\n'.join((\n",
    "    r'$\\mathrm{N}=%1.1f\\pm%1.1f$' % (fit.params['amplitude'], err_amp),\n",
    "    r'$\\mu=%1.2f\\pm%1.2f$' % (fit.params['center'], err_mu),\n",
    "    r'$\\sigma=%1.2f\\pm%1.2f$' % (fit.params['sigma'], err_std),\n",
    "    r'$\\chi^2/\\mathrm{dof}=%1.2f$' % (fit.redchi)))\n",
    "    \n",
    "    \n",
    "    props = dict(boxstyle='round',facecolor = 'white', alpha=0.5)\n",
    "\n",
    "    ax[i].text(0.05, 0.95, textstr, transform=ax[i].transAxes, fontsize=10,\n",
    "    verticalalignment='top', bbox=props)\n",
    "    # The parameters should go in a text box on the plot\n",
    "    #print(dim, pars)\n",
    "    \n",
    "    ax[i].set_xlim(-1,1)\n",
    "    ymax = ax[i].get_ylim()[1]\n",
    "    ax[i].set_ylim(0, ymax*1.1)\n",
    "    ax[i].text(0.95, ymax*1.07, 'MicroBooNE In Progress', verticalalignment='top', horizontalalignment='right', fontsize=12)\n",
    "\n",
    "\n",
    "plt.tight_layout()\n",
    "#plt.savefig(\"overlaysplitvertex.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dk = d['on']['df_daughters'].query('python_select & muon_candidate & contained & track_range_mom_mu > 1.52')\n",
    "prange = dk.track_range_mom_mu\n",
    "pmcs = dk.mcs_trackMom_MuFwd_rm10.where(dk.mcs_trackIsBestFwd_Mu_rm10==1, dk.mcs_trackMom_MuBwd_rm10)\n",
    "phitsY = dk.track_mcs_ll.where(dk.mcs_trackIsBestFwd_Mu_rm10==1, dk.mcs_trackMom_MuBwd_rm10)\n",
    "ptrackscore= dk.track_score.where(dk.mcs_trackIsBestFwd_Mu_rm10==1, dk.mcs_trackMom_MuBwd_rm10)\n",
    "\n",
    "print(\"Items of notice are: 21 and 32\")\n",
    "for i,h in enumerate(phitsY): \n",
    "    h = str(h)\n",
    "    i = str(i)\n",
    "    print(i + \" \" + h)\n",
    "        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.15+"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}