07_linthompsamp_modeling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "from striatum.bandit import linthompsamp\n",
    "from utils.data import *\n",
    "from utils.policy import *\n",
    "plt.style.use('seaborn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_DIR = os.path.abspath('../../Data/yahoo!_front_page_today_module_user_click_log/R6')\n",
    "DATA_FILE_NAMES = ['ydata-fp-td-clicks-v1_0.2009050{}.gz'.format(str(d + 3)) for d in range(7)]\n",
    "MODEL_DIR = os.path.abspath('models')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_paths = []\n",
    "for data_file in DATA_FILE_NAMES:\n",
    "    data_paths.append(os.path.join(DATA_DIR, data_file))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_file_paths = []\n",
    "for data_file_name in DATA_FILE_NAMES:\n",
    "    data_file_paths.append(os.path.join(DATA_DIR, data_file_name))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_fit = 1600000\n",
    "policy_learner = PolicyLearner(linthompsamp.LinThompSamp, context_dimension=41, delta=0.25, R=0.001,\n",
    "                               epsilon=(1.0 / np.log(n_fit)), random_state=0, seed=0, reset_freq=100000)\n",
    "param_grid = {'data_size': [0.01, 0.05, 0.1, 0.2, 0.3, 1.0]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parameters: {'data_size': 0.01} | Average Reward: 4.6639% | Computation Time: 10:36:31\n",
      "Parameters: {'data_size': 0.05} | Average Reward: 4.7803% | Computation Time: 10:23:51\n",
      "Parameters: {'data_size': 0.1} | Average Reward: 4.8524% | Computation Time: 10:31:58\n",
      "Parameters: {'data_size': 0.2} | Average Reward: 4.9071% | Computation Time: 10:28:19\n",
      "Parameters: {'data_size': 0.3} | Average Reward: 4.7304% | Computation Time: 10:30:07\n",
      "Parameters: {'data_size': 1.0} | Average Reward: 5.0219% | Computation Time: 10:55:43\n"
     ]
    }
   ],
   "source": [
    "res = search_param(policy_learner, param_grid, parse_data, data_paths=data_paths, share_context=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "if not os.path.isdir(MODEL_DIR):\n",
    "    os.mkdir(MODEL_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(os.path.join(MODEL_DIR, 'model_linthompsamp.pkl'), 'wb') as file:\n",
    "    pickle.dump(res, file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(os.path.join(MODEL_DIR, 'model_linthompsamp.pkl'), 'rb') as file:\n",
    "    res = pickle.load(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Parameter: {'data_size': 1.0}\n",
      "Cumulative CTR: 5.0219%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAGACAYAAAADLH61AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XlcVPX+x/H3wAw7uIJb4o5l5kU0y33LtMW8WilS3MxuP7W8mZVhmmtkWmqLW1Ze7Voq3vRqZYuZW2pm6lUj9yXcBRWFGRQY5vz+4DpJKmIyHoTX8/Hw8eAsc873w8jw5vs953wthmEYAgAAgGm8zG4AAABASUcgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQyAxxmGoQ4dOqhu3bpKSEgwuzl/yuDBg1W3bl33v/r166t169Z65513zG6aJOmZZ55R3bp1zW4GgD+JQAbA41avXq2DBw9Kkj755BOTW3N9pkyZopkzZ2rSpEkKDw/XtGnTtH79erObBeAmRyAD4HFz5syRv7+/HnvsMe3evVs//vijJOno0aOqV6+ehg4d6t73iSee0N13362srCxt3rxZ3bp1U8OGDfXXv/5Va9eulSQtXLhQdevW1UsvvaRGjRpp8uTJ2rJli7p166Y77rhDjRs31gsvvKDz589LklatWqVOnTqpYcOGio+PV/v27RUbGytJOnHihPr166eoqCi1a9dOs2bNyreWyMhINWrUSFFRUQoPD5ck5eTkSJK+//573X///WrYsKF69uypxMREpaWlqX79+ho7dqwkadKkSapbt64WLVokSerbt69atGghl8uladOmqUWLFqpfv77atm2rf//733leM3jwYDVs2FALFy7Uvn37FBMTo6ioKPXt21dnzpwpjLcKgEkIZAA86tChQ1q9erU6duyoXr16yWKxaPbs2ZKkypUrq23btvr666+VkZGhw4cP66efftLDDz+sc+fOqV+/fgoNDdXo0aNVs2ZN9e/fX8nJye5j79+/X/Hx8ercubPmzJkjSZo4caI6d+6sJUuWaN26dTp9+rSef/55WSwWjR07VmlpaTp8+LD7GIMGDdKBAwc0fPhwdevWTW+88YZWrFhxxXqaN2+uBg0aqEmTJlq6dKn+/ve/q3nz5kpKStKAAQN0xx13aPTo0fL399ezzz4rHx8f3X333Vq5cqUkuUPl+vXrdf78ef3444/q2LGjjhw5os8//1xt27bVpEmT5O/vr/Hjx+c5d0pKiuLj49WmTRu99NJL2rFjh1555RXddddd2rx5c6G8XwDMYTW7AQCKt7lz58rlcqlDhw4KCgpSZGSkVqxYoUOHDqlq1ap67LHHtGzZMn3zzTfuoBQdHa0tW7bozJkzWrlypTvMSNLGjRvdX0dHR+u+++6TJL322mtasWKFNm/erG3btkmSzp49q61btyojI0N///vf1bFjR7Vp00ZffvmlJCkjI0MbNmyQYRiKi4tzH3fVqlVq27btZeuZMWOG/vvf/2ry5Mlq1qyZBgwYIElas2aNsrOztWjRInfvlyTt3r1bHTt21KuvvqrExERt27ZNt99+u3788UetW7dO58+fV6dOnVS1alV99NFHWrlypZYuXaozZ85c0uvVu3dvNW/eXHa7Xdu3b1e3bt306KOPSpK+++47bdq06U+9RwDMRyAD4DGZmZlasGCBJOnZZ5/Ns23OnDmKi4tTs2bNVLNmTS1atEiHDx9Wy5YtVbVqVe3atUtS7sXqrVu31rlz52S1WlWjRg2tXr1aklS6dGn38WJjY5Wenq7+/fvr9ttv14svvijDMNzDiZebtjcnJ0eGYahNmzbq16+fcnJydP78eVWpUuWKNTVo0EAtWrRQSkqKEhISVKFCBQ0ZMkROp1OSNGLECNWrV08ZGRny9fVV9erVVbVqVY0cOVLjxo2TJL300kt68sknNXPmTIWGhqpRo0ZKTEzU448/rvvuu08PPfSQsrKy3MHxgovrleQ+pyRZLJZ83gkARR1DlgA8ZsmSJTpz5ox69eqlmTNnuv+FhYXps88+U0ZGhiQpJiZGP/30k44cOaKePXtKkqKiohQQEKBly5bpxIkT+uijj9S7d29lZ2e7j+/llfsRlpaWpq1bt8pqtcpisbh7qHJychQVFSV/f3/NmDFDS5cu1auvvuoOacHBwYqMjNTGjRu1f/9+LV26VL1799b27duvWltcXJwqVaqkTz75RDt27FDz5s3l5eWlr7/+WidOnND48ePVv39/Wa1WlSlTRk2aNNGGDRt0xx13qGnTpipXrpw2bNigjh07ysvLSxs3btS5c+fk5+eno0ePuoc2L7T14nqDgoLUsGFDLVu2TAkJCZoxYwa9Y8BNjkAGwGM+/fRT+fj4qG/fvmrWrJn7X0xMjNLS0rR48WJJUteuXRUQEKDKlSurdevWkqSyZctqxowZ8vPzU1xcnA4ePKhx48apUqVKl5wnJCREzz77rI4fP67hw4erdOnSstls2r17t8qWLav33ntPLpdLI0aMUPXq1WWz2WSz2SRJ7733npo0aaKxY8dq8eLF6tu3r3sYND+BgYEaNWqUcnJyNGrUKNWqVUuTJk3SmTNn9PLLLys7O1sTJ05UQECAJKlTp06SpGbNmslisahZs2Z51j/00EO68847tWDBAk2dOlWRkZGScoc8L+ett97SHXfcoXHjxmndunXu4wC4OVmMy/XjA0AxcebMGQ0aNEiVKlVS+/btlZSUpNdff129evXSK6+8YnbzAEASgQxACfDhhx9q7ty5Sk5Olo+Pj1q0aKHRo0dfck0WAJiFQAYAAGAyj11D5nK5NHz4cPXo0UOxsbFKSkq6ZJ/Tp0/r3nvvVWZmpiTp/Pnz+sc//qGYmBg9/fTTOn36tKeaBwAAUGR4LJAtW7ZMWVlZSkhI0Isvvuh+SvUFP/zwg3r37q2TJ0+6182dO1cRERGaM2eO/vrXv2rq1Kmeah4AAECR4bFAtmnTJrVs2VJS7lQjiYmJeU/s5aWZM2fmuYbj4te0atXKPb0KAABAceaxB8Pa7XYFBQW5l729veV0OmW15p6yefPml31NcHCwpNxbytPT0696HqczR1ardyG1GgAA4MbzWCALCgqSw+FwL7tcLncYK8hrHA6HQkJCrnqe1NSM62toAYSGBisl5erhsLiifuovqfWX5Nol6qd+6vdE/aGhwZdd77Ehy6ioKPf0Jlu2bFFERESBXrNq1SpJ0urVq9WoUSNPNQ8AAKDI8FgPWYcOHbR27VpFR0fLMAyNGTNGM2fOVHh4uNq3b3/Z1/Ts2VNxcXHq2bOnbDabJkyY4KnmAQAAFBkeC2ReXl4aPXp0nnW1atW6ZL/ly5e7v/b399d7773nqSYBAAAUScxlCQAAYDICGQAAgMkIZAAAACYjkAEAAJjMYxf1AwAAafPmjVq8eIFGjXojz/oRI17Rq6+Ols1m88h5U1NTNWxYnCRp797dqlo1XL6+furU6X5t3bpF7dvfq7vvbuaRcxc2l8ulKVPe0b59e+Xl5SWr1aYBA15UlSq3FOp5vvrqCyUl/abu3XtqypQJevbZFwv1+PkhkAEAYII/BrTCVqZMGU2e/IEkqX///9OgQUNUrVp1SdLWrVs8eu7C9tNP63TyZIreeSd3juvVq1dq0qSJGjt2okfOV65ceY0cOfKGPhiXQAYAgAkeeaSzPv30M40f/4ZsNpuOHz+mU6dOasiQkapb91YtX75MCQmfysvLSw0aRKpfv38oOfmExo8fq6ysTKWlnVWvXk+rVas2io3trqpVq8lms2nUqDEFOv/ixQs1Z86/ZLfb9dJLg1WvXn3NnfuJvv9+qby9vdW06V3q1auvZsyYriNHDuvMmTNKTz+rrl0f1cqVy3XoUJKGDh2lcuXKadiwwSpXrpxSUpJ1113N1KfPs1q1ark++eRjWa1WVapUWa++OkoOh0OvvTZMDodDOTk5evrpfmrU6E498US0IiOjtG/fXknS2LET80y/GBZWUTt37tD33y9Vo0ZN1LJlazVtmjsF44oVy7Rw4b9lGIYkKT7+Te3fv1effDJLNptNyckn1KXLw9q8eaP27t2tRx/tqa5dH9Hjjz+qBg0ideDAfoWEhGjkyN+/b8eOHdWzzw7TlCkzLtu2wMBATZgwTrt2bVfZsuV07NhRjRv3tipVqvyn/z8QyAAAJUqrVknauTOr0I53660+Wr262nUdo2LFSnr55aH6/PP/6PPPF6pPn2f1z39O10cfzZafn59ee22Yfv55vSSLoqMfU1RUY/3yy1bNmDFdrVq10blz59Sr11OKiLi1wOesW/dW9er1d3311Rf66qsv5evrp+XLv9P77/9T3t7eGj16iNau/UGS5Ovrq4kTJ2n27Fn68ce1evPNt7Vkyef6/vul6t69p44fP6qJEycpMDBIzzzzd+3atVPfffetevSI0T33dNTXX38ph8Ohjz+eocaN71L37j2VkpKsZ575uxISFsnhcOieezpq4MCXNWrUq1q/fq3uuaeju621atVWXNyrWrx4od55Z7xCQ8P0j38MVMOGjXTo0EG99da78vPz05tvvq4NG35U+fKhSk5O1qxZc7Rz5w4NHz5YCQmLlJKSrCFDBqlr10d0/vx53XvvfYqMjNLUqe9q8eIFCgkpdcn36XJt8/X1VVraWX344b+Umpqqnj27Xtf7LxHIAAAlTEHD042cy7FOnbqSpLCwCvrll606fPiQzpxJ1UsvPSdJysjI0JEjR9SgQaQ+/niGlixZLMkip9PpPkZ4ePVrOmfdurdJksqWLafMzPNKSvpNt99+h3ve6caNG+vAgX2S5A56wcFBql69xv++DlFWVqYkqVatCHeYqVevvg4e/E3/+MdAzZ49S4sWLVC1atXVqlUbJSUd0L33dpIkhYaGKSAgUGfOpP7vHL9/D7Ky8gbmvXv3KDy8mkaNGiPDMPTzzz9p+PBX9Pnn36pMmbKKjx+hgIAAJSX9pvr1G0iSatasJavVquDgYFWuXEU2my1Pm61WqyIjoyRJ9ev/RevXr3W/9o/+2LZjx46pfv07JOUODV/r9/5yuMsSAACTWSyWPMuVKlVRWFgFvfPOVE2e/IEeeaSHbr+9vj766H116vSAhg17TVFRjfM9xrWes1q16tq+PVFOp/N/oednVa1a7X/75n+spKQDOn/+vHJycrR9e6KqV6+pzz//j5566v80efIHMgxDq1evVLVqNdzXr6WkJCs9Pe2iXqkrn2Tjxp/0/vuTlZOTI4vFoho1asrPz18Oh0MzZkzXqFFjFBf3qnx9fd1Dl1drs9Pp1J49uyVJv/yyVTVqXDqb0O/yHqxmzVpKTPxFkpSWlqZDhw7mf7ICoIcMAAAP27DhJz31VKx7ecSI+Hz3L1OmjHr0eEz9+/+fcnJyVKlSZbVr10Ft27bXu++O1+zZMxUWVkFnzpwptDbWqlVb7drdo379npJhGLrrrjvVqlUb7d27+6qvtdlsGjYsTqdPn1abNu1Vp06ETpw4rueff1alSpVSQECAmjVroebNW+qNN0Zr5crvlZmZqZdfHurukcvPI49Ea8qUd9W792MKCAiUl5eXhg0bpcDAQN1xx1/Uu/fj8vf3V3BwsE6eTCnwtVyffvqxTpw4rgoVKurpp/tp2bJvC/S6Zs1aaP36derbt7fKli0nPz+/AtWRH4txIUrepG5Ed/KN7LYuiqif+ktq/SW5don6qb9g9R87dlQjRgzRBx/M8nyjCtGFmyp8fX0vuz2/+pOSftOePbt0zz0ddfbsGcXG9tBnn30hHx+fq543NDT4suvpIQMAALgGYWEVNG3ae5o/f65cLpf69ftHgcJYfghkAADgT6tUqfJN1zsmSZ999sWffq2/v3+hPwONi/oBAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGRWTx3Y5XJp5MiR2rVrl3x8fBQfH69q1aq5t8+fP1/z5s2T1WpVv3791LZtWx09elQvv/yyDMNQqVKlNGHCBPn7+3uqiQAAAEWCx3rIli1bpqysLCUkJOjFF1/U2LFj3dtSUlI0e/ZszZs3TzNmzNDEiROVlZWlWbNm6b777tOnn36qOnXq6LPPPvNU8wAAAIoMjwWyTZs2qWXLlpKkyMhIJSYmurdt27ZNDRs2lI+Pj4KDgxUeHq6dO3fqtttuU1pamiTJbrfLavVYBx4AAECR4bHEY7fbFRQU5F729vaW0+mU1WqV3W5XcHCwe1tgYKDsdrsqVqyoCRMm6Msvv1RWVpb69+9/1fOUKRMgq9XbIzVcLDQ0+Oo7FWPUT/0lVUmuXaJ+6qf+G8VjgSwoKEgOh8O97HK53D1ef9zmcDgUHBys4cOH64033lDLli21cuVKxcXF6YMPPsj3PKmpGZ4p4CKhocFKSUn3+HmKKuqn/pJaf0muXaJ+6qd+T9R/pZDnsSHLqKgorV69WpK0ZcsWRUREuLc1aNBAmzZtUmZmptLT07Vv3z5FREQoJCTE3XMWFhbmHr4EAAAozjzWQ9ahQwetXbtW0dHRMgxDY8aM0cyZMxUeHq727dsrNjZWMTExMgxDAwcOlK+vr4YNG6bRo0fL5XLJMAwNHz7cU80DAAAoMiyGYRhmN+J63IjuVLptqZ/6S2b9Jbl2ifqpn/qLxZAlAAAACoZABgAAYDICGQAAgMkIZAAAACYjkAEAAJiMQAYAAGAyAhkAAIDJCGQAAAAmI5ABAACYjEAGAABgMgIZAACAyQhkAAAAJiOQAQAAmIxABgAAYDICGQAAgMkIZAAAACazmt2Aou7YMacMI1sWi9ktAQAAxRU9ZFfRsOEBNW++2+xmAACAYoxAdhUul3T6tNPsZgAAgGKMQFYAhmF2CwAAQHFGIAMAADAZgawA6CEDAACeRCADAAAwGYGsAOghAwAAnkQgAwAAMBmBrAAMusgAAIAHEcgAAABMRiArADrIAACAJxHIAAAATEYgKwB6yAAAgCcRyAAAAExGICsAesgAAIAnEcgAAABMRiArAHrIAACAJxHIAAAATEYgKwCe1A8AADyJQAYAAGAyAlkB0EEGAAA8iUAGAABgMgJZAdBDBgAAPIlABgAAYDICGQAAgMkIZAXAkCUAAPAkAhkAAIDJCGQFQA8ZAADwJAIZAACAyQhkBcDUSQAAwJMIZAAAACYjkBUAHWQAAMCTCGQAAAAmI5AVAD1kAADAkwhkAAAAJiOQFQA9ZAAAwJMIZAAAACYjkBUAPWQAAMCTCGQAAAAmI5ABAACYzGOBzOVyafjw4erRo4diY2OVlJSUZ/v8+fPVrVs3de/eXStWrJAkZWRk6OWXX1ZMTIweffRRbdu2zVPNAwAAKDKsnjrwsmXLlJWVpYSEBG3ZskVjx47VtGnTJEkpKSmaPXu2FixYoMzMTMXExKh58+aaMWOG6tSpozfffFM7d+7Uzp071aBBA081EQAAoEjwWA/Zpk2b1LJlS0lSZGSkEhMT3du2bdumhg0bysfHR8HBwQoPD9fOnTu1Zs0a2Ww2PfXUU5o6dar79QAAAMWZx3rI7Ha7goKC3Mve3t5yOp2yWq2y2+0KDg52bwsMDJTdbldqaqrS0tI0Y8YMLVq0SOPGjdObb76Z73nKlAmQ1ertqTLcQkODr75TMUb91F9SleTaJeqnfuq/UTwWyIKCguRwONzLLpdLVqv1stscDoeCg4NVunRptWvXTpLUtm1bffDBB1c9T2pqRiG3PC+LJfexFykp6R49T1EWGhpM/dRvdjNMUZJrl6if+qnfE/VfKeR5bMgyKipKq1evliRt2bJFERER7m0NGjTQpk2blJmZqfT0dO3bt08RERFq1KiRVq1aJUn6+eefVbt2bU81DwAAoMjwWA9Zhw4dtHbtWkVHR8swDI0ZM0YzZ85UeHi42rdvr9jYWMXExMgwDA0cOFC+vr7q06ePXn31VfXo0UNWq1Xjxo3zVPMK7EIPGQAAgKdYDOPmjhue7k6tWHGPXC4pObmOR89TlNFtTf0ltf6SXLtE/dRP/cViyLK4sFjMbgEAACjuCGQAAAAmI5BdBT1kAADA0whkAAAAJiOQXQU9ZAAAwNMIZAAAACYjkF2FhS4yAADgYQQyAAAAkxHIroIOMgAA4GkEMgAAAJMRyK6CHjIAAOBpBDIAAACTEcgAAABMRiArZK+/flKffHLW7GYAAICbiNXsBhR113INWVjYHvfXL7yQrBMnavMcM5jOMAxNmpSq77/P0I8/nnOv79QpUL6+FrVoEaDAQIuysgytXp2hhQvtCg+36t//rqLq1W0mthwomNTUHDkcLgUGeslikWw2ixwOl4KCvOTvb+FzGDcFAlkhOXfOdcm6t99O1QsvlDWhNSjJDhzI0hNPHNPOnVn57vfNNw5J0uLF9ku2HTzo1F13JbmXW7b0l81mUXi4Td7e0uzZacrKMi573Ftv9ZGvr0Vbt2aqc+cglS3rrWrVbEpNzVGbNgGqXdtHlSrx0YPLMwxDdrtLKSk5WrfunBYtSpePj0U1a/rogw/OXLJ/UJCX7PZLP3+v1dixoapWzaaaNW3/+39OiMONxafiVRT0D6sPP8z9oJg7t7Latw9UWNgejR17ikCGG8Zud6lmzX2XrI+PL6+nny79p3oJDh/O1okT3tq3z679+7Plchn64gt7njBWu7ZNe/dm6y9/8dXWrZm6804//fe/mZKkL77IG/YmTUq97HnuvNNPP/98Xl27BqlRIz89+WRp2Wzm/UI0DEMJCelKTExVZma2Pv74z1+GULmyVUePOi+7rWJFb0VG+rnDsSQ9/HCw9u/Pks1mUYMGvtqzJ1t2u0snTjiVnJzj/t43beqfp8czIMCijIzcbS1a+Cs2tpSaNfNXWJi36T1ETqehpKRsbd+eqaAgLx0/7tTSpQ4dOJCtU6dydOJEziWvqVrVqkOHfv++DR8eoB49gmWxSNHRIbLZLPL399Itt1hVqpTXFWvMysoNeGlpLu3enanffsuWw2Fox45Mbdx4XoYh5eQY+vXXTH35pV3btmUqLc2lihW91aJFsJo181GHDoGqUIFfl/Asi2EYl/8z9yaRkpLu0eNXr75XGRmGkpPr5LvfheHKC/sNHHhCn36apv/+t7qqVLm5h31CQ4M9/n0uyq5U/4WhwPj4U1d87YoV4br9dl9PNk9Op6HKlfe6l4cOLacBAwrvD4HCfv9dLkNHjjh19qxLixalq3ZtH9lsUr9+J/LsV6eOTXff7a/y5b01eHA5j4WK06dztHHjOT3++LHLbu/Vq5RmzcoNZOXKeeveewN1+HC2+vYtozVrMlS1qk3Z2YbWrz+njh0D5XAYuu02HzkcLu3alSU/P4uSk3MUEOCls2dzlJKSo88+S5efn0U9eoTIYpFmzTqrRo38VK+ej8qV89bnn+cG4LAwbyUnXxpWJOnll8sqOTlHs2ad1d/+FqJDh5xasSLjinX26lVKx4451aSJnzp3Dla5cl4KCrpykJGu/t4bhqFDh5xavDhdBw86tWSJXSdPXr69F9x+u4/Kl7dq1aoMtW8foLAwqw4ezNZjj4WofHlv/eUvfipTxjvfY3ia3e5SYmKmFi06px07MrRx4znZbLmBd9CgsjpzxiV/f4vuuy9It9xiLRKh1xP47PdM/aGhwZddTyC7ioIEssWL0/X008fVqJGfvv66qnv9H0PazYofytz6f/75nB544PA1vz4kxEuffVZFkZF+hdqu9PQc9ep1TD/8kNtLUr68t7Zvr1mo55DMef9TUpxatMiuJUvsWrcut76xY0MVG1vqunrODMPQV1859OSTlw9fkvTAA4F67bVQ3XKLrVj8309KytYPP2Ro794sTZ166ZDfxfz8LDp//vp+JQQEWPTMM2Xk62vR3Xf7q0YNm8qW9ZbVevMFlgvvv2EY2rEjS//611kFBXnpvfcu39MrSfffH6hvvnHo5ZfL6dZbfZSZaahaNZsqVbKqQgVveXndPN+H4vD//3rcNIHsyy+/1IMPPnhdjSoMRSGQXSl4XXyR//HjtW+qH8SLlbQfyuRkp5577oSWL79yb4MkzZhRUZ07X/4H64Ldu7PUokXutVjly3tr2bKqqlz5z/WY/vJLptq3P3iF89RU6dKe6VUw+/03DEOLFtn1+usndfCgU507B+mBB4LUoUOAgoMvX/PZszmaOzdNw4efzPfY8+ZVVqtWAVcMC2bX7mlOp6HU1Bzt3ZstPz+LMjJc+uYbh6pUsSoxMVOhoX6SnJoyJVUtWvjLy0sKCfFWSIiX6tb1UWxsKQUFFd+b9Qv6/mdkuLRvX7ZWrcpQqVJemj37rP7yFz/9+mvusOjFKle2yjCkY8ecuueeANWq5SNJWrv2nGJjQ9S4sb/KlPFSlSpW03vdivv//6spUoFs2bJlGjFihEqXLq2pU6eqWrVq2rp1q+Lj43XkyBGtW7eu0Bt6rTz9n6VGjb1yOP5cILt42+ef36K77/b3TCM9rDj/UJ4+naNbb91/1f06dw5S//5l1LDhn+vlmjv3rAYMSJYkVajgrdOnc9S/fxkNGFBWPj4WWa0W5eQYOnvWpZQUp1q1Oqj8/lSqWdOmFSvC5e/v+V+GRen9//Zbu8aOPaVff829YcHLS3IV8HruqChfTZ1aUTVr+hT4fEWpdjNQf+HW73C4dPy4U7/8kqm9e7Pk7++lzExDc+ac1cGDudfLVa9u02+/ZUv6/brAsmW91KZNoI4ezVZQkJdatw5Q9eo2nT9vqHFjP/n75w4/+/gUboCEuZLqAAAf10lEQVTj/S9Cgaxjx44aNGiQjh49qu3bt6t69eqaPn26Hn/8cfXp00dBQUGF3tBrVVQC2dixoerdu/QVt0vSsWO1b8o7d4rTD6VhGPryS7ueeur4Zbe/8EJZPflkqTwX8BZm/fv2ZenZZ49r8+bMa3pdXFxZPf10aYWE3Phra4rq++90Gtq1K0vz56dp+fIM3Xabjx5+OFinTuWoU6fcuzuvV1Gt/UahfvPqdzoNnTqVo717s3T+vKHkZKf+/e90rVlzThZLbo97Ssql1+uFhHjJ19fi3lazpk3792erTBkv3X67rxwOl06ezJG3t0W1atlUq5aP/PwsKl3aW3Xq2FSqlLd8fS2y212qUCFQvr5ZCgnxUnCw10057Hw9bnQgy/e2ER8fH91zzz2SpBYtWujw4cP64osvdMsttxR6A4uq3C7jK3dV9O6dey2K03nlfYYMKacxY07pnnsOasWKaoXdRBTA5MmpGj360uGrX36pcUPvnqpVy0fffBPuXna5DLlcucOkoaFWWa0yfZjiZmG1WnT77b4aNSpUo0aZ3RqgcFmtFlWoYM3z+dSzZ6kr7u90GnI6DTkchlJSnNq1K0u//pqpChWsWrv2nNq2DVB4uE3+/rl3p1os0qFD2dq3L1tLlth17pxLlStbdeaMS0lJ2XnCns0mZWf/fq7mzf0VEeGjJk381LChn6pXt920l+QUJfn+JvL2/v0vTD8/P02fPl2BgYEeb9TN5Msvc2/r79Tpyr2Fzz9fVmPG5A6z/PzzOd155805dHmzSUzMVLt2ea+58vOz6MCBWkWmp9LLyyIvL/3p68oAQMoNcFarRX5+uXcD33qrr7p0ye2Jeeqpy4/e1K+fewd4//5lLrv94h6iC3dHJyfn6LffsnTgQLaWLHEoPv6UjhzJ+1iX8HCrDh50ystL6tYtWH5+FiUmZmrLlkzFxIT87y5kQ99/79Bdd/mrYUNfnT9vKCTES4GBudfPVahgVUBA8b0+8XLyDWQX/6UeHBxcIsNYfp0VF4/2hofn/wt148bqatz4Nz3wwOGb/q7Lom7v3iw1a5b0h3U1TRnuA4DiwMvLoqpVbapa1aZGjfJeS5uVZej4cad+/TVTTqehwEAv/fDDOZUr563y5b117pxLFou0ZUum9u3LUkCARQcPOvXzz+f166+5l29ceIZefm65xarDh50qXdpL998fpP37s5SW5tJtt/mqS5cghYR4qU4dH5Uvf3M+hiTfQHb06FG98sorl3x9wRtvvOG5lt0Evv4692GOx47Vvuq+4eE2vf9+RfXte1zR0Uc0b14VTzevRDl/3qXBg1M0Z06ae13PniF6990KJrYKAIo/H5/cWTwu7pho1+7SDpwJEwr+eWwYuTc5pabm6PRpl06fztGJE07NnHlWwcG5z/Tbti1TGRmGtm/P0ldf2XXunKEyZbyUmupS6dJeOnPm9zt+Hn44WGXL5j5HsFIlq6pWtd6Qm6KuRb6BbPDgwe6vmzRp4vHGFEX5hexevXKvHyvo8Fe3bsHq2/e4li/PUFjYHnrKrpPD4VKNGpc+mf6++wL18ceVTWgRAKAwWCy5NxqULu2tGjV+X//441e+ju4Cu92l5GSntm7N1IYN5zRr1lkdP+7UggXp2rz5vHbvzlJ6ukulSuUOj1aubNW5c4buvNNPQ4aU92BVV2HkY+HChfltLhJeemmwodyr7g1JxnffrTK++25VnnUvvTTYSE5OMypUqOhe16BBpJGcnGbExvbKs++2bbuM2bMT8qyThhrJyWl51t17bydD2mRILfOsT05OM8aPfzfPutmzE4xt23b94ZhdjeTkNKNBg0j3ugoVKhrJyWk3pKbx49+9bE3JyWn/q+339YZhFKim2NheN6wm6RtDejvPurfeeqfANRX0fYqN7WUYhnFTvE/XUtO1vE8bN24sdjUV9H0yDKPY1XQt75NhGMWupmt5n6ZPn17sarqW9ykqKqrY1XTx+xQaWtFYtizZ6Nr1hRta05Xk+9iLrl276j//+c+VNhcJnr4luU6dfTp71pXvM8autafr0KFsNWr0m559toxGjDAxjRdQUbn1/bvvHHrssaPu5U6dAvWvf3m+J6yo1G+Wklx/Sa5don7qp/4i89gLXFlmZu7Y9OzZla75tVWr5o6zT5mSqn/8o0yhPC+pODMMQxUq/D5X4xdf3KK77uJOVQBA8ZFvINuzZ4/at29/yXrDMGSxWPT99997rGFFxZWuITt+PPcZLR06/Lk7T3/5pYbuuOOAYmOPasmSqld/QQk1dGiKPvwwd/69fv1Ka9SoUJNbBABA4cs3kFWrVk0ffPDBjWrLTSUzM3ek988+DK9CBauio4M1b17J7Q7OT2amS1Wr/n7B/sGDteTnV7TuiAEAoLDkG8hsNpuqVCnZj2e4Ug/Zyy8nX/exhw8vr3nz0nX0aDYPBr2I3e5SzZq5Yez++wM1axZ3TAIAird8uxyioqJ09uxZnT592r1uw4YNeZaLuysFsoiIgk9QfCXly+fm4cjI32S3F3CG5BLgQhg7dKgWYQwAUCLkG8geeeQRPfDAA0pMTHSvW7t2rbp06aKdO3d6vHFFSU5O3ptRZ806WyjH7dYt926LCyGkpFu1KkOStHBhFfn6MkQJACgZ8v2NN27cOE2YMEGtWrVyrxs4cKDGjBmjsWPHerxxRcGF6Rf+8x/PXOv1/vsVNXp07qMvwsL2KJ+nkBR7x4459eijRyRJLVoEmNwaAABunHwDWVpamu66665L1rds2VKpqakea1RRdPas54YU+/Yto1tvzR0CvfjxDp7kchWd4JeamqOwsD36y18OSJIOH776VFQAABQn+QYyp9Mpl+vSIOJyuZSdne2xRhUlF64hGz/+tPbsydIjjxxWcrJTFSt66/vvwwvtPKtXV3N/3apV0nUHJsMwlJiYqbCwPZf9V7Hi3kvWmdE7Fxa2R3Xr7pckPfFEKZ04UVs+PjffpLAAAFyPfO+yvPPOOzV58mQ999xzedZPnTpV9evX92jDippTp3LUvHmSJKl+/dyenMKeTD45uY7CwvZo584sVay4V+vXV1PNmgW7eWDz5vPq1OlQgc/1/vsV5XQa6t//hHvdhd65vXtrKiTEsw+rnT8/Lc+5mdcTAFCS5RvIXnjhBf3f//2fFi1apFtvvVW+vr7avn27ypYtq2nTpt2oNprq5MncB8BaLNIfO5BKlSr8i86Tk+tozpyzev75ZN19d9Jlg4phGBox4qTef//MVY934kRt93Vwl9O9e4gkKSvL0Ntvn9aECadVu/Z+HT1aW1Zr4fdUtWyZpF27stzLPHUfAAAp37kspdxf/uvXr9eOHTvk5eWl+vXrq3HjxjeqfVfl6Xm2LsxXWbq0l86cyTt8e+xYbXl7e2Z4bffuLLVokdsj16ZNgObPr6ITJ5x66KHDOnDg0uHirVtrqGLF3F6tc+cMBQT8ubCYnW2oSpXcnrILYbAw5vNauzZDXbsecS+/+GJZxcWVu65j3ijM51Zy6y/JtUvUT/3UX6TmsrRYLGratKmaNm1a6I26mfwxjEnyWBiTcp9z1qdPaU2ffkYrV2a4g+EFK1aE6/bbfS/72oCAP98um83intYpLGzPdQ8lHjyYrcaNf3MvX8swLAAAJQUPeirCXnstVMnJddSp0+/zZSYl1dKJE7WvGMYKQ4UKVs2bl/tA1ocfPnzNr3e5DD3xxFGFhe1xh7HgYC8lJ9chjAEAcBlX7SGD+f71rxv/tPp27QK1ZMkteuCBw0pIOK127Qo2tdMfe/IefzxEEydW8EQTAQAoNghk12jBgipq2bJkPLT0zjv99cYboYqO/k0bNlRX9epXDmUDB57Qp5+muZcPHKilwEA6YAEAKAgC2TUqKWHsgqeeKq0NG7LUpMlvevfdMPXsWeqSfS7uFVuzplqhzPMJAEBJQhfGNWjdumSFsQsWLKgpSRowINn9ENkpU1LdX0vSkCHllJxchzAGAMCfQA/ZNZg+vaLZTTCFxWJRcnIdbd+eqTZtDkqSRo066d6+fHm46tf33E0GAAAUdwSya1C2rGefXl/U1avn634MhstlyMuLKY4AACgMDFkW0Kef3vg7HYsywhgAAIWHQFZAHToEXn0nAACAP4FABgAAYDICGQAAgMkIZAAAACYjkAEAAJiMQAYAAGAyAhkAAIDJPBbIXC6Xhg8frh49eig2NlZJSUl5ts+fP1/dunVT9+7dtWLFijzbfv75Z7Vu3dpTTQMAAChSPPak/mXLlikrK0sJCQnasmWLxo4dq2nTpkmSUlJSNHv2bC1YsECZmZmKiYlR8+bN5ePjo2PHjumf//ynnE6np5oGAABQpHish2zTpk1q2bKlJCkyMlKJiYnubdu2bVPDhg3l4+Oj4OBghYeHa+fOncrMzNSIESM0cuRITzULAACgyPFYD5ndbldQUJB72dvbW06nU1arVXa7XcHBwe5tgYGBstvtGj16tHr37q0KFSoU+DxlygTIavX8HJOhocFX36kYo37qL6lKcu0S9VM/9d8oHgtkQUFBcjgc7mWXyyWr1XrZbQ6HQzabTRs3btTBgwc1ZcoUnT17VgMHDtTbb7+d73lSUzM8U8AfpKSk35DzFEWhocHUT/1mN8MUJbl2ifqpn/o9Uf+VQp7HAllUVJRWrFih+++/X1u2bFFERIR7W4MGDfTOO+8oMzNTWVlZ2rdvnxo0aKBvv/3WvU/z5s2vGsYAAACKA48Fsg4dOmjt2rWKjo6WYRgaM2aMZs6cqfDwcLVv316xsbGKiYmRYRgaOHCgfH19PdUUAACAIs1iGIZhdiOuh6e7U8PC9kiSkpPrePQ8RRnd1tRfUusvybVL1E/91H8jhyx5MCwAAIDJCGQAAAAmI5ABAACYzGMX9RcXTzxRShaL559zBgAASi4C2VUEBlpUvbq/2c0AAADFGEOWAAAAJiOQXcXN/VAQAABwMyCQFYDFYnYLAABAcUYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGRWTx3Y5XJp5MiR2rVrl3x8fBQfH69q1aq5t8+fP1/z5s2T1WpVv3791LZtWx09elRDhgxRTk6ODMPQ6NGjVbNmTU81EQAAoEjwWA/ZsmXLlJWVpYSEBL344osaO3ase1tKSopmz56tefPmacaMGZo4caKysrL07rvv6vHHH9fs2bPVp08fTZw40VPNAwAAKDI81kO2adMmtWzZUpIUGRmpxMRE97Zt27apYcOG8vHxkY+Pj8LDw7Vz507FxcUpODhYkpSTkyNfX19PNQ8AAKDI8Fggs9vtCgoKci97e3vL6XTKarXKbre7g5ckBQYGym63q2zZspKk/fv3a9y4cZoyZcpVz1OmTICsVu/CL+B/AgLOSpJCQ4OvsmfxRv3UX1KV5Nol6qd+6r9RPBbIgoKC5HA43Msul0tWq/Wy2xwOhzugrV+/XqNGjdKbb75ZoOvHUlMzCrnleWVkZEmyKSUl3aPnKcpCQ4Opn/rNboYpSnLtEvVTP/V7ov4rhTyPXUMWFRWl1atXS5K2bNmiiIgI97YGDRpo06ZNyszMVHp6uvbt26eIiAitX79er7/+uj766CPdcccdnmoaAABAkeKxHrIOHTpo7dq1io6OlmEYGjNmjGbOnKnw8HC1b99esbGxiomJkWEYGjhwoHx9fTVmzBhlZ2dr8ODBkqQaNWpo9OjRnmoiAABAkeCxQObl5XVJmKpVq5b76+7du6t79+55tn/++eeeag4AAECRxYNhr8IwzG4BAAAo7ghkBWCxmN0CAABQnBHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAAAAkxHIAAAATEYgAwAAMBmBDAAAwGQEMgAAAJMRyAAAAExGIAMAADAZgQwAAMBkVk8d2OVyaeTIkdq1a5d8fHwUHx+vatWqubfPnz9f8+bNk9VqVb9+/dS2bVudPn1aL730ks6fP6+wsDC98cYb8vf391QTAQAAigSP9ZAtW7ZMWVlZSkhI0IsvvqixY8e6t6WkpGj27NmaN2+eZsyYoYkTJyorK0tTp07Vgw8+qDlz5qhevXpKSEjwVPMAAACKDI8Fsk2bNqlly5aSpMjISCUmJrq3bdu2TQ0bNpSPj4+Cg4MVHh6unTt35nlNq1attG7dOk81r8DOnzfkxcAuAADwII8NWdrtdgUFBbmXvb295XQ6ZbVaZbfbFRwc7N4WGBgou92eZ31gYKDS09Ovep7Q0OCr7nM9Zs707PFvFp7+Phd11F9y6y/JtUvUT/3Uf6N4rO8nKChIDofDvexyuWS1Wi+7zeFwKDg4OM96h8OhkJAQTzUPAACgyPBYIIuKitLq1aslSVu2bFFERIR7W4MGDbRp0yZlZmYqPT1d+/btU0REhKKiorRq1SpJ0urVq9WoUSNPNQ8AAKDIsBiGYXjiwBfusty9e7cMw9CYMWO0evVqhYeHq3379po/f74SEhJkGIb69Omjjh076uTJk4qLi5PD4VCZMmU0YcIEBQQEeKJ5AAAARYbHAhkAAAAKhvsHAQAATEYgAwAAMBmBLB8ul0vDhw9Xjx49FBsbq6SkJLOb9KdkZ2dr0KBBiomJ0SOPPKLvv/9eSUlJ6tmzp2JiYjRixAi5XC5J0uTJk/XII48oOjpa27Ztk6RC2ddsp06dUuvWrbVv374SV/v06dPVo0cPdevWTf/+979LVP3Z2dl68cUXFR0drZiYmBL1/m/dulWxsbH5ts1TNV9u3xvt4vp37NihmJgYxcbG6qmnntLJkycl5c4Y061bN3Xv3l0rVqyQJJ0+fVq9e/dWTEyMnn/+eZ07d65Q9r3RLq7/gi+++EI9evRwL5eU+k+dOqV+/frpscceU3R0tA4ePFgoNRV6/Qau6NtvvzXi4uIMwzCM//73v0bfvn1NbtGf89lnnxnx8fGGYRjG6dOnjdatWxt9+vQx1q9fbxiGYQwbNsxYunSpkZiYaMTGxhoul8s4cuSI0a1bN8MwjOve12xZWVnGM888Y9x7773G3r17S1Tt69evN/r06WPk5OQYdrvdeO+990pU/d99953x3HPPGYZhGGvWrDH69+9fIur/4IMPjAcffNB49NFHr9g2T9V8pX1vpD/W/9hjjxnbt283DMMw5s6da4wZM8ZITk42HnzwQSMzM9NIS0tzf/3aa68ZCxYsMAzDMKZPn27MnDmzUPY1s37DMIzt27cbf/vb39zrSlL9cXFxxpIlSwzDMIwff/zRWLFiRZGsnx6yfOQ328DNpFOnThowYIB72dvbW7/++quaNGki6fdZETZt2qQWLVrIYrGocuXKysnJ0enTp697X7ONGzdO0dHRCgsLk6QSVfuaNWsUERGhZ599Vn379lWbNm1KVP01atRQTk6OXC6X7Ha7rFZriag/PDxckyZNci/fyJqvtK+Z9U+cOFG33XabJCknJ0e+vr7XNGNMYexrZv2pqakaP368hgwZ4l5XkurfvHmzTpw4oV69eumLL75QkyZNimT9BLJ8XGm2gZtNYGCggoKCZLfb9dxzz+n555+XYRiyWCzu7enp6ZfUe2H99e5rpoULF6ps2bLuHxpJJaZ2KfeDODExUe+++65GjRqll156qUTVHxAQoCNHjui+++7TsGHDFBsbWyLq79ixo/tB3NKN/T9/pX1vpD/Wf+GPsc2bN+uTTz5Rr169rmnGmMLY90a6uP6cnBwNHTpUQ4YMUWBgoHufklK/JB05ckQhISGaNWuWKlWqpA8//LBI1k8gy0d+sw3cbI4dO6a//e1v6tKlizp37iyviybovDArwpVmULjefc20YMECrVu3TrGxsdqxY4fi4uLy/LVenGuXpNKlS6tFixby8fFRzZo15evrm+eXY3Gvf9asWWrRooW+/fZbLV68WIMHD1Z2drZ7e3Gv/4Ib+fN+pX3N9tVXX2nEiBH64IMPVLZs2WuaMaYw9jXLr7/+qqSkJI0cOVIvvPCC9u7dq9dff73E1C/lfg62a9dOktSuXTslJiYWyfoJZPnIb7aBm8nJkyfVu3dvDRo0SI888ogkqV69evrpp58k5c6K0LhxY0VFRWnNmjVyuVw6evSoXC6XypYte937munTTz/VJ598otmzZ+u2227TuHHj1KpVqxJRuyQ1atRIP/zwgwzD0IkTJ3Tu3Dk1bdq0xNQfEhLi/jAsVaqUnE5nifm/f7EbWfOV9jXT4sWL3Z8DVatWlXRtM8YUxr5madCggZYsWaLZs2dr4sSJql27toYOHVpi6pdyPwcvtPPnn39W7dq1i2T9PBg2H5ebbaBWrVpmN+uaxcfH6+uvv1bNmjXd64YOHar4+HhlZ2erZs2aio+Pl7e3tyZNmqTVq1fL5XLplVdeUePGjXXgwAENGzbsuvYtCmJjYzVy5Eh5eXlddz03U+1vvvmmfvrpJxmGoYEDB+qWW24pMfU7HA4NGTJEKSkpys7O1t/+9jfVr1+/RNR/+PBhvfDCC5o/f36h1HG9+5pV/9y5c9W0aVNVqlTJ3Wt555136rnnnrumGWOud1+z6p8/f/4V15WU+o8cOaJXX31V586dU1BQkCZMmKBSpUoVufoJZAAAACZjyBIAAMBkBDIAAACTEcgAAABMRiADAAAwGYEMAADAZAQyAKY6fPiw6tatq7Vr1+ZZ365dOx0+fPiGt+fo0aPq2LGjunTpkucJ2wsXLlSTJk3UpUuXPP+2bt1aaOfetm2b3nrrrUI7HoCbx8352HkAxYrNZtOwYcP0+eef55l2xwwbNmxQ/fr1NWHChEu2tWvXTmPHjvXYuffu3atTp0557PgAii56yACYLiwsTM2aNdO4ceMu2fbTTz8pNjbWvTx48GAtXLhQhw8fVpcuXTRw4EB17txZcXFxmjdvnnr06KFOnTpp3759+Z7zwIEDio2NVefOndWjRw9t27ZNO3bs0DvvvKMffvhBw4cPL3D7+/fvr2+//da93K1bN23fvl1JSUl68skn1bVrV/Xs2VPbt2931xAfH6+ePXuqXbt2WrBggdLS0vTee+9p+fLlmjZtmnbu3Knu3burW7du6tmzp3777bcCtwfAzYdABqBIGDx4sNasWXPJ0GV+du3apaefflqLFy/W5s2bdeTIESUkJOjBBx9UQkJCvq8dNGiQYmNj9cUXX+iVV17RgAEDVKtWLT333HNq166dRo8efclrli9fnme48tFHH5UkdenSRUuWLJEk/fbbb8rMzFS9evUUFxenQYMG6T//+Y9ee+01DRw40H2s48ePa86cOZo2bZrefPNNhYSEuM/dr18/ffzxx3ryySe1cOFCde/eXVu2bCnw9wXAzYchSwBFQlBQkF577TX30GVBlC9fXvXq1ZMkVaxYUU2bNpUkVa5cOd/rzxwOhw4ePKh7771XkhQZGalSpUpp//79+Z7vSkOWrVu31ujRo2W32/Xll1/qoYceksPhUGJiol555RX3fhkZGUpNTZUkNW/eXBaLRRERETpz5swVj/nDDz+oXbt2atu27VW+GwBuZgQyAEVGixYtLhm6tFgsuniGt+zsbPfXPj4+eV5f0LkjLzdjnGEYysnJudYmu9vRtm1bLV++XN98842mT58ul8slHx8fLV682L3f8ePHVbp0aUmSr6+vpNz6LqdTp05q2LChVqxYoVmzZmnlypWKj4//U+0DUPQxZAmgSLkwdJmcnCxJKlOmjA4dOqTMzEydOXNGmzZtuu5zBAUF6ZZbbtHSpUslSVu2bNHJkydVp06dP33MLl26aObMmSpdurSqVKmi4OBgVa9e3R3I1q5dq8ceeyzfY3h7e8vpdEqSnn/+ef3yyy+Kjo7WgAED3NefASie6CEDUKRcGLp86qmnJEl16tRR69at9cADD6hKlSpq1KjRNR3v3XffVVhYmHr27Jln/VtvvaWRI0dq0qRJstlsmjRp0iU9bn904Rqyiz355JP661//qkaNGik9PT3PeS6c46OPPpLNZtPbb799xR4xSWrQoIEmT56s8ePHq2/fvho6dKimTJkim82mkSNHXlPdAG4uFuNyffcAAAC4YRiyBAAAMBmBDAAAwGQEMuD/261jAQAAAIBB/tbD2FMUAcBMyAAAZkIGADATMgCAmZABAMyEDABgFsk0zOweKg/5AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "best_param = res['param'][np.argmax(res['avg_reward'])]\n",
    "cum_ctr = np.max(res['avg_reward'])\n",
    "print('Best Parameter: {0}\\nCumulative CTR: {1:0.4%}'.format(best_param, cum_ctr))\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "_ = ax.plot(res['seq_avg_reward'][np.argmax(res['avg_reward'])], color='mediumblue', \n",
    "            label='Linear Thompson Sampling', linewidth=1)\n",
    "ax.axhline(y=cum_ctr, color='black', linestyle='--', linewidth=1)\n",
    "ax.set_xlabel('Num. of Events')\n",
    "ax.set_ylabel('CTR')\n",
    "ax.set_ylim([0.0, 0.1])\n",
    "ax.legend(loc='best')\n",
    "_ = ax.set_title('Average Reward', fontsize=12, fontweight='bold')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}