eda_evaluation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#匯入相關套件\" data-toc-modified-id=\"匯入相關套件-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>匯入相關套件</a></span></li><li><span><a href=\"#事前準備\" data-toc-modified-id=\"事前準備-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>事前準備</a></span></li><li><span><a href=\"#EDA\" data-toc-modified-id=\"EDA-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>EDA</a></span></li><li><span><a href=\"#Evaluation\" data-toc-modified-id=\"Evaluation-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Evaluation</a></span><ul class=\"toc-item\"><li><span><a href=\"#Resampler\" data-toc-modified-id=\"Resampler-4.1\"><span class=\"toc-item-num\">4.1&nbsp;&nbsp;</span>Resampler</a></span></li><li><span><a href=\"#Resampler-+-Classifier\" data-toc-modified-id=\"Resampler-+-Classifier-4.2\"><span class=\"toc-item-num\">4.2&nbsp;&nbsp;</span>Resampler + Classifier</a></span></li><li><span><a href=\"#Classifier\" data-toc-modified-id=\"Classifier-4.3\"><span class=\"toc-item-num\">4.3&nbsp;&nbsp;</span>Classifier</a></span></li><li><span><a href=\"#Classifier-+-n_estimator\" data-toc-modified-id=\"Classifier-+-n_estimator-4.4\"><span class=\"toc-item-num\">4.4&nbsp;&nbsp;</span>Classifier + n_estimator</a></span></li><li><span><a href=\"#AdaBoostClassifier-+-max_depth\" data-toc-modified-id=\"AdaBoostClassifier-+-max_depth-4.5\"><span class=\"toc-item-num\">4.5&nbsp;&nbsp;</span><code>AdaBoostClassifier</code> + <code>max_depth</code></a></span></li><li><span><a href=\"#GradientBoostingClassifier,-XGBClassifier-+-learning_rate\" data-toc-modified-id=\"GradientBoostingClassifier,-XGBClassifier-+-learning_rate-4.6\"><span class=\"toc-item-num\">4.6&nbsp;&nbsp;</span><code>GradientBoostingClassifier</code>, <code>XGBClassifier</code> + <code>learning_rate</code></a></span></li><li><span><a href=\"#Best-Model\" data-toc-modified-id=\"Best-Model-4.7\"><span class=\"toc-item-num\">4.7&nbsp;&nbsp;</span>Best Model</a></span><ul class=\"toc-item\"><li><span><a href=\"#f1-score\" data-toc-modified-id=\"f1-score-4.7.1\"><span class=\"toc-item-num\">4.7.1&nbsp;&nbsp;</span>f1-score</a></span></li><li><span><a href=\"#balanced-accuracy\" data-toc-modified-id=\"balanced-accuracy-4.7.2\"><span class=\"toc-item-num\">4.7.2&nbsp;&nbsp;</span>balanced accuracy</a></span></li></ul></li></ul></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 匯入相關套件"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.154778Z",
     "start_time": "2020-04-16T15:09:07.789923Z"
    }
   },
   "outputs": [],
   "source": [
    "# Import built-in packages\n",
    "from math import isnan\n",
    "from functools import reduce\n",
    "\n",
    "# Import 3-rd party packages\n",
    "import sqlalchemy\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from plotnine import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 事前準備"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.165749Z",
     "start_time": "2020-04-16T15:09:12.157771Z"
    }
   },
   "outputs": [],
   "source": [
    "def print_info(info, width=61, fillchar='='):\n",
    "    \"\"\"\n",
    "    印出格式化的資訊\n",
    "    \"\"\"\n",
    "    temp_width = width - (width-len(info))//2\n",
    "    print(info.rjust(temp_width, fillchar).ljust(width, fillchar))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.178714Z",
     "start_time": "2020-04-16T15:09:12.170744Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_connector(user, host, database, password=None, port='5432', protocol='postgres'):\n",
    "    \"\"\"\n",
    "    取得連線引擎，預設為連線至 PostgreSQL，埠號預設為 5432。\n",
    "    \"\"\"\n",
    "    print_info(\"GETTING CONNECTOR START!\")\n",
    "    user_info = f'{user}:{password}' if password else user\n",
    "    url = f'{protocol}://{user_info}@{host}:{port}/{database}'\n",
    "    engine = sqlalchemy.create_engine(url, client_encoding='utf-8')\n",
    "    print_info(\"DONE!\")\n",
    "    return engine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.191680Z",
     "start_time": "2020-04-16T15:09:12.182705Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_tables(engine, table_names):\n",
    "    \"\"\"\n",
    "    依照 `tables_names` 的順序，取得 tables，並依序儲存於 `list` 當中，回傳型態為 `list`，每個 element 為 `DataFrame`。\n",
    "    \"\"\"\n",
    "    print_info(\"GETTING TABLES START!\")\n",
    "    rslt = []\n",
    "    for tn in table_names:\n",
    "        query = f'SELECT * FROM {tn}'\n",
    "        exec(f'{tn} = pd.read_sql(query, engine)')\n",
    "        # exec(f\"{tn} = pd.read_csv('{tn}.csv', encoding='utf8')\") # from current working directory\n",
    "        print(\n",
    "            f'{format(tn, \"26s\")} 總共有 {eval(f\"{tn}.shape[0]\"):9,} 筆資料和 {eval(f\"{tn}.shape[1]\")} 個欄位')\n",
    "        exec(f'rslt.append({tn})')\n",
    "    print_info(\"DONE!\")\n",
    "    return rslt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.206639Z",
     "start_time": "2020-04-16T15:09:12.193674Z"
    }
   },
   "outputs": [],
   "source": [
    "def merge_tables(tables, table_names, how):\n",
    "    \"\"\"\n",
    "    合併所有 tables，回傳型態為 `DataFrame`。\n",
    "    \"\"\"\n",
    "    print_info(\"MERGING TABLES START!\")\n",
    "    # 分別處理 post_{shared, comment_created, liked, collected}_{train, test} 四個 tables\n",
    "    # groupby 每篇文章，將前十小時的分享數、評論數、愛心數、收藏數加總起來\n",
    "    for idx, (table, tn) in enumerate(zip(tables, table_names)):\n",
    "        if len(tn.split('_'))==2: continue                  # for handling posts_{train, test} table\n",
    "        col_name = f\"{tn.split('_')[1]}_count\"              # tn.split('_')[1] is either {shared, comment, liked, collected}\n",
    "        mapper = {'count': col_name}\n",
    "        exec(f\"tables[{idx}] = table.groupby(['post_key'], as_index=False).sum().rename(columns=mapper)\")\n",
    "    # 將 tables 合併起來並回傳。\n",
    "    total_df = reduce(lambda left, right: pd.merge(left, right, on=['post_key'], how=how), tables)\n",
    "    print_info(\"DONE!\")\n",
    "    return total_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:09:12.217615Z",
     "start_time": "2020-04-16T15:09:12.208635Z"
    }
   },
   "outputs": [],
   "source": [
    "def preprocess_total_df(total_df):\n",
    "    \"\"\"\n",
    "    預處理剛合併好的 total_df 以符合後續建模需求，回傳型態為 `DataFrame`。\n",
    "    \"\"\"\n",
    "    print_info(\"PREPROCESSING TOTAL_DF START!\")\n",
    "    total_df.set_index('post_key', inplace=True)                                    # post_key 欄位設為索引\n",
    "    total_df['created_at_hour'] = pd.to_datetime(total_df['created_at_hour'])       # 將 created_at_hour 欄位轉換成 datetime 型態\n",
    "    total_df['weekday'] = total_df['created_at_hour'].dt.dayofweek                  # 擷取出發文的 weekday\n",
    "    total_df['hour'] = total_df['created_at_hour'].dt.hour                          # 擷取出發文的 hour\n",
    "    total_df.fillna(0, inplace=True)                                                # NaN 值補 0\n",
    "    total_df['is_trending'] = 0+(total_df['like_count_36_hour']>=1000)              # 轉換成 is_trending 類別欄位\n",
    "    total_df = total_df.drop(['created_at_hour'], axis=1)                           # drop 掉不必要的欄位\n",
    "    # 將計次欄位轉換成 int 型態\n",
    "    col_names = ['shared_count', 'comment_count', 'liked_count', 'collected_count']\n",
    "    for cn in col_names:\n",
    "        total_df[cn] = total_df[cn].astype(dtype='int')\n",
    "    print_info(\"DONE!\")\n",
    "    return total_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:33.911175Z",
     "start_time": "2020-04-16T15:09:12.222599Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "===================GETTING CONNECTOR START!==================\n",
      "============================DONE!============================\n",
      "====================GETTING TABLES START!====================\n",
      "posts_train                總共有   793,751 筆資料和 3 個欄位\n",
      "post_shared_train          總共有   304,260 筆資料和 3 個欄位\n",
      "post_comment_created_train 總共有 2,372,228 筆資料和 3 個欄位\n",
      "post_liked_train           總共有 3,395,903 筆資料和 3 個欄位\n",
      "post_collected_train       總共有 1,235,126 筆資料和 3 個欄位\n",
      "============================DONE!============================\n",
      "====================MERGING TABLES START!====================\n",
      "============================DONE!============================\n",
      "================PREPROCESSING TOTAL_DF START!================\n",
      "============================DONE!============================\n"
     ]
    }
   ],
   "source": [
    "# Get engine\n",
    "engine = get_connector(\n",
    "    user=\"candidate\",\n",
    "    password=\"dcard-data-intern-2020\",\n",
    "    host=\"35.187.144.113\",\n",
    "    database=\"intern_task\"\n",
    ")\n",
    "# Get tables from db\n",
    "table_names_train = ['posts_train', 'post_shared_train', \n",
    "                     'post_comment_created_train', 'post_liked_train', 'post_collected_train']\n",
    "tables_train = get_tables(engine, table_names_train)\n",
    "# Merge tables\n",
    "total_df_train = merge_tables(tables_train, table_names_train, how='left')\n",
    "# Preprocess total_df\n",
    "total_df_train = preprocess_total_df(total_df_train)\n",
    "\n",
    "engine.dispose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:33.954059Z",
     "start_time": "2020-04-16T15:13:33.917158Z"
    }
   },
   "outputs": [],
   "source": [
    "cv_results = pd.read_csv('./outputs/cv_results.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:34.832540Z",
     "start_time": "2020-04-16T15:13:33.958050Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23755717ba8>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp = total_df_train.drop(columns=['weekday', 'hour', 'is_trending'])\n",
    "sns.heatmap(temp.corr(), cmap='YlGnBu')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:34.844817Z",
     "start_time": "2020-04-16T15:13:34.837836Z"
    }
   },
   "outputs": [],
   "source": [
    "mapper = dict(zip([0, 1, 2, 3, 4, 5, 6], ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:35.326509Z",
     "start_time": "2020-04-16T15:13:34.847809Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x237490fb518>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 觀察不同時段下的發文數\n",
    "num_articles_heatmap_df = total_df_train.groupby(['weekday', 'hour']).size().reset_index().rename(columns={0:'count'})\n",
    "num_articles_heatmap_df = num_articles_heatmap_df.pivot(index='weekday', columns='hour', values='count')\n",
    "num_articles_heatmap_df = num_articles_heatmap_df.rename(mapper=mapper, axis=0)\n",
    "num_articles_heatmap_df = num_articles_heatmap_df.reindex(['Monday', 'Tuesday', 'Wednesday', \n",
    "                                                           'Thursday', 'Friday', 'Saturday', 'Sunday'])\n",
    "plt.figure(figsize=(20, 5))\n",
    "plt.title(f'Number of Articles by Day of Week / Hour of Day')\n",
    "sns.heatmap(num_articles_heatmap_df, cmap='OrRd', cbar=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:35.909633Z",
     "start_time": "2020-04-16T15:13:35.328564Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2375e9e5cc0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 觀察不同時段下的熱門文章比例\n",
    "num_pops_heatmap_df = total_df_train.groupby(['weekday', 'hour'])['is_trending'].sum().reset_index()\n",
    "num_pops_heatmap_df = num_pops_heatmap_df.pivot(index='weekday', columns='hour', values='is_trending')\n",
    "num_pops_heatmap_df = num_pops_heatmap_df.rename(mapper=mapper, axis=0)\n",
    "num_pops_heatmap_df = num_pops_heatmap_df.reindex(['Monday', 'Tuesday', 'Wednesday', \n",
    "                                                   'Thursday', 'Friday', 'Saturday', 'Sunday'])\n",
    "pct_pops_heatmap_df = num_pops_heatmap_df/num_articles_heatmap_df\n",
    "plt.figure(figsize=(20, 5))\n",
    "plt.title(f'Percentage of Popular Articles by Day of Week / Hour of Day ')\n",
    "sns.heatmap(pct_pops_heatmap_df, cmap='Blues', cbar=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:36.363990Z",
     "start_time": "2020-04-16T15:13:35.911629Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x237625a2ef0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 觀察不同時段下，前 10 小時愛心平均數\n",
    "num_likes_10_heatmap_df = total_df_train.groupby(['weekday', 'hour'])['liked_count'].mean().reset_index()\n",
    "num_likes_10_heatmap_df = num_likes_10_heatmap_df.pivot(index='weekday', columns='hour', values='liked_count')\n",
    "num_likes_10_heatmap_df = num_likes_10_heatmap_df.rename(mapper=mapper, axis=0)\n",
    "num_likes_10_heatmap_df = num_likes_10_heatmap_df.reindex(['Monday', 'Tuesday', 'Wednesday', \n",
    "                                                           'Thursday', 'Friday', 'Saturday', 'Sunday'])\n",
    "plt.figure(figsize=(20, 5))\n",
    "plt.title(f'Average Likes within 10 hours by Day of Week / Hour of Day')\n",
    "sns.heatmap(num_likes_10_heatmap_df, cmap='Purples', cbar=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:36.773335Z",
     "start_time": "2020-04-16T15:13:36.365984Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2376129c400>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 觀察不同時段下，前 36 小時愛心平均數\n",
    "num_likes_36_heatmap_df = total_df_train.groupby(['weekday', 'hour'])['like_count_36_hour'].mean().reset_index()\n",
    "num_likes_36_heatmap_df = num_likes_36_heatmap_df.pivot(index='weekday', columns='hour', values='like_count_36_hour')\n",
    "num_likes_36_heatmap_df = num_likes_36_heatmap_df.rename(mapper=mapper, axis=0)\n",
    "num_likes_36_heatmap_df = num_likes_36_heatmap_df.reindex(['Monday', 'Tuesday', 'Wednesday', \n",
    "                                                           'Thursday', 'Friday', 'Saturday', 'Sunday'])\n",
    "plt.figure(figsize=(20, 5))\n",
    "plt.title(f'Average Likes within 36 hours by Day of Week / Hour of Day ')\n",
    "sns.heatmap(num_likes_36_heatmap_df, cmap='YlGn', cbar=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:36.797427Z",
     "start_time": "2020-04-16T15:13:36.775331Z"
    }
   },
   "outputs": [],
   "source": [
    "# 去除用不到的欄位資訊\n",
    "cv_results = cv_results.drop(columns=[col for col in cv_results.columns if 'split' in col])\n",
    "# 將欄位內容化簡\n",
    "def transform(elem, mapper):\n",
    "    if type(elem)==float and isnan(elem):\n",
    "        return elem\n",
    "    for sub_str in mapper:\n",
    "        if sub_str in elem:\n",
    "            return mapper[sub_str]\n",
    "    return elem\n",
    "# resampler\n",
    "mapper = {\n",
    "    'SMOTE': 'SMOTE',\n",
    "    'NearMiss': 'NearMiss'\n",
    "}\n",
    "cv_results['param_resampler'] = cv_results['param_resampler'].apply(transform, args=(mapper,))\n",
    "# classifier\n",
    "mapper = {\n",
    "    'AdaBoostClassifier': 'AdaBoostClassifier',\n",
    "    'XGBClassifier': 'XGBClassifier',\n",
    "    'GradientBoostingClassifier': 'GradientBoostingClassifier'\n",
    "}\n",
    "cv_results['param_classifier'] = cv_results['param_classifier'].apply(transform, args=(mapper,))\n",
    "# classifier__base_estimator\n",
    "mapper = {\n",
    "    'max_depth=1': 'DecisionTreeClassifier(max_depth=1)',\n",
    "    'max_depth=2': 'DecisionTreeClassifier(max_depth=2)',\n",
    "    'max_depth=3': 'DecisionTreeClassifier(max_depth=3)'\n",
    "}\n",
    "cv_results['param_classifier__base_estimator'] = cv_results['param_classifier__base_estimator'].apply(transform, args=(mapper,))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resampler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:37.430148Z",
     "start_time": "2020-04-16T15:13:36.799307Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884558871573)>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results.groupby(['param_resampler'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()\n",
    "temp.reset_index(inplace=True)\n",
    "\n",
    "(ggplot(temp, aes(x='param_resampler', y='mean_test_f1_score', fill='param_resampler'))\n",
    " + geom_bar(stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + scale_fill_brewer('qualitative', 2)\n",
    " + geom_text(aes(label='mean_test_f1_score'), format_string='{:.2f}')\n",
    " + ggtitle(f'Average F1 Score by Resampler')\n",
    " + labs(fill=f'Resampler')\n",
    " + xlab('Resampler')\n",
    " + ylab(f'Average F1 score'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:38.056082Z",
     "start_time": "2020-04-16T15:13:37.432144Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884558718762)>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(ggplot(temp, aes(x='param_resampler', y='mean_test_recall', fill='param_resampler'))\n",
    " + geom_bar(stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + scale_fill_brewer('qualitative', 2)\n",
    " + geom_text(aes(label='mean_test_recall'), format_string='{:.2f}')\n",
    " + ggtitle(f'Average Recall by Resampler')\n",
    " + labs(fill=f'Resampler')\n",
    " + xlab('Resampler')\n",
    " + ylab(f'Average Recall'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:38.590069Z",
     "start_time": "2020-04-16T15:13:38.059074Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (152294750826)>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(ggplot(temp, aes(x='param_resampler', y='mean_test_precision', fill='param_resampler'))\n",
    " + geom_bar(stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + scale_fill_brewer('qualitative', 2)\n",
    " + geom_text(aes(label='mean_test_precision'), format_string='{:.2f}')\n",
    " + ggtitle(f'Average Precision by Resampler')\n",
    " + labs(fill=f'Resampler')\n",
    " + xlab('Resampler')\n",
    " + ylab(f'Average Precision'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:39.126727Z",
     "start_time": "2020-04-16T15:13:38.593062Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAs0AAAHICAYAAAC1atevAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdd1hUx/s28HuXshTpoNJEQUDsXzXYFSs24osFRVExxt6jxt5LiCUaO4oFFaOxJnaDLbEkGpWQWGLvWEB6kTbvH/5247oLyyq4iPfnurh058yc85zC8uzsnDkSIYQAERERERHlSarrAIiIiIiIijsmzUREREREGjBpJiIiIiLSgEkzEREREZEGTJqJiIiIiDRg0kxEREREpAGTZiIiIiIiDZg0ExERERFpwKSZiIiIiEgDJs1EH0BkZCQkEgm2bNmi61CKRFhYGCQSCU6fPq3rUOj/BAUFQV9fX9dh0Ftu3boFiUSCOXPm6DoUItISk+YSJCMjA9bW1pBIJJg9e7auw/moZGdnQyKRKP1IpVKULVsWTZo0waZNm8Anzhe+M2fOKI73uXPndB0OvQd5Mvjmj5GRETw9PTFu3DgkJCToOkQiovfCbogSZNeuXYiPj0fFihWxfv16TJkyBRKJRNdhfVSqV6+OcePGAQBycnLw6NEjbNy4EX369MHVq1cREhKi4whLlrCwMFhbW0MIgbCwMNSvX1/XIdF7at68Ofr27QsAiIuLw6FDh7Bw4UJERkbiwoUL7P0moo8We5pLkLCwMFSqVAkLFizAvXv3EBkZqeuQkJGRgZycHF2HUWD29vYICgpCUFAQ+vTpg8mTJ+PMmTPQ09NDWFiYrsMrUZKTk7Fjxw706NEDPXr0wPbt25GcnKzrsApECIHU1FRdh1Esubu7K36HRo4cicOHD6N9+/aIiorCwYMHdR1eiZWSkqLrEIhKPCbNJcStW7dw6tQp9O3bF+3bt4ednZ3aJK9u3bqwsrLCq1evVJZdv34dEokEgwcPVio/duwY2rRpAysrK8hkMlSuXBkLFixQSYYbNWqEihUr4u7du+jWrRtsbW1hbGyMmJgYAMCKFSvQunVrODo6wtDQEGXKlEHXrl1x9epVtfsUHh6OatWqQSaTwdHREWPHjsXff/+tdjygEAKrV6/GZ599BlNTU5iamqJhw4b4+eeftTqO6tjY2EAmk8HQ0FCp/Nq1axgyZAiqVKkCc3NzGBsbo3r16li8eDFyc3M1rjc7Oxtz585F06ZNUbZsWRgaGsLR0RF9+/bFo0ePVOpKJBJ8+eWXOHPmDJo2bQpTU1NYWVmhZ8+eiI2NVVl/cnIypk2bhipVqsDY2BhWVlbw9vbGqlWrlOplZmYiJCQE1apVg7GxMczNzdG6dWu145MzMzMxdepUuLi4wMjICJUrV0ZoaGhBDqOKrVu3IjU1FX379kXfvn2RmpqKbdu25Vl/z549aNGiBaysrGBsbAw3Nzf0798fL1++VKp36tQp+Pn5wc7ODjKZDC4uLujZsyfu3r0LQPlYvk3d2HP5eO0TJ05g7ty5cHd3h0wmw+LFiwEAv//+O/r27QsPDw/Fteft7Y1Nmzap3Q9N5+XcuXOQSCSYPHmy2vb9+/eHVCrFnTt38jm6/3n+/Dl69+4NW1tbmJiYoGHDhjh16pRSHS8vL5QrV07tdSsfQjNjxowCbU+dli1bAgBu3LihsiwmJgZDhw6Fi4uL4n2hd+/eePDggVK9jIwMzJw5E15eXjA1NYWZmRnc3d3Rr18/ZGVlKeodPnwY3bt3h6urK4yNjWFhYQEfHx8cOHBAZdvycd+xsbH44osvYGdnh1KlSqFNmza4ffs2AGDv3r2oU6cOjI2N4ejoqPbbJicnJ7Rs2RIXL15EixYtYGZmBktLS3Tt2hX3798v8HHasWMHmjRpong/qVWrFtavX5/n9qKiotCuXTtYWlrCysqqwNshonfD78lKiLCwMEilUvTq1QsGBgYICgrCihUrEBsbC1tbW0W94OBgDBkyBD/99BMCAgKU1hEeHq6oI7dmzRoMGjQItWvXxoQJE2BhYYHTp09j/Pjx+Ouvv1RubEtKSkLjxo1Rr149zJw5E0lJSTAxMQEAzJ8/H40bN0bLli1hZWWFGzduICwsDEePHkVUVBQqVKigWM/333+PUaNGoXLlypg5cyb09fWxdetWlT/2cr169cLWrVvx//7f/0NQUBAAYOfOnejYsSPWrFmD/v37F+g4ZmVlKRLQnJwcPH78GEuWLEFaWhqmTJmiVPfYsWM4ffo0OnbsCBcXF6SlpWH//v346quvcPfuXSxdujTfbWVkZGDhwoXo0qULPv/8c5iZmeGvv/7Chg0bcOzYMURHR8PS0lKpzcWLF7F7924EBwejR48euHDhAtavX4/ExETs379fUS8+Ph6NGzfGlStX4O/vr0i0/v77b+zevVvxwSgrKwu+vr44ffo0evTogcGDByMtLQ2bNm1Cs2bNsG/fPrRp00ax3p49e2Lnzp1o2bIlxo4dixcvXmDKlClwcXEp0PF909q1a1G9enXUqlULAFCtWjWEhYWpPVeTJ0/GvHnz4OHhgZEjR8LBwQH379/Hzz//jCdPnsDa2hoAEBoaisGDB6Ns2bLo378/KlSogCdPnuDw4cO4evWq0jWmra+++grp6ekIDg6Gra0typcvD+D1sKh///0X3bt3h7OzMxISErBt2zb06dMHsbGx+OqrrxTrKMh5qV+/PmrUqIENGzZg1qxZ0NPTU7RPSUnBtm3b0KJFC7i6umqMWQiB1q1bw8bGBtOmTUNcXBxWrVqFVq1a4ciRI2jWrBkAYMCAAfjqq69w+PBhtGvXTmkd8veWL7744p2P3a1btwC8/gD6pvv376NBgwZIT09Hv3794O7ujkePHmHVqlU4evQo/vzzTzg5OQEABg8ejI0bN6JXr14YOXIkhBC4e/cu9u3bh1evXsHAwAAAFL8PwcHBcHBwwLNnzxAeHo4OHTpg586d6Ny5s8ox8vX1hZOTE2bMmIHHjx9j8eLF8PX1xbRp0/D1119j0KBB+OKLL7B161ZMnDgRFSpUQLdu3ZTW8+DBA7Ro0QKdOnVCly5dcOXKFYSGhuLs2bO4fPkySpcune8xkl/jLVq0wPTp02FkZITDhw+jX79+uHPnjkpHwb179+Dj44NOnTrh22+/xbNnz7Q/MUSkHUEfvaysLFG2bFnRvn17RVl0dLQAIL777juluvHx8cLIyEi0bdtWqTwnJ0c4OTkJLy8vRdmjR4+ETCYTnTt3VtlmSEiIACBOnz6tKGvYsKEAICZPnqw2zpSUFJWy6Ohooa+vL0aMGKEoi4uLE8bGxsLd3V0kJycrytPT00WtWrUEADF79mxF+Y4dOwQAsXTpUqV15+bmirZt2woLCwu1235TVlaWAKD2x9jYWCxfvrxA+yOEEN26dRP6+vri2bNnirJffvlFABCbN29WlOXk5Ii0tDSV9ocOHVI5d/L4pFKp+P3335Xq9+vXTwAQt27dUpQNHDhQAFAbd05OjuL/CxYsEADEzz//rFTn1atXonr16sLd3V1RduTIEQFAdOnSReTm5irKb968KWQymQAgfvvtN7XH5G1RUVECgFi8eLGi7LvvvhMARHR0tFLdc+fOCQCicePGao+XfH/u378vDAwMhIeHh3j58mWe9eTHsl+/fip11J2ntWvXCgDCw8ND7TlXV5adnS0aNGggrKysRHZ2tqK8oOdl5cqVAoD46aeflOrIY9m+fbtK+7f17NlTABD+/v5K5+vOnTvCyMhIVKpUSVH+8uVLYWRkJPz9/ZXWkZiYKExMTFTeL9S5efOmACD69OkjXrx4IV68eCH+/fdfsXjxYmFgYCAsLCxEbGysUpv27dsLa2trce/ePaXyW7duCVNTU/Hll18qyszNzYWfn5/GONSdj+TkZOHm5iaqV6+uVC4/RkOHDlUqX7RokQAgzMzMxP379xXlaWlpwtbWVjRq1EipvqOjo8r1LIQQW7duFQDEwIEDFWXy4/Tme9iFCxcEAPHVV1+pxD5o0CChp6endIzk21u7dm1+h4KIChmT5hJgz549AoDYuXOnUnnt2rVF5cqVVeoHBAQIPT098eTJE0WZPCH69ttvFWWLFy8WAMShQ4cUfwTlP//8848AIKZOnaqo37BhQyGRSERiYmK+8ebm5orExETFury8vETdunUVyyMiIgQAsWTJEpW2mzdvVvmD07FjR2FqaipiYmJU4gwNDRUAxLFjx/KNSZ5I1alTR/zyyy/il19+EUePHhXh4eGibdu2QiKRqPxBfFNGRoaIi4sTL168EBs2bFAcNzl1ydibcnJyRHx8vCLuUqVKiW7duqnE16RJE5W227ZtU9pedna2MDc3Fx4eHkrJkjo1atQQFStWVDluL168EJMnTxYAxO3bt4UQQgwYMEAAEJcuXVJZT3BwsFZJ89ChQ4WBgYF4/vy5ouz58+fCwMBA6QOUvO7bH9DUkX8A2LJlS7713jVpXrZsmcb9SktLE7GxseLFixdi5syZAoC4du2aEEK785KUlCRKlSolOnTooFTu7e0t7OzsxKtXrzTGIk8Iz58/r7Ksb9++AoD4559/FGW9e/cW+vr64unTp4oyefK+Z88ejduTJ4Pqfry9vcWff/6pVD8uLk5IJBKlJPvNHx8fH+Hs7KyoX65cOeHi4qLyoSo/KSkpivPx5ZdfColEIlJTU1WO0Z07d5TanT9/XgAQvXv3Vllnu3btRJkyZZTKHB0dhaWlpcp5yc3NFa6ursLW1lblOL35HjZy5EjF+Xj7OBw8eFAAEOvWrVPanp2dndIHLSIqehyeUQKsXbsWpqam8PLyUnwNCgC+vr6YN28ezp07pzQrQXBwMH788UdERERg7NixAF4PzdDT00OvXr0U9a5duwYAaNu2bZ7bfvsrwbJly8Lc3Fxt3ZMnT2L27Nk4d+4c0tPTlZa5u7sr/i8fq1mpUiWVdagru3btGlJTU2Fvb1/gOPNiY2OjGH8p17t3bzRv3hzjxo2Dn58f3NzcAACpqamYNWsWtm/frnbc4ttjbdXZtWsXFi5ciEuXLiEzM1Nje3Vfycu/8o6LiwPwel+TkpLg6+urcfaUa9euITMzE3Z2dnnWefbsGVxdXRVjPL28vFTqVKlSJd/tvCk9PR0RERFo2LAhEhMTkZiYqFjWoEEDbNmyBfPnz4dMJgPw3zjYmjVr5rteeb3//e9/BY5FGx4eHmrLX7x4gWnTpmHv3r14+vSpynL5edTmvJiZmaFHjx5Yt24dnjx5AgcHB0RHR+P8+fMYO3asyvj6/FSuXDnPslu3binO3aBBg7Bp0yZs3LgR48ePB/D6vcXe3h4dOnQo8PY6dOiAkSNHIjc3F/fu3cOiRYsQExOjGKYl9++//0IIgfDwcMXQsLe9uZ/Lly9HUFAQqlevDhcXFzRu3Bi+vr7o2rWr4loBgLt372LKlCk4fPiw2t+hhIQEpVikUqnK8CL5+GB1v29WVlaK37U3VaxYUeW8SCQSeHl54cCBA0hMTISFhYXa/ZS/11atWlXtckD1PaxixYqQSnlbEtGHxKT5I/f48WMcOXIEOTk5eSYub0/l1bp1azg4OCA8PBxjx45FUlIS9uzZA19fX6XEU35T0IYNGxTjCt/m6Oio9PrtP4xyf/zxB1q2bImKFSti3rx5qFChAkxMTCCRSDB8+HClG3nkCjpdXm5uLqytrbF9+/Y86+T3x6gg2rZtixMnTuDEiROKpLlbt244ePAg+vfvjyZNmsDGxgb6+vq4cOECJk2apPFmwJ07d6Jr167w9vbG4sWL4ezsDGNjYwBA165d1bZ/c3zr28T/zSMttJhPWgiBypUr4/vvv8+zzptJV17nRJtt7ty5EwkJCTh58qTSh6U37d69G4GBgVqtu6D18ruusrOz81ym7trOzc1Fy5Ytcf36dYwYMQK1a9eGlZUV9PT0sG/fPixdulRxHrU5RsDrJHbNmjXYsGEDJk+ejLVr1wJAgcfnF8Sbx0I+lnrdunUYP348Ll68iMuXL2PSpElaTRPn6Oio9MGzY8eOqFKlCvz9/REdHa1ILOXHJTAwMM/x0m8mhX5+frh//z4OHTqEU6dO4cSJE9iyZQtmzZqFc+fOwcbGBklJSWjUqBFSU1MxcuRIVKtWDebm5pBKpQgLC8P27dtVfq/kc7Krk9fvm7bnUr6dvMhjOnjwoGJs9tsqVqyo9Dqv91oiKjpMmj9y69evR05ODr7//ns4ODioLF+9ejW2b9+OJUuWwMzMDMDrPwRBQUGYP3++4g+j/AanN8l71qytrVV6X7UVERGBnJwcHD16FOXKlVOUCyHw4sULpRve5L07165dQ+vWrZXWI++ReTvOQ4cOoU6dOio3zhUWeVKflJQE4HWv7oEDB9C3b1+V2SPUxajOpk2bYGJigpMnTyqSZfk23ux91Za8tz8qKgpCiHz/WLu7u+P58+do1qxZvgk5ALi5ueHYsWO4du2aSm9uXjOgqBMWFgYbGxusXr1a7fIhQ4YgLCxMkTR7enoiMjISUVFRaNiwYZ7r9fT0BABcvnxZbe+qnJ6eHiwsLNT2QhZ0Rgq5qKgoREdHY+bMmZg2bZrSskOHDim91ua8AK97zL29vbFu3TqMHj0aW7ZsQdOmTfPs8c7L1atX8dlnn6mUAVB8AJQbOHAghgwZgpMnT2Lbtm2QSCTo16+fVtt7W5kyZTBjxgwMHz4cS5cuVXy75e7uDolEgoyMjAK/v1haWiIwMFBxbSxZsgSjR4/GmjVrMHHiRERGRuLJkycIDw9H7969ldq+PWtMYbt16xYyMzOVepuFELh27RpsbW3z/AYOeP0eFhkZCUdHR1SvXr1I4ySid8fvdj5iQgisX78e5cuXx4gRI9ClSxeVn0GDBqmdykv+8AH5V6PW1tb4/PPPlep069YNhoaGmD59uto5adPS0go8r648IXu7h2bFihUqX3W2adMGRkZGWLFihdLcoxkZGViyZInKuvv06QMhBMaPH6+2B+h97yrPzc3Fnj17AAC1a9fOd3+SkpLUxqiOnp4epFKpSs/XrFmz3uvpg3p6eggMDMTNmzexcuVKleVvbk8+w0NeD21589h16tQJADBv3jyl+G7dupXvVHFvunHjBn799Vf4+/urvV67dOmCTp064cSJE4oEVj4bysSJE5GRkaGyTnksAQEBMDAwwMyZMxEfH5/vfnt4eODMmTNK60tPT8fy5csLtB9yeV0HDx8+VJkqTJvzIjdw4EDcvXsXgwcPRkJCAgYMGKBVfADwzTffKMV39+5dbNu2DZ6eniofLoKCglCqVCl8//33+OGHH9CyZcsCzdKhSf/+/eHk5ISQkBDFe0bp0qXRpk0b/PTTT/j111/VtpNff9nZ2WqfKCj/fZR/AMrrfFy+fBn79u177/3IT0JCgsp53bZtG+7cuaP43cmLPMGfMGGC2m/dEhISVIZvEdGHx57mj9gvv/yCe/fuKZ5gp067du1gYmKiMpVXpUqV4O3tjY0bNyI5ORlDhw5VGhcIAOXKlcOqVaswYMAAeHp6ok+fPqhQoQLi4uJw7do17NmzBwcOHECjRo00xtq5c2csXboUrVu3xoABA2BsbIxTp07h2LFjKtOAWVtbY+7cuRgzZgzq1q2L3r17Q19fHxEREYqvLt/spQsICMDhw4exZs0aREVF4fPPP0eZMmXw5MkTXLhwAZGRkSpjqPMSExOjmEZPCIGYmBj8+OOPuHjxIjp27IimTZsCeN3j1apVK4SHh8PAwADe3t54/Pgx1q1bh7Jlyxaox7Jr167Yu3cvfHx80Lt3b+Tm5uLQoUO4devWe8+5+s033+DUqVMYNmwYjh07hqZNm0JfXx///PMPbt++jaNHjwIARo8ejWPHjmHKlCk4efIkWrVqBSsrKzx8+BCnT5/Go0ePFGOFfX194e/vj507d8LX1xd+fn6IjY3FypUrUaVKFVy8eFFjXPK5w7t27ZrvcQkNDcW6deswd+5c1KtXD19//TXmz5+PGjVqoEePHnB0dMTDhw+xd+9eREREoGrVqihXrhyWLFmCYcOGoWrVqggODkaFChXw9OlTHD58GBMnTkT79u0BACNGjECvXr3g4+ODoKAgpKSkYOPGjVof9ypVqsDLywvffPMNEhMTUblyZdy+fRtr1qyBu7s7Lly48E7nRa579+4YM2YMNm3aBGtra5Xp0jSRz+fcqlUrdOzYEXFxcVi5ciWysrKwfPlyld5uMzMzBAYGFvpQEJlMhgkTJmDYsGFYvHixold+9erVaNy4MVq0aIGePXuidu3akEqluHfvHg4cOIAGDRogLCwMCQkJcHZ2hp+fH2rWrImyZcvi8ePHCA0NhaGhIXr06AEAaNy4McqUKYNRo0bh1q1bKFeuHK5cuYKwsDBUr169QNfou3J3d8f06dMRHR0Nb29v/PPPPwgNDYW9vT1mzpyZb9u6detixowZmDFjBqpXr47u3bvD0dERz549Q3R0NH7++WfcvHkzz2FyRPSBfLBbDqnQde3aVQAQf/zxR4HqvX3X+apVqxR3t1+4cCHP9mfPnhWdO3cWZcqUEQYGBqJMmTKiQYMGYs6cOUpTezVs2FC4ubnluZ49e/aI2rVrCxMTE2FlZSU6duworl27lme7devWicqVKwtDQ0Ph6Ogovv76a3H27FkBQCxatEilfkREhGjatKmwsLAQhoaGwtnZWbRr106sWbMm3+MjRN5TzpmYmIhatWqJ7777TmRlZSm1kd+R7+DgIGQymahUqZJYuHChYiaSN2dgyGv2jLVr14oqVaoIIyMjYWdnJ4KCgsSjR4+Eo6OjaNGihUp8BZ3xQQghEhISxIQJE4S7u7swNDQUVlZWom7dumL16tVK9bKzs8Xy5cuFt7e3KFWqlDAyMhIVKlQQnTt3Fjt27FCqm5GRISZNmiScnJyEoaGhqFSpkli1apVihon8Zs/IzMwUZcqUEdbW1irH8u147OzshIODg9J0bdu3bxeNGzcWZmZmwsjISLi5uYmBAweqTC/3yy+/CF9fX2FpaSkMDQ1FuXLlRFBQkLh7965SvYULFwoXFxdhYGAg3NzcxKJFi8TRo0fznD0jr327d++eCAgIEHZ2dsLIyEjUrFlTbNy4Mc92BT0vcsOHDxcAxKhRo/I8Zur07NlT6OnpiWfPnolevXoJGxsbYWRkJOrXry+OHz+eZ7uLFy8KAMLOzk5kZmYWeHvyWSHenF7tTRkZGcLR0VFYWFgonbPY2Fgxfvx44enpKWQymTAzMxNeXl5i4MCBipk/MjIyxIQJE4S3t7ewsbFR/H4HBASIixcvKm0nOjpatG3bVlhZWQlTU1NRv359sX//fsVsMA8fPlQ5Rnnty5szXOTXRv77+ueff4rmzZsLU1NTYW5uLjp37qwyM0d+6z548KBo06aNsLa2FgYGBsLBwUE0b95cfPfddyIjI0Nle0T0YUmEeI/vgYk+sB9++AE9evRQ+5ACopJo7NixWLRoEa5cuZLvWO3C8s8//6BatWoYN24c5s+fX+TbKwmcnJxQqVIlREZG6joUIipCHNNMxVJGRobKuET5E/RkMplimARRSZacnIz169ejSZMmHyRhBl7fXCeRSN5p/DQRUUnGMc1ULJ08eRJjxoxBly5d4OTkhJiYGERERODGjRuYNWuW0qPBiUqav//+G1FRUdi8eTPi4+MxderUIt1eSkoK9u/fj6tXr2LDhg3o0aOHyhRnRESfOibNVCxVrFgRnp6eWLduHWJjY6Gvr4+qVauqnUqKqKTZvn075s6dCwcHByxZsuS9p3zU5OnTpwgMDESpUqXQsWNHLFu2rEi3R0T0MeKYZiIiIiIiDTimmYiIiIhIAybNREREREQaMGkmIiIiItKASTMRERERkQZMmomIiIiINGDSTERERESkAedpBhAbG6vrED4pEokExsbGSE9PV3nqH1Fh4rVGHwqvNd3gg67oQ2JPM31wUqkUJiYmkEp5+VHR4rVGHwqvNaKSj7/dREREREQaMGkmIiIiItKASTMRERERkQZMmomIiIiINGDSTERERESkAZNmIiIiIiINmDQTEREREWnApJmIiIiISAMmzUREREREGjBpJiIiIiLSgEkzEREREZEG+roOoLDt378fx48fx71791C/fn2MGzdO1yERERER0UeuxCXN1tbWCAgIQFRUFJKTk3UdDhERERGVACUuaW7QoAEA4M6dO0yaiYiIiKhQlLikuSBiY2MRGxureC2VSmFnZ6fDiD4tenp6Sv8SFRVea/Sh8FojKvk+yaR5165dWLt2reJ1cHAwhg0bpsOIPk3m5ua6DoE+EbzW6EPhtUZUcn2SSXPnzp3RtGlTxWupVIr4+HgdRvRp0dPTg7m5OZKSkpCTk6PrcKgE47VGHwqvNd2wsrLSdQj0Cfkkk2ZbW1vY2toqXsfGxvJNTgdycnJ43OmD4LVGHwqvNaKSq8QlzfI3rNzcXOTm5iIzMxNSqRT6+iVuV4mIiIjoAylxmeT27duxbds2xeszZ86gefPmGDVqlA6jIiIiIqKPmUQIIXQdhK69OZMGFT09PT1YWVkhPj6eX2NSkeK1Rh8KrzXdeHOoJVFR42O0iYiIiIg0YNJMRERERKQBk2YiIiIiIg2YNBMRERERacCkmYiIiIhIAybNREREREQaMGkmIiIiItKASTMRERERkQZMmomIiIiINGDSTERERESkAZNmIiIiIiINmDQTUbF3+/ZtBAQEwMXFBV5eXpg0aRLS09M1tsvMzMT48eNRuXJlODs7o3Xr1vj111+V6qSkpKBv376oXbs2nJ2dUalSJXTr1g2XL18uqt0hIqKPEJNmIirWEhMT4e/vj5SUFKxfvx4zZszArl278NVXX2lsO2nSJKxYsQIjRoxAeHg4XFxcEBgYiL/++ktRJzMzE8bGxhg3bhx++OEHLFq0COnp6fD398ft27eLcteIiOgjIhFCCF0HoWuxsbG6DuGToqenBysrK8THxyMnJ0fX4VAxt3TpUixatAiXLl2CjY0NAGDXrl0YNGgQzpw5A3NH7UcAACAASURBVA8PD7XtYmJi8L///Q+LFy9GUFAQcnJyIISAj48PypUrh82bN+e5zZSUFHh6emLs2LEYPXp0kewXlSx8X9MNW1tbXYdAnxD2NBNRsRYZGYkmTZooEmYA6NChA2QyGSIjI/Nsd+XKFeTk5MDX11dRJpFI4OPjgxMnTiAzMzPPtqamppDJZMjOzi6cnSAioo8ek2YiKtZu3rwJd3d3pTKZTIby5cvjxo0bebZ79eoVAMDQ0FCp3NDQEK9evcKDBw+UynNzc5GdnY2nT59i2rRpkEql6Nq1ayHtBRERfeyYNBNRsZaQkAALCwuVcgsLCyQkJOTZzs3NDQBw/vx5pfKLFy8CAOLj45XKQ0JCYG9vj2rVqmHXrl344YcfUL58+feMnoiISgp9XQdARIXj5dgyug6haOTkIv3gXLx8OE+pOOtuLrKe5r3fpQF4lwW+6tcNCxtLUMEc2H0LOHvp9W0cKcvb46WdRFHfP02gYXsJXqQDP954ge7/rx02tpagio1E7fo/dtYLn+k6BCKijwp7momoWDOXAUmZqvcrJ2e+XpafkIYSWMuAwEMC9bYLRFwXGFLjdRJsa6xct4yJBNVsJWjuLMHK5hI4mAJLoz75+6SJiOj/sKeZiIo1NwvgdqJyWWaOwINkoLN7/r3AjqUk2NlBgkcpAhnZQAVzYMNVwM749bK8SCUSeFkLRL0ojD0gIqKSgD3NRFSsNXGU4PcYID7jv17fXx4AmblAU8eCrcOplAQVLSXIygV23hTo6p5//axcgehYwNnsPQInIqIShT3NRFSsdfcEtlwHhpwQGFIdiMsAQi4I+LkCbpb/9RZPOpOLvbeBq73/6wvYck2glCFgbwo8TgE2XBWQ6QH9q/7XbvsNgehYgQb2EtgZAy/SgW03Xvdkz6pfMsczExGR9pg0E1GxZm4oQXhrYM55geEnBYz0gA4VgLG1lRPaXAHkvDUEOTMXWB4l8DQNsJQBrcsBI/8ngYnBf20rWgJH7wNzzwskZb4eulHNFtjZXoJK1kyaiYjoNT4REHwi4IfGJ2cVjRI7ewYVCc6eUbj4vqYbfCIgfUgc00xEREREpAGTZiIiIiIiDZg0ExERERFpwKSZiIiIiEgDJs1ERERERBowaSYiIiIi0oBJMxERERGRBkyaiYiIiIg0YNJMRERERKQBk2YiIiIiIg2YNBMRERERacCkmYiIiIhIA31dB1AcGBoaQiaT6TqMT4ZEIgEAmJqaQgih42hKjpe6DoA+KmZmZroOoUTh+xpRycekGUBmZiYyMzN1HcYnQ09PD4aGhkhNTUVOTo6uwyH6JCUnJ+s6hBKF72u6wQ4v+pA4PIOIiIiISAMmzUREREREGjBpJiIiIiLSgEkzEREREZEGTJqJiIiIiDRg0kxEREREpAGTZiIiIiIiDZg0ExERERFpwKSZiIiIiEgDJs1ERERERBowaSYiIiIi0oBJMxERERGRBkyaiYiIiIg0YNJMRERERKQBk2YiIiIiIg2YNBMRERERaaB10jx27Fhcv369KGIhIiIiIiqWtE6af/zxR1SpUgUNGjTA+vXrkZqaWhRxEREREREVG1onzffv38eBAwfg5OSEoUOHwt7eHl9++SXOnj1bFPEREREREemc1kmzRCJBmzZt8OOPP+Lx48eYPXs2Ll68iMaNG6Ny5cpYtGgRnj9/XhSxEhERERHpxHvdCGhtbY2RI0di48aNaNy4Ma5fv45x48bB2dkZffr0wYsXLworTiIiIiIinXnnpDkxMRGrVq1CnTp1UKtWLSQlJWHFihV48uQJVq1ahd9++w3du3cvzFiJiIiIiHRCX9sGx48fx7p167B3717o6+sjMDAQoaGhqF27tqLOF198AWdnZ/j5+RVqsEREREREuqB10tyyZUvUrVsXy5YtQ/fu3WFiYqK2noeHBwIDA987QCIiIiIiXdM6af7rr79QrVo1jfVcXFywYcOGdwqKiIiIiKg40XpMc/ny5RETE6N2WUxMDFJSUt47KCIiIiKi4kTrnuYvv/wSZmZmCAsLU1k2ffp0pKSkYOvWrYUSHBERERFRcaB1T/Ovv/6K9u3bq13Wrl07nDp16r2DIiIiIiIqTrROmuPj42FmZqZ2mampKeLi4t47KCIiIiKi4kTrpNnV1RWRkZFqlx07dgzly5d/35iIiIiIiIoVrZPmL7/8Et999x3mz5+P2NhYAEBsbCwWLFiAxYsXo3///oUeJBERERGRLml9I+Do0aNx+/ZtTJw4ERMnToS+vj6ys7MBAIMGDcKYMWMKPUgiIiIiIl3SOmmWSCRYsWIFRo0ahePHjyMuLg42NjZo3rw53N3diyJGKqbkH57++OMPmJiYwN/fH1OnToWxsXGebR48eKD09Mg3GRoa4vHjx0pl169fx5w5c3D27Fnk5OTA3d0d8+bNg7e3d6HuCxEREVF+tE6a5dzd3Zkkf8ISExPh7+8PJycnrF+/HrGxsZg2bRri4+OxatWqPNuVKVMGR44cgbm5OZKSkpCbmwshBLp3745GjRop1b1y5Qr8/PzQqlUrrFmzBvr6+oiOjkZ6enpR7x4RERGRkndOmjMyMnDnzh1kZGSoLKtVq9Z7BUXFX3h4OBITE3HixAnY2NgAAPT19TFo0CCMHj0aHh4eatvJZDJ89tlnsLKyQnx8PHJycnDmzBkkJSWhU6dOSnXHjRuHVq1aITQ0VFHm4+NTZPtERERElBetk+bMzEwMGTIEmzdvVoxlfltOTs57B0bFW2RkJJo0aaJImAGgQ4cOkMlkiIyMzDNpVmfXrl0wMzODr6+vouzGjRu4cOECZs6cWahxExEREb0LrZPmmTNn4siRI9i4cSN69uyJFStWwNTUFFu2bMHt27exbNkyrYNISUnBihUrcOnSJRgbGyMgIADt2rVTqXfy5EmsXLlS8VoIgVevXmHChAlo0KAB/v77b0yZMgUymUxRp0uXLggICNA6JsrfzZs3ERgYqFQmk8lQvnx53Lhxo8DrycrKwv79+9GuXTsYGRkpyv/8808AQFJSEnx8fHD9+nU4ODhg8ODBnKGFiIiIPjitk+YdO3ZgxowZCAgIQM+ePeHt7Y3atWujd+/eCA4Oxr59+9QmvPkJDQ1FTk4ONmzYgJiYGEybNg1OTk6oXr26Uj0fHx+lr+f//PNPLFy4UOnGMgsLC2zatEnb3SItJSQkwMLCQqXcwsICCQkJBV7PsWPHEB8fj86dOyuVP3/+HAAwePBgDBkyBHXq1MHhw4cxadIkWFlZoUuXLu+3A0RERERa0DppfvToETw8PKCnpwcjIyPEx8crlvXs2ROBgYH53gj2toyMDJw5cwZLliyBiYkJ3Nzc0Lx5c0RGRqokzW+LjIxEo0aNlHqWi5ua+xbqOoQikS1ysfT6b9iwL0up/MXLx5DmJBR4v1+u+glSc1OMSroAyb6LivLkq2cBAFl13bHRLRsb438H6lrC6I+KGD5rCubI7hXWrhQrUX5jdR0CERERqaF10mxvb6/oSaxQoQJOnjyJli1bAoBWX8vLyacYK1eunKLM1dUVe/fuzbddcnIyzp8/j3nz5qmU9+7dGwYGBqhVqxZ69+6t8tjv2NhYxYNZAEAqlcLOzk7r2D9lUhMj5Kaq3gSam5YBfXsbNS1U5WZkIuOv2zBpUh0SqfJzdiSlXk9bJ/NyUSqXVSqHjOjbENk5kOjrvWP0xZeeXsnbJyqeeK0VLvnx5HElKrm0Tpp9fHzw22+/wc/PD/3798fYsWNx7do1GBoaYu/evejRo4dW68vIyFCZ19fU1FTjtGInT55E2bJlUalSJUWZk5MTvv/+ezg5OeHly5dYuXIllixZgqlTpyq13bVrF9auXat4HRwcjGHDhmkV96dO394G2TFxSmUiKxvZzxNg0ij/bwjkMi7dgMjMgkndyirLDPJLvCWS1z8lkJWV1Tu3fVGIcVDJ9z7XGuXN3Nxc1yEQURHROmmeO3euopd21KhREEJg586dSE9Px4gRIzBt2jSt1mdkZKSSIKempub7gAzg9dAMeQ+3nJWVleIPga2tLQYMGIBBgwbh1atXSkM4OnfujKZNmypeS6VSpWEmpJlRdVck7zuLnJR06P1fr3D6pRtAdg6MqrsWaB1pv1+FXmlLGLo5qCwzrOgIiakRXl29B6Nq/60v4+p9GDjYQqKn9RPgPwq8DulD4bVWuPT09BTzz3MGqQ+HH/7oQ9Iqac7Ozsbjx4/h7OysKBs9ejRGjx79zgE4OjoCAB4+fKhY7927d+Hi4pJnmzt37uDBgwdo1qxZvuuWSqUQQkAIoVRua2sLW1tbxevY2Fi+yWnJ1KcmUo5dxMulu2D2eQPkJqUhcdtxGNerDAOH/45t/PqDSDv7DxzDvlZqn5OUhlfX7sOsbV2165fo68HcrwESd5yExMQIhq72yIi6jVfRt2E9zL9I902XeB3Sh8JrrWjk5OTw2BKVUFp110mlUtSvXx/R0dGFFoCRkREaNmyIiIgIpKWl4e7duzh27BhatGiRZ5vIyEjUrl1b5RNmdHQ0nj17BiEE4uPjsWbNGtSsWVNpKjMqHFITI9iOC4REZoCXy/e+TpjrVoZVcBvlikIAuUKlffqF60BOLozrqQ7NkCvV+jNYBDRD2m/RiFuyE6+u3oVVv/YwrlXwOaCJiIiICoNWPc1SqRSurq5aTSlWEAMHDsTy5csRHBwMExMT9OzZEzVq1AAABAQEYPr06ahSpQqA1/P6njp1CsOHD1dZz507d7BkyRIkJyfD1NQUtWrVQp8+fQo1VvqPQVlr2I7plm8dq37tYdWvvUp5qRa1UKqF5idHlmpVB6Va1XnnGImIiIgKg0S8PXZBg02bNuG7777DoUOHYG9vX1RxfVBvzqRR2ErqlHNUNN5nyrmXY8sUYiRU0lkvfKbrEEoUPT09WFlZIT4+nsMzPqA3h1oSFbV3erjJs2fP4OrqiurVq6N06dKQvDGTgUQiwU8//VSoQRIRERER6ZLWUxCkpKSgUqVKqFevHkxMTJCSkoLk5GTFT1JSUlHESUREVORu376NgIAAuLi4wMvLC5MmTdI4BeqDBw9gbW0NiUQCa2tr2NnZKX7kN7vLLVq0CF26dIGrqyvs7OwQFRVVlLtDRIVI657mEydOFEUcREREOpWYmAh/f384OTlh/fr1iI2NxbRp0xAfH5/vk27LlCmDI0eOKKacy83NhRAC3bt3R6NGjZTqhoeHo3z58mjatCn2799f1LtERIVI66SZiIioJAoPD0diYiJOnDgBG5vXD1jS19fHoEGDMHr0aHh4qJ+5RyaT4bPPPlMa03zmzBkkJSWhU6dOSnWjoqIglUpx5swZJs1EHxmtk+ZZs2ZprKPtA06IiIh0LTIyEk2aNFEkzADQoUMHyGQyREZG5pk0q7Nr1y6YmZnB19dXqVwqLZkPZiL6FGidNC9YsEClLC0tDUIIyGQyGBgYMGkmIqKPzs2bNxEYGKhUJpPJUL58edy4caPA68nKysL+/fvRrl07PieAqATR+iPvmzf9yX/S09Oxb98+eHh44MyZM0URJxERUZFKSEiAhYWFSrmFhYVWzyc4duwY4uPj0blz58IMj4h0rFDGNBsaGqJ9+/Z49uwZBg0axMSZiIg+Sm9OoSqn5eMMsHPnTtjZ2aFJkyaFFRYRFQOFeiOgk5MTp88hIirh5s/9XdchFAlDA1McOXQVGcnK+3fn9lNkpFkWaL8zM9Nx8MBhVKvaGotCLuRZ78HDqwCATev/xtEyGe8XeDH39eR6ug6BqFAU2h0Jd+/exbfffgs3N7fCWiUREdEHY23thJcvHymVZWdnISHxKWysnQu0jpu3zyEr+xW8PJsWRYhEpENa9zSbmZmpfH2VlZWFzMxMmJiYYPfu3YUWHBER0YfiWr4Ozp3fjvT0JBgbmwN4nQTn5GTBtUKdAq3j2vVTsLSwh4O9Z1GGSkQ6oHXSPGbMGJWk2cjICE5OTmjbti2sra0LLTgiIqIPpUa1Nrj0137s2TcH9b27Iy09ASd+XQevSj5KPc2Hf1mKf64ew9iRPym1T0tLxP2Hf6FunS55buPho7+Rlp6E2LgHAID7D6ORmPQcFualUbaMe9HsGBEVCq2T5hkzZhRBGERERLplZFQK3TrPxbETofhp/zzoG8jg5dkETRoFK9XLzc2BELkq7a/fPI3c3Bx4Vcp7aMaZc1vx8PE/ite/nt4IAKji1RztfEcXyn4QUdHQOml++PAhXrx4gVq1aqksu3TpEkqXLg0nJ6dCCY6IiOhDsrZyRNdO+T/Eq53vaLUJbq0a7VGrRvt823bv+s17xUdEuqP1jYCDBw/G5s2b1S7bunUrhg4d+t5BEREREREVJ1onzX/88QeaN2+udlmzZs1w7ty59w6KiIiIiKg40TppTklJgYGBgfqVSaVITk5+76CIiIiIiIoTrZNmLy8v7NmzR+2yvXv3wtOT0+wQERERUcmi9Y2Ao0aNQnBwMPT09PDFF1/AwcEBT548wYYNGxAWFob169cXRZxERERERDqjddLcu3dvPHv2DDNnzkRoaKii3NjYGCEhIejTp0+hBkhEREREpGtaJ80AMG7cOAwcOBDnzp1DXFwcbGxsUL9+fZibmxd2fEREREREOvdOSTMAmJubw9fXtzBjISIiIiIqlrS+EXDZsmWYMGGC2mUTJkzAihUr3jsoIiIiIqLiROukeeXKlXBzc1O7zMPDAytXrnzvoIiIiIiIihOtk+b79+/D3d1d7TJXV1fcu3fvfWMiIiIiIipWtE6azc3NcffuXbXL7ty5AxMTk/cOioiIiIioONE6aW7dujVmzpyJhw8fKpU/evQIs2fPRtu2bQstOCIiIiKi4kDr2TNCQkJQr149eHp6onnz5oqHmxw/fhx2dnb45ptviiJOIiIiIiKd0bqn2cHBAVFRURg9ejTi4uJw8uRJxMXFYcyYMbh8+TIcHR2LIk4iIiIiIp15p3mara2tMXfu3MKOhYiIiIioWNK6pzkv0dHRGD9+PMqVK1dYqyQiIiIiKhbe+YmAwOvp57Zu3YqIiAhcu3YNhoaG8PPzK6zYiIiIiIiKBa2T5ri4OPz444+IiIjAuXPnIISARCLBxIkTMXbsWFhaWhZFnEREREREOlOg4Rnp6en44Ycf0KFDBzg4OGDo0KFISEjAnDlzcOnSJQgh0Lp1aybMRERERFQiFainuXTp0khLS4OTkxNGjRqFHj16oEaNGgCAxMTEIg2QiIiIiEjXCtTT/OrVKwghYG5uDisrK/YoExER0SdpxowZkEgkih9bW1s0atQIBw8e1HVoRWbjxo2QSCSIjY3VdSg6VaCe5qdPn2LHjh2IiIjA5MmTMXnyZNSrVw89e/ZEq1atijrGImdoaAiZTKbrMIhgZmb2zm1fFmIcVPK9z7VGpI2SeK0ZGxvj+PHjAIAnT54gJCQEfn5++O2339CgQQMdR0dFpUBJs7W1NQYOHIiBAwfi4cOHiIiIwNatWzFs2DDo6elBIpHg119/RZ06dWBiYlLUMRe6zMxMZGZm6joMIiQnJ+s6BPpE8FqjD6UorzVddXhJpVLUq1dP8bp+/fpwdHREeHg4k+Z8CCGQmZn50XZUaj1Ps7OzMyZMmIDo6GhER0djzJgxcHZ2xrRp01C2bFkEBwcXQZhERERExZO9vT3s7Ozw4MEDRZkQAgsXLoSHhwdkMhlcXV2xePFipXaPHj1CQEAAypQpAyMjI1SoUAGjR49WLL9+/Tq6d+8OZ2dnmJiYoHLlyli0aBFyc3MVde7duweJRIJNmzahf//+sLS0hJ2dHRYuXAgA2LZtGzw9PWFubo5OnTohISFB0fbkyZOQSCQ4ePAgOnXqBFNTU9jb22PevHka9/nVq1eYNGkSXFxcIJPJ4OXlha1btyrVCQ4ORtWqVXHw4EHUqFEDMpkMP//8s3YHtxh5r3maq1atipCQEISEhOC3335DREQEdu7cWVixERERERV7KSkpePnyJdzc3BRlI0eORFhYGCZPnoy6devi7NmzGD9+PIyNjTFo0CAAQO/evfHkyRMsXboUZcqUwYMHD/Dnn38q1vH48WN4enqiZ8+eMDMzQ1RUFKZPn47U1FRMmzZNKYYpU6aga9eu2LFjB/bu3Ytx48YhNjYWp06dwvz585GUlIThw4fj66+/xpo1a5TaDhgwAIGBgdi9ezciIyMxefJkWFtbK+JUJyAgAKdPn8b06dPh5eWFgwcPIigoCFZWVmjbtq2i3pMnTzBy5EhMmTIFzs7OcHZ2fq9jrUvvlTS/qXHjxmjcuDGWL19eWKskIiIiKpays7MBADExMZgwYQLMzc0xcuRIAMDt27exfPlyrF69GgMGDAAAtGzZEikpKZg5cyYGDBgAqVSK8+fP45tvvkG3bt0U6+3du7fi/y1atECLFi0AvO65btSoEdLS0rB8+XKVpLlBgwZYtGgRAKB58+bYtWsXli9fjvv378PGxgYA8Ndff2HdunUqSXPz5s2xYMECAICvry+ePn2KuXPnKuJ824kTJ/Dzzz/jyJEjaN26NQCgVatWePz4MaZPn66UNMfHx+Pw4cPw9vbW9hAXO4X2GG05ff1Cy8OJiIiIip3U1FQYGBjAwMAA5cqVw/bt27F582a4u7sDACIjIwEAnTt3RnZ2tuKnRYsWePr0KR4+fAgAqFWrFhYuXIhVq1bh1q1bKtvJyMjA9OnTUbFiRchkMhgYGGDy5MmIiYlBSkqKUt2WLVsq/q+npwdXV1fUrFlTkTADgIeHBxISElTa+vv7K73u1KkTHj16hEePHqnd/6NHj8La2hrNmzdX2b/Lly8jJydHUdfW1rZEJMxAESTNRERERCWZsbExLly4gD/++ANbtmyBvb09evXqhZiYGABAbGwshBCwtbVVJNcGBgZo06YNACiS5u3bt6NFixaYPHky3N3dUalSJezevVuxnfHjx2PBggXo378/Dh48iAsXLmDKlCkAXifUb3p7OmBDQ0O1Zerali5dWu1r+f68LTY2Fi9fvlTaNwMDAwwaNAjZ2dlK7d5e98eM3cJEREREWpBKpahTpw4AwNvbG5UqVYK3tzdmzZqFVatWwdraGhKJBKdPn1Ykqm/y9PQE8PoGwvXr1yMsLAwXL17EnDlz0K1bN/z7779wdXXFjh07MHDgQIwfP17R9sCBA4W+P8+fP1f72t7eXm19a2tr2NnZ5Tk39ZuJskQiKaQodY9JMxEREdF7qF27NgIDA7FhwwZMnz5dMQ45Li4Ofn5+GttLpVJ89tlnmDNnDn7++WfcunULrq6uSE9PV0q6c3JysG3btkKPf8+ePUpDNHbv3g0HBwc4OTmprd+yZUvMnz8fhoaGqF69eqHHU1wxaSYiIiJ6T1OnTsUPP/yAJUuWICQkBEOHDkWvXr0wbtw41K1bF1lZWbhx4wZOnDiBvXv3IjExEb6+vujVqxc8PT2RlZWFpUuXwtLSErVq1QLw+ua6tWvXonLlyrCzs8OKFSvw6tWrQo/9+PHjGDduHFq1aoVffvkFW7ZswYoVK9TeBCiPy8/PD23atMHXX3+N6tWrIzU1FVeuXMGtW7cQFhZW6DEWBwVKmmfNmlXgFUokEkydOvWdAyIiIiL62Hh6eiIwMBCrVq3CxIkTsXTpUnh6eiI0NBSzZs2CqakpPD09ERAQAAAwMjJCtWrVsGzZMjx48ADGxsaoU6cOjh49CltbWwDAsmXLMGjQIAwfPhwmJiYIDg6Gv78/+vfvX6ixh4aGIjQ0FCtXroSZmRlmz56NIUOG5Ntm586dCAkJwcqVK3H//n1YWFigatWq6Nu3b6HGVpxIhBBCU6W3H4GZnZ2t+KSjr6+vmHZFfmdnUlJSEYRadIryWeo19y0ssnVTyRPlN/ad274cW6YQI6GSznrhs3duO3/u74UYCZV0X0+up7nSO5Inl/RuTp48iWbNmuHChQuKMdqUtwLNnpGcnKz4OXHiBOzt7REaGoq4uDhkZmYiLi4Oq1evhr29PY4dO1bUMRMRERERfVBaj2keNmwYxo4dq/TVgJWVFQYMGICsrCwMHToU58+fL9QgiYiIiIh0Seuk+a+//oKrq6vaZW5ubvjnn3/eOygiIiIiKlo+Pj4owChd+j9aJ83ly5fH6tWr4evrqzT3nhACK1euhIuLS6EGSERERKRrnpunITU7s9DX+6hvSKGvk4qG1klzSEgIunTpAnd3d/j5+aF06dJ4/vw59u3bh/v372Pnzp1FEScRERERkc5onTR37NgRFy5cQEhICH766SfExMTA3t4e3t7e2LlzJ2rWrFkUcRIRERER6cw7PdykZs2aRfJEGiIiIiKi4qhAU87l5eHDhzh79ixSU1MLKx4iIiIiomLnnZLmNWvWwNHRES4uLmjcuDH+/fdfAIC/vz++//77Qg2QiIiIiIqHKlWqIDIyUtdh6ITWSfOSJUswfPhwBAUF4dChQ0pTlfj4+ODHH38s1ACJiIiIPiU+Pj4wMDDArVu3FGXXr19XmrWsMG3cuBESiUTl8dxRUVGQSCSoV++/pzpeuXIFLVu2LJI4ijutk+Zly5Zh6tSp+Pbbb1UOmqenp6LXmYiIiIjejbm5OaZNm1bk28nKygIAuLq6Ys+ePUhPT1cs27BhAzw9PYs8ho+F1knz48eP0aBBA7XLDAwMOL6ZiIiI6D0NHz4c+/btQ1RUlMqyzMxMTJw4EeXLl4etrS169OiB+Ph4xfLAwEDY29vDwsICjRs3xt9//61YFhwcjIEDB+Lzzz9HqVKlsGvXLgCAnZ0dGjRogD179gB4nUxv374dvXr1Utp2+fLlcfjwYQDA+fPn8dlnn8Hc3BylS5fG0KFDAQAZGRno1asXbGxsYGlpidq1a+PRo0eFe4B0QOuk2cXFJc/HZP/xxx/w8PB476CISUoJAwAAIABJREFUiIiIPmVly5bFiBEjMHnyZJVlEydOxOXLl3H+/Hk8fPgQhoaGGDZsmGJ569at8e+//+L58+eoXbs2AgMDldpHRERg9OjRSE5ORseOHRXlwcHBCA8PBwDs378fNWrUgKOjY54xjhw5EiNGjEBSUhLu3r2LoKAgAEB4eDgSEhLw8OFDxMXFISwsDKampu91PIoDrZPm/v37Y86cOVi3bh2SkpIAvP40cuDAASxYsAADBw4s9CCJiIiIPjVff/01fv/9d/z222+KMiEEQkNDsXjxYpQuXRrGxsaYPXs2duzYgezsbABA3759YW5uDplMhhkzZuDKlSuIi4tTrKNDhw5o1qwZJBIJjI2NFeV+fn6IiorC48ePsXHjRgQHB+cbn3zc9YsXL2Bqaor69esryuPi4nDz5k3o6enhf//7H6ysrArxyOiG1knz2LFj8cUXX2DAgAGws7MDADRs2BAdO3ZEr169MGTIkEIPkoiIiOhTY2FhgfHjx2PixImKshcvXiA1NRX169eHpaUlLC0tUa1aNUilUjx79gw5OTmYMGEC3NzcYG5ujvLlywMAYmNjFesoV66c2u0ZGBigW7du+O6773D27Fn4+/vnG9+6dev+f3t3Hhdluf9//DXDvikguIKipKiVmluuuVKa2gn04IILHjUts1XNApVcykxP5r501FJMczezLLeTpaVltngyI01USBk3FEG2+f3hz/k2YY2jA0P4fj4ePB7c931d13zuedxNby6vuW9+/vln6tatS6NGjSxLO/r160enTp3o06cPlSpV4rnnnuPq1au3+W443y093GTmzJk888wzbNu2DZPJRGBgIB06dKBmzZqOrk9ERETkjjVixAjefPNNNm/eDEBQUBBeXl4cPHjQEoh/b9myZaxbt45PPvmE6tWrk5GRgb+/v9XdzozGP58zjYuLo3HjxgwZMgRPT8+/rK1mzZqsWLGCgoICNm3aRExMDKdPnyYgIIBx48Yxbtw4UlJS6NKlC+Hh4ZY1z39XtxSa4dq3LB977DFH1iIiIiIiv+Pl5cW4ceMsa5uNRiNDhw7l2WefZe7cuVSqVIkzZ86wd+9e/vGPf3Dp0iU8PDwoV64cWVlZJCQk2PV6DRs2ZPv27dSpU8dm2+XLl/PQQw8RHBxMmTJlMJvNuLq6snPnToKCgqhbty6+vr64urri4uJyS+dfkti9PGPVqlW8/vrrNzw2bdo0Vq9efdtFiYiIiMg1gwYNsloTPGXKFOrVq0erVq3w8/OjRYsWlps09O/fnxo1alClShXq1KlDkyZN7H69du3aUbFiRZvttm7dyt13342vry/Dhw9nxYoV+Pn58dtvv9GjRw/Kli1LnTp1aNasGYMGDbK7jpLGYP79fP1NuO+++xg4cCBPPfVUoWNz5sxh8eLFfP311w4rsDj8fp2PozV4f1qRjS2lz8FuI2+577mRFRxYiZR2gdNO33LfqZO/cGAlUtqNjm9mu9EtCgoKKrKx/yhi2Tgy83IcPu7JgVMcPqYUDbtnmo8cOcI999xzw2N169blyJEjt12UiIiIiEhJYndo9vT05PTpG89QpKWl4ep6y8ukRURERERKJLtDc5s2bZgyZUqhJ/9lZmYydepU2rZt66jaRERERERKBLunhV955RWaN29OeHg4PXr0oHLlyqSmprJmzRpycnJYuXJlUdQpIiIiIuI0dofm2rVrs3//fsaPH8/atWs5e/Ys5cqVIzIykvHjx3PXXXcVRZ0iIiIiIk5zSwuQ77rrLpKSkhxdi4iIiEiJ9M/whuTk5zm7DHEifWtPRERExIZJLR51dgniZHaH5oKCAt566y3WrFnDyZMnyc7OtjpuMBj45Zdf7Brz8uXLzJkzhwMHDuDl5UVMTAwPP/zwDds+8sgjeHh4YDAYgGu3uUtMTLQc37x5M2vWrCErK4tGjRrx5JNP4u3tbd9JioiIiIj8jt2h+YUXXmD69Om0bNmS1q1b4+7ufttFLFiwgPz8fJYsWUJaWhrjxo0jJCSEevXq3bD9G2+8QUhISKH933zzDStXrmTChAlUrFiRN954gwULFvDss8/edo0iIiJy50r9zxAKcrJtN7RTyOPLHD6mFA27Q3NSUhKJiYmMGzfOIQVkZ2fz+eefM2PGDLy9vQkPD6d9+/Zs27btT0Pzn9mxYwcdOnSgRo0aAMTGxvL888/zxBNP4OHh4ZB6RURE5M6TuX815uxLjh9Yoflvw+7QnJ2dTcuWLR1WwKlTpwCoWrWqZV+NGjXYsGHDn/ZJSEggPz+fmjVrEhcXZ+l7/PhxGjVqZGlXrVo1CgoKSE1NpXr16pb9JpPJ6tHZRqOR4OBgh52TyK1ycXFxdglyh9C1JsVF15qUFnaH5tjYWN5//306dOjgkAKys7Px8vKy2ufj40NWVtYN27/yyitERESQm5vLunXrGDduHHPnzsXb25vs7Gx8fHwsbQ0GA97e3oXGWrt2LYsWLbJsx8XF8eSTTzrkfERuR0BAwC33TXdgHVL63c61JmIPXWtSWtgdmps1a0ZCQgKnT58mMjISf3//Qm2io6NvejxPT89CoTYzM7NQkL7unnvuAcDNzY2+ffuyc+dOfvzxRxo1aoSnpydXrlyxan/lypVCY3Xv3p02bdpYto1GI+fPn7/pmkWKiq5DKS661qS4FOW1pkAuxcnu0NyvXz/g2lKIVatWFTpuMBjIz8+/6fGqVKkCwIkTJwgNDQXg2LFjVKtW7ab6X7+LBlxbjnHs2DFLID5+/DhGo5HKlStb9QkKCiIoKMiybTKZ7KpZpKjoOpTiomtNiouuNSkt7A7Nx44dc2gBnp6etGzZkqSkJJ566ilOnz7N9u3bGT16dKG2KSkp5ObmEhYWRl5eHmvXriUnJ4eIiAgA2rdvz7///W/atGlDhQoVSEpKolWrVvoSoIiIiIjcFqO9HapVq2bzx15Dhw4Frq0tfvnll4mNjaV+/foAxMTEcOjQIQAuXLjAtGnT6NWrF//617/46aefePnll/H19QXgvvvuo2fPnrz88svExcXh4uJiGVtERETk72LPnj20bt0af39//P39ady4MVu2bOHXX3/FYDBQq1Ytq/a5ubmUL18eg8Fg9QyNffv2ERkZSZkyZfDz8yMyMpJ9+/YBsHv3bnx9fS0/BoMBHx8fy/bu3btJTEzEzc3Nqp2vry9nzpwp1vejJLitJwJeuXKl0MNNAAIDA+0ax9fXlzFjxtzw2HvvvWf5vV69esybN+8vx+ratStdu3a16/VFRERESoqMjAy6dOnCjBkziI2NJTc3ly+//NLqTiQGg4E9e/bQokUL4NrD3QIDA0lP/7+vhX/55Zd06NCB8ePHs2bNGgAWLlxI+/bt2b59O61bt+by5cvA/92Y4dChQ4SFhVnG2L59O927d2flypXFcOYlm90zzWazmUmTJhESEoKfnx/BwcGFfkRERETk1hw5coTc3FwGDBiAq6srXl5etG3bltatW1vaDBgwgKVLl1q2ly5dyoABA6zGGTVqFH369GHUqFGULVuWsmXLMmrUKHr37n3DZbDy1+wOzW+88QbTp09n+PDhmM1m4uPjGTduHLVq1SIsLMzqVm4iIiIiYp9atWrh4eFBbGwsmzdvtpo9vq5v376sX7+e7Oxs0tPT2bNnD1FRUZbjV65c4fPPP6dXr16F+vbu3ZvPPvvsT2/vKzdmd2j+z3/+w8svv2z5C+XRRx9l/PjxHDp0iDp16pCcnOzwIkVERETuFGXKlOGzzz7Dw8OD4cOHU7FiRdq3b88vv/xiaVO+fHmaN2/Ohg0bWL58OdHR0Xh6elqOnz9/noKCgkJ3EAOoVKkSBQUFnDt37qbqWbdunWVttb+/v+UGDHcau0Pzr7/+SoMGDXBxccHNzY0LFy5cG8hoZPjw4Vb/VCAiIiIi9qtTpw6LFy/m+PHj/PLLL7i5uVlu+3tdXFwcS5cu5e233yYuLs7qWEBAAEajkdTU1EJjp6WlYTQab/o+19HR0Vy4cMHy89NPP93yef2d2R2ay5UrZ1k0XrVqVQ4cOGA5lp6eXujhIiIiIiJy68LCwhgxYgTff/+91f5u3bpx4MABsrOzad68udUxb29vWrRoYXVDhetWrVpFixYt8Pb2LtK6Sxu7757RsmVL9u/fz8MPP0yfPn1ITEzkt99+w83NjUWLFjns8doiIiIid6LDhw/z/vvv07NnT0JDQ0lPT+ett94qFIzd3Nz4+OOP//R5FK+99hqRkZHUrFmTIUOGYDab+c9//sPy5cvZtm1bcZxKqWJ3aE5MTOTUqVMAvPTSS1y4cIF3332XrKwsIiMjmTVrlsOLFBEREblT+Pn58dVXX/Hmm29y4cIFypQpQ/v27Zk3bx5Xr161atugQYM/HadFixZs376dhIQEEhMTAbj//vvZvn07zZo1u+l61q5da3kmxnV79+7l3nvvvfmTKgUMZrPZ7OwinM1kMhXZ2A3en1ZkY0vpc7DbyFvue25kBQdWIqVd4LTTt9x36uQvHFiJlHaj428+nNkrKCioyMb+o5+H+WPOvuTwcWst1WPG/y7sXtMsIiIiInKnuanlGY888shND2gwGNi4ceMtFyQiIiIiUtLcVGjOyMjAYDAUdS0iIiIiIiXSTYXmXbt2FXEZIiIiIiIll9Y0i4iIiIjYYPct565LTk7myJEjZGdnFzoWHR19W0WJiIiIlCQ1519wdgniZHaH5oyMDKKjo9m5cycA1+9Y9/s1z/n5un2KiIiIiJQedi/PeOGFF0hLS2P37t2YzWbWr1/Prl27GDRoENWrV+eLL3T/ThEREREpXeyeaf7oo4+YPHky999/PwCVK1emSZMmPPDAA4wcOZLp06ezcuVKhxcqIiIi4iz/nrqLnBzH/0v6mIQODh9TiobdofnMmTOEhobi4uKCj48PZ8+etRzr3Lkz3bt3d2iBIiIiIs527lwWOVe1/PROZvfyjNDQUMtjp2vWrMmmTZssx/bs2YOnp6fjqhMRERERKQHsnmmOjIxk27ZtREVF8eyzzzJgwAC+/PJL3N3d2bdvH88//3xR1CkiIiIi4jR2h+bXXnuNK1euANCvXz98fX1Zs2YNWVlZzJ49m6FDhzq8SBERERERZ7I7NHt7e+Pt7W3ZjoqKIioqyqFFiYiIiIiUJLf9RMALFy6wf/9+jh075oh6RERERMSJDAYDhw8fdnYZAOzatYuKFSs6uwzAjtD8wQcfMHjwYAYMGMDmzZuBa0s1KleuTLNmzbjrrrvo3LkzmZmZRVasiIiIiDhOWFgYH330kbPL+Fu4qeUZK1asoG/fvtSoUQNfX1+SkpJISEhgypQpPPbYY9StW5fvv/+et956i9dee40JEyYUdd0iIiIi4mS5ubm4ubk5u4xicVMzzdOnT+df//oXycnJHDx4kDfffJNJkyYxfvx4Zs6cybBhw5gzZw7x8fG89957RV2ziIiISKkVFhbGq6++yr333kvZsmXp3r07Fy5cAKB3795UqlSJsmXL0rp1a77//ntLvy1btlC3bl38/PyoXLkyr7zyCgAmk4muXbsSEBBAYGAgLVu25OrVq/Tr14+UlBSioqLw9fUlPj7eMtann35K7dq1KVu2LH379iUnJwf4v+US//73v6lcuTKPPPKI5bXr1atH2bJladasGfv27bM6n9/PZs+fP5+2bdtatnfs2GGpu3///kRHR5OYmGj1nsyaNYtKlSpRvnx5Xn/9dce80Xa6qdB85MgRevXqZdmOjY2loKCA1q1bW7V74IEHOH78uGMrFBEREbnDLF26lI0bN3Ly5EmuXr3K008/DcCDDz7ITz/9xJkzZ2jUqBG9e/e29Bk0aBALFizg0qVLHDp0iAcffBC4NvkZEhLCmTNnOH36NFOnTsXFxYVly5ZRtWpV1q9fz+XLl5k8ebJlrHXr1rFnzx5+/vln9u7dy/Llyy3HTCYTx48f5+jRo6xbt46ff/6ZHj168Oqrr3L27FkGDRpE586dOX/+vM3zPHfuHI8++igJCQmcP3+eTp068f7771u1MZlMpKamcvz4cTZv3kx8fDzJycm39f7eipsKzZmZmZQpU8ay7efnB2B1Fw0ALy8vy18iIiIiInJrnnzySWrUqIGfnx+TJ09m5cqVFBQUMHDgQMqUKYOHhweJiYkcOnTI8nRmNzc3/ve//5GRkUFAQACNGze27E9LS+PXX3/Fzc2Nli1b4ur61yt0x4wZQ2BgIOXLl6dLly4cOHDAcsxsNvPqq6/i6emJl5cXq1at4qGHHqJLly64uroyZMgQQkND+eCDD2ye5+bNm6lVqxZ9+vTB1dWVPn36UL9+fas2RqORCRMm4O7uTtOmTalduzYHDx609y29bTf9RUCDwXBT+0RERETk9lStWtXye7Vq1cjJySE9PZ0xY8YQHh5OmTJlCAsLA7A8qXnt2rV8+OGHVKtWjdatW7N7924ARo0axV133UWnTp2oXr06EydOxGw2/+Xr//6OFd7e3ly+fNmyXa5cOauJ01OnTllquS4sLIxTp07ZPM/U1FRCQ0Ot9v1xOzAw0Grd9B/rKS43fZ/mPn364OXlZbWvZ8+eVo/NzsrKclxlIiIiIneolJQUq9/d3Nz46KOPWLduHZ988gnVq1cnIyMDf39/SwBu0qQJGzZsIC8vj3nz5hEdHU16ejp+fn5Mnz6d6dOn87///Y/27dvTsGFDunTpcksToEaj9ZxrlSpVrGaiAX799Vd69OgBgK+vr+XBeAC//fab5ffKlStz4sQJq74nTpwoNNtcEtzUTPOAAQNo1aoVjRo1svwMGDCAli1bWu1r1aoV/fv3L+qaRUREREq1uXPncuzYMS5dukRCQgI9e/YkMzMTDw8PypUrR1ZWFgkJCZb2OTk5JCUlcfHiRVxdXfH19cXFxQW4tgQiOTkZs9lMmTJlcHFxsRyrUKECv/zyy23VGhMTw9atW9m6dSt5eXksXryYlJQUHn74YQDuu+8+VqxYQU5ODv/73/9YunSppW+XLl04cuQIq1atIi8vj1WrVvHtt9/eVj1F5aZmmpcsWVLUdYiIiIjI/9e/f38eeeQRUlJSaN++PW+++Sbu7u5s3bqVKlWqUK5cOSZOnGjVZ9myZYwYMYK8vDxq1qzJypUrAfj5558ZMWIE6enplC1blkGDBtGpUycAXnzxRZ566ini4+MZMWJEoTFvRq1atVi5ciUjR44kJSWFiIgIPvjgAwIDAwGYOHEiffr0ITAwkKZNm9K/f38+/fRT4NpSj3Xr1jFixAgGDx5MVFQUnTp1wsPD43beviJhMNta1HIHuL4WqCg0eH9akY0tpc/BbiNvue+5kRUcWImUdoHTTt9y36mTv3BgJVLajY5vVmRjBwUFFdnYf5Qw5kNyruY7fNypb3QttC8sLIz58+dbgu2dpn79+owaNYq+ffs6uxQrt/0YbRERERGRW7V161bOnTtHbm4u8+fP5+jRoyXyD4ab/iKgiIiIiIij7d+/n9jYWLKzs6lZsyYbNmwo1n9FuFkKzSIiIiIlyK+//ursEopVQkKC1ZcaSyqFZsDd3b1ELjiXO8/1BwfdinMOrENKv9u51kTsoWtNSovbCs1ms5m0tDTKly9v88kyJVlOTo6eZCglwqVLl5xdgtwhdK1JcSnKa00TXlKcbinpbt26lfHjx/PNN9+Qn5/Pvn37aNiwIY899hht2rQhNjbW0XWKiIiIOM2kKZ2dXYI4md13z3j33Xd5+OGHqVatGjNnzqSgoMByLDw8XPd0FhEREZFSx+7QPHHiRJ555hlWrVrF4MGDrY7dfffd/PDDDw4rTkRERESkJLA7NB89etTyWMQ/8vHx4eLFi7ddlIiIiIhISWJ3aK5YsSKHDx++4bHvvvuOatWq3XZRIiIiIiIlid2huU+fPiQmJrJ9+3bLPoPBwA8//MDUqVNL3CMPRURERERul913z0hMTOTQoUNERkZSrlw5ADp37kx6ejpdu3ZlzJgxDi9SRERERMSZ7A7N7u7ubNy4kZ07d/LJJ59gMpkIDAykY8eOdOzYsShqFBERERFxqlt+Ikm7du1o166dI2sRERERESmR7A7NKSkpf3rMaDRStmxZPTJTREREREoVu0NzWFgYBoPBZptnnnmGESNG3HJhIiIiIiIlhd2hecmSJYwdO5bw8HD+8Y9/UL58ec6cOcP69es5duwYo0aN4ssvv+TZZ58FUHAWERERkb89u0PzZ599xkMPPcSiRYus9j/zzDMMHjyYgwcPsnz5csqWLcvcuXMVmkVERETkb8/u+zSvXr2amJiYGx7r2bMn69atA6Br164cO3bs9qoTERERESkB7A7NLi4ufPvttzc8dvDgQYxGo6Wdt7f37VUnIiIiIlIC2L08IzY2lrFjx5KTk0O3bt0IDg4mPT2djRs3MmnSJIYOHQrA119/TZ06dRxesIiIiIhIcbM7NE+fPh1XV1cmTZrE2LFjLfs9PDwYPnw4r732GgAtW7bkoYceclylIiIiIiJOYndodnNz49///jdjx47l+++/Jy0tjUqVKnHvvfcSEBBgaffAAw84tFAREREREWe55ScCBgQEKBiLiIiIyB3hlkNzcnIyR44cITs7u9Cx6Ojo2ypKRERERKQksTs0Z2RkEB0dzc6dOwEwm80AVk8JzM/Pd1B5IiIiIiLOZ/ct51544QXS0tLYvXs3ZrOZ9evXs2vXLgYNGkT16tX54osviqJOERERERGnsTs0f/TRR8THx3P//fcDULlyZR544AEWLlzIo48+yvTp0x1epIiIiIiIM9kdms+cOUNoaCguLi74+Phw9uxZy7HOnTvz0UcfObRAERERERFnszs0h4aGYjKZAKhZsyabNm2yHNuzZw+enp6Oq05EREREpASw+4uAkZGRbNu2jaioKJ599lkGDBjAl19+ibu7O/v27eP5558vijpFRERERJzG7tD82muvceXKFQD69euHr68va9asISsri9mzZ1seoy0iIiIiUlrYFZpzcnL48MMPadCgAUFBQQBERUURFRVVJMWJiIiIiJQEdq1pdnd3JzY2lhMnThRVPSIiIiIiJY7dXwSsXbu2QrOIiIiI3FHsDs2vvvoqkyZN4uuvvy6KekREREREShy7vwg4evRoTCYTTZs2JSgoiPLly1s9QttgMPDtt986tEgREREREWeyOzQ3atSIxo0bF0UtIiIiIiIlkt2heenSpUVQhoiIiIhIyWX3mubfM5vNpKamkpeX56h6RERERERKnFsKzVu3bqVZs2Z4enoSGhrKd999B8Bjjz1GUlKSQwsUEREREXE2u5dnvPvuu/Tt25cePXowcOBAHn/8ccux8PBwlixZQmxsrF1jXr58mTlz5nDgwAG8vLyIiYnh4YcfLtTu8OHDvPvuuyQnJwMQERHB4MGDqVy5MgDff/89CQkJeHh4WPr06NGDmJgYe09TRERERMTC7tA8ceJEnnnmGaZPn05+fr5VaL777rt544037C5iwYIF5Ofns2TJEtLS0hg3bhwhISHUq1fPql1mZiYdO3Zk9OjRuLu7k5SUxKRJk5g7d66lTdmyZXnnnXfsrkFERERE5M/YvTzj6NGjN5wFBvDx8eHixYt2jZednc3nn39O37598fb2Jjw8nPbt27Nt27ZCbRs1akTr1q3x8fHBzc2NRx99lJMnT5KRkWHvaYiIiIiI3DS7Z5orVqzI4cOH6dChQ6Fj3333HdWqVbNrvFOnTgFQtWpVy74aNWqwYcMGm31/+OEHAgICKFOmjGXfpUuX6N+/P25ubjRs2JD+/fvj5+dn1c9kMmEymSzbRqOR4OBgu+oWKQouLi7OLkHuELrWpLjoWpPSwu7Q3KdPHxITE6lduzZt27YFrj3Q5IcffmDq1KlWyzVuRnZ2Nl5eXlb7fHx8yMrK+st+v/32GwsWLGDo0KGWfSEhIbz55puEhIRw7tw55s6dy4wZMxg7dqxV37Vr17Jo0SLLdlxcHE8++aRddYsUhYCAgFvum+7AOqT0u51rTcQeutaktLA7NCcmJnLo0CEiIyMpV64cAJ07dyY9PZ2uXbsyZswYu8bz9PQsFJAzMzMLBenfM5lMjBs3jh49etCqVSvL/oCAAMt/nEFBQTz22GMMGzaMq1evWn05sHv37rRp08aybTQaOX/+vF11ixQFXYdSXHStSXEpymtNgVyKk92h2d3dnY0bN7Jz504++eQTTCYTgYGBdOzYkY4dO9pdQJUqVQA4ceIEoaGhABw7duxPl3mcPXuW+Ph4HnzwQf7xj3/85dhGoxGz2YzZbLbaHxQURFBQkGXbZDKRn59vd+0ijqbrUIqLrjUpLrrWpLSwOzSbzWYMBgPt2rWjXbt2t12Ap6cnLVu2JCkpiaeeeorTp0+zfft2Ro8eXajt2bNneemll2jbti09evQodPy7776jQoUKlC9fngsXLrBw4UIaNGiAp6fnbdcpIiIiIncuu0NzSEgIPXv2pHfv3jRp0sQhRQwdOpTZs2cTFxeHt7c3sbGx1K9fH4CYmBjGjx/P3Xffzccff0xaWhrr169n/fr1lv5z5swhODiYo0ePMmPGDC5duoSPjw8NGzZkwIABDqlRRERERO5cBvMf1y7Y8Pzzz7N69WpOnTpF9erV6dOnD7169aJu3bpFVWOR+/2dNBytwfvTimxsKX0Odht5y33PjazgwEqktAucdvqW+06d/IUDK5HSbnR8syIb+/dLLUWKmt33aZ4+fTopKSns3LmTBx98kAULFnDvvffSoEEDXnvtNY4fP14UdYqIiIiIOI3dofm6Bx54gLlz55KamsqHH37Ifffdx5QpUwgPD3dkfSIiIiIiTnfLofk6s9lMTk4OV69eJS8vD4PB4Ii6RERERERKjFsKzWazme3btzNkyBAqVKjAI488wrEVAuHhAAAZ+ElEQVRjx3jllVc4efKko2sUEREREXEqu++e8fTTT/Pee+9x+vRp7rnnHkaOHEnv3r0JCwsrgvJERERERJzP7tC8efNm/vWvf9G7d2/uueeeoqhJRERERKREsTs0//LLL395PDk5mbvuuuuWCxIRERERKWlu+4uAAGfOnGHmzJncf//9REREOGJIEREREZESw+6Z5usuX77MunXrSEpKYseOHeTn59OkSRNmzZrlyPpERERERJzOrtCcl5fHli1bSEpKYvPmzWRnZ1OlShUKCgpYvXo13bt3L6o6RURERESc5qZC86effkpSUhJr1qzh/PnzBAUFMXDgQGJjY6lTpw6BgYEEBwcXda0iIiIiIk5xU6G5bdu2GAwGOnTowHPPPUdkZCQuLi4AXLx4sUgLFBERERFxtpsKzQ0aNODgwYPs2rULo9HImTNniIqKws/Pr6jrExERERFxupu6e8aBAwf48ccfeeGFF0hOTiYuLo4KFSoQExPDxo0b9ehsERERESnVbvqWcxEREUycOJHk5GT27NnDoEGD+PTTT4mLiwPgzTff5NNPPy2qOkVEREREnOaW7tPcrFkzZs2aRWpqKlu2bCE2NpZPPvmEdu3aUaNGDUfXKCIiIiLiVLd8n2YAo9FIp06d6NSpE1lZWWzcuJEVK1Y4qjYRERERkRLBIU8EBPDy8qJXr15s2rTJUUOKiIiIiJQIDgvNIiIiIiKllUKziIiIiIgNCs0iIiIiIjYoNIuIiIiI2KDQLCIiIiJig0KziIiIiIgNCs0iIiIiIjYoNIuIiIiI2KDQLCIiIiJig0KziIiIiIgNCs0iIiIiIjYoNIuIiIiI2KDQLCIiIiJig0KziIiIiIgNCs0iIiIiIjYoNIuIiIiI2KDQLCIiIiJig0KziIiIiIgNrs4uoCRwd3fHw8PD2WWI4Ofnd8t9zzmwDin9budaE7GHrjUpLRSagZycHHJycpxdhgiXLl1ydglyh9C1JsWlKK81TXhJcdLyDBERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAaFZhERERERGxSaRURERERsUGgWEREREbFBoVlERERExAZXZxfgaJcvX2bOnDkcOHAALy8vYmJiePjhh51dloiIiIj8jZW60LxgwQLy8/NZsmQJaWlpjBs3jpCQEOrVq+fs0kRERETkb6pULc/Izs7m888/p2/fvnh7exMeHk779u3Ztm2bs0sTERERkb+xUjXTfOrUKQCqVq1q2VejRg02bNhg1c5kMmEymSzbRqOR4ODg4ilS5C+4uLg4uwS5Q+hak+Kia01Ki1IVmrOzs/Hy8rLa5+PjQ1ZWltW+tWvXsmjRIst2XFwcTz75ZJHUdHLglCIZV+SPApbmO7sEuUNMfaOrs0sQESl2pSo0e3p6FgrImZmZhYJ09+7dadOmjWXbaDRy/vz5YqlRrs06lClThoyMDPLzFfSk6Ohak+Kia805AgICnF2C3EFKVWiuUqUKACdOnCA0NBSAY8eOUa1aNat2QUFBBAUFWbZNJpM+5JwgPz9f77sUC11rUlx0rYmUXqXqi4Cenp60bNmSpKQkrly5wrFjx9i+fTsdOnRwdmkiIiIi8jdWqmaaAYYOHcrs2bOJi4vD29ub2NhY6tev7+yyRERERORvrNSFZl9fX8aMGePsMkRERESkFClVyzNERERERIqCQrOIiIiIiA0KzSIiIiIiNig0i4iIiIjYoNAsIiIiImKDQrOIiIiIiA0KzSIiIiIiNig0i4iIiIjYoNAsIiIiImKDQrOIiIiIiA0KzSIiIiIiNhjMZrPZ2UXIncVkMrF27Vq6d+9OUFCQs8uRUkzXmhQXXWsipZ9mmqXYmUwmFi1ahMlkcnYpUsrpWpPiomtNpPRTaBYRERERsUGhWURERETEBpfExMREZxchdx4vLy8aN26Mt7e3s0uRUk7XmhQXXWsipZu+CCgiIiIiYoOWZ4iIiIiI2KDQLCIiIiJig0Kz/C0MHz6cgwcPOrsMEREAHnnkEU6ePOnsMgD4/vvv6d+/v7PLECn1XJ1dgJQcL730Ej/++CNz5syhcuXKAJw8eZInnniCTZs2Ofz1tm/fzptvvsmDDz7Ik08+adl/9OhRnnnmGWrVqsW0adMAmDNnjsNfX0qmH3/8kbfffpvjx48DUKlSJfr06UNoaChDhgyhcuXKzJ8/39I+Ly+PgQMHcvHiRdasWYO7uzsAR44cYdmyZRw5cgSAWrVq0a9fP2rVqsWhQ4d4+eWXLWNkZ2fj4eGBwWAAYPz48Xz77besXr0aNzc3q/oWLlyIv79/kb4HUrIMHjyYxx9/nEaNGjm7FBFxIoVmseLt7c2KFSsYOXJkkb5OXl4eABUrVmTv3r0MGTIEDw8P4FqYrlKlSpG+vpRMV65cYcKECQwePJi2bduSl5fHkSNHMBr/7x/FDAYDP/74I3Xq1AFg//79+Pr6cvHiRUubn376ibFjx9KrVy/GjBkDwNatW0lISGDixIncfffdvPfeewDk5OTQo0cPZs+eTYUKFSxjfPvtt7Ro0YJRo0YVx6lLKZaXl4erq/53K/J3p/+KxUqXLl3YuHEjR48epUaNGlbHcnNzWbFiBbt37yYrK4v77ruPYcOG4evrC8Drr7/ODz/8wNWrVwkLC2PYsGGEhYUBMGPGDNzc3Dh//jzfffedZWa5TJkyVK1alS+++II2bdqQl5fH7t276dq1K/v27bO89u9neo4cOcL8+fM5deoU7u7utGzZkmHDhpGTk8Ps2bP5+uuvKSgooGLFisTHx+uRtn8jp06dIj8/nw4dOgDg4uLCvffeC8Dp06cBaNeuHdu3b7eE5u3bt9O+fXuWLVtmGWfJkiU88MADREdHW/ZFR0eTmprK0qVLefXVV4vrlKSIDB48mIceeohPP/2U9PR06tevz4gRI/D19f3Lz6KvvvqKJUuWYDKZ8PLy4uGHHyYmJoaMjAxmzJjBjz/+iMFgICQkhMmTJzNr1izS09N59dVXMRqNdOvWjX79+gHwww8/MHnyZM6fP0+TJk146qmncHNz4/vvv+f1118nOjqaDRs2EBYWRmJiIl999RVvv/026enphISE8Nhjj1GrVi3L+fx+NvvDDz9k9+7dvPLKK8C1P+IWLlyIyWSiWbNmZGVlERYWRp8+fSzvyebNm1m9ejUFBQVERUVZXf8icvu0plmsBAQE0LVrV5YvX17o2DvvvMPRo0eZNm0aixcvxtXVlQULFliO33fffcybN49ly5YRHh5uWVpx3X//+18eeeQRVq1axf3332/Z3759e3bs2AFcmzWsXr065cqV+9MaFy1aRLdu3Vi1ahWLFi2ibdu2AOzYsYPMzEwWL17M8uXLefLJJ/H09Lydt0OKWZUqVXBzc2P69Ons37/favb4unbt2vHFF1+Qk5PDxYsXOXz4MM2bN7ccv3r1KocPH6Z169aF+rZu3Zoff/yRq1evFul5SPHYsWMH8fHxLF68mNzcXBYtWgT89WfRrFmzeOKJJ1i1ahWzZ8/mvvvuA2D9+vUEBQWxbNky3nnnHeLi4jAajTz33HMEBwfz4osv8t5771kCM8DevXuZOnUq8+fP56effmLXrl2WYxkZGaSnp7Nw4UJefPFFUlNTmTJlCv379ycpKYnIyEhefvllLl++bPM8L126xCuvvEJMTAwrVqygYcOG7N+/36pNRkYGZ8+e5a233mLs2LEsX76c1NTU23l7ReQPFJqlkOjoaH766ScOHTpk2Wc2m9m6dSuDBw/G398fDw8PYmNj+fzzz8nPzwegY8eOeHt74+bmRu/evUlJSSEjI8MyRpMmTahXrx4Gg8GyFAOgadOmHD16lLNnz1pmDf+Ki4sLaWlpXLx4EU9PT2rXrm3Zf+nSJVJTU3FxcSE8PNwyCy5/D97e3kyZMgVXV1fmz5/PgAEDiI+PJy0tzdKmbNmy1K5dmy+++IJdu3bRvHlzq3XHly9fpqCg4IZ/eAUGBlJQUHBTQQWuhaLevXtbfh5//PHbP0lxmC5dulCxYkW8vb3p168fu3fvpqCg4C8/i1xcXDhx4gRXrlzB19eXmjVrAuDq6sq5c+c4c+YMrq6u1K1bFxcXl798/R49euDn54e/vz+NGzfml19+sTrev39/3N3d8fDwYPfu3TRs2JAmTZrg4uLCQw89RFBQUKHweyP79++nSpUqtGnTBhcXF9q0aUP16tWt2hiNRmJjY3Fzc6NWrVpUqVKFY8eO2fN2iogNWp4hhfj4+BAdHc0777zDiBEjALh48SLZ2dmMHj3aqq3BYODChQv4+/uzfPlyPv/8cy5evGhZg5qRkUGZMmUA/nSZhKurK61bt2bDhg0cPnyY0aNHs3v37j+t76mnnmLFihUMHz6c4OBgYmJiaN68Oe3atePs2bNMmzaNy5cv88ADD9C/f/9CX+SSki00NJSnn34auLYkY+7cubzxxhs8//zzljbt27dn69atXLhwoVCQ9fX1xWg0cvbsWUJCQqyOnTt3DqPReNN/TDVv3lxrmkuw33+mBAcHk5eXx8WLF9m0adOffhZdnzF+++23qVatGv369ePuu+8mKiqKd999l8TEREvw7tmzp+XLoTcSEBBg+d3Dw4Pz589btv38/KwmB86ePUv58uWt+pcvX55z587ZPM9z584V+vz847avr6/VumkPDw+ys7Ntji0iN0+hWW6oa9euvP/++5ZZkDJlyuDu7s6MGTOsvix13c6dO9m7dy8TJkygQoUKXLlyhd69e1u1+f2Xuf6oQ4cOPPfcczz44IOWux/8mcqVKzNy5EgKCgrYt28fU6dO5Z133sHX15devXrRq1cv0tPTmTBhAhUrVqRLly638A5ISVChQgW6du1aaKlP06ZNmTdvHr6+vtSuXduy3hmuhYXatWvz+eefU79+fat+n332GbVr17YKM/L3ZTKZLL+np6fj6urKgQMH/vKzqGbNmsTHx5Ofn8+HH37Iq6++yvLly/H29mbQoEEMGjSIlJQUEhISCA8Pp0mTJrdU2x/Ddrly5QrNRJ85c4YWLVoA4OnpabVs6PcBPDAw0Opcr5/79XXaIlI8tDxDbsjDw4NevXqxZs0a4Frg7dSpE//5z38sMyMXLlzgiy++ACArKws3Nzf8/PzIycm54ZrovxIeHs7EiROtvtTyZ3bu3GmZQfLy8sJsNmM0Gvnuu+/49ddfyc/Px9PTE6PR+JdBXUqekydPsm7dOtLT0zGbzVy4cIGPP/6YiIgIq3aurq68/PLLvPTSSzccZ8CAAezcuZP169dz5coVMjMz2bBhA7t27SIuLq4YzkSKw5YtW/jtt9+4cuUKSUlJtGrViqtXr/7pZ1Fubi67du0iMzMTFxcXy+cEXFsCkZqaitlsxtvb2+rzw9/fn99+++22am3VqhXffPMNBw4cID8/n08++YT09HQaN24MQI0aNfj000/Jzc0lJSXF8j0PgMaNG3Pq1Cl2795Nfn4+u3fv1tILESfQTLP8qcjISNavX8+lS5eAa0Fk9erVjBkzhosXL1K2bFlat25Ns2bNaNeuHQcOHGDgwIH4+fkRGxtr9+vVq1fvptp98803LF68mKtXrxIUFMTzzz+Pt7c358+fZ968eZw9exYPDw+aN29OZGSk3XWI83h5eZGcnMz7779PZmYm3t7e3HvvvTz++OPk5uZatf3j3V1+r06dOkycOJGkpCRWrlwJXLtP88SJEy1r4G/Gnj17iImJsdo3depUzfCVEO3atWPy5Mmkp6dTr149hgwZYplt/rPPop07d7Jw4ULy8/OpXLmyZflNamoqCxYsICMjA29vbyIjIy13sujRoweLFi1i+fLldOnShb59+9pda5UqVRg1ahRLliwhPT2dKlWqMG7cOPz8/ACIjY1l+vTpxMbGUrNmTdq1a2f5Xsn1ZSULFy5k9uzZNGvWjIYNG2rpmUgxM5jNZrOzixAREbHHnf7AkaeeeoqoqCjatWvn7FJE7hj6t2sREZES7sCBA1y6dIm8vDw+/PBDTp8+fcf+wSDiLFqeISIiUsL9/PPPTJ8+ndzcXCpVqsRLL71kuTORiBQPLc8QEREREbFByzNERERERGxQaBYRERERsUGhWURERETEBoVmEREREREbFJpFRERERGxQaBaRv5SYmIjBYLD8BAUF0apVK7Zs2eLs0orM0qVLMRgMmEwmZ5ciIiIlhEKziNjk5eXF3r172bt3LwsXLiQnJ4du3bqxZ88eZ5cmIiJSLPRwExGxyWg00qxZM8t28+bNqVKlCm+//TYtWrRwYmUlm9lsJicnBw8PD2eXIiIit0kzzSJit0qVKhEcHExKSopln9lsZtq0adSqVQsPDw9q1KjBG2+8YdXv5MmTxMTEUKFCBTw9PalevTrPPvus5fjhw4fp1asXoaGheHt7U7duXaZPn05BQYGlza+//orBYOCdd95hyJAh+Pv7ExwczLRp0wBYuXIlERERlClThujoaC5cuGDpu2vXLgwGA1u2bCE6OhofHx8qVarEK6+8YvOcr169yksvvUS1atXw8PCgTp06rFixwqpNXFwc99xzD1u2bKF+/fp4eHiwadMm+95cEREpkTTTLCJ2u3z5MufOnSM8PNyy7+mnn+att94iPj6e+++/nz179vDCCy/g5eXFsGHDAOjfvz+pqanMnDmTChUqkJKSwldffWUZ49SpU0RERBAbG4ufnx8HDx5k/PjxZGZmMm7cOKsaEhIS+Oc//8nq1avZsGEDo0aNwmQy8d///pepU6eSkZHBiBEjGD16NAsXLrTq+9hjj9G7d2/WrVvHtm3biI+PJzAw0FLnjcTExPDZZ58xfvx46tSpw5YtW+jbty8BAQF07tzZ0i41NZWnn36ahIQEQkNDCQ0Nva33WkRESgiziMhfGD9+vNnHx8ecm5trzs3NNaekpJj79OljDgwMNB85csRsNpvNycnJZoPBYF6wYIFV31GjRpkrVqxozs/PN5vNZrOPj4955syZN/W6BQUF5tzcXPPkyZPNlSpVsuw/duyYGTD37NnTsi8vL89coUIFs4+Pj9lkMln2P//882Z/f3/L9s6dO82AuV+/flav1bdvX3NISIilziVLlpgBc3p6utlsNpt37NhhBsxbt2616vfPf/7T3KRJE8v2gAEDzID5yy+/vKlzFBGRvw8tzxARmzIzM3Fzc8PNzY2qVauyatUqli1bRs2aNQHYtm0bAN27dycvL8/y06FDB3777TdOnDgBQMOGDZk2bRrz5s0jOTm50OtkZ2czfvx47rrrLjw8PHBzcyM+Pp60tDQuX75s1bZjx46W311cXKhRowYNGjSgXLlylv21atXiwoULhfpGRUVZbUdHR3Py5ElOnjx5w/P/+OOPCQwMpH379oXO75tvviE/P9/SNigoiKZNm9p8T0VE5O9FoVlEbPLy8mL//v18+eWXLF++nEqVKtGvXz/S0tIAMJlMmM1mgoKCLOHazc2NTp06AVhC86pVq+jQoQPx8fHUrFmT2rVrs27dOsvrvPDCC7z++usMGTKELVu2sH//fhISEoBrgfr3/P39rbbd3d1vuO9GfcuXL3/D7evn80cmk4lz585ZnZubmxvDhg0jLy/Pqt8fxxYRkdJBa5pFxCaj0Ujjxo0BaNq0KbVr16Zp06ZMmDCBefPmERgYiMFg4LPPPrME1d+LiIgArn2BcPHixbz11lt8/fXXTJo0iZ49e/LTTz9Ro0YNVq9ezdChQ3nhhRcsfT/44AOHn8+ZM2duuF2pUqUbtg8MDCQ4OPhP7039+6BsMBgcVKWIiJQkCs0iYrdGjRrRu3dvlixZwvjx4+nQoQMAZ8+epVu3bjb7G41GmjRpwqRJk9i0aRPJycnUqFGDrKwsq9Cdn5/PypUrHV7/+vXrrZZorFu3jsqVKxMSEnLD9h07dmTq1Km4u7tTr149h9cjIiIln0KziNySsWPH8u677zJjxgymTJnC8OHD6devH6NGjeL+++8nNzeXI0eOsHPnTjZs2MDFixd56KGH6NevHxEREeTm5jJz5kz8/f1p2LAhAJGRkSxatIi6desSHBzMnDlzuHr1qsNr37FjB6NGjSIyMpJPPvmE5cuXM2fOHIzGG69Yi4yMpFu3bnTq1InRo0dTr149MjMzOXToEMnJybz11lsOr1FEREoWhWYRuSURERH07t2befPm8eKLLzJz5kwiIiJYsGABEyZMwMfHh4iICGJiYgDw9PTk3nvvZdasWaSkpODl5UXjxo35+OOPCQoKAmDWrFkMGzaMESNG4O3tTVxcHFFRUQwZMsShtS9YsIAFCxYwd+5c/Pz8mDhxIk888cRf9lmzZg1Tpkxh7ty5HD9+nLJly3LPPfcwcOBAh9YmIiIlk8FsNpudXYSISHHYtWsX7dq1Y//+/ZY12iIiIjdDd88QEREREbFBoVlERERExAYtzxARERERsUEzzSIiIiIiNig0i4iIiIjYoNAsIiIiImKDQrOIiIiIiA0KzSIiIiIiNig0i4iIiIjYoNAsIiIiImKDQrOIiIiIiA3/D7I0FI3hkI/fAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884547166951)>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(ggplot(temp, aes(x='param_resampler', y='mean_test_balanced_accuracy', fill='param_resampler'))\n",
    " + geom_bar(stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + scale_fill_brewer('qualitative', 2)\n",
    " + geom_text(aes(label='mean_test_balanced_accuracy'), format_string='{:.2f}')\n",
    " + ggtitle(f'Average Balanced Accuracy by Resampler')\n",
    " + labs(fill=f'Resampler')\n",
    " + xlab('Resampler')\n",
    " + ylab(f'Average Balanced Accuracy'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resampler + Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:40.422624Z",
     "start_time": "2020-04-16T15:13:39.129719Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884550063342)>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results.groupby(['param_resampler', 'param_classifier'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()\n",
    "temp.reset_index(inplace=True)\n",
    "\n",
    "(ggplot(temp, aes(x='param_resampler', y='mean_test_f1_score', fill='param_classifier'))\n",
    " + geom_bar(position='dodge', stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + geom_text(aes(label='mean_test_f1_score'), position=position_dodge(width=0.9), format_string='{:.2f}')\n",
    " + ggtitle(f'Average F1 Score by Resampler and Classifier')\n",
    " + labs(fill=f'Classifier')\n",
    " + xlab('Resampler')\n",
    " + ylab(f'Average F1 score'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.065998Z",
     "start_time": "2020-04-16T15:13:40.424618Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884545924818)>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results.groupby(['param_classifier'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()\n",
    "temp.reset_index(inplace=True)\n",
    "\n",
    "(ggplot(temp, aes(x='param_classifier', y='mean_test_f1_score', fill='param_classifier'))\n",
    " + geom_bar(stat=\"identity\")\n",
    " + ylim(0,1)\n",
    " + geom_text(aes(label='mean_test_f1_score'), format_string='{:.2f}')\n",
    " + ggtitle('Average F1 Score by Classifier')\n",
    " + labs(fill='Classifier')\n",
    " + xlab('Classifier')\n",
    " + ylab(f'Average F1 score'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classifier + n_estimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.662689Z",
     "start_time": "2020-04-16T15:13:41.068991Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223371884546480871)>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results.groupby(['param_classifier', 'param_classifier__n_estimators'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()\n",
    "temp.reset_index(inplace=True)\n",
    "\n",
    "(ggplot(temp, aes(x='param_classifier__n_estimators', y='mean_test_f1_score', color='param_classifier'))\n",
    " + geom_line()\n",
    " + geom_point()\n",
    " + ylim(0,1)\n",
    " + ggtitle('Average F1 Score by Classifier and Number of Estimators')\n",
    " + labs(color='Classifier')\n",
    " + xlab('Number of Estimators')\n",
    " + ylab('Average F1 score'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## `AdaBoostClassifier` + `max_depth`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.698592Z",
     "start_time": "2020-04-16T15:13:41.669672Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean_test_precision</th>\n",
       "      <th>mean_test_recall</th>\n",
       "      <th>mean_test_specificity</th>\n",
       "      <th>mean_test_f1_score</th>\n",
       "      <th>mean_test_balanced_accuracy</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>param_classifier</th>\n",
       "      <th>param_classifier__base_estimator</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">AdaBoostClassifier</th>\n",
       "      <th>DecisionTreeClassifier(max_depth=1)</th>\n",
       "      <td>0.738579</td>\n",
       "      <td>0.436288</td>\n",
       "      <td>0.996339</td>\n",
       "      <td>0.548524</td>\n",
       "      <td>0.716314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTreeClassifier(max_depth=2)</th>\n",
       "      <td>0.759336</td>\n",
       "      <td>0.443006</td>\n",
       "      <td>0.996670</td>\n",
       "      <td>0.559510</td>\n",
       "      <td>0.719838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTreeClassifier(max_depth=3)</th>\n",
       "      <td>0.755862</td>\n",
       "      <td>0.441223</td>\n",
       "      <td>0.996619</td>\n",
       "      <td>0.557159</td>\n",
       "      <td>0.718921</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                        mean_test_precision  \\\n",
       "param_classifier   param_classifier__base_estimator                           \n",
       "AdaBoostClassifier DecisionTreeClassifier(max_depth=1)             0.738579   \n",
       "                   DecisionTreeClassifier(max_depth=2)             0.759336   \n",
       "                   DecisionTreeClassifier(max_depth=3)             0.755862   \n",
       "\n",
       "                                                        mean_test_recall  \\\n",
       "param_classifier   param_classifier__base_estimator                        \n",
       "AdaBoostClassifier DecisionTreeClassifier(max_depth=1)          0.436288   \n",
       "                   DecisionTreeClassifier(max_depth=2)          0.443006   \n",
       "                   DecisionTreeClassifier(max_depth=3)          0.441223   \n",
       "\n",
       "                                                        mean_test_specificity  \\\n",
       "param_classifier   param_classifier__base_estimator                             \n",
       "AdaBoostClassifier DecisionTreeClassifier(max_depth=1)               0.996339   \n",
       "                   DecisionTreeClassifier(max_depth=2)               0.996670   \n",
       "                   DecisionTreeClassifier(max_depth=3)               0.996619   \n",
       "\n",
       "                                                        mean_test_f1_score  \\\n",
       "param_classifier   param_classifier__base_estimator                          \n",
       "AdaBoostClassifier DecisionTreeClassifier(max_depth=1)            0.548524   \n",
       "                   DecisionTreeClassifier(max_depth=2)            0.559510   \n",
       "                   DecisionTreeClassifier(max_depth=3)            0.557159   \n",
       "\n",
       "                                                        mean_test_balanced_accuracy  \n",
       "param_classifier   param_classifier__base_estimator                                  \n",
       "AdaBoostClassifier DecisionTreeClassifier(max_depth=1)                     0.716314  \n",
       "                   DecisionTreeClassifier(max_depth=2)                     0.719838  \n",
       "                   DecisionTreeClassifier(max_depth=3)                     0.718921  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results[cv_results['param_resampler']=='passthrough'].groupby(['param_classifier', 'param_classifier__base_estimator'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## `GradientBoostingClassifier`, `XGBClassifier` + `learning_rate`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.735220Z",
     "start_time": "2020-04-16T15:13:41.701808Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean_test_precision</th>\n",
       "      <th>mean_test_recall</th>\n",
       "      <th>mean_test_specificity</th>\n",
       "      <th>mean_test_f1_score</th>\n",
       "      <th>mean_test_balanced_accuracy</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>param_classifier</th>\n",
       "      <th>param_classifier__learning_rate</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">GradientBoostingClassifier</th>\n",
       "      <th>0.025</th>\n",
       "      <td>0.790585</td>\n",
       "      <td>0.395179</td>\n",
       "      <td>0.997518</td>\n",
       "      <td>0.526909</td>\n",
       "      <td>0.696348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.050</th>\n",
       "      <td>0.780465</td>\n",
       "      <td>0.423388</td>\n",
       "      <td>0.997177</td>\n",
       "      <td>0.548966</td>\n",
       "      <td>0.710282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.100</th>\n",
       "      <td>0.778204</td>\n",
       "      <td>0.434859</td>\n",
       "      <td>0.997062</td>\n",
       "      <td>0.557939</td>\n",
       "      <td>0.715961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">XGBClassifier</th>\n",
       "      <th>0.025</th>\n",
       "      <td>0.754734</td>\n",
       "      <td>0.404196</td>\n",
       "      <td>0.996884</td>\n",
       "      <td>0.526422</td>\n",
       "      <td>0.700540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.050</th>\n",
       "      <td>0.776283</td>\n",
       "      <td>0.406060</td>\n",
       "      <td>0.997226</td>\n",
       "      <td>0.533204</td>\n",
       "      <td>0.701643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.100</th>\n",
       "      <td>0.783911</td>\n",
       "      <td>0.419787</td>\n",
       "      <td>0.997256</td>\n",
       "      <td>0.546763</td>\n",
       "      <td>0.708522</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                            mean_test_precision  \\\n",
       "param_classifier           param_classifier__learning_rate                        \n",
       "GradientBoostingClassifier 0.025                                       0.790585   \n",
       "                           0.050                                       0.780465   \n",
       "                           0.100                                       0.778204   \n",
       "XGBClassifier              0.025                                       0.754734   \n",
       "                           0.050                                       0.776283   \n",
       "                           0.100                                       0.783911   \n",
       "\n",
       "                                                            mean_test_recall  \\\n",
       "param_classifier           param_classifier__learning_rate                     \n",
       "GradientBoostingClassifier 0.025                                    0.395179   \n",
       "                           0.050                                    0.423388   \n",
       "                           0.100                                    0.434859   \n",
       "XGBClassifier              0.025                                    0.404196   \n",
       "                           0.050                                    0.406060   \n",
       "                           0.100                                    0.419787   \n",
       "\n",
       "                                                            mean_test_specificity  \\\n",
       "param_classifier           param_classifier__learning_rate                          \n",
       "GradientBoostingClassifier 0.025                                         0.997518   \n",
       "                           0.050                                         0.997177   \n",
       "                           0.100                                         0.997062   \n",
       "XGBClassifier              0.025                                         0.996884   \n",
       "                           0.050                                         0.997226   \n",
       "                           0.100                                         0.997256   \n",
       "\n",
       "                                                            mean_test_f1_score  \\\n",
       "param_classifier           param_classifier__learning_rate                       \n",
       "GradientBoostingClassifier 0.025                                      0.526909   \n",
       "                           0.050                                      0.548966   \n",
       "                           0.100                                      0.557939   \n",
       "XGBClassifier              0.025                                      0.526422   \n",
       "                           0.050                                      0.533204   \n",
       "                           0.100                                      0.546763   \n",
       "\n",
       "                                                            mean_test_balanced_accuracy  \n",
       "param_classifier           param_classifier__learning_rate                               \n",
       "GradientBoostingClassifier 0.025                                               0.696348  \n",
       "                           0.050                                               0.710282  \n",
       "                           0.100                                               0.715961  \n",
       "XGBClassifier              0.025                                               0.700540  \n",
       "                           0.050                                               0.701643  \n",
       "                           0.100                                               0.708522  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results[cv_results['param_resampler']=='passthrough'].groupby(['param_classifier', 'param_classifier__learning_rate'])['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### f1-score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.747188Z",
     "start_time": "2020-04-16T15:13:41.738213Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'classifier': AdaBoostClassifier(algorithm='SAMME.R',\n",
      "                   base_estimator=DecisionTreeClassifier(ccp_alpha=0.0,\n",
      "                                                         class_weight=None,\n",
      "                                                         criterion='gini',\n",
      "                                                         max_depth=2,\n",
      "                                                         max_features=None,\n",
      "                                                         max_leaf_nodes=None,\n",
      "                                                         min_impurity_decrease=0.0,\n",
      "                                                         min_impurity_split=None,\n",
      "                                                         min_samples_leaf=1,\n",
      "                                                         min_samples_split=2,\n",
      "                                                         min_weight_fraction_leaf=0.0,\n",
      "                                                         presort='deprecated',\n",
      "                                                         random_state=None,\n",
      "                                                         splitter='best'),\n",
      "                   learning_rate=1.0, n_estimators=100, random_state=None), 'classifier__base_estimator': DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
      "                       max_depth=2, max_features=None, max_leaf_nodes=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=1, min_samples_split=2,\n",
      "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
      "                       random_state=None, splitter='best'), 'classifier__n_estimators': 100, 'resampler': 'passthrough'}\n"
     ]
    }
   ],
   "source": [
    "print(cv_results[cv_results['rank_test_f1_score']==1]['params'].values[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.770127Z",
     "start_time": "2020-04-16T15:13:41.750181Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_test_precision</th>\n",
       "      <th>mean_test_recall</th>\n",
       "      <th>mean_test_specificity</th>\n",
       "      <th>mean_test_f1_score</th>\n",
       "      <th>mean_test_balanced_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.759668</td>\n",
       "      <td>0.44419</td>\n",
       "      <td>0.996667</td>\n",
       "      <td>0.560527</td>\n",
       "      <td>0.720429</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_test_precision  mean_test_recall  mean_test_specificity  \\\n",
       "15             0.759668           0.44419               0.996667   \n",
       "\n",
       "    mean_test_f1_score  mean_test_balanced_accuracy  \n",
       "15            0.560527                     0.720429  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results[cv_results['rank_test_f1_score']==1]\n",
    "temp[['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### balanced accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.780100Z",
     "start_time": "2020-04-16T15:13:41.773118Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'classifier': GradientBoostingClassifier(ccp_alpha=0.0, criterion='friedman_mse', init=None,\n",
      "                           learning_rate=0.1, loss='deviance', max_depth=3,\n",
      "                           max_features=None, max_leaf_nodes=None,\n",
      "                           min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                           min_samples_leaf=1, min_samples_split=2,\n",
      "                           min_weight_fraction_leaf=0.0, n_estimators=100,\n",
      "                           n_iter_no_change=None, presort='deprecated',\n",
      "                           random_state=None, subsample=1.0, tol=0.0001,\n",
      "                           validation_fraction=0.1, verbose=0,\n",
      "                           warm_start=False), 'classifier__learning_rate': 0.025, 'classifier__n_estimators': 120, 'resampler': SMOTE(k_neighbors=5, n_jobs=None, random_state=None, sampling_strategy='auto')}\n"
     ]
    }
   ],
   "source": [
    "print(cv_results[cv_results['rank_test_balanced_accuracy']==1]['params'].values[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-04-16T15:13:41.803039Z",
     "start_time": "2020-04-16T15:13:41.783093Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_test_precision</th>\n",
       "      <th>mean_test_recall</th>\n",
       "      <th>mean_test_specificity</th>\n",
       "      <th>mean_test_f1_score</th>\n",
       "      <th>mean_test_balanced_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0.199325</td>\n",
       "      <td>0.958312</td>\n",
       "      <td>0.908746</td>\n",
       "      <td>0.330003</td>\n",
       "      <td>0.933529</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_test_precision  mean_test_recall  mean_test_specificity  \\\n",
       "46             0.199325          0.958312               0.908746   \n",
       "\n",
       "    mean_test_f1_score  mean_test_balanced_accuracy  \n",
       "46            0.330003                     0.933529  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = cv_results[cv_results['rank_test_balanced_accuracy']==1]\n",
    "temp[['mean_test_precision', 'mean_test_recall', 'mean_test_specificity', 'mean_test_f1_score', 'mean_test_balanced_accuracy']]"
   ]
  }
 ],
 "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": "3.6.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}