6. Отток клиентов банка.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Содержание<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Подготовка-данных\" data-toc-modified-id=\"Подготовка-данных-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Подготовка данных</a></span></li><li><span><a href=\"#Исследование-задачи\" data-toc-modified-id=\"Исследование-задачи-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Исследование задачи</a></span></li><li><span><a href=\"#Борьба-с-дисбалансом\" data-toc-modified-id=\"Борьба-с-дисбалансом-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Борьба с дисбалансом</a></span></li><li><span><a href=\"#Тестирование-модели\" data-toc-modified-id=\"Тестирование-модели-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Тестирование модели</a></span></li><li><span><a href=\"#Чек-лист-готовности-проекта\" data-toc-modified-id=\"Чек-лист-готовности-проекта-5\"><span class=\"toc-item-num\">5&nbsp;&nbsp;</span>Чек-лист готовности проекта</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Отток клиентов"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Из «Бета-Банка» стали уходить клиенты. Каждый месяц. Немного, но заметно. Банковские маркетологи посчитали: сохранять текущих клиентов дешевле, чем привлекать новых.\n",
    "\n",
    "Нужно спрогнозировать, уйдёт клиент из банка в ближайшее время или нет. Вам предоставлены исторические данные о поведении клиентов и расторжении договоров с банком. \n",
    "\n",
    "Постройте модель с предельно большим значением *F1*-меры. Чтобы сдать проект успешно, нужно довести метрику до 0.59. Проверьте *F1*-меру на тестовой выборке самостоятельно.\n",
    "\n",
    "Дополнительно измеряйте *AUC-ROC*, сравнивайте её значение с *F1*-мерой.\n",
    "\n",
    "Источник данных: [https://www.kaggle.com/barelydedicated/bank-customer-churn-modeling](https://www.kaggle.com/barelydedicated/bank-customer-churn-modeling)\n",
    "\n",
    "Признаки:\n",
    "- RowNumber — индекс строки в данных\n",
    "- CustomerId — уникальный идентификатор клиента\n",
    "- Surname — фамилия\n",
    "- CreditScore — кредитный рейтинг\n",
    "- Geography — страна проживания\n",
    "- Gender — пол\n",
    "- Age — возраст\n",
    "- Tenure — количество недвижимости у клиента\n",
    "- Balance — баланс на счёте\n",
    "- NumOfProducts — количество продуктов банка, используемых клиентом\n",
    "- HasCrCard — наличие кредитной карты\n",
    "- IsActiveMember — активность клиента\n",
    "- EstimatedSalary — предполагаемая зарплата\n",
    "- Exited — факт ухода клиента"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Подготовка данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt \n",
    "from sklearn.utils import resample\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import RandomizedSearchCV, GridSearchCV\n",
    "from sklearn.metrics import f1_score, confusion_matrix, roc_auc_score, ConfusionMatrixDisplay, precision_score\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RowNumber</th>\n",
       "      <th>CustomerId</th>\n",
       "      <th>Surname</th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>15634602</td>\n",
       "      <td>Hargrave</td>\n",
       "      <td>619</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101348.88</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>15647311</td>\n",
       "      <td>Hill</td>\n",
       "      <td>608</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>41</td>\n",
       "      <td>1.0</td>\n",
       "      <td>83807.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112542.58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>15619304</td>\n",
       "      <td>Onio</td>\n",
       "      <td>502</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>8.0</td>\n",
       "      <td>159660.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113931.57</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>15701354</td>\n",
       "      <td>Boni</td>\n",
       "      <td>699</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>39</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93826.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>15737888</td>\n",
       "      <td>Mitchell</td>\n",
       "      <td>850</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>43</td>\n",
       "      <td>2.0</td>\n",
       "      <td>125510.82</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084.10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   RowNumber  CustomerId   Surname  CreditScore Geography  Gender  Age  \\\n",
       "0          1    15634602  Hargrave          619    France  Female   42   \n",
       "1          2    15647311      Hill          608     Spain  Female   41   \n",
       "2          3    15619304      Onio          502    France  Female   42   \n",
       "3          4    15701354      Boni          699    France  Female   39   \n",
       "4          5    15737888  Mitchell          850     Spain  Female   43   \n",
       "\n",
       "   Tenure    Balance  NumOfProducts  HasCrCard  IsActiveMember  \\\n",
       "0     2.0       0.00              1          1               1   \n",
       "1     1.0   83807.86              1          0               1   \n",
       "2     8.0  159660.80              3          1               0   \n",
       "3     1.0       0.00              2          0               0   \n",
       "4     2.0  125510.82              1          1               1   \n",
       "\n",
       "   EstimatedSalary  Exited  \n",
       "0        101348.88       1  \n",
       "1        112542.58       0  \n",
       "2        113931.57       1  \n",
       "3         93826.63       0  \n",
       "4         79084.10       0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('/datasets/Churn.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на вид данных."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 14 columns):\n",
      " #   Column           Non-Null Count  Dtype  \n",
      "---  ------           --------------  -----  \n",
      " 0   RowNumber        10000 non-null  int64  \n",
      " 1   CustomerId       10000 non-null  int64  \n",
      " 2   Surname          10000 non-null  object \n",
      " 3   CreditScore      10000 non-null  int64  \n",
      " 4   Geography        10000 non-null  object \n",
      " 5   Gender           10000 non-null  object \n",
      " 6   Age              10000 non-null  int64  \n",
      " 7   Tenure           9091 non-null   float64\n",
      " 8   Balance          10000 non-null  float64\n",
      " 9   NumOfProducts    10000 non-null  int64  \n",
      " 10  HasCrCard        10000 non-null  int64  \n",
      " 11  IsActiveMember   10000 non-null  int64  \n",
      " 12  EstimatedSalary  10000 non-null  float64\n",
      " 13  Exited           10000 non-null  int64  \n",
      "dtypes: float64(3), int64(8), object(3)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Я решил провести небольшой эксперимент и заполнять не нулями, а предсказывать значения. Спойлер: показатели точности ухудшились, ниже я вернусь к заполнению пропусков нулями. Данный блок эксперимента удалять не стал :)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Имеем пропуски в Tenure (количество недвижимости у клиента). Пропуски скорее всего означают нулевые значения. #После замены пропусков скорректируем тип данных.\n",
    "\n",
    "#df['Tenure'] = df['Tenure'].fillna(0).astype(int)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Удалим столбцы не содержащие признаки для обучения модели и округлим значения."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['RowNumber', 'CustomerId', 'Surname', 'CreditScore', 'Geography',\n",
       "       'Gender', 'Age', 'Tenure', 'Balance', 'NumOfProducts', 'HasCrCard',\n",
       "       'IsActiveMember', 'EstimatedSalary', 'Exited'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns # для копипаста"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>619</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101349</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>608</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>41</td>\n",
       "      <td>1.0</td>\n",
       "      <td>83808</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112543</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>502</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>8.0</td>\n",
       "      <td>159661</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113932</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>699</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>39</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93827</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>850</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>43</td>\n",
       "      <td>2.0</td>\n",
       "      <td>125511</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CreditScore Geography  Gender  Age  Tenure  Balance  NumOfProducts  \\\n",
       "0          619    France  Female   42     2.0        0              1   \n",
       "1          608     Spain  Female   41     1.0    83808              1   \n",
       "2          502    France  Female   42     8.0   159661              3   \n",
       "3          699    France  Female   39     1.0        0              2   \n",
       "4          850     Spain  Female   43     2.0   125511              1   \n",
       "\n",
       "   HasCrCard  IsActiveMember  EstimatedSalary  Exited  \n",
       "0          1               1           101349       1  \n",
       "1          0               1           112543       0  \n",
       "2          1               0           113932       1  \n",
       "3          0               0            93827       0  \n",
       "4          1               1            79084       0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.drop(['RowNumber', 'CustomerId', 'Surname'], axis=1, inplace=True)\n",
    "df[['Balance', 'EstimatedSalary']] = df[['Balance', 'EstimatedSalary']].round(0).astype('int')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.duplicated().sum() # глянем на явные дубли."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Исследование задачи"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbMAAAG9CAYAAACf55j3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAACoIUlEQVR4nOydd3wVVfbAv+e1FNIgoZcVIXQF6Si6iIo0xb6WtaIgdvy5rqu7a9fVtaC4ClgW3bWsFVFAFkFXVKqKiHRFJKGEkkrIq+f3x0yS91JfCCGQ3O/nk0/ezD33nnumnTn3npkRVcVgMBgMhqMZR313wGAwGAyG2mKcmcFgMBiOeowzMxgMBsNRj3FmBoPBYDjqMc7MYDAYDEc9xpkZDAaD4ajHODODwWAwHPUYZ2ZARH4RkdPDltuKyM8i8nh99stgMBiixTgzQwQi0hz4FJinqnfWd38MBoMhGowzM5QgIinAf4HlwE1h62NEZIqIbLf/pohIjF02TEQyRORuEdljR3mXhdWdKSLTRGSBiOSLyP9E5Ddh5d3ssn0iskFELirTp/tExC8iBSKyX0RURFx22VARWWW3WyAiIREZFqb3obB2OouIlmn332HLz9ttd7aXW4nIf0Ukx27bLyL3VbLd7rPrnhu27gZ73bVh664RkXUiki0i84u3g4ik2bZfEL5N7d8OEXlPRKaU0ZkhIgfsvvmKbRGRP4jIe2VknxWRZyrp+y9h7RSISJGIfB5WriJyix2p7xGRv4uIwy67SkS+DJO905Y/3V6+VUR22u2uLt43Ye12Dlt+SERmhi2/Y9fNFZEvRKRnWFnJvhWRVBFZKyKT7OWmIvKxiOy2t/PHItIurO7n9vZqEbbu7TL7vuyxMzf8uDMcmRhnZigmAZgHuIBrNPI9Z/cAg4E+QG9gIPDnsPJWQBrQFrgSmCEiXcPKLwMetGVWAa8DiEgTYAHwBtACuBh4XkR6hNV1AG+pagLQk0ieAD4Akuzy7QdhNyLSBRhVZvVtQBBobbf9n2qaWQ9cG7Z8FbApTMc44G7gPKA5sBh4E0BV9wBjgKdE5MQy7T4JCHB72W4DI+2+PRK2/t/ASLFuTLAvwBcDr1XR97NUNcFu66YKys8F+gN9gXHANWUFRKQZcAuQE7b6I6ArkAg8b9sSLfOAdKzj4lvsY6aMzuJj9g1VfcFe7QD+CfwG6AAcAJ4rU/UnrOMUEUkDulTWCRE5FTi+Bv021BPGmRmKeQEoANoBJ5Upuwx4QFWzVHU3cD9weRmZv6iqV1X/B8wBwiOsOar6hap6sRzjEBFpD4wFflHVf6pqQFW/A94DLgyr6wF8VfTbiXVhrw2PYDnbsjiI/hz5BmglIu1EpC+wi0jnej3wqKquU9WArbNPcXSmqpuxLrCzsS7iiMgtWE7uMlUNldEXRwXbRVV3AF9Qug1HAntU9Zso7aiIx1R1n6r+CkwBLqlA5m7gFSA3rC8/q2rxsmA5pahQ1VdUNd8+Zu4DeotIcphIDDALWKeqD4XV26uq76lqoarmAw8Dvy3T/GuUHr9XAP+qqA8iIsDjwF+j7beh/jDOzFDMeuAs4E7gJRGJCytrA2wNW95qrysmW1X3V1G+rfiHqhYA++zy3wCD7KG8HBHJwXKcrcLqNgOyK+nzzcDZQJFdt00lcpUiIoOxoodXyxQ9CRQC+XbbF1E9/wSuxorQXipT9hvgmTA792Fd4NuGyZwB/ILlMJpj2ZcIdA5vSKwh3hRgdyX9eBX4vf3791Rysa4B28J+l9232A75IuDvZSuKyF1Y2/FB4OMyxd+GbY87wuo4ReRvIvKTiORhbROwIvtibgSaACeGH6siEi8i00Vkq133CyBFRJxhdXcDG0XkZCynVlnUehGwB1hUSbnhCMI4M0MxD6tqkaq+iHXxCo9UtmNdjIvpQGTU0dQeMqysvH3xD3toqJldvg34n6qmhP0lqOqksLpdgI0VdVhVV2BdXO9R1RQObpjxceBPqhos0/ZurKHAeXbbb0fR1r+BS4FTsaLTcLYBE8vYGqeqXwPYQ6tXYw13PoQ1xDkW+BMw3Y4SiukD5ANbKunHLOB4Eellt1FuiK6GtA/7XXbfgnWsPG5HQhGo6t+AeKxh17eLhz9t+hZvC6wh42IuxRrOPB1IBo6x14dvg6+Bk4EVWNFXMf+HdXMySFWTgFMqqAvWzcZUYLO9r8vitu36YwVlhiMQ48wMFXEdMEFEBtrLbwJ/FpHm9hzDX7Eu3OHcLyIe+253LPBOWNlosZI1PFgXiKWqug3rTr2LiFwuIm77b4CIdBeLcVhzNfMq6qRYySIdgKcP0s7hQEhVy0YMiMgxWBeyG6JtTFVzsKKzJ+2hxHCmAX8qTmQQkWQRudD+LcB04M/2hXUJsE9VN6jqTMALTLRlHVgR2ztlHXBYP4qAd7HmIpfbw4O14Q92YkV74FYi5w87A4Ps/kcgIj3CkibigBBQFIW+RCyb92I5wkcqkFlqb+NbgEtEZEhY3QNAjj2Pd28lOv6LNexZ2bFzOfC1qq6Oor+GIwDjzAzlUNWfsRzWP20H9BCwElgN/IB1EXgorMpOrKHA7VhRwPWquj6s/A2si8o+oB/2EJh9Jz8CK0Fhu93OY1jzISNtHZfZji8CEWmKdSG6rgLHUcwtYmX9ZWBFWYjIkrDy1ljDqhUxHfibqm6tpLxCVPVxVS07xIiqfoBl21v28NcaSpNOxgOKNedUEdcD94lIKyyneBnwe7EzELHmq34nYVmkWEONx1H7IUaAD7HmBFdhRZwvh5W1xHLC/grq3QxkYc2j3QNcZDva6ngNK+LOBNYCSysTtJNnbgZesYdfp2A5zj12vU8qqRdS1WuKI+MKaAr8JYq+Go4QxHyc01AbxEq3/reqtqukfCaQoap/rqj8cCMiv6jqMfXdj4PF3p4zVfXzMut/D7jsSA4R6YA1D9pKVfNqoU+BdDtBxWA4YjHPTRgaG19WL3JEsw9rCK4s+7HPZ3so8nasRxoO2pEZDEcTxpkZGhWq+vvqpY5cVLXs82bF6z+Akmf3dmEN0408jF0zGAAQkVew5s2zVLVXBeUCPAOMxsp0vUpVo35so1K9ZpjRYDAYDIcKETkF65nV1ypxZqOx5jlHYyUPPaOqg2qr1ySAGAwGg+GQoapfYA2HV8Y4LEenqroU6znA1rXVa5yZwWAwGA4nbYl8ED+DyJcHHBRmzqyekUmDD/s4b2HLtOqF6gBx1PatUweJ8/Dfs4m7fu4TQ4UVZcjXPc7kmHrRG8qv6k1ndai3oH62c/zjn9T6JKr1NWfasonAhLA1M1R1Rq3aPAQYZ2YwGAyGqLEdV22cVyaRb5VpZ6+rFWaY0WAwGBoR4pBa/R0CZgNX2G/5GQzk2i/IrhUmMjMYDIZGRF0P94vIm8AwIM1++869WO+6RFWnAXOxMhk3Y6XmX30o9BpnZjAYDI2IunZmqlrRJ4LCyxXrqweHFOPMDAaDoRFRb4lYdYyZMzMYDAbDUY+JzAwGg6EREflpvIaDcWYGg8HQiGiow4zGmRkMBkMjoqE6MzNnZjAYDIajHhOZGQwGQyOioUZmxpkZDAZDI8I4M8Nh4+XL72HscSeRlZ/NcQ9eVqHMMxfdzuieQyj0ebnqtQf5btsGAK4YPJo/j7IeqH9o3j95bencGul2j5qAI70/+L34Zk1Bd/xUTkZad8JzzmRwewhtWol/nv2atrgEPBf8EUlpiebswvfO36Bof1R6XWdOwJHeD/xe/B8+g+6sWK/77Ntsvd8QmG/rjU3AfcGdSHJLNHcX/ncfi0qv49gTcI24DsRBcNUCgkveixRwunCfPRlp1QkO5OP/4O9obhbEJeI+74842nQmuHpRaT+iQDqegOu08Yg4CK7+lOCy98vpdI25FUfLTuiBfPyzn4C83RCbiPucPyCtOhNa8xmBT1+MWmcxtdm3jh4n4R52KdK8Pd4Xb0e3b47O3t/0wTXsGnA4CK5ZSGjFB+XsdZ55C46Wx6IH8gnMfQrydiMtO+M8/Xq7ESG45D/oT8trZG99HFMA7rMn4ew2APxevG8/iWaW31bStjMxF/0fuGMIrl+Bf/YLVt0RV+DsOQQ0hBbk4Hv7STSvqq+p1JyG6syOiDkzEWklIm+JyE8i8o2IzBWRLgfZ1kwRucD+/ZKI9LB/311G7h4R+VFEVovIKhGp9cfhDhUzl8xh5NTJlZaP6jmE9BbtSb/3Qia88SgvXHInAE3jk7h3zHgGPTaegY9dw71jxpMSnxi1Xkd6f6RZG7zPTsD30XN4xtxQoZxn7I34PpqK99kJSLM2ODr3A8A19EJCW77HO3UCoS3f4xp6YXR6O/dDUtvge24i/o//gXvMpArl3KNvwP/xc/iem4ikhuu9gNCW1fj+MZHQltW4TrqgeqXiwDVyIv637sc3/SacPU9G0tpHiDj7nIEWFeB74XoCy2fjGn6lVRDwEfjf6wQWzozKvnCd7tMn4H/nQXwv34Kj+1AktV3ktjjudCjaj+/FGwiu/AjXsCusgqCPwOI3CXz+as10Frdby32rWVvx/ecRQlt/rJG9ruHX4Z/1MP5Xb8PRdSg0K2Nvz9PAW4D/nzcR+vZjnEMvt/Tt/ZXAG3cSeP0OAh88iOv060Giv1zVyzEFOLoNwJHWhqLHr8H33jN4zr2pQjnPuTfje+8Zih6/BkdaGxxd+wPg/9+7FD09iaIpNxJctxzX6RXfzNaGI+DdjHVCvTsz+xPaHwCfq2onVe0H/AloGSZzUBGkql6rqmvtxRJnJiJDsD7r3VdVjwdOJ/L7OjXmYPtYEYs3r2Lf/rxKy8f1PqUk4lq25UdS4hNolZTKmT0GsWDdcrIL88gpzGfBuuWM7DE4ar3OroMIfr8IAM3YALFNIKFppFBCU4iJs8qB4PeLcHYbXFI/sGohAIFVC0vWV4ej6+BSvZkbIKYyvfFWua3X0dVq39FlEMHvF9rrF5asrwppk47u24nm7IJQgODaxTi6DIzsV/oggqutfoXWfYXjmOOtAr8XzVgHgZp9fkRap6M5OyDX0hla9yWOzpE6nekDCa75zNK54WscHcJ0ZtZcZ0m7tdy3uicD3VuzF5tLq85ozs5Sezd8iaPTgAgZR6eBhNZ+DkBo0xIcHY6zCgI+0JDdeQ9ozb5aUh/HFICzxxAC31r1Qr+uR+ISILFZpFBiMyQ2ntCv6y1Tv12Is+eJVpm3sFTOE1tjuxsz9e7MgFMBv/0CSgBU9XvAKSKLRWQ2sFZEnCLydxFZYUdTE8FyhiLynIhsEJFPgRbF7YjI5yLSX0T+BsTZEdjrQGtgj6p6bX17VHW7XWeAiHwtIt+LyHIRSRSRWBH5p4j8ICLficiptuxVIjJbRBYBC0WkiYi8Ytf7TkTG1cUGa5vSnG3ZWSXLGdlZtE1pXun6aJGkVDRvT8my5u1FklIrkNlboYwkpEBBtlVQkG0tR6M3sYze/L1IYmo1MntKZA5GrySmovllbC2ns1mpTg2h3v0QF32kW05nQrNInRXYSUJqhE68hbXSWaK7lvv2oChjLwX7kISy9obJFNsba9krrdJxXTEF9+VPEVg4vdS5RUF9HFMAjuRUNGd3aZs5u3Ekp5aXyd1TqYz7zCuJvftfuE44Ff9//xWV3prQUCOzI2HOrBfwTSVlfYFeqrpFRCZgfSpggIjEAF+JyH+BE4CuQA+saG4t8Ep4I6p6l4jcpKp9AEQkAfiriGwEPgX+o6r/ExEP8B/gd6q6QkSSgAPArVYzepyIdAP+GzYM2hc4XlX3icgjwCJVvUZEUoDlIvKpqkY32N7QqK+bSnMz2yDQnZsIvHYbNGuL68ybCfzyHQTr56OYh/OY8s9/Ff/8V3Gd+jvcJ56Ff8G/D2n75g0g9cNyVd1i/x4BHF88HwYkA+nAKcCbqhoEtttRUpWoaoGI9ANOxooM/yMid2E51R2qusKWywMQkaHAVHvdehHZChQ7swWqWjxDOwI4W0TusJdjgQ7AunD9tmO2vtR6Skfo0YKakJmzm/ZNS+u0a9qCzJzdZObsZliXvhHrP9/4bZVtOQeMwdXvTABCmZuQpNKvUJe9U4fyd+vhMlqQYw3dFGRDQlN0f07levuPxtnX1rvd0lt8vbCipjJ68/dG9i0xrUSmJnoj2kssY2s5nfusfuXvBXEgMU3gQH61bVeqs2BfpM4K7KTAslMLLJ3ExB+0zkO5bw+KMvaS0MyyqwIZLdhXam9RGXv3ZYKvCEnrgO4qn8RRTH0dU64hZ+EaNNLSu20jEjYaIinNCeVG6g3l7kWS06qUAQh+t4iYax489M7sCI6uasORMMz4I9CvkrLwiEaAm1W1j/3XUVX/e7BKVTWoqp+r6r3ATcD5B9lU2T6eH9bHDqq6rmwFVZ2hqv1VtX9NHRnA7NWLuWLwaAAGdexJ7oECdubtZf7aZYzoPoiU+ERS4hMZ0X0Q89cuq7Kt4Io5eKfdgnfaLQTXL8HZe7hlSLuu1pBP8VBLMQXZ4D1glQPO3sMJbrB0BDcsw9XnNABcfU4rWV+h3pVz8c24Fd+MWwltWFqqt21VegutcltvaMNSAEIbl+PsfZq9/jRCG6u2GUC3b0KatUaSW4DDhbPHyYQ2RmbLhTYtx3m81S9H95MI/bK62nar1LljE9K0Ndg6Hd2HEtq8IkImuHkFzl6nWjq7nkjo1x8OWt+h3LcHg+7cbNmbZNvbdSj688oImdDPK3D0GAaAI30IoW1rrIKkFqUJH4nNkWZtrUzSquytp2MqsOQjiqbcSNGUGwn8uARXX6ueo0M39MB+yC+TjZi/Dy0qxNGhGwCuvqcRXLvE6mtamxIxZ48hhLJqNZVfIQ11mFG0nicY7QSQpcDL9ue4EZHjgXHAIFUda6+bgPVBtwtV1W8P82UCZwIT7bIWWMOM16nquyLyOXCHqq4UkWyghV23KxBS1U122w8BKcDtwHpKhxkTsYYZbwF6qup4W+8CrMjsEqC/qt5kt/MIkITldFVETlDV76q0f9LgcjvgjWseYFiXvqQlpLArbx/3fvwibqcVRE9fbKU2P3fxHYzsMZhCXxFXv/YQ39iTyVcPGcvdI62su4c/mcnMJXPK6SxsmVZuXTHu0ddbGV1+L74Pp5SkYMdc/yzeabdYfW7T2UrfdnkIbf4G/1x7ujMuEc+FdyHJzdHcLCs1/0BBqa1VnAiuUdfj6NTXSqOe/Qy6w9LrmfAMvhm3WvVbd8Y97rYSvYFPppfodV/wRyTJ0mulUZfqxVnxPZujUz9cZ4y30sa/X0jwq3dwnXIpoR2bCW1aDk437nGTkZbHQlE+/g+esBJGgJgbZ1hRhNNlZR++eR+6p/TCI+5KdB7bF+dwOzX/h4UEl76Lc+gl6M7NlmNzunGNuQ1Hy45oUQH+2U9aCRSAZ+J08MRZOr378b99P7o3I6L9UGHlw3C12beObkPwjJ4I8clQVEBo5xZ8//5r6SZOjqlQpxzTF9ewq63HH35cRGj5eziHXExo12bLsTnduEbegrSw7A3MfRpyd+Ho/lscA86FYABUCS57p8LU/FB+5QkxdXlMhQqq2M7n3Iizaz/wefG98xShjE0AxN72D4qmWJ/xcrRLx3PR/4HbQ3D9SvwfPm/17fI/42jeDlTR7F343p8aER3HP/5Jrb1JysNn1uqin3PP/CPSo9W7MwMQkTbAFKwIrQj4BZgFjAtzZg7gIeAsrAhoN3AOkIc1BHgG8CvgB16pwJk9BpwNfAs8ZddJAQJYXzydoKp7RGSAXRaH5chOt2VeAPrbv29X1c9E5CoinVmcbceJWFHvluL+V2p7Bc6srqnKmdUl9XZXV4kzq0sqc2Z1TVXOrC6pzJnVNVU5szrVW4Uzq0sOhTNr+ujIWl1zsv9U+z7UBUeEM2vMGGd2GDDOrM4xzuzwcCicWbPHRtXqmrPvj/OOSGd2pCeAGAwGg+EQciTPe9WGIyEBxGAwGAyGWmEiM4PBYGhENNTIzDgzg8FgaEQYZ2YwGAyGox7jzAwGg8Fw1NNQX2dlEkAMBoPBcNRjIjODwWBoRJhhRoPBYDAc9RhnZjAYDIajnobqzMycmcFgMBiOekxkZjAYDI0IRwMNYYwzMxgMhkaEs4Gm5htnVs/Uxxvs43ftOew6AQrTUqsXqgOcKYf/je6BrXmHXSeAxLvrRa/rN0n1ote7ene96I2/58p60XsocDbQOTPjzAwGg6ER0VAjswY6emowGAyGxoSJzAwGg6ERUQ/fqj0sGGdmMBgMjYiGOsxonJnBYDA0IowzMxgMBsNRT0PNZmygo6cGg8FgaEyYyMxgMBgaEc6GGZgZZ2YwGAyNCTPMaDAYDIajHqdIrf6qQ0RGisgGEdksIndVUN5BRD4Tke9EZLWIjD4UdhlnZjAYDIZDgog4gX8Ao4AewCUi0qOM2J+Bt1X1BOBi4PlDodsMMxoMBkMjoo6HGQcCm1X1ZwAReQsYB6wNk1Gg+GWeycD2Q6HYODODwWBoRNRxAkhbYFvYcgYwqIzMfcB/ReRmoAlw+qFQbJzZEYx71AQc6f3B78U3awq646dyMtK6E55zJoPbQ2jTSvzzZlgFcQl4LvgjktISzdmF752/QdH+KvW9fPk9jD3uJLLysznuwcsqlHnmotsZ3XMIhT4vV732IN9t2wDAFYNH8+dRVwPw0Lx/8trSuTWzdfREHF1sW99/umJb23TGc95kcHkIbVyJf+50ABw9h+IefimS1h7v9Mno9s1R6fxyaz6PfbmdYAjO69GUa/u1iCj3BUPc/WkGa7MOkBLr5O9ndqBtkocfdhVy/2eZgHWLecPAFpx2bHLN7D3nBpzdB4DPi/etJ9DM8n2WdunEXHwHuD0E163AP8sajXGPvBJnzyGgihbk4Hvr72jevup11sM2Xrwhm0c+3kIoBBcMaMF1w9pFlPsCIf749ibWZu4nJd7FU5d2oW3TWHyBEPfN+ok1GftxCNx9VkcG1nQb1+L8cfQ4CfewS5Hm7fG+eHtU9n6xZBMPP/kJoVCIC8f1ZcKVJ0eU//P1r3ln9rc4nQ6apTThkb+Mo23rFAD+PnUB//tqIwA3jP8to8/oVSNba0JtIzMRmQBMCFs1Q1Vn1KCJS4CZqvqkiAwB/iUivVQ1VJt+mTmzahCRc0RERaTb4dTrSO+PNGuD99kJ+D56Ds+YGyqU84y9Ed9HU/E+OwFp1gZH534AuIZeSGjL93inTiC05XtcQy+sVufMJXMYOXVypeWjeg4hvUV70u+9kAlvPMoLl9wJQNP4JO4dM55Bj41n4GPXcO+Y8aTEJ9bM1tQ2eKdch+/DqXjOurFiW8+6Ad+sZ/FOuQ5JbYMj3bJVs7bie/NhQlvXRK0zGFIe/mI7z4/tyIeXpjNvUy4/7SuKkHl/bTZJMU7mXt6Vy/uk8fSSnQB0bhbLWxd15t2L05l21jE88HkmgZBGb2+3ATjS2lL06NX43pmC5/xbKrb3/Jvxvf00RY9ejSOtLY5uAwDwf/YORU9eT9FTkwiuXYbrjN9Xr7OetvGDs39mxtU9+GhyH+Z8v4fNuwojZN5dsYvkOBfz/9CXK4a24Yl5WwF4Z8UuAGbf1oeXx/fgsTm/EKrJNq7l+aNZW/H95xFCW3+MztZgiAcen8tLz1zGnP/cyMfz17D556wIme5dW/PeqxP46I0bOHN4D/4+dQEAn3+5kbUbdjDr39fz9j+v4+V/f01BQVFFag4JtU0AUdUZqto/7C/ckWUC7cOW29nrwhkPvA2gqkuAWKDW38Iyzqx6LgG+tP8fNpxdBxH8fhEAmrEBYptAQtNIoYSmEBNnlQPB7xfh7Da4pH5g1UIAAqsWlqyvisWbV7Fvf+Xf4RrX+5SSiGvZlh9JiU+gVVIqZ/YYxIJ1y8kuzCOnMJ8F65Yzskf1+kps7T6Y4KowW+MqszW+1NZVi3B2H2LV2b0N3VP2fKmaH7IK6ZDsoX2yB7fTwaj0ZD7bEmn7Z1vyOLtbCgBndEpmWUYBqkqc24HLvrv1BhWo2Z2us9eJBL6xLmShX9cjcU0gsVmkUGIzJLYJoV/XAxD4ZgHOXidaZd4wh+CJxYoPq9FZD9t49bYCOqTG0b5ZLB6Xg9G901i0LjKCXLQum3F9rYj4zF6pLP0pF1Xlp6wDDLIjsdQED0lxLtZkFkStu7bnj+7JQPdGb+/qHzP5TbtmtG/bDI/bxZgRvVj4xYYImcH9OxIX6wGgz3Ht2JllHW+bt+ym/wm/weVyEh/noWvnlnyxJLrI9whkBZAuIh1FxIOV4DG7jMyvwGkAItIdy5nV+sN0xplVgYgkAEOx7iQuttc5ROR5EVkvIgtEZK6IXGCX9ROR/4nINyIyX0RaH7TupFQ0r/Qjmpq3F0lKrUBmb4UykpACBdlWQUG2tVxL2qY0Z1t26d1mRnYWbVOaV7o+WiQpFc0tPZY1d08UtpaXqQlZBQFaJZR+yLJlgptd+/2RMvv9tEqwLj4uh5DgcZJTFARg9c5CznljI+e9uYm/DmtT4tyiwZGciuZE2utITq1aJidSxj3qKmL/8jquvsPxf/JatTrrZRvneWmV7ClZbpnkYVeuL0JmV56X1in2NnYKibFOcgoDdGsdz2frsgkElYx9RfyYWcDOMnWrorbnT03ZtTuPVi1LP1DaskUSu3ZXfmP47uxvOWVIZwC6pbdk8ZLNHCjysS9nP8u+2VLi6OqCukzNV9UAcBMwH1iHlbX4o4g8ICJn22L/B1wnIt8DbwJXqWr0YXclmDmzqhkHfKKqG0Vkr4j0AzoCx2ClnbbA2mGviIgbmAqMU9XdIvI74GHgmrKNho85Tx17HOP7dah7S2p9qBjCOb5VPLMu7cLP+4q4Z2EGQzskEuM6fPeG/nkz8c+biWv4xbiHno1//r8Om+7DwXn9WvJT1gEu/Mf3tEmJoU+HRBrKs74fzvueNeu28+9p1hzz0MGd+WHtdi4e/zLNmjahz3HtcdShsXX9CRhVnQvMLbPur2G/1wInHWq9xplVzSXAM/bvt+xlF/COPVm5U0Q+s8u7Ar2ABWLdvTiBHRU1ao8xzwA4cN/YEjfjHDAGV78zAQhlbkKSSoeRy95FQvk7yXAZLcixhlEKsiGhKbo/p+bWlyEzZzftm5YmSbRr2oLMnN1k5uxmWJe+Ees/3/htlW05B47B1X8kAKHMjUhyaSQnyWlR2Fpepia0SHCxs6A0EttV4KdlE3ekTBM3Owt8tEpwEwgpBb4gKbHOCJljm8US73aweV8RPVvEV6rPddJZuAZZz4aGtm1AUiLtDeVG2hLK3Rspk1JeBiD47UJirn24QmdW79s4KSYimtqV56NlWKQG0DIphh05PlolxxAIKvlFQVLiXYgIfxrbsUTukhd+4Ji0uCr1Hcrzp6a0bJ7Ezl2l0dSurDxaNk8qJ/f18p+Y9s/F/HvaVXg8pZffSdecwqRrTgHg//78Lh07HHxEXB0N9a35ZpixEkSkGTAceElEfgH+AFxE5RMkAvyoqn3sv+NUdURNdAZXzME77Ra8024huH4Jzt7DrYbbdbXmSYqHDYspyAbvAasccPYeTnDDMqutDctw9TkNAFef00rW14bZqxdzxWDrgjyoY09yDxSwM28v89cuY0T3QaTEJ5ISn8iI7oOYv7ZqfcHlc/A+fzPe528muG4pzj5hthbtr8TWwlJb+wwnuG7pQdvSq0U8W3O9ZOT58AdDzNuUy7BjIi8+wzomMXt9DgALfsplYNsERISMPF9Jwsf2PB9bsr20SfSUVRFB4KuPKHpqEkVPTSKw5mtc/c4AwNGhG1q0H/LLZCPm70OL9uPoYOUdufqdQXDN1wBIWpsSMWevEwllbaMi6nsbH9cuga17DpCxrwhfIMTc7/dwavfIucFTuzflw2+tIer5a/YyuFMyIsIBX5BCnzWk+9WmHJwOoXPLym8W4NCePzW2tUcbftm2l22Z2fj8Aeb8dw3DT+4aIbN2ww7++ujHvPDEJaQ2SyjtdzBEdo41D7p+0042bN7FSYM6HVQ/osHpkFr9HanIIRiqbJDYQ4H9VHVi2Lr/AYuAAcDZQHOsYcYJWJOca4HLVXWJPezYRVWrTIcKj8zK4h59vZVd5ffi+3BKSXpwzPXP4p1mZcBJm85WarHLQ2jzN/jnTrMqxyXiufAuJLk5mptlpeYfsCbQ43ftqVDfG9c8wLAufUlLSGFX3j7u/fhF3E7r7nH64g8AeO7iOxjZYzCFviKufu0hvrETFK4eMpa7R14JwMOfzGTmkjnl2i9Mq/xu0z12kpU5V5w2XmzrDVPxPn9zqa3nTQZ3jJU2Psey1dF9CJ4x10OTZCgqILTjZ3yvlYxq4Gxe8R39F7/k8fiXOwgqnNu9KRP6t+C5Zbvo2SKOUzsm4Q2E+NOn21i/u4jkWCePj+hA+2QPH23I5uVvduNyCA6BiQPKp+YHtlY95+E+7yacXe208beeIJSxCYDY21+g6KlJll3t0vFc/AcrNX/9Cvwf/AMAz5V/wdG8PWgIzc7C9+4zJRGFxLsrVljH2zjmhBblFQL/W5/Nox9vIaTKef1bcv2p7Xh2wa/0apvA8B7N8Pqt1Px12/eTHO/iyUu60L5ZLJnZRVz7ylocIrRI8vDQ+Z1o2zS2XPve1ZXnDdTm/HF0G4Jn9ESIt+3duQXfv0vtjZtc/tGV/321kUee+oRgSDn/rBOYdM0pPDN9Eb26t+G0U7px1Y2vsvGnLJqnWo6sdatkpj15KV6vn3OvsB6BSGgSw/13jaV7l0qm25MvqbU3Oe/jK2t10X9/7KtHpEczzqwS7OHDx1T1k7B1twDdsaKwYVgPB4ott0BE+gDPYj3V7gKmqOqLVempypnVFZU5s7qmKmdWl1TmzOqS6pxZXVGVM6tLKnNmdU1VzqwuqciZHRYOgTO7cM5VtbrmvDNm5hHpzMycWSWo6qkVrHsWrCxHVS0QkVRgOfCDXb4KOOVw9tNgMBhqQl0ngNQXxpkdHB+LSArgAR5U1Z313B+DwWCIioaaAGKc2UGgqsPquw8Gg8FgKMU4M4PBYGhEHMkZibXBODODwWBoRJhhRoPBYDAc9ZgEEIPBYDAc9TTUyKyB+miDwWAwNCZMZGYwGAyNiDr+0nS9YZyZwWAwNCIcDXSY0Tgzg8FgaESYyMxgMBgMRz0N9DEzkwBiMBgMhqMfE5nVM1IPt0n19fb6+D0H/6HH2lDoTKte6BDjbNnksOsEkDIfDz1cFH2VWS96nWlVf+OsrjjweP182Tvu4Utq3YYZZjQYDAbDUY+jgY4zGmdmMBgMjYiGGpmZOTODwWAwHPWYyMxgMBgaEQ10lNE4M4PBYGhMNNRhRuPMDAaDoRFh3gBiMBgMhqOehhqZmQQQg8FgMBz1mMjMYDAYGhEmAcRgMBgMRz0N9eOcxpkZDAZDI8JEZgaDwWA46jEJIAaDwWAwHKGYyOwIxnXmBBzp/cDvxf/hM+jOn8rJSOtOuM++DdweQpu+ITB/hlUQm4D7gjuR5JZo7i787z4GRfur1ekePRFHl/7g9+J7/2l0RwU623TGc95kcHkIbVyJf+50ABw9h+IefimS1h7v9Mno9s3V6nv58nsYe9xJZOVnc9yDl1Uo88xFtzO65xAKfV6ueu1Bvtu2AYArBo/mz6OuBuChef/ktaVzq9VXzt5RE3Ck2/bOmlKxva074Tlnsr2NV+KfZ21jR4+TcA+7FGneHu+Lt0dlr3Q8Addp4xFxEFz9KcFl70cKOF24xtyKo2Un9EA+/tlPQN5uiE3Efc4fkFadCa35jMCnL9bIzi+35vPYl9sJhuC8Hk25tl+LiHJfMMTdn2awNusAKbFO/n5mB9omefhhVyH3f2a9EV+BGwa24LRjk2uk233ODTi7DwCfF+9bT6CZ5beTtEsn5uI7wO0huG4F/lnPW3VHXomz5xBQRQty8L31dzRvX7U65dgTcJ1+LeJwEFy1gODSCrbz2NtwtLa386wnIDcL4hJxn3sn0rozoR8WEfhvzbYzgHvMRBxdB1jH1HtPodsrOYfOv906pjaswD/HOodcI6/B2W0QBAPovh343ns6qvO2JjgaaAjTQM0qRURSRWSV/bdTRDLDlj313b/KcHTuh6S2wffcRPwf/wP3mEkVyrlH34D/4+fwPTcRSW2Do3M/AFxDLyC0ZTW+f0wktGU1rpMuqF5nen8ktQ3eKdfh+3AqnrNurFDOc9YN+GY9i3fKdZbOdEunZm3F9+bDhLauidrOmUvmMHLq5ErLR/UcQnqL9qTfeyET3niUFy65E4Cm8UncO2Y8gx4bz8DHruHeMeNJiU+MWi/Y9jZrg/fZCfg+eg7PmBsqlPOMvRHfR1PxPjsBaVa6jTVrK77/PEJo64/RKRQH7tMn4H/nQXwv34Kj+1AktV1kn447HYr243vxBoIrP8I17AqrIOgjsPhNAp+/WiMbAYIh5eEvtvP82I58eGk68zbl8tO+ogiZ99dmkxTjZO7lXbm8TxpPL9kJQOdmsbx1UWfevTidaWcdwwOfZxIIadS6Hd0G4EhrS9GjV+N7Zwqe82+pUM5z/s343n6aokevxpHWFke3AQD4P3uHoievp+ipSQTXLsN1xu+rVyoO3CMm4n/7AXwzbsbR4+Ty27n3GVBUgG/aJILLZ5du54CPwBdvEFg0M2obI9rt0h9Ja4v3qWvxzXoWz9k3VWzvuBvxzXoG71PXImltrRtIILT5O7zPTsI79UZCezJx/faig+pHVThFavV3pNLgnZmq7lXVPqraB5gGPF28rKq+Q6lLRA7Zx6QcXQcT/H4RAJq5AWKaQELTSKGEphATb5UDwe8X4eg62KrfZRDB7xfa6xeWrK8KZ/fBBFfZOjM2QFwVOjNsnasW4ew+xKqzexu6p2bftVq8eRX79udVWj6u9yklEdeyLT+SEp9Aq6RUzuwxiAXrlpNdmEdOYT4L1i1nZI/qbQzH2XVQ6TbO2ACxldkbV2rv94twdrP06J4MdG/09krrdDRnB+TuglCA0LovcXQeGNmn9IEE13wGQGjD1zg6HG8V+L1o5joI1PyQ/SGrkA7JHtone3A7HYxKT+azLZHb/LMteZzdLQWAMzolsyyjAFUlzu3AZWcMeIMK1Oxi5ux1IoFvFlj2/LoeiWsCic0ihRKbIbFNCP26HoDANwtw9jrRKvMWlsp5YrHiw6qRNulo9g7ICdvOXQZF9it8O6//GscxYds5Yx0E/DWys6Td7oMJfmedd7rNPqYSyxxTifY5ZI8wBL9biLO7dUyFNn8HoZD1e9t6JOnQf4vPIbX7O1Jp8M6sIkSkn4j8T0S+EZH5ItLaXv+5iDwmIstFZKOInGyvv0pEngur/7GIDLN/F4jIkyLyPTBERH5v118lItMP1sFJYiqat6dkWfP3Iomp1cjsKZGRhBQoyLYKCrKt5ep0JqWiubtL28vdgySllpfJK/3IpuaVlzmUtE1pzrbsrJLljOws2qY0r3R9TbBsCdt+eXujsLe8TNT6Epqh+VXvUxLC+qQh62IeV7OIsyxZBQFaJbhLllsmuNm1P/JinbXfT6sEa6DC5RASPE5yioIArN5ZyDlvbOS8Nzfx12FtSpxbNDiSU9GcyGPKkZxatUxOpIx71FXE/uV1XH2H4//ktWp1SkKzCs6d8g70UG9nAElKizyH8vaUc0iWTFj/csvLALj6jSC0cWWt+9RYaIzOTICpwAWq2g94BXg4rNylqgOB24B7o2ivCbBMVXsDe4HfASfZkWAQKDcRJCITRGSliKx8eeXW2tgSPdGPDBkMERzfKp5Zl3bhrQs78dI3u/EGQodVv3/eTIoevIzAt4twDz37sOquL1zDfgehIMHvPzvkbTuldn9HKo0xASQG6AUsEGv81wnsCCsvnin+BjgmivaCwHv279OAfsAKu+04IKtsBVWdAcwAKHrgrBI34+w/GmffMwEIbd9k3cHZZZKYiubvjWwnf2/EHZ0kppXIaEGONURWkA0JTdH9ORV23jlwDK7+Iy2dmRuR5NLoRpLTIqISKB+ZSFJ5mUNJZs5u2jctTVZo17QFmTm7yczZzbAufSPWf77x22rbcw4Yg6ufvY0zN0VuvzJRGFRkb3mZaNGCfUhi+P4qv08psPapFuwFcUBMPBzIPyh9xbRIcLGzoDQS21Xgp2UTd6RMEzc7C3y0SnATCCkFviApsZGDCsc2iyXe7WDzviJ6toivVJ/rpLNwDRoNQGjbBiQl8pgK5UbaHMrdGymTUl4GIPjtQmKufRj//H9Vaa8W7CtzXqSi+WWSRvItGc2v/XZ2DhqLa4B9TGVsijyHktIiokSwo7XksP4lR8o4TzgdZ9eBeF+5+6D6Ux1H8lBhbWiskdmPYfNmx6nqiLByr/0/SKmzDxC5rWLDfhepajCs7VfD2u6qqvdF27Hgyrn4ZtyKb8athDYsxdl7uNVo267WMEjxsGExBdngLbTKAWfv4YQ2LAUgtHE5zt6n2etPI7RxWcU6l8/B+/zNeJ+/meC6pTj72DrbdbWyqCrT2c7W2Wc4wXVLozWxxsxevZgrBlsXxkEde5J7oICdeXuZv3YZI7oPIiU+kZT4REZ0H8T8tRXbGE5wxRy8027BO+0WguuXlG7jdlVt4wOl9vYeTnBD9XoqQndsQpq2huQW4HDh6D6U0OYVkf3bvAJnr1MBcHQ9kdCvPxyUrnB6tYhna66XjDwf/mCIeZtyGXZMUoTMsI5JzF6fA8CCn3IZ2DYBESEjz1eS8LE9z8eWbC9tEqvOmwp89RFFT02i6KlJBNZ8javfGZY9HbqhRfuhAseiRftxdOgGgKvfGQTXfA2ApLUpEXP2OpFQ1rZq7dXtFWznTcsjZIKblpdu524nEtp68Ns5uOxjvM/djPe5mwmuW4LzBOu8k/Zdwbsf8sscU/n2OdTePqZOOK3kHHKk98N1ygV4/3U/+L3UBQ01AURUG8/4k4jcBxQCE4DLVXWJiLiBLqr6o4h8DtyhqitFJA1YqarHiMhQ4HFgKNAW+BE4W1U/F5ECVU2w2+8BfIg1zJglIs2ARFWtdCwxPDIri2vU9Tg69bVS82c/g+6wUpo9E57BN+NWy6bWnXGPu81Kk9/8DYFPrBRf4hJxX/BHJKk5mptlp+YXAKD+yoeJ3GMnlTwO4Hv/6ZJ085gbpuJ9/mZLZ3FqvjvGSs2fMw0AR/cheMZcD02SoaiA0I6f8b3215K24/eUv9t+45oHGNalL2kJKezK28e9H7+I22ndQ0xf/AEAz118ByN7DKbQV8TVrz3EN3aiwNVDxnL3yCsBePiTmcxcMqdCmwpbVj6J7h59vZWd6Pfi+3BKqb3XP4t32i2l9p4zuWQb++fa9nYbgmf0RIi37d25Bd+/LXsd8e4K9TmO7YtzuJ2a/8NCgkvfxTn0EnTnZsuxOd24xtyGo2VHtKgA/+wnrYQRwDNxOnjiwOkC7378b9+P7s2IaF9iK56i/eKXPB7/cgdBhXO7N2VC/xY8t2wXPVvEcWrHJLyBEH/6dBvrdxeRHOvk8REdaJ/s4aMN2bz8zW5cDsEhMHFAxan5ga2VJ/G4z7sJZ1f78Ye3niCUsQmA2NtfoOgpK0vX0S4dz8V/sFLz16/A/8E/LJuv/AuO5u1BQ2h2Fr53n4mIjJ1pFUeIjk79cJ5+DSJO6xGIr9/FefIl6I6w7XzWbThaHWul5n/4pJUwAngmzYAYezsX7cf/1n3ltnOooPJEHPdZN0SeQ5mWvTE3TcX7nH0OtU3Hc/5kcMVYj3t89IIlc/tL4HTDAWt7hrZtwP9hyXQ9cQ/PrbU3eWXtDbW66F/T4/kj0qM1RmdWAHwKPAskY0VfU1T1xSqcmQD/xhpCXAc0Be4r68xsHb8D/oQVyfmBG1W10tClKmdWV1TlzOqSipzZ4aAqZ1ZXVObM6prKnFldU5Uzq0sqc2Z1TVXOrC4xzqxyGtWcWZkhv1MqKB8W9nsP9pyZWh6/wid6wx2Zvfwf4D+17qzBYDDUAUdyEkdtaIxzZgaDwdBocYjU6q86RGSkiGwQkc0iclclMheJyFoR+VFE3jgUdjWqyMxgMBgaO3UZmdnP1f4DOAPIwMrsnq2qa8Nk0rGmYk5S1WwRaVFxazXDODODwWBoREQTXdWCgcBmVf0ZQETeAsYBa8NkrgP+oarZAKpa7vGlg8EMMxoMBoPhUNEWCH9+IsNeF04XoIuIfCUiS0Vk5KFQbCIzg8FgaETUNjITkQlYjzcVM8N+EUS0uIB0YBjQDvhCRI5T1Zza9Ms4M4PBYGhE1NaZhb/BqAIygfZhy+3sdeFkYL0C0A9sEZGNWM5tBbXADDMaDAZDI8Ihjlr9VcMKIF1EOtqf2LoYmF1GZhZWVIb9PG8X4Oda21XbBgwGg8FgAFDVAHATMB/rBRNv229XekBEit8SPR/YKyJrgc+AP6hqrd+oYIYZDQaDoRFRx9mMqOpcYG6ZdX8N+63A7fbfIcM4M4PBYGhE1LUzqy+MMzMYDIZGhHFmBoPBYDjqcTTQVAnjzOob5+E/sJwpMYddJ0Ch8/C/vR4gftee6oUOMa9Py65eqA449+5O9aJX6ukrAfX11tyY33asF72GyjHOzGAwGBoRZpjRYDAYDEc9xpkZDAaD4agnigefj0oaplUGg8FgaFSYyMxgMBgaEWaY0WAwGAxHPcaZGQwGg+Goxzgzg8FgMBz1mAQQg8FgMBiOUExkZjAYDI0IB2aY0WAwGAxHOWbOzGAwGAxHPQ11zsw4M4PBYGhEmMjMcNhwHHsCrhHXgTgIrlpAcMl7kQJOF+6zJyOtOsGBfPwf/B3NzYK4RNzn/RFHm84EVy8iMH9GjfR+uTWfx77cTjAE5/VoyrX9WkSU+4Ih7v40g7VZB0iJdfL3MzvQNsnDD7sKuf+zTAAUuGFgC047NrlGut2jJuBI7w9+L75ZU9AdP5WTkdad8JwzGdweQptW4p9n2efocRLuYZcizdvjffF2dPvmavW9fPk9jD3uJLLysznuwcsqlHnmotsZ3XMIhT4vV732IN9t2wDAFYNH8+dRVwPw0Lx/8trSuRXWr4p+z9xDm9G/JVBYxNKr7iL7u7XlZIbNe4m41s0Rl5Pdi79h5Y33o6EQnqbJnPSfp0k4pi0Fv2Ty5UW34c/Jq1ane8xEHF0HWNv4vafQ7RVs4zad8Zx/u7WNN6zAP2c6AI5eQ3EPv8zaxtMmo5mborbVPXoiji72vn3/6Yr3bZvOeM6bDC4PoY0r8c+19fYcinv4pUhae7zTJ0e1bwGk4wm4ThuPiIPg6k8JLns/UsDpwjXmVhwtO6EH8vHPfgLydkNsIu5z/oC06kxozWcEPn0xajsXr93DI+9tJBRSLhjSlutGHBNRvmJzNo++t5GN2wt48qpenHlCSwDWZeRz/3/WU1AUwOkQJo44htH9WkWt12DRYOJNEQmKyCoR+V5EvhWRE6OoU3A4+lYjxIFr5ET8b92Pb/pNOHuejKS1jxBx9jkDLSrA98L1BJbPxjX8Sqsg4CPwv9cJLJxZY7XBkPLwF9t5fmxHPrw0nXmbcvlpX1GEzPtrs0mKcTL38q5c3ieNp5fsBKBzs1jeuqgz716czrSzjuGBzzMJhDRq3Y70/kizNnifnYDvo+fwjLmhQjnP2BvxfTQV77MTkGZtcHTuB4BmbcX3n0cIbf0xap0zl8xh5NTJlZaP6jmE9BbtSb/3Qia88SgvXHInAE3jk7h3zHgGPTaegY9dw71jxpMSnxi1XoA2o04hMf0YPkofwfIJf2HAC/dVKPflRbcyr8845vYaS0zzpnS4cCQAPe6awK6FS/ioy5nsWriEnndNqFano0t/JK0t3qeuxTfrWTxn31ShnGfcjfhmPYP3qWuRtLaWEwJ011Z8bzxE6Jc1NbLVkd4fSW2Dd8p1+D6ciuesGyvWe9YN+GY9i3fKdUhqGxzpYfv2zYcJba2BXnHgPn0C/ncexPfyLTi6D0VS20X267jToWg/vhdvILjyI1zDrrAKgj4Ci98k8PmrNbIzGFIefGcDMyb14aN7hjDnm51s3hF5eWnTNJZHf9+DMf1aRqyP9Tj42+U9+fieIbx4wwk8+v5G8gr9NdJfExwitfo7Umkwzgw4oKp9VLU38Cfg0fru0MEgbdLRfTvRnF0QChBcuxhHl4ERMo70QQRXLwIgtO4rHMccbxX4vWjGOgj4aqz3h6xCOiR7aJ/swe10MCo9mc+2RN7tf7Ylj7O7pQBwRqdklmUUoKrEuR24HNZB7g0q1DBbytl1EMHvLXs0YwPENoGEppFCCU0hJs4qB4LfL8LZbbBVZ08GujezRjoXb17Fvv2VRzPjep9SEnEt2/IjKfEJtEpK5cweg1iwbjnZhXnkFOazYN1yRvYYXCPdbcedxpbXZgGwd9n3eFKSiG3VvJxcIH8/AOJy4fS4UbVuENqNO42fX7Xq//zqLNqdc3q1Op3dBxP8biEAus3exolltnFiU4iJt8qB4HcLcXa3t/Hubeiemm3jEr2rwvZtXGX7Nr50365ahLP7kIPWK63T0ZwdkGudQ6F1X+LoHHkOOdMHElzzGQChDV/j6BB2DmXW/BxavTWXDmlxtE+Lx+NyMLpfSxb9sDtCpm1qHF3bJpZzCB1bNOGYFvEAtEiOITXBw76CunRmjlr9HakcuT2rHUlANoCIJIjIQjta+0FExpUVrkxGRI4RkXUi8qKI/Cgi/xWROLuss4h8GhYJdrLX/0FEVojIahG5v6Ydl8RUNL/0Y5KatxdJTC0j0wzNs2U0hHr3Q1zNooOyZBUEaJVQ+oHFlgludu2PPKGy9vtpleABwOUQEjxOcoqCAKzeWcg5b2zkvDc38ddhbUqcWzRIUmqpPdg2J6VWILO3SplDSduU5mzLzipZzsjOom1K80rX14T4ti0p3LazZLkwYyfxbVtWKHvqJy9xftbX+PP3s+3d+QDEtkylaKd1oSzauZvYltVvB0lKQ3NLL66atwdJSqtAJmw/5JaXqSmSlBqpN3dPFPu2vEyNdCY0izyH8sufQySkRpxDeAtrdQ5l5Xhp1TS2ZLllSiy7crw1bmf1L7n4gyE6pMUddF+qw0RmRz5x9jDjeuAl4EF7fRFwrqr2BU4FnhQpt0eqkkkH/qGqPYEc4Hx7/ev2+t7AicAOERlhyw8E+gD9ROSUQ2/qkcfxreKZdWkX3rqwEy99sxtvIFTfXWoQfDbyWt5vPRRHjIeWwyuJADX6IV3DkUtWrpc//utHHr6sJ44a3AwaLBpSAsgBVe0DICJDgNdEpBfWmNcjtlMJAW2BlsDOsLqVyQBsUdVV9u9vgGNEJBFoq6ofAKhqka13BDAC+M6WT8Bybl+Ed1REJgATAJ4bdzzjBxxTUmbdRZbeDUtSKpq/N7w6mr/PuovO3wviQGKawIH86LdUBbRIcLEzbGhjV4Gflk3ckTJN3Ows8NEqwU0gpBT4gqTEOiNkjm0WS7zbweZ9RfS0h04qwjlgDK5+ZwIQytwUEQGUvVOH8pFYRTKHksyc3bRvWpoA065pCzJzdpOZs5thXfpGrP9847fVtpd+w6V0vu4iAPau+IH49qUT/PHtWlGYuavSuiGvj8wPF9Ju3Gns/PRrinbtJbZVcysqa9Wcoqx9FdZzDhqLa4C9jTM2IcmlEaQkpUVEw2BHRMlh+yG5vEw0OAeOwdXfmt8LZW6M1JucFsW+LS9TE7RgX+Q5lFj+HKJgr6WnwDqHiImv1TnUIiWGndmlc8y7copomRITdf2CAwGun7aK28Z2ok/HmiVP1ZSG+tB0Q4rMSlDVJUAa0By4zP7fz3Z2u4DYMlWqkgkfKwhS9Q2AAI/ac3d9VLWzqr5cQf9mqGp/Ve0f7sgAdPsmpFlrJLkFOFw4e5xMaOPyCJnQpuU4jx8OgKP7SYR+WV1Fl6KjV4t4tuZ6ycjz4Q+GmLcpl2HHJEXIDOuYxOz1OQAs+CmXgW0TEBEy8nwlCR/b83xsyfbSJtFTpb7gijl4p92Cd9otBNcvwdnbskfadbWGfAqyIysUZIP3gFUOOHsPJ7hhWa3trozZqxdzxeDRAAzq2JPcAwXszNvL/LXLGNF9ECnxiaTEJzKi+yDmr62+H5uef4N5J5zDvBPOIWPWp3S84hwAUgf1xp+bXzJsWIyrSXzJPJo4nbQZM4y89T8DkDF7EcdeadU/9spzyPhwYYU6g8s+xvvczXifu5nguiU4TzjNaq99V/Duh/wy2zg/G7yFVjngPOE0guuWVr+xyupdPgfv8zfjff5mguuW4uwTtm+L9leybwtL922f4QeltxjdsQlp2hrsc8jRfSihzSsi+7h5Bc5epwLg6HoioV9/OGh9AMd1SGLr7gNk7DmALxBi7je7OPW46IaffYEQN7/0PeMGti7JcKxLGuowY0OKzEoQkW6AE9gLJANZquoXkVOB31RQJRqZElQ1X0QyROQcVZ0lIjG2vvnAgyLyuqoWiEhbwK+qWVW1F9l4iMD8GbgvuQ8cDoLfL0T3bMN1yqWEdmwmtGk5wVULcI+bjGfSNCjKx//BEyXVY26cYd1lOl04uwzC9+Z96J5t1ap1OYS7T27D9bO3EFQ4t3tTOqfG8tyyXfRsEcepHZM4r3tT/vTpNkb/awPJsU4eH9EBgO927Oflb3bjcggOgXt+24amcdEfWqFNK9H0/sTc8qKVvv3hlFJ7rn8W77RbAPDNed5KzXd5CG3+htCmlQA4ug3BM3oixCcTc+m9hHZuwffvv1ap841rHmBYl76kJaSw7ZHZ3Pvxi7idVp+nL/6AuWu+ZnSvE9n8wLsU+oq4+rWHAMguzOPBua+w4o+vAPDA3JfJLqw+LT6c7XP/R5vRv+WszQsIFh5g6dV3l5SN+m4W8044B1eTOH47+wUcMR7EIez6bBmbpr0FwNq/zWDo21PoNP4C9m/dzpcX3VatztCGFWiXAcTc/nJJinwxMTdNxfvczQD4Zj+P5/zJ4IohtGkloY32Nu4xBM/YSdAkmZgr7iO042d8M/9Svd6NK9Au/YmZ/FJ5vTdMxfu8rfej563UfHcMoY0rS/dt9yF4xlxv6b3c1vta1fsWDRH49EXcF95rpeb/sBDduw3n0EvQnZsJbV5BaPWnOMbchue659GiAvyznyyp7pk4HTxx4HThSR+I/+370b0ZVap0OR38+cKuXPv8d4RUOW9wG9JbJ/DsnJ/o1SGJ4cc154etudz80mryCv18tmYPU+f+zMf3DOGT73axcnMOOfv9zFq2HYBHft+T7u1qNw9eGUdyEkdtEG0g4+0iEgSKb68EuFtV54hIGvAR1pDfSmAwMEpVfxGRAlVNqEzGbutjVe1l67gDSFDV+0QkHZiOFQH6gQtV9WcRuRW41q5bAPxeVcs/WGNT9PC4w74DHInu6oXqgGCZVP/DRfyumg+V1ZbXp2VXL1QHnHt3p3rRSz3N8TiqGQGoK9x92tSLXseIf9R6Q2/MeaJW15wuKXcckeFZg4nMVNVZyfo9wJBKyhKqkwF6hck/EfZ7EzC8gjafAZ6JuuMGg8FgqDUNxpkZDAaDoXqkgQ4zGmdmMBgMjQhHw8z7M87MYDAYGhMmMjMYDAbDUU9DzWZsmFYZDAaDoVFhIjODwWBoREgDjWGMMzMYDIZGREMdZjTOzGAwGBoRDTUya5hWGQwGg6FRYSIzg8FgaESYYUaDwWAwHPWY58wMBoPBcNRj3gBiqBPEffgPrMDWmn2y5FDhbNmkXvTW1xvsL7u+6WHXeSC+fr6III3sy8iFs9fXi96EEbVvo6FGZg3TKoOhnqkPR2YwNGaMMzMYDIZGhEMctfqrDhEZKSIbRGSziNxVhdz5IqIi0v9Q2GWGGQ0Gg6ERIVT46cdD07aIE/gHcAaQAawQkdmquraMXCJwK7DsUOk2kZnBYDA0Iuo4MhsIbFbVn1XVB7wFjKtA7kHgMeCQfX7eODODwWBoRAiOWv1VQ1tgW9hyhr2uVL9IX6C9qs45lHYZZ2YwGAyGqBGRCSKyMuxvQg3qOoCngP871P0yc2YGg8HQiKjtG0BUdQYwo5LiTKB92HI7e10xiUAv4HMRAWgFzBaRs1V1ZW36ZZyZwWAwNCLq+DmzFUC6iHTEcmIXA5cWF6pqLpBW2hf5HLijto4MjDMzGAyGRkVdvgFEVQMichMwH3ACr6jqjyLyALBSVWfXlW7jzAwGg8FwyFDVucDcMuv+WonssEOl1zgzg8FgaEQ01NdZGWdmMBgMjQjzCRiDwWAwHPU01C9NG2d2BCIdT8B12nhEHARXf0pw2fuRAk4XrjG34mjZCT2Qj3/2E5C3G2ITcZ/zB6RVZ0JrPiPw6Ys11u0+5wac3QeAz4v3rSfQzM3l+9cunZiL7wC3h+C6FfhnPW/VHXklzp5DQBUtyMH31t/RvH1HtL39nrmHNqN/S6CwiKVX3UX2d2vLyQyb9xJxrZsjLie7F3/DyhvvR0MhPE2TOek/T5NwTFsKfsnky4tuw59T9RcJXr78HsYedxJZ+dkc9+BlFco8c9HtjO45hEKfl6tee5Dvtm0A4IrBo/nzqKsBeGjeP3lt6dwK61eE49gTcI24DsRBcNUCgkveixRwunCfPRlp1QkO5OP/4O9obpZVdOL5OHufARoi8N8XCf38XdR65dgTcJ1+LeKw9S6tYN+OvQ1Ha3vfznoCcrMgLhH3uXcirTsT+mERgf/WbN/W5zHlueBGXD0HoT4v3n89TihjUzkZR/t0Yi6/E3HHEPhxGb53/2HVHX0FrhPHoAU5APhmv0xw7fIa96EqGmpkVmOr7BdDPhm2fIeI3HeoOmQ/kLfe/lsuIkPDyk4WkR9FZJWIdBeRA/bvtSIyTWoxGCwiv4hIWvWS5eodIyKXVi8ZbYMO3KdPwP/Og/hevgVH96FIarsIEcdxp0PRfnwv3kBw5Ue4hl1hFQR9BBa/SeDzVw9KtaPbABxpbSl69Gp870zBc/4tFcp5zr8Z39tPU/To1TjS2uLoNgAA/2fvUPTk9RQ9NYng2mW4zvj9EW1vm1GnkJh+DB+lj2D5hL8w4IX7KpT78qJbmddnHHN7jSWmeVM6XDgSgB53TWDXwiV81OVMdi1cQs+7qn92dOaSOYycOrnS8lE9h5Deoj3p917IhDce5YVL7gSgaXwS944Zz6DHxjPwsWu4d8x4UuITozNUHLhGTsT/1v34pt+Es+fJSFr7CBFnnzPQogJ8L1xPYPlsXMOvtKqmtcfZ42R8M27C/+Z9uEZOhGhPM3HgHjER/9sP4JtxM44eJ5fft73PgKICfNMmEVw+u3TfBnwEvniDwKKZ0ekqq7eejilnj4E4mrej8P4r8L75FDEX31qhXMzvbsP7xlMU3n8FjubtcPYYWFLm/+xdDvxtIgf+NvGQO7KGzMFc/L3AeQdz4a8OERkLTASGqmo34HrgDRFpZYtcBjyqqn2AA8BP9u/jgR7AOWXaOxyR5zGEPUdRW6R1OpqzA3J3QShAaN2XODoPjJBxpg8kuOYzAEIbvsbR4XirwO9FM9dBwHdQup29TiTwzQKr3V/XI3FNILFZpFBiMyS2CaFfre85Bb5ZgLPXiVaZt7BUzhMLaLU669PetuNOY8trswDYu+x7PClJxLZqXk4ukL/f6qvLhdPjRtWyq9240/j5Vav+z6/Oot05p1erc/HmVezbX3n0Nq73KSUR17ItP5ISn0CrpFTO7DGIBeuWk12YR05hPgvWLWdkj8FR2Slt0tF9O9EcaxsH1y7G0SVyGzvSBxFcvQiA0LqvcBxjbWNHl4EE1y6GYADNzUL37UTapEevN3sH5ITt2y6DImQi9u36r0v04veiGesg4I9KV4TeejymXMefRGD5f612f1mHxCUgSZHnkCQ1Q2LjCf2yDoDA8v/iOv6kg9J3MNTx66zqjYPpWQDr6e9yt5ciMlNELghbLrD/DxOR/4nIhyLys4j8TUQusyOvH0Skk13lj8AfVHUPgKp+C7wK3Cgi1wIXAQ+KyOvhelU1AHwNdBaRq0RktogsAhaKSDMRmSUiq0VkqYgcb/cpVUT+a0d6LwFirz9GRNaE2VASeYpIZxH5VES+F5Fv7X7/DTjZjhAni0hP265Vts7ozvxifQnN0Pw9pbbl70USUyOFElLRPFtGQ5YTiYvyLr0KHMmpaM7uUt25e3Akp1YtkxMp4x51FbF/eR1X3+H4P3mtWp31aW9825YUbttZslyYsZP4ti0rlD31k5c4P+tr/Pn72fbufABiW6ZStNPaFkU7dxPbMrXCujWhbUpztmVnlSxnZGfRNqV5peujQRJTI7dxXvltLInNIraxevdDXKJVNy98/+wpv38q05vQrEzdvUgFN0eHet/W5zElKWmEskvPj1DObiQlrbxMTrjMnggZ9ynnEPenF4m57A6IS6h1n8pS15+AqS8Otmf/AC4TkeQa1OmNFWl1By4HuqjqQOAl4GZbpifwTZl6K4GeqvoSMBvL2UVMNohIPHAa8IO9qi9wgar+Frgf+E5VjwfuBoqvsPcCX6pqT+ADoEMUNrwO/ENVewMnAjuAu4DFqtpHVZ+2bXzGjhj7Y71os9HgnzeTogcvI/DtItxDz67v7hwyPht5Le+3HoojxkPL4ZVERFp9JGowVIV/8UcU3nc5B/42Ac3bR8x51x9yHSKOWv0dqRxUz1Q1D8spVDypUjErVHWHqnqBn4D/2ut/wBqqOxg6icgq4CtgjqrOs9cvUNXizIOhwL/sfi8CUkUkCTgF+Le9fg6QXZUi+/s7bVX1A7tOkaoWViC6BLhbRP4I/EZVD1TQVsmLOl9a9ktEmRbsQxJL79Ksu+q9kQ0U7EWSbBlxQEw8HMivqvuV4jrpLGJvf4HY219A8/YhYXf7kpxGKDdSdyh3b6RMSnkZgOC3C3Eed3K1+g+3vek3XMqo72Yx6rtZHNixm/j2rUrK4tu1ojBzV6V1Q14fmR8upN240wAo2rW3ZFgytlVzirKqT3apjsyc3bRv2qJkuV3TFmTm7K50fTRYkUnYNk4qv401f1/ENpaYJnAg36qbFL5/0srvn8r0FuwrUzcVzS+zjcrorc2+jdB7GI8p9ynjiLtrOnF3TUdz9+JoWnp+OFKaozl7IuQ1Zw+OlHCZtBIZzc+2IkVV/F/NwfGbbgfVp6oQrd3fkUpt3OwUYDzQJGxdoLhNOxnDE1bmDfsdClsOUZpVuRboV0ZPP+DHSvrwkx0RnaCq94Wt3x+dCRVSYoNNbE0qq+obwNlYc3pzRWR4BTIzVLW/qva/dtAxkWU7NiFNW0NyC3C4cHQfSmjzigiZ4OYVOHudCoCj64mEfv2BgyXw1UcUPTWJoqcmEVjzNa5+Z1jtduiGFu2HCi4+WrQfRwfrJHP1O4Pgmq8BkLQ2JWLOXicSytpGdRxuezc9/wbzTjiHeSecQ8asT+l4xTkApA7qjT83v2TYsBhXk/gShyVOJ23GDCNv/c8AZMxexLFXWvWPvfIcMj5ceND9Kmb26sVcMXg0AIM69iT3QAE78/Yyf+0yRnQfREp8IinxiYzoPoj5a6P7rqFu34Q0a43Y29jZ42RCGyMTC0KbluM83jpUHd1PIvTLamv9xuU4e5wMTheS3AJp1hrdXj47r1K9Zfftpki9wU3LS/dttxMJbT34fVui9zAfU/4vPixJ2Ais/grXwBFWu8d0Rw/sL5fRq3n70KJCHMd0B8A1cASB1V8BRMyvuXoPJbTjl4PuV2PjoBMkVHWfiLyN5dBesVf/guV83sa6oLtr2OzjwGMiMlJV94pIH+AqYFCVtapmMVbiyIMiMgzYo6p5IvIFVuLGQyIyCmhqy+8CWohIKlAAjAU+UdV8EckQkXNUdZaIxGC9eywf603QAIjIscDPqvqsiHTASk5ZFHVvNUTg0xdxX3ivlVb8w0J07zacQy9Bd24mtHkFodWf4hhzG57rnkeLCvDPLkkuxTNxOnjiwOnCkz4Q/9v3o3ujG+kMrVtOqPtAYv80E/xefG89UVIWe/sLFD01CQD/e1PxXPwHKzV//QpC660LhXvMeBzN21tzLtlZ+N595oi2d/vc/9Fm9G85a/MCgoUHWHr13SVlo76bxbwTzsHVJI7fzn4BR4wHcQi7PlvGpmlvAbD2bzMY+vYUOo2/gP1bt/PlRbdVq/ONax5gWJe+pCWksO2R2dz78Yu4ndZpOH3xB8xd8zWje53I5gfepdBXxNWvPQRAdmEeD859hRV/tE61B+a+THZh1Y8BlKAhAvNn4L7kPnA4CH6/EN2zDdcplxLasZnQpuUEVy3APW4ynknToCgf/wfWvtc92wiu+wrPxOcgFCIwf7oVOUSrd8GLuC++FxEnwdWfonu24Tz5EnSHvW+//xTHWbfhuf4FK0X+w7B9O2kGxBTv20H437ovun1bj8dU8MdlOHsOIv7ef6H+Irz//ntJWdxd0znwt4kAeN9+hpjf26n5a5eXZC16zpmAo10nUNB9O/G++XRUemtEtPuvMuTQdONQI1rDcX4RKVDVBPt3S2AL8Liq3mcvfwjEAZ8AN6pqgu1E7lDVsXa9z+3llRWUTQJuw0qFywf+T1W/sMtmAh+r6rsicoz9u1eZ/l0F9FfVm+zlZljO9ligEJigqqttZ/Um1ofjvgZGAP1UdY+I3IL1Se9M4GfgF9u+dGA61luf/cCFWB+imw+kAjOBGKw5QT+wE7g0bMizHN7Hzz3sgXtwV20C14PH2bJJ9UJ1wHt/LP/sWF1z2fVNqxeqAw60qziBpa4RRz1d4Zz1o9f/a5Q3EoeYhOcW1t7g4ILaXXOcZxyR7qzGzsxwaDHOrO4xzqzuMc7s8HBInFlgfu2uOa4zj0hnduSmphgMBoPBECXmdVYGg8HQmKjtnNkRinFmBoPB0JgIGWdmMBgMhqMdE5kZDAaD4aingTozkwBiMBgMhqMeE5kZDAZDY6KBRmbGmRkMBkNjwiSAGAwGg+Gox0RmBoPBYDjqaaDOzCSAGAwGg+Gox0RmBoPB0JhooJGZcWYGg8HQiFAN1qr+EfmWYYwzq3dChf7DrlPia/qZuUOkN9ZZL3rPvbvTYdd5oJ62cVxG5V/KrksK01LrRW994UyLq+8uHDwNNJvRzJkZDAaD4ajHRGYGg8HQmDBzZgaDwWA46jHOzGAwGAxHPcaZGQwGg+Gop4E6M5MAYjAYDIajHhOZGQwGQ2OigabmG2dmMBgMjYkGOsxonJnBYDA0JowzMxgMBsNRTwN1ZiYBxGAwGAxHPSYyMxgMhsaESQAxGAwGw1FPAx1mNM7sCMU9agKO9P7g9+KbNQXd8VM5GWndCc85k8HtIbRpJf55MwBw9DgJ97BLkebt8b54O7p9c/R6R0/E0cXW+/7TFett0xnPeZPB5SG0cSX+udMtvT2H4h5+KZLWHu/0yVHr/XJrPo99uZ1gCM7r0ZRr+7WIKPcFQ9z9aQZrsw6QEuvk72d2oG2Shx92FXL/Z5kAKHDDwBacdmxy1LYCuMdMxNF1gGXve0+h2yux9/zbre28YQX+Oba9vYbiHn6ZtZ2nTUYzN1Wrz3HsCbhGXAfiILhqAcEl70UKOF24z56MtOoEB/Lxf/B3NDfLKjrxfJy9zwANEfjvi4R+/i5qO1++/B7GHncSWfnZHPfgZRXKPHPR7YzuOYRCn5erXnuQ77ZtAOCKwaP586irAXho3j95bencqPVC/RxT9aq3ns7dqGmgzuyInDMTkYIyy1eJyHMH2VYXEZkrIptE5FsReVtEWtZWNkrdM0XkgprWc6T3R5q1wfvsBHwfPYdnzA0VynnG3ojvo6l4n52ANGuDo3M/ADRrK77/PEJo648115vaBu+U6/B9OBXPWTdWrPesG/DNehbvlOuQ1DY40sP0vvkwoa1rotYZDCkPf7Gd58d25MNL05m3KZef9hVFyLy/NpukGCdzL+/K5X3SeHrJTgA6N4vlrYs68+7F6Uw76xge+DyTQEijt7dLfyStLd6nrsU361k8Z99Usb3jbsQ36xm8T12LpLW1LpCA7tqK742HCP0Spb3iwDVyIv637sc3/SacPU9G0tpHiDj7nIEWFeB74XoCy2fjGn6lVTWtPc4eJ+ObcRP+N+/DNXIiSPSn78wlcxg5dXKl5aN6DiG9RXvS772QCW88yguX3AlA0/gk7h0znkGPjWfgY9dw75jxpMQnRq23Po6petdbD+fukYSIjBSRDSKyWUTuqqD8dhFZKyKrRWShiPzmUOg9Ip3ZoUJEYoE5wAuqmq6qfYHngeZl5FzRylah65BFuc6ugwh+vwgAzdgAsU0goWmkUEJTiImzyoHg94twdhts1dmTge7NrLne7oMJrgrTG1eZ3vhSvasW4ew+xKqzexu6p2Z6f8gqpEOyh/bJHtxOB6PSk/lsS16EzGdb8ji7WwoAZ3RKZllGAapKnNuBy2F9KtAbVGr62UBn98EEv1to9X2bvZ0Ty9ibaNtrRynB7xbi7D74oOyVNunovp1ozi4IBQiuXYyjy8AIGUf6IIKrrX0QWvcVjmOOt9Z3GUhw7WIIBtDcLHTfTqRNetS6F29exb79eZWWj+t9SknEtWzLj6TEJ9AqKZUzewxiwbrlZBfmkVOYz4J1yxnZY3DUeuvjmKpXvfV07taIUKh2f1UgIk7gH8AooAdwiYj0KCP2HdBfVY8H3gUePxRmHXXOTETOEpFlIvKdiHxaHDmJyG9FZJX9952IJAKXAktU9aPi+qr6uaqusaO92SKyCFhYjewxIrLYjta+FZETbZ3D7PWzgbVi8Zx9V/IpEDleFq2NSalo3p6SZc3biySlViCzt0qZg9Kbu7u0zdw9UegtL1MTsgoCtEoo/ZBlywQ3u/ZHfrA0a7+fVgkeAFwOIcHjJKfI+lru6p2FnPPGRs57cxN/HdamxLlFgySlRdqbtwdJSqtAJmxf5JaXiVpfYiqaX2a/JpbZvonNSve9hlDvfohLtOqGHxP5e8rVrQ1tU5qzLTurZDkjO4u2Kc0rXR8t9XFM1bveejh3a0RIa/dXNQOBzar6s6r6gLeAceECqvqZqhbai0uBdofCrCN1zixORFaFLTcDZtu/vwQGq6qKyLXAncD/AXcAN6rqVyKSABQBvYBvqtDTFzheVfeJyFNVyGYBZ6hqkYikA28C/cPa6KWqW0TkPKAr1h1JS2At8ErZxkRkAjABYOrY4xjfr0MVXTRUxfGt4pl1aRd+3lfEPQszGNohkRjXUXePZjAcPuo2m7EtsC1sOQMYVIX8eGDeoVB8pDqzA6rap3hBRK6i1Hm0A/4jIq0BD7DFXv8V8JSIvA68r6oZItXepS9Q1X1R9McNPCcifYAg0CWsbLmqFvfhFOBNVQ0C2+2orxyqOgOYAXDgvrEK4BwwBle/MwEIZW6KuPsveycH5e/mKpKJBufAMbj6j7T1bkSSS++6JTktCr3lZWpCiwQXOwtKI7FdBX5aNnFHyjRxs7PAR6sEN4GQUuALkhLrjJA5tlks8W4Hm/cV0bNFfKX6nIPG4hpgb+eMTZH2JqVF3FWDfbeeHLYvksvLRIvm70USy+zX/DLbN3+f1Y/8vSAOJKYJHMi36oYfE4lp5erWhsyc3bRvWjqQ0K5pCzJzdpOZs5thXfpGrP9847dVtlVfx1S96a2nc/egqaUzC78Zt5lhX9Nq2s7vsa7rv61Vh2yOxlvYqcBzqnocMBGIBVDVvwHXAnHAVyLSDfgR6FdFW/vDflclOxnYBfTG2vieSto4aIIr5uCddgveabcQXL8EZ+/hAEi7ruAthILsyAoF2eA9YJUDzt7DCW5YVnO9y+fgff5mvM/fTHDdUpx9wvQW7a9Eb2Gp3j7DCa5bWmO9xfRqEc/WXC8ZeT78wRDzNuUy7JikCJlhHZOYvT4HgAU/5TKwbQIiQkaeryThY3uejy3ZXtokesqqiLR32cd4n7sZ73M3E1y3BOcJp1n2tu8K3v2QX8befNve9ra9J5x20Pbq9k1Is9ZIcgtwuHD2OJnQxuURMqFNy3Eeb+0DR/eTCP2y2lq/cTnOHieD04Ukt0CatUa3V589GS2zVy/misGjARjUsSe5BwrYmbeX+WuXMaL7IFLiE0mJT2RE90HMX1v1cVZfx1S96a2nc7e+UNUZqto/7C/ckWUC4VlN7ex1EYjI6cA9wNmq6j0U/RLV6LO/DhciUqCqCWHLV2FNGN4kIt8B16rqNyLyT6Cjqg4TkU6q+pMt/y7wb2A+8ANwq6rOsctOAfZhOaX+qnqTvT6uCtnxQIaqPikiVwOvWKOcMgy4Q1XH2vLnYTnY0VjzZWuB61T13cpsLY7MyuIefb2V4eT34vtwSkmKbsz1z+Kddou1Xdp0ttJ7XR5Cm7/BP3caAI5uQ/CMngjxyVBUQGjnFnz//mtp48HK97l77CQro6s4nblY7w1T8T5/c6ne8yaDO8ZKZ55j6+0+BM+Y66GJrXfHz/heK9XrbB5Xoc4vfsnj8S93EFQ4t3tTJvRvwXPLdtGzRRyndkzCGwjxp0+3sX53EcmxTh4f0YH2yR4+2pDNy9/sxuUQHAITB1Scmh/cVViBVtves26ItNdOr4+5aSre52x726bjOX8yuGKsNOqPXrDs7TEEz9hJkfbO/ItVJ95doT5Hp364zhgPDgfB7xcS/OodXKdcSmjHZkKbloPTjXvcZKTlsVCUj/+DJ6yEEcB50oU4e58GoRCBBS8R+ql8hBSXsatCvW9c8wDDuvQlLSGFXXn7uPfjF3E7rYGZ6Ys/AOC5i+9gZI/BFPqKuPq1h/jm1/UAXD1kLHePtLIqH/5kJjOXzCnXfmFa5XM+dXlMVUWd6nVWPupTl+du3H0f1yzLqQJ03UO1uuhL9z9X2gc7EW4jcBqWE1sBXKqqP4bJnICV+DFSVQ/ZHdnR6MzGAU8D2cAiYIDtzKYCpwIhrCjrKlX12hHaFKAT4AdWA7diZduUODNbT2WyScB7WI8zfYI1N5dQgTMTrMjxDOBXu41XDsaZ1SlVOLO6pDJnVtdU5czqisqcWV1TmTOra6pyZg2SKpxZXXJInNmPD9TOmfX8a5V9EJHRWNdRJ9b172EReQBYqaqz7eS444AddpVfVfXs2vQJjlBn1pgwzqzuMc6s7jHO7PBwSJzZD/fVzpkdd1/9GF8NR+OcmcFgMBgMERyp2YwGg8FgqAtq8JacownjzAwGg6ExYd6abzAYDIajHhOZGQwGg+Gop4FGZiYBxGAwGAxHPSYyMxgMhsZEA43MjDMzGAyGRkRtny0+Ih8ywzgzg8FgaFyYyMxgMBgMRz0N1JmZBBCDwWAwHPWYyMxgMBgaE+Y5M0Nd4EyOOew6Xb9Jql6oDij6qtxnjQ4L9fHSX3HUzzR5fb3wN37PYfy4ZBj1ZW/m/F/qRW/n+w5BIw10mNE4M4PBYGhMNFBnZubMDAaDwXDUYyIzg8FgaEyYOTODwWAwHPU00GFG48wMBoOhMWGcmcFgMBiOehroMKNJADEYDAbDUY+JzAwGg6ExYYYZDQaDwXDUY5yZwWAwGI56GuicmXFmBoPB0JhooJGZSQAxGAwGw1GPicwMBoOhEaFBM8xoOEzIb/rgGnYNOBwE1ywktOKDSAGnC+eZt+BoeSx6IJ/A3KcgbzfSsjPO06+3GxGCS/6D/rQ8ar2LN2TzyMdbCIXgggEtuG5Yu4hyXyDEH9/exNrM/aTEu3jq0i60bRqLLxDivlk/sSZjPw6Bu8/qyMBjk2tks/ucG3B2HwA+L963nkAzN5ffLu3Sibn4DnB7CK5bgX/W81bdkVfi7DkEVNGCHHxv/R3N2xed3tETcXTpD34vvvefRnf8VF5vm854zpsMLg+hjSvxz50OgKPnUNzDL0XS2uOdPhndXr7P5do69gRcp1+LOBwEVy0guPT9SAGnC9fY23C07oQeyMc/6wnIzYK4RNzn3om07kzoh0UE/vtiVPbVp60vX34PY487iaz8bI578LIKZZ656HZG9xxCoc/LVa89yHfbNgBwxeDR/HnU1QA8NO+fvLZ07hFvL0Da5D8Rf+LJaFERWQ/eg3fjunIyzSbeQuKos3EmJvHzaQMjyhJOO5Nm429AVfFt3sCue/9YY7urpYHOmUU9zCgiBVHI9BERFZGRUcheJSJtwpZfEpEe0fanTFu/iMjiMutWiciag2mvgvZnisgFh6Kt6pU5cA2/Dv+sh/G/ehuOrkOhWaRTcfQ8DbwF+P95E6FvP8Y59HIAdO+vBN64k8DrdxD44EFcp18PEt0uDoaUB2f/zIyre/DR5D7M+X4Pm3cVRsi8u2IXyXEu5v+hL1cMbcMT87YC8M6KXQDMvq0PL4/vwWNzfiFUgxPG0W0AjrS2FD16Nb53puA5/5YK5Tzn34zv7acpevRqHGltcXQbAID/s3coevJ6ip6aRHDtMlxn/D46ven9kdQ2eKdch+/DqXjOurFivWfdgG/Ws3inXIektsGR3g8AzdqK782HCW2N8jATB+4RE/G//QC+GTfj6HEyklpm3/Y+A4oK8E2bRHD5bFzDrrAKAj4CX7xBYNHM6HTVt63AzCVzGDl1cqXlo3oOIb1Fe9LvvZAJbzzKC5fcCUDT+CTuHTOeQY+NZ+Bj13DvmPGkxCfWwNr6sTd+yMm423fg1wtHk/W3+2h+518qlNv/5edkjL+43Hp3uw40veJaMiZezrbLzmHPlMei1l0jglq7vyOUQz1ndgnwpf2/Oq4CSpyZql6rqmtroTtRRNoDiEj3WrRzSBGRGkW/0qozmrMTcndBKEBow5c4Og2IkHF0Gkho7ecAhDYtwdHhOKsg4AO1J3edHtDoD7zV2wrokBpH+2axeFwORvdOY9G6yOhm0bpsxvVtAcCZvVJZ+lMuqspPWQcYZEdiqQkekuJcrMms9t6nBGevEwl8s8Cy59f1SFwTSGwWKZTYDIltQujX9Zap3yzA2etEq8wb5nQ9sUB0dju7Dya4ahEAmrEB4ppAQtNIoYSmEBNvlQPBVYtwdh9i1dm9Dd0T/TfapE06mr0Dcux9u+5LHF0GRfYpfSDBNZ8BEFr/NY5jjrcK/F40Yx0E/FHri2j3MNsKsHjzKvbtz6u0fFzvU0oirmVbfiQlPoFWSamc2WMQC9YtJ7swj5zCfBasW87IHoNrpLs+7G1yyqnkz5sNgPfH1TgSEnGmppWT8/64muDePeXWJ427gNx33yKUb22zYHZ0owsGixo7MxFpLSJfFEc+InKyvV6AC7Gc1BkiEhtW548i8oOIfC8if7OjnP7A63Y7cSLyuYj0F5HrReTvYXWvEpHn7N+/F5Hldp3pIuIM69rbwO/s35cAb4a14RSRv4vIChFZLSIT7fXDROR/IvKhiPxs9+0yW8cPItIprP3TRWSliGwUkbFRtLtYRGYDNXPQCc3Q/LADvWAfkpBauYyGrIt5rHXnKq3ScV0xBfflTxFYOL3UuVVDVp6XVsmekuWWSR525foiZHbleWmdYsm4nEJirJOcwgDdWsfz2bpsAkElY18RP2YWsLNM3apwJKeiObtLljV3D47k1KplciJl3KOuIvYvr+PqOxz/J69FpVeSUtHcSL2SlFpeJq/0w5OaV14mWiShGZpXum81fy9SgdMukSnet3E1i0oq1H2YbY2GtinN2ZadVbKckZ1F25Tmla6vCfVhr6t5SwK7dpYsB3bvwtW8ZdT13e1/g7vDb2g7/V+0e/F14gefdNB9qQoNaa3+jlQOJjK7FJivqn2A3sAqe/2JwBZV/Qn4HBgDICKjgHHAIFXtDTyuqu8CK4HLVLWPqh4Ia/894Nyw5d8Bb9nR1u+Ak2zdQeCyMvXOs3+fBXwUVjYeyFXVAcAA4DoR6WiX9QauB7oDlwNdVHUg8BJwc1gbxwADbbum2c66qnb7AreqapeyG1BEJtiOceVLS7aULa4VunMTgdduw//mH3EOPA+cdf+V5fP6taRlsocL//E9j368hT4dEjncH1r2z5tJ0YOXEfh2Ee6hZx9e5QbDIUBcLtztf0PmDVez86930vyu+3Ek1P5GphwNdJjxYBJAVgCviIgbmKWqq+z1lwBv2b/fAq7AcjCnA/9U1UIAVa0ydlbV3XaUNBjYBHQDvgJuBPoBK6wgkDggK6zqXiBbRC4G1gHhEz4jgOPD5r2SgXTAB6xQ1R0AIvIT8F9b5gfg1LA23lbVELBJRH62+1VVu8tVtUJPpaozgBkAvqfPjzw6CvYhiWFDEwnN0IIyn6S3ZbRgnzUnFhMPRfmRMvsywVeEpHVAd5Wf+C5Li6SYiGhqV56PlmGRGkDLpBh25PholRxDIKjkFwVJiXchIvxpbMcSuUte+IFj0uKq1Oc66Sxcg0YDENq2AQm785bkNEK5kTaHcvdGyqSUlwEIfruQmGsfxj//XxXqdQ4cg6u/NaUbytyIJEfqDb9TB9C8vRF365JUXiZatGAfklS6byUxFc0vczrkWzKav7d03x4os2+jpD5tjYbMnN20b9qiZLld0xZk5uwmM2c3w7r0jVj/+cZvq22vPuxNPv9iks62Tv+idWtwtWxVUuZq3pLA7l1RtxXI2kXRj6shGCCwIxP/tl9wt/8N3nWHZOq/lKB5zgwAVf0COAXIBGaKyBX2cN/5wF9F5BdgKjBSRA72tuIt4CK7zQ9UVQEBXrUjuT6q2lVV7ytT7z/APwgbYrQR4Oawuh1VtdhpecPkQmHLISKdfdlbkuI+Vdbu/hpZXNzozs1I09aQ1AIcLhxdh6I/r4yQCf28AkePYQA40ocQ2mYf7EktShM+EpsjzdqiuVlEw3HtEti65wAZ+4rwBULM/X4Pp3aPHAI7tXtTPvzWam/+mr0M7pSMiHDAF6TQFwTgq005OB1C55bxVeoLfPURRU9NouipSQTWfI2r3xmWPR26oUX7oYKLvBbtx9GhGwCufmcQXPM1AJJWMvWKs9eJhLK2Vao3uHwO3udvxvv8zQTXLcXZZ7jVRruuULQfCrIjKxRkg7fQKgecfYYTXLe0StsqQ7dvsvZtsr1vuw8ltCky2zS4aTnOXtY9lKPbiYS2/nBQuqB+bY2G2asXc8Vg64ZmUMee5B4oYGfeXuavXcaI7oNIiU8kJT6REd0HMX/tsmrbqw97c997i21XXsC2Ky9g/xeLSBxljQrE9Dye0P6CCufGKmP/FwuJ62vNjzuSU3C3PwZ/ZuXH8sHSUIcZaxyZichvgAxVfVFEYrCG03YCq1X1zDC5V7GGCxdgObnXVbVQRJrZ0Vk+UJmz+wC4BzgBKM5NXQh8KCJPq2qWiDQDElV1a5l6rYH5hCWX2MuTRGSRqvpFpAuWM64JF9o2dQSOBTYconYj0RCBRS/hPu8vIA6CPy5C927DOeRiQrs2oz+vJLRmIa6Rt+C++jm0qIDA3KcBcLTtjmPAuRAMgCqBRS+Wj9gqweUU/nz2sVz7ylpCqpzXvyXpLeN5dsGv9GqbwPAezbigf0v++PYmzvz7tyTHu3jyEmsEdd9+P9e+shaHCC2SPDx2UecamRxat5xQ94HE/mmmlUb91hMlZbG3v0DRU5MA8L83Fc/Ff7BS89evILR+BQDuMeNxNG8PGkKzs/C9+0x0ejeuQLv0J2bySyXp28XE3DAV7/PWKLPvo+et9G13DKGNKwltsm4uHN2H4BlzPTRJJuby+wjt+Bnfa3+tXKGGCCx4EffF9yLiJLj6U3TPNpwnX4Lu2Exo8wpC33+K46zb8Fz/gpWa/+GTJdU9k2ZATBw4XXjSB+F/6z50b8aRaSvwxjUPMKxLX9ISUtj2yGzu/fhF3E7rkjN98QfMXfM1o3udyOYH3qXQV8TVrz0EQHZhHg/OfYUVf3wFgAfmvkx2YeWJJEeKvYVff0H8iSfzm3fmEfIeIOuh0mzG9q++y7YrrQgu9cbbSRwxGomN5ZgPPyVv9vvse/l5Cpd+RfzAE+nwxodoKMje554klJdbI7sbM6JRZryJSIGqJojIlcAfAD9QgDWc+FdgmapOC5M/G5ikqqNE5C5bzgfMVdW7ReR84BHgADAEmAfcoaor7fofAz1U9diwNn8H/AkrovQDN6rqUjsa7K+qe8JkjwE+VtVeIuIAHsKaSxNgN3AOlrO8Q1WLEzo+L+6DiAwrLhORmUARVtJKEnC7qn4cbbtVUW6Y8TDg+k3S4VYJQNFXtfPzB4vE1/28YVkc9aATIFR4cNmOtSV+T90NR1ZFYVrdJahUReanW6sXqgM6L1lT69lo//SLa3XNcU986zDPiEdH1M7MUDcYZ1b3GGdW9xhndng4JM7shd/VzplN+s8R6czMG0AMBoOhEWFeZ2UwGAyGox/z1nyDwWAwGI5MTGRmMBgMjQkzzGgwGAyGo50j+Vmx2mCGGQ0Gg6ExUcevsxKRkSKyQUQ2249llS2PEZH/2OXL7Meoao1xZgaDwWA4JNhvg/oHMAroAVxSwae9xgPZqtoZeBo4JN+6Mc7MYDAYGhN1G5kNBDar6s+q6sN6NeG4MjLjgFft3+8Cp9lfXakVZs7MYDAYGhF1PGfWFgh/oWQGMKgyGVUNiEgukApE/yLLCjDOzGAwGBoTtXxrvohMACaErZphfwmkXjHOzGAwGBoRtY3Mwj9hVQGZQPuw5XaUf/l6sUyGiLiwPp1V6/ehmTkzg8FgMBwqVgDpItJRRDzAxcDsMjKzgSvt3xcAi/QQvCTYRGb1TCjfV73QIca7enf1QnWAM63qb5zVneIj8r2oDYr6euFvfb3guOjc9HrRe0iow4em7Tmwm7A+j+UEXlHVH0XkAWClqs4GXgb+JSKbgX1YDq/WGGdmMBgMjYk6fmhaVecCc8us+2vY7yLgwkOt1zgzg8FgaESYt+YbDAaD4ejHvM7KYDAYDIYjExOZGQwGQ2Oils+ZHakYZ2YwGAyNiIb61nzjzAwGg6Ex0UATQMycmcFgMBiOekxkZjAYDI0IM8xoMBgMhqMe85yZwWAwGI56TGRmMBgMhqOeUAONzEwCiMFgMBiOekxkdgTjOnMCjvR+4Pfi//AZdOdP5WSkdSfcZ98Gbg+hTd8QmG9/Zig2AfcFdyLJLdHcXfjffQyK9ler0z1qAo70/uD34ps1Bd1RsU7POZNtnSvxz7N0OnqchHvYpUjz9nhfvB3dvjkqO+XYE3Cdfi3icBBctYDg0vcjBZwuXGNvw9G6E3ogH/+sJyA3C+IScZ97J9K6M6EfFhH474tR6SvR2/EEXKeNR8RBcPWnBJdVoHfMrTha2npnPwF5uyE2Efc5f0BadSa05jMCn0avtz50FuMePRFHF3vfvv90xfu2TWc8500Gl4fQxpX4504HwNFzKO7hlyJp7fFOnxz1vq0PvS9ffg9jjzuJrPxsjnvwsgplnrnodkb3HEKhz8tVrz3Id9s2AHDF4NH8edTVADw075+8tnRuhfUroj73bU1oqMOM1UZmIhIUkVVhf3dVIXuOiPQIW35ARE6vbSdFJEVEbjiIeveJyB3278Eissy2YZ2I3FdN3WEi8vFBdrnWODr3Q1Lb4HtuIv6P/4F7zKQK5dyjb8D/8XP4npuIpLbB0bkfAK6hFxDashrfPyYS2rIa10kXVK8zvT/SrA3eZyfg++g5PGMq3uSesTfi+2gq3mcnIM1KdWrWVnz/eYTQ1h+jN1QcuEdMxP/2A/hm3Iyjx8lIarvIfvU+A4oK8E2bRHD5bFzDrrAKAj4CX7xBYNHM6PWF6z19Av53HsT38i04ug8tr/e406FoP74XbyC48qNSvUEfgcVvEvj81SNfZ3G76f2R1DZ4p1yH78OpeM66sUI5z1k34Jv1LN4p11nHU3rYvn3zYUJb1xzxemcumcPIqZMrLR/VcwjpLdqTfu+FTHjjUV645E4AmsYnce+Y8Qx6bDwDH7uGe8eMJyU+MTql9bhva4qGQrX6O1KJZpjxgKr2Cfv7WxWy5wAlzkxV/6qqn9a2k0AKUGNnVoZXgQmq2gfoBbxdy/YisL+YeshwdB1M8PtFAGjmBohpAglNI4USmkJMvFUOBL9fhKPrYKt+l0EEv19or19Ysr4qnF0HlerM2ACxlemMs8ptnc5uVtu6JwPdW/ajslUjbdLR7B2QswtCAULrvsTRZVBkv9IHElzzGQCh9V/jOOZ4q8DvRTPWQcBfI50A0jodzdkBuWF6Ow+sXO+Gr3F0CNObuQ4CNfsWXX3oLGm3+2CCq8L2bVwVx1Pxvl21CGf3IVad3dvQPTXbt/Wld/HmVezbn1dp+bjep5REXMu2/EhKfAKtklI5s8cgFqxbTnZhHjmF+SxYt5yRPao/b6B+921N0aDW6u9I5aDnzETkbyKyVkRWi8gTInIicDbwdzv66SQiM0XkAlv+FxF51C5bKSJ9RWS+iPwkItfbMgkislBEvhWRH0RknK3ub0Anu+7fbdk/iMgKW//9Yf26R0Q2isiXQNewLrcAdgCoalBV19ryA0VkiYh8JyJfi0h4HaqSEZGrRGS2iCwCForIayJyTli918NsqNn2TUxF8/aULGv+XiQxtRqZPSUykpACBdlWQUG2tVydzqQy7eXtRZJSK5DZW6VMTZCEZhXY2SxSKDFMRkPgLYS4KO+Yq9KbX/X2JSH1kOqtD50lupNS0dzSj7Jq7p4o9m15maNFb1W0TWnOtuyskuWM7CzapjSvdH001Oe+rSka0lr9HalEE03EiciqsOVHgU+Bc4FuqqoikqKqOSIyG/hYVd8FECn3hd9fVbWPiDwNzAROAmKBNcA0oAg4V1XzRCQNWGq3eRfQy46qEJERQDowEBBgtoicAuzH+mppH9u2b4FvbN1PAxtE5HPgE+BV+yNx64GT7S+kng48Apxfpt9VyfQFjlfVfSLyW2AyMEtEkoETKf08eP1y5B6DBoPBUGuicWYHip1IMfaQWhHwsj2vFO3c0mz7/w9AgqrmA/ki4hWRFCxn9IjtmEJAW6BlBe2MsP++s5cTsJxbIvCBqhba/SzWh6o+ICKv2/UuBS4BhgHJwKsiko51yXdXoK8qmQWqus/W8T8ReV5EmmM5u/dUNVC2MRGZAEwAeO6s4xjf/zcAOPuPxtn3TABC2zchSWklPkgSU9H8yE/Ea/5eJCmttN3EtBIZLcixhm8KsiGhKbo/pwKzwDlgDK5+ts7MTZHtlbljhvKRWEUyNUEL9pWxIRXN3xcplG/JaP5eEAfExMOB/IPWWaI3sazeMnYUWNtXCw6N3sOt0zlwDK7+IwEIZW5EkkujDElOi2Lflpc5kvVGS2bObto3bVGy3K5pCzJzdpOZs5thXfpGrP9847dRtVkfx9PBciQPFdaGgxpmtC/QA4F3gbFYkU40eO3/obDfxcsu4DKgOdDPdqC7sCK3sgjwaNg8XmdVfTmKfv+kqi8ApwG9RSQVeBD4TFV7AWdVoq8qmbIpgq8BvweuBl6ppB8zVLW/qvYvdmQAwZVz8c24Fd+MWwltWIqz93DL2LZdrSGJ4mHDYgqywVtolQPO3sMJbVgKQGjjcpy9T7PXn0Zo47IKt0lwxRy8027BO+0WguuXlOpsV5XOA1a5rTO4oeK2o0G3b0KatobkFuBw4eg+lNCm5ZF93LQcZ69TAXB0O5HQ1h8OWl+J3h0V6N28IlLv5hWlerueSOjX2uk93DqDy+fgff5mvM/fTHDdUpx9wvZt0f7Kj6fifdtnOMF1S48avdEye/Virhg8GoBBHXuSe6CAnXl7mb92GSO6DyIlPpGU+ERGdB/E/LXRHdv1cTwdLI15mLEcIpIAxKvqXBH5CvjZLsrHio4OlmQgS1X9InIqUHylL9vufOBBEXldVQtEpC3gB74AZorIo1i2nQVMt/s8BpirqooVxQWBHFtn8QzzVVX0qzqZYmYCy4GdxfNyB0No00ocnfvjuWmGlZo/+5mSMs+EZ/DNuBUA/9wXcI+7zUpp3vwNoc3WqGrgq3dxX/BHnH3OQHOzrNT8KHRqen9ibnnRSqP+cEpJWcz1z+KddgsAvjnPW6n5xTo3rQTA0W0IntETIT6ZmEvvJbRzC75//7VqpRoisOBF3Bffi4iT4OpP0T3bcJ58CbpjM6HNKwh9/ymOs27Dc/0LVkrzh0+WbotJMyAmDpwuPOmD8L91H7o3o/oNrCECn76I+8J7rVTqHxaie7fhHHoJutPWu/pTHGNuw3Pd82hRAf7ZYXonTgdPsd6B+N++v3q99aHTJrRxBdqlPzGTXypJkS8m5oapeJ+/GQDfR89bKfLuGEIbV5bu2+5D8Iy5HpokE3P5fYR2/IzvtWr2bT3pfeOaBxjWpS9pCSlse2Q29378Im6ndambvvgD5q75mtG9TmTzA+9S6Cvi6tceAiC7MI8H577Cij9a96APzH2Z7MLKE0kiqMd9W1NCR7BDqg1iXdurEBAJYg0LFvMJ8AzwIVaEIsATqvqqiJwEvIgVdV0A/AV7Dk1EfgH6q+oeEbnK/n2TreMXoL/d/kdYw4YrgcHAKFX9RUTeAI4H5qnqH0TkVuBau04B8HtV/UlE7sGap8oCfgW+VdUnROQtrPmtQiAA3KOq80VkCFam435gjt3OMSIyDLhDVcdWIRNhR9g2+wSYparTqty4QNEDZx32I6u+7q4cMfX0WKOz3NxtgyWUf3gy4o4U4vfU3XBkVRR1bF0vemPu/KDWB/POCwbX6gLQ6t2lR+QJVa0zM9QMEYnHcv59VTW3OnnjzA4Dxpk1WIwzqzk7zh1UqwtA6w+WHZEnlHmd1SHEznRcB0yNxpEZDAbD4cbMmRmqxX5A/DfVChoMBkM9cSQ7pNpgnJnBYDA0IkxqvsFgMBgMRygmMjMYDIZGxJH8suDaYJyZwWAwNCIa6jCjcWYGg8HQiGioCSBmzsxgMBgMRz0mMjMYDIZGREN9nZVxZgaDwdCIMHNmBoPBYDjqaahzZsaZGQwGQyPCRGaGOiFU4D/sOuPvqZ+PXx94/F/1ojfmtx3rRW/h7PWHXaczLe6w6wTInP9LvegtOje9XvTGbtlRL3obphs6NBhnZjDUAfXhyAyGaDDDjAaDwWA46jHOzGAwGAxHPWbOzGAwGAxHPQ31OTPzBhCDwWAwHPWYyMxgMBgaEQ30pfkmMjMYDIbGRChUu7/aICLNRGSBiGyy/zetQKaPiCwRkR9FZLWI/C6ato0zMxgMhkZEfToz4C5goaqmAwvt5bIUAleoak9gJDBFRFKqa9g4M4PBYDAcLsYBr9q/XwXOKSugqhtVdZP9ezuQBTSvrmEzZ2YwGAyNiHpOZmypqsWvT9kJtKxKWEQGAh7gp+oaNs7MYDAYGhGHYN5rAjAhbNUMVZ0RVv4p0KqCqveEL6iqikilrlVEWgP/Aq5U1Wp7bZyZwWAwNCJq68xsxzWjivLTKysTkV0i0lpVd9jOKqsSuSRgDnCPqi6Npl9mzsxgMBgaEfWcADIbKH7T+ZXAh2UFRMQDfAC8pqrvRtuwicyOYNxnT8LZbQD4vXjffhLN3FxORtr+f3vnHSZVdf7xz7sVkKUXBUGJIM1GEcRCrBjAXmKNBQy2SCyJJtGfDWPvqLHEGHtiF4MVG0aNFAtFihhRVKRJWdq2+f7+OHdgdpldMOy5w+yez/PMw5x778z33MvOfe855y2dKfzlRZBfSMXMiZSN+Yv77KBTyO05AJRAK5dR+tQtaMWPNeqN//AL/nzLqyQSCY49vDcjTt2n0v6HHv+Ap8d8TG5uDi2abcW1/3c47bdpBsBNo9/g3fdnA3DO8J8z5KCdftq5Dj2TnK7uXEufvRV9v+EUubXrTMHRF0J+AYlZEykbex8Aeb8YRm63/lBRjn6cT+mzt8HaVRvVfO/zxVz77GwSCXHMgPb8etD2lfZPnLOU656dzezvV3LLaTtxcC83vT/j22Ku+udMVq4tJzfHOHPQ9gzpk25WJT0Fx5xLXs/+qLSEkkdvJPHtFxsck9OhC4W/uhjLL6R8+keUPnO3++yQU8jbcyhauQyA0jEPUvH5hE3SzR88gpwufd01fuF2ND/NNd5mBwqOuMBd4y8mUfaKewDP6bEX+fueiLXuQMkDF6LvN/xbrI5WF/yRRnvug9auZeGoSymZPWODY1qcOZKiwYeRW9SE/x7Qr9K+xgccTIvh5yCJ0jmzWHDFJRvVtE69yDtgOGY5VEwZR8VHz1U+IDePvKG/JaftDmhNMWVjboYVi6BBEflH/B7bujOJaW9TPu6BTT7PB391KYfsvBcLi5ey86iT0h5zxy8vZEjPAawuLeG0R0bxybxZAJyyxxAuG3w6ANe88hCP/OflTdbNMq4HnjKz4cDXwC8BzKwvcJakM6JtA4GWZnZa9LnTJH1a0xfXq5GZmVWY2acpr3RuoanHv2xmzaLXOf+D3pVm9rv/pa853XYnp1U71t44jNJn76DgyN+kPa7gyPMoffYO1t44jJxW7cjp2heAsnefYe1tZ7P29nOpmDGBvAPT/7iSVFQkuPrGl/nrHScx9p/n8q/XpjHnv5VnALp33YZnHx7BS0+cw8H79+Cm0W8A8M6/Z/P5rPm88NhZPPXQr3nwsQ9YuXLtpp/rjn2xVu0pufUMSl+4k4LDqjnXw8+l9IU7KLn1DKxVe3J2dOeamPMJJXeeTcnoc0ks/o68n/9yo5oVCTHq6Vncf/ZuvHTpAMZO/oE581dWOqZd8wZcd3IPhvapvEbdoCCH63/Vk39dOoAHzunFdc/NZsXqTSvlk9ujHzmtt2X1VadQ8uStFB7/27THFR53PiVP3Mrqq04hp/W25PZYf4Mve/sZ1lx/JmuuP3OTDVlOl75Yi3aU3DmC0pfuomBo+j/ngkPOpfSl0ZTcOQJr0Y6czn0A0MKvKf3ntSS+nr5JekkaDdiH/A4d+ebYISy8/kpaX/x/aY9b9e93+Hb48Rtsz9+2I81POYNvz/wV8046gsW337BxUcsh/8ARlD09itIHR5LTfW+s5baVDsnZ+UBYu4rSB86hYtJL5O17ittRUUr5e09S/s7Dab64Zv7+4Vh+MfqCavcP7jmALm060OWKYxnxxHX85YSLAWjeqAlXDB1O/xuG0++GYVwxdDjNGhX9ZP1NJZMjM0lLJB0gqYukAyX9GG2fFBkyJD0mKV/SbimvTzf23fXKmAFrqlyg62s6WNIQScuAZsBPNmabQ26PAZR//CYAiW9mYg0bQ1GLygcVtcAaNCLxjSs3Uv7xm+T23NPtK1m9/riCBqCaXZimTP+O7bZtQYf2LSjIz2PooJ14c/ysSsfs0bcTDRsUALDbztvyw8IVAMz5ahF9e21HXl4ujRoW0LVzW8Z/uOlP7rnd96DiE3eumjcLGmwFRVViKYuaQ2Ejtx+o+ORNcrvvAThjlvyVJebNxJq02qjmlK+X07FVQzq0akRBXg5D+rTlramLKh3TvmVDurYvIses0vZObbZi+zaNAGjTtJCWjQv4cRPr0uXtshflE153fZ07A2vYGGtS+f/VmkT/r3PdCKZ8wuvk7bLXJn1/deR27U/FZ28BoG+ja9y4yjVu3BwKG7r9QMVnb5HbzV1jLf4WLfnuJ+tuNXA/il8ZA0DJ9CnkNC4it+WG/z8l06dQsWTxBtubHH4My5/5B4li97dWsbTm2QUA26YLWjYfli+ARDmJGf8mp3Pl0V5ul35UTHsbgMSsD8jpuIvbUVaCvpsB5aU/6TwB3pvzKT+uWlHt/sN3HbhuxPXRV9Np1qgxWzdpycE9+vPGjAksXb2CZauLeWPGBH7RY4+frL+pZHia0Rv1zZhtgJk1NbNZZtY1aj9pZr+O3s81s1a4ofEO0Wjupmjf781sYhShflXK911qZrPN7N9A1/+1XzlNW6Jl62+uWraInKYtNzxm+eJqj8k/+FQa/OlR8nrtR9nrNRfGXLBoBVu3bbKu3bZNExYsqv6H+cyYjxk4oDMA3bq05b0P57BmbSk/LlvFR5O/WmfoNgVr0gotTznXFYs3MEjumJRzXb7hMQB5fQaRmD1po5oLl5WwdfMG69ptmzVgwbKSTe5zkilzl1NWkaDjJhbFtGatSCxdf66JZYuwZq02PGZZ6jGLKx2TP/AIGv7xAQpP+h00bLxpuk1aohUp12/FEqxJyzTHLKnxmJ9KXuu2lC/4YV27fNEC8lrX6I1difwO25HfcTva3/co2z7wOI322LhRt8YtUHHKuRYvwYqqnEfjlOuhhHv4a+hvNATQvllr5i1dP9vx7dKFtG/Wutrtvqirxqy+rZk1NLNPU9rXSfqnmf0G+LuZ3QE0l1R1ovwPwE6SdgMws0FAF6AfYMAYMxsIrAKOB3bDXduPgcn+Tqdmyl57mLLXHiZvv+PI3/NQyt54rFa+98VXPmPajO957F43x7/3Hp2Z+vn3HD/8QVo034rddu5ATo5t5Ftqn7x9j4NEBRWfvR2L3sLlJVzy6HSuO7lnbOdb9t5LlL7yGCAKDjmdwqPOouTxm2PRzgSWl0d+h+347pzTyWvTlvZ/eZh5Jx9JYmVxpruWtWzJBmlzqG/GbE3SIKUi6Q0zOxa4G9h1E75nUPT6JGo3xhm3IuB5SasBzGxMug+nxmmMHtSDYbt2ACBvwKHk9f8FAIl5s7GUpzNr1prE8iWVviexfAnWtFWNxwBUfPIWhcNG1WjM2rZuwg8L1o+mFixcQdvWTTY47oMJX3LvQ+/x2L2nUVCw/s/n7GEDOXvYQAAuuuwZOnWs+Yk+t/8h5O1+sDuPb7/Amqaca5NWlUYREI3WUs+1aeVjcnsdSG7XfpT87U816iZp06yQH5auX9dbsGwtbZsVbtJnAVauKeesez/l/EN2YLdOTWs8Nn/g4eTtOQSAxNezyGnemuT9JKdZa7SsyrkuW0xOyv99TrNW645R8dJ128veH0uDs/5crW7u7kPJ6xNd4+++qDSSrToKgw1HYumO2RSaHn08TQ47BoC1M6aR13a9c0xe67aUL1qwyd9VvnABa6dPgYpyyud/R9m8ueR32I6SGdOq/YxW/ogVpZxrUUtUXOU8Vi5xf2crl4DlQGEjWOPXQH63bBEdmrdZ1962eRu+W7aI75YtYt8de1fa/s7sj732pS5S76cZAcwsB+iOywm2QeLLdB/BjeqSa2+dJT24qXqS7pfUV1LfpCEDKP/wJdbefi5rbz+X8ukfktf7AAByOnZDa1ZBcZX1guIf0drV5HTsBkBe7wOo+PxD18FW7dYdlttjAImF82rs08492jF33hLmfbeU0rJyxr4+jf33qTxL+vms+Vx+3b/4y80n0LLF+umtiooES5e5NbqZX/zArDkL2Kv/DjXqVXz0L0ruOo+Su86jYsaH5PZy52odukLJKki5abtzXQolq91+ILfXAVTMcOEnOV36kDfwGEoevQrKNm2qcOeOTfh60Rq+XbyG0vIEL09ewH47b9rUTml5gvP++hmH99tmnYdjTZSNf3Gdw0b5lPfJ6zfI9Xv77mjNqg28TLUi+n/dvjsAef0GUT7lfYBK62t5u+5NYv7canUrJo6l5N6RlNw7koqZH5K76/7uO7bt6qbVVla5xiuXQskatx/I3XV/KmZ9tNHzq8ryZ//BvFOPYd6px7Bq/FsUDT4MgMKeu5BYtTLt2lh1rBr/Jg177w5ATtNm5HfYnrLvav5b1vwvsObbQNM2kJNHTve9ScyZWOmYijkTyd1pP/e9Xfck8c3Un3KK/xNjprzHKXu4h5r+nXqyfM1KflixhNc+/4hB3fvTrFERzRoVMah7f177/Kdf900lTDPWbS4AZgB/Ah4yswGSUlf0i3GjriSvAaPM7HFJK82sPVAGjMdNV16Hu7aHAvf9Lx1KzJxAotvuNLjkb1BaQunTt67b1+D8u1l7+7kAlL1wFwW/vAjyC6iYOYnETPejzR88jJzW24KEli6g9LnRNerl5eVy+e+HcMbIR6lIiKMP7UWXHdpwx31vsVP3dhwwsBs33vk6q9eU8ts/PgXANls35d5bTqS8vIKTzvwbAI23KuSmq48iLy9308911kS04+4UXvigcxt/7rZ1+wp/M5qSu84DoHTMPRQcfQHkFZL4YtK6tbH8Q8+G3HwKh7lRSmLeLMpevKvm883N4bJju3LGPZ+QkDhqj3Z02aYxd479kp06NmH/nVsz9evlnPfXKaxYXcbb0xYz+uX/8q9LB/DqJwuYNGcZy1aV8cJH3wNw7ck96b7txtdcKqZ/RG7P/jS64lFUtpaSx25at6/hH+5jzfVnAlDy1B0Unhy55n8+YZ3XYsERI8jZdgcQ6McfKHnytrQ6G1zjLyahLn0pHPmAu8Yv3r7+Gp91JyX3jnTXeOw9zjU/r4DEnMkkvnDXOKfbAAqGnAmNmlJ44hUkfviK0scu36ju6g/G02jPfdju6VdIlKxh4TXrvRk7PPwM8051I7iW515I0aAhWIMGbP/iOFaMeY4fH7yH1f95n0b99qTjEy+iRAVL7rqFxIrlNYsqQfm4B8g/9grnmj/1TbRkHrl7n4B+mENizkQSU8aRM/R8Cn59D1q7krIxt6z7eMGZ90FBQ8jNo6BLP8qeugot+Xaj5/rEsKvZd8fetGrcjHnXjuGKfz1Afq67xd733vO8PO0Dhuy0J3OufobVpWs5/ZFrAFi6egWjXv4bEy9xv6GrX36Qpas3fc35p6KNOINlK1ZXTywdZlYBpD6CvQo8BLwA9JNUbGa3AsWSrjCzuUBfSYvN7AlgF+AVSb83s98CZ0TfsxI4WdKXZnYpLhhwIfAN8LGkahc1Vl/8i9j/AxpdeurGD/LAmhtrdkLxReHPO8WuuXrMzNg1AXI30RGltvnutbkZ0e1wZJeM6Db4av7GD/KA/vKfzV6cHdem62bdcw5cOCv+BfFNoF6NzCRVN1zonnLMhSnvt095f2KV77oDuCONxp+B6hcyAoFAIFDr1CtjFggEAvWdLXnda3MIxiwQCATqEcGYBQKBQCDrCcYsEAgEAllPXTVmIc4sEAgEAllPGJkFAoFAPaKujsyCMQsEAoF6RDBmgUAgEMh6EnU0T0YwZoFAIFCPqKsjs+AAEggEAoGsJ4zMAoFAoB5RV0dmwZgFAoFAPSIYs0AgEAhkPXXVmNWrEjB1DTMbIen++qBbn861vunWp3PNpG5dJziAZDcj6pFufTrX+qZbn841k7p1mmDMAoFAIJD1BGMWCAQCgawnGLPsJlPz7pnQrU/nWt9069O5ZlK3ThMcQAKBQCCQ9YSRWSAQCASynmDMAoFAIJD1BGMW2CIxx8lmdnnU7mhm/TLdr0AgsGUSjFmWYWYNzaxrpvsRA/cAA4ATonYxcHdc4ma2t5mdHr1vbWadYtDcwcwKo/f7mtlIM2vmWzduzCzXzGZmSPtQMwv3vTpISGeVRZjZocDNQAHQycx2A66WdJhn3bbAtUA7SYPNrAcwQNKDHmX7S+ptZp8ASFpqZgUe9dZhZlcAfYGuwENAPvAYsJdn6WeBvmbWGefx9iLwBDDEh5iZHVXTfknP+dCVVGFms8yso6RvfGjUwHHA7Wb2LPA3SV6NqpldWNN+Sbf61K9PBGOWXVwJ9APeAZD0aRwjBuDvuJv6pVF7NvBPwKcxKzOzXEDgRkdAXFnljgR6AR8DSPrezIpi0E1IKjezI4HRkkYnjbknDo3+bQPsCbwVtfcDPgC8GLOI5sB0M5sArEpu9P1gJulkM2uCG/H/3cyE+9t+UlKxB8nk301XYHdgTNQ+FJjgQa/eEoxZdlEmabmZpW6LI7ailaSnzOyPANENt8Kz5p3A80AbM/szcAxwmWfNJKWSFN3oMLOtYtItM7MTgFNZb2jyfYlJSk6jvg70kDQ/am+De4Dxyf95/v5qkbTCzJ4BGgLn4x5efm9md0oaXctaVwGY2Xigd9JgmtmVwNja1KrvBGOWXUw3sxOBXDPrAozEPUH7ZpWZtWT9KGkPYLkvsWhN4yvgYuAAwIAjJM3wpVmFp8zsPqCZmf0aGAY8EIPu6cBZwJ8lfRWNuh+NQbdD0pBFLAA6+hSU9K6ZbQd0kTTOzBoBuT41AczscOA0oDPwCNBP0sJI/3OgVo1ZCm2B0pR2abQtUEsEY5ZdnIeb6ivBraW8BlwTg+6FuOmRHczsfaA1bqTkBUkJM7tbUi8gdkcBSTeb2UHACtz00OWS3ohB+iBJI1P68ZWZrY1B900zew14MmofB4zzKRg9JIwAWgA7AO2Be3EPLz45ErhN0vjUjZJWm9lwj7qPABPM7PmofQTwsEe9ekfIAJIlROtH4yTtlyH9PNyN3YBZkso8690MfAg8p5j/SKMR0XxJa6N2Q6CtpLmedT+W1LvKtk8io+6VaJ1uYNQcL+n5mo6vBb1Pceu/HyXPz8ymStrZo2amf0O9gX2i5nhJPtdD6x1hZJYlRB5gCTNrKsnbFF860ni97Whmy4GpkhZ6kj0TNyIsj0YnBkhSE096qTyNc4hIUhFt292HWLROdiLOQ3VMyq4i4EcfminaucB0Sd1wa5RxUSKpNLn+Gz0seX1oyeRvKKIRsELSQ8lwD0lfZaAfdZJgzLKLlcBUM3uDyh5gI6v/SK0wHBfz9XbU3heYjLv5Xi2p1td1JMXhPVgdeZLWrW9EN12fYQEfAPOBVsAtKduLgSkedTPpJv+umf0JaBhN6Z4DvBSDbkZ+QxkM96g3BGOWXTyHX3fp6sgDuktaAOvizh4B+gPj8eCkYGYD022vutbhiUVmdpikMVFfDgcW+xKT9DXwtZmdBHxfZXpzW2CuL+2ITLjJ/wH3kDQVNwp/GfirR70kmfoNZSrco94Q1syyjGiEsGPU9L52FWl+LqlHSttwU1M9fK3pmFnqU3oD3PrKZEn717ZWGu0dgMeBdrjpzXnAKZLmeNadBOyZHBVG/9fvS/IyvZmi+/N02yW961m3AOiGm16clToarmuY2QRJ/ZLrolG4x4eSdsl03+oKYWSWRZjZvjgPqLm4m2wHMzs1htHKO2b2L9y6EcDR0batgGU+BCUdmto2sw7A7T600mh/CexhZo2j9so4dIl/ejOp49VopcPMhuK8F7/E/S13MrMzJb3iWbcLcB3QA/eQBICkn/nUJX24Rxwj0XpDGJllEWY2GThR0qyovSMuc0Efz7oGHAXsHW1aivPuO9enbpo+TE8dIXrUKsQZ7O1JeeCTdLVn3TdwmT9SpzdHSvLqrh7FDY4GuuNSpeUCq3w625jLzXhIcrQbjYbHRo4o3jCzfwNXALfhAtNPB3IkXe5TN9I+CBiEM96vxRTuUW8II7PsIj9pyAAkzTYzbxkiUnRkZv8F9gCOxQU0P+tT08xGs967LQfYjWi9IQZexAWFT8bF9MXFWcDjZnYXKdObMejeBRyPG3n3jTR3rPETm09xlWnb/+IcXnzTUNKbZmbRWuWV0UOiV2NmZjdIugR4I822QC0QjFl2McnM/orzggI4CZjkSywa+Z0QvRbj8jFaTHE6qedVjhuBvh+DLsC2kn4Rk9Y6Mji9iaQ5ZpYrqQJ4yFxOyD/Wtk5KmMckM3sZeAr30HIsMLG29dJQEmWY+cLMfgN8BzSOQfcgoKrhGpxmW+B/JBiz7OJs4FxcGiuA93ClUnwxM9JInQ66wKPeOiRlMjvCB2a2s6SpcYpaVLstpQ34n94EVkdrc5+a2Y24MAFfZVJS10IXAEnnk0W4XIm++S0u3mskMArYH5cL0wtmdjYu7OBnZpYaZlEExPVwVi8Ia2ZZRORwsTZ6ek4GvBZKWu1J7wjc9NNewKvAP4C/SoqjttdeuCoB2+EeupJB074X6jGzz3G5+77CTTMmtb16npnZRSnNBsAhwAxJwzzrboczLAXABUBT4B7f3pv1ATNrigt9uA4XjpCkWJLXgPj6RjBmWYSZ/Qc4MDn9FE1HvS5pz5o/udm6WwGH46Yb98fFmD0v6XWPmjNxN9bJuAwcAEha4kszRXu7dNujNZbYiBxRXpO0r6fvbw20lvR5le09gYWSFvnQjTQ64XKNbk9lJxsvsW1RqEe1NzuPuk3ksvS3qEY3GLRaIkwzZhcNUtdRJK00l+3bK5JW4RIbP2FmzXHrG5cA3owZsNy3m3Z1JI2WmbUhxX07AzTCBU37YjTpp6lb4BJan+hR+wVcPbyXiKdO3c0xaKTjCdwIezLOmKbWbxLgfaahvhBGZlmEuYz150n6OGr3Ae6SNCCzPat9zOx6nIv4c6R4FCbP3bP2Ybi0Uu2AhbipzhmSenrWncr60UMurjrB1ZLu8qQ3SVLfavZNk7STD93o+z+S1N/X9wfqH8GYZRFmtjtu3ep73BPe1sBxkiZntGMeMLO302xWTBlAPsNNp46T1MvM9gNOluSzREjV6c1yYIGkco96syR1/an7akn7RKALbnQf28NKpoKmzWy4pAdT2rnAZYqKdwY2nzDNmEVImmhm3XDJSiGmdFaZICb3/+ook7TEzHLMLEfS22Z2uy+xlPWUqnFWTczM57rKHDMbIunlKv0ZjIv78snOwK9wDw3JaUZFbZ88xPqg6f2IgqY9awIcYGZH4/JRtoz6EXvmlbpMGJllAdGIbJ6kH6L2KbgMFV8DV9bFRWRzyYyvBdpJGmxmPYABqU+3HrXH4YonXofLZL8Q2N2Xo42ZfcX69ZSOuAwrBjQDvvHlPRqNUsbisvYnR/d9cRUSDpE024dupD0H6BF3PkYzmyypj6XUTktui0H7OOBuXDLnE2OMm6wXxPFEEth87iMquW4um/z1OI/C5cD9GeyXT/6Oq6TdLmrPBs6PSftwYA3Om/JVXP7AQ2v8xGYgqVM0zTUOOFRSK0ktcY4D3pxsJH2BGyG9i/Mq3D56v4tPQxYxDWes46ZS0LS5oqTeg6ajB4ff4jLnfA38Kg7nrfpEGJllAWb2maRdo/d3A4skXRm1P5W0Wwa7V6uYWZ6kcjObKGl3S8nKX9fOtSqWptJyum21rJmR6stm9g6wCy7rR+qamc+yM8lZjhk4QzoKF1N3o6T/eNadCZybTKWFKzw7zLdTUX0irJllB7nJmzxwADAiZV9d+z+cAPQGVplZSyLvvigZrtfqwGZWTPpYpLiqXH9vZpdROV3Z9z4Flbnqy1fEqLUOScmUWSvN7EJgmeJ5ou8naUXUBwG3WOUyR4HNJEwzZgdP4irzvoib/noPwMw64/kGnwGScTgXAmOAHaKQhEdwQbbekFQkqUmaV1EMhgxcUHpr4Pno1Sba5ptk9eUHzezO5MunoFzZmbm45Nnv4kZo3jwZzezyyHkKMyuMvGW/BBaY2YEedS8GiAKnj62y+zRfuvWRMM2YJUQjk21wGT9WRdt2BLaS9ElGO1eLmNm3wK1RMwcoxBm4EqBC0q3VfdZDXyoFTUv6JibdIicXT6JhM0ubm9BnfkxzNb1GAC0k7RCtKd0rT+VuzGw6sJMkmdkI3EPCgbjqAA9L6udJ92NJvau+T9cObB51bYqqziLpP2b2qKTnU7bNNrNHcS7OdYVc3IK8Vdke22J5dUHTgO+g6Z1xI9AWUXsxcKqkaZ70kumsHq6yvSfuvH1yLq56+EfgnFGihwdflKZMJx4M/EMux+kMM/N5H7Rq3qdrBzaDYMyyi0o302jx3rtLcczMl/8s8RtjFK52W6Wg6Rh07wMulPQ2kKwsfj/gK/dmJtNZlchV0gac4w815E6sDT0z2wmXUHk/4Hcp+3w+KKma9+nagc0grJllAWb2x8g5YRczWxG9inFPzy9muHu1zZbwtFoml9B4XdA0Lv7KN1slDRmApHeArTzqdZY0vupGSe/hPA198q6Z/QloaK4C89O4PI2+OB94BlfW6DZJXwGY2RDA5zT9rim/16q/X29eqvWRsGaWRZjZdZJqvWDiloSZtch0EHjcQdMpus/jnCAejTadDPSRdKQnvUyms8rBZcMYhHuAeQ1XXijckAL/E8GYZQFm1k3STDNLu1jsO59dfcNcyZs1uJmLk3CxSI/Lc/kZcxUJrgL2xk1BvQdcJWmpJ72xwN3VpLMaKWmwD91MELnhV0ucjkUBP4Q1s+zgIuDXOKeEqsSRz65ekfQWBRLRDX+J7xFDtP75XMzBy+cDY83sl6RJZ+VD0CpXW94A+SuAWhT92xXYHRf2AS6zywRPmoEYCSOzQCAiCn+4HvgR5wTyKG6aMQc4RdKrnvXfBI6KM3jZXAHQE4FkuZfpwBOS1nrS+xT3APYEbo1sTep+eS6AambjgaGSiqN2ETBW0kCfugH/BGOWBZjZUTXtl/RcXH2py5jZJOBPuGnF+4HBUUhEN+DJZFotj/ovAr2AN3DJaAGQNNKz7lbA2igbyI5AN+AVearIEF3PE3Cjos9xhu11eSx3k6I9C5d7siRqFwJTfK4PBuIhGLMswMweit62wblpvxW19wM+kORlSqi+kZr70cxmSOqesu+TGIxZ7MHLke5kYB+gOfA+LhtHqaSTfOpG2slM8jdIuikGvUuBX+IyrIBz9HlK0rW+tQN+CWtmWYCk0wHM7HVc2Yz5UXsbXHb5QO2QSHm/pso+32tmR+BSWU2V9JpPrXTyklab2XDgHkk3RtOBfsTM2gPHA0fiyt1cwHrj4hVJfzazV3DGG+D0upRBpz4TjFl20SFpyCIW4OpfBWqHXc1sBc5VvGH0nqjdoPqPbR5mdg8uIP4DYJSZ9ZM0ypde+i7YAJznZrKadq4noXdxzhhP4QpjJj1EC2IMy2gErJD0kJm1NrNOybizQPYSphmzCDO7C1dq/slo03HAHEleE/AG/GJm04BdozWrRsB7iqFYZIr+QFxGjPcl3WBmPwPO97FWZ2ZzWT/KTb35JCsT/Ky2NavoX4Hz2OwqaUczawc8LWkvn7oB/wRjlmWYKyaY9Lwan5qrMZCdhAS08RFNn/YCPtb6OnlTPIYEBGIiTDNmHx8DxZLGmVkjMytKuhkHspZuKfFXhit7M4X1oxWvN9oo4fDFuKnO1CoB3uIXzSVlPAnoJGmUmXUEtpbkO+arNMqcn6yT5zNdWCBGgjHLIlLLZgA7AO2Be3EFOwPZS282dDiJk8eBf+ICpc8CTgUWeda8B+dwsz8upq8YeBYX0OyTp8zsPqBZ9HsaBvzVs2YgBsI0YxYRTZH0Az5KmSKZKikkLM1iktOKUYmf2Mv5mNlkSX1Sp9vMbKIkb4Yl5Zw/Sflb/kzSrr40U7QPIiUnpKQ3fGsG/BNGZtlF3GUzAvFQYGYnAnumC5CPISg+GRw938yGAt8T1VTzqRml8EpO97WmcmiEF8zsBkmX4ALTq24LZDHBmGUXVctmnIPfshmBeDgLt37UDJcVIxUBvo3ZNWbWFJcDdDTQBBf75ZM7cbFlbczsz8AxwGWeNQEOAqoarsFptgWyjDDNmEVEi+ZnEMpm1EnMbLikBzPdj7iI0lodgPtbflPSDI9aZ+Me/n4GfJmyqwgXkhBH8dWAR4IxyxKiKZnpkrplui+B2sfM2gC/AXpEm6bjyrMs9Kg5mhqmqX3mhDSzO4F/SPrAl0YVvaa4dF3XAX9I2VWc6fp5gdohVJrOEiRVALMiF+ZAHcLM9sLlQxTwSPQCmBDt88UkXOmXycBhKe+TL59MBi4zsy/N7GYz81rJW9JySXMlnRBl5l+Du96Nw2+qbhBGZllEVL6iF67+UmpW9cMy1qnAZmNm/wHOrpoj0Mx2A+6T1D+GPnhPpFyNbgvgaFyuxo6SunjWOxS4FWiHqyC+HTBDUk+fugH/BAeQLMDMOgNtgf+rsmsfYP6GnwhkGU3SJbuV9GlUbysOMvVU2xlXcmY7wNuaWQrXAHsA4yT1MrP9gLBeVgcI04zZwe24xKjvpr6AF3ElLALZjZlZ8zQbW1BHf6NmdqOZfQFcDUwD+kqq6snpgzJJS4AcM8uR9DYuV2Mgywkjs+ygraSpVTdKmmpm22egP4Ha5TbgdTP7HS5dGUAf4IZonxfMrJj1I7JGVaoESFITX9o4j8IBkhZ71EjHMjNrDIwHHjezhaRM2Qeyl7BmlgWY2RfVrSWY2RxJnePuU6B2MbNDWJ8fEZw3402S6lQcoZl1kzTTzNImUpb0cbrttai/FbAWZ7BPwlUVfzwarQWymGDMsgAzexJ4S9IDVbafARwk6bjM9CwQ+GmY2f2SRpjZ22l2y2dy4yr9aELKzFRwz89+gjHLAsysLS5bQinrXab7AgXAkZJ+yFTfArWHmXUCzgO2p/KNts55q5pZA0lrN7bNg+6ZwFW40VmCmOqoBfwTjFkWEXle7RQ1p0t6K5P9CdQuZvYZ8CAwlZQ8hZGzT50iXc22OOq4RU4nmVirC3gmOIBkEZHnVbrpmUDdYK2kOzPdCZ+Y2da40kUNzawXbmQELh9koxi68CWwOgadQMyEkVkgsIUQZc7vArwOlCS3+3aKiBMzOxU4DTdNPpH1xmwF8LDvCgGRAX0I+IjK19hb6q5APARjFghsIZjZdcCvcKOH5DRjbE4RcWJmR0t6NgO6E4B/s+FU7sNx9yVQu4RpxkBgy+FY4GeSSjPdkRjoY2ZvSloGEAWNXyTJdxmYfEkXetYIZIA6mV0gEMhSpuFqmtUHBicNGYCkpcCQGHRfMbMRZraNmbVIvmLQDXgmjMwCgS2HZsBMM5tI5fWcOueaD+SaWaGkEgAzawgUxqB7QvTvH1O2CVfnLJDFBGMWCGw5XJHpDsTI48CbZvZQ1D6d9aVvvCGpk2+NQGYIDiCBQCAjmNkvgAOj5huSXvOotb+kt8zsqHT7fXtRBvwTRmaBwBZClcS/BUA+sMpzwt+MIelV4NUoX+JRZjZW0lBPcj8H3gLSZeYXEIxZlhNGZoHAFoiZGXA4sIekP2S6P7WNmRUAQ4ETgYOBZ4HnfCdWNrNOkr7a2LZA9hGMWSCwBZOpCtC+MLNBOCeMQbhsNv8ERkvaPib9dGm0JkvqE4d+wB9hmjEQ2EKosp6Tg8uS4TXxbgZ4FXgP2Ds5GjKzO3yLmlk3XHmdplWucxOggW/9gH+CMQsEthxS13PKgbm4qca6RG/geGCcmf0X+AeQG4NuV+AQXPhD6nUuBn4dg37AM2GaMRAIZAQz2xM35Xg08BnwvKT7PWsOkPShT41AZgjGLBDIMGZ2eQ27JWlUbJ3JAGaWAxwAnCBpmGetG4FrgDW4Kc9dgAskPeZTN+CfkM4qEMg8q9K8AIYDl2SqUz4xs70il3xwHo2DcUUzfTNI0grclONcoDPw+xh0A54JxiwQyDCSbkm+gPuBhriMGP+g7qZZ+guw2sx2BS7CVQrwngEEF7sHLizgaUnLY9AMxEAwZoHAFkCU8PYaYArOMau3pEskLcxw13xRLrfGcThwl6S7gaIYdF8ys5lAH1w6rdbUPY/ReklYMwsEMoyZ3QQchRuV3S1pZYa75B0zexe3ZnU6MBBYCHwmaecYtFsAyyVVRFOdRZJ+8K0b8EsYmQUCmecioB1wGfC9ma2IXsVmtiLDffPFcbjKAMMjQ7ItcJMvMTO7OKV5gKQKAEmrgFBlug4QRmaBQKDOk5r5o2oWkHRZQQLZRwiaDgQCsVElmXKlXbgwBF9Jla2a9+nagSwkGLNAIBAbkuJw8kgrXc37dO1AFhKmGQOBQJ3HzCpw8XuGC31YndwFNJCUX91nA9lBMGaBQCAQyHqCN2MgEAgEsp5gzAKBQCCQ9QRjFggEAoGsJxizQCAQCGQ9wZgFAoFAIOv5f8D9rWX1kF70AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 6))\n",
    "sns.heatmap(df.corr(), annot=True, fmt='.2f', cmap='RdYlGn')\n",
    "plt.title('Корреляция между признаками')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Видно явное влияние возраста на целевой признак. Также из всех признаков заметно влияние признаков 'IsActiveMember' и  'Balance'. Данный факт не поможет модели обучения, но было интересно это увидеть (я попробовал удалять признаки с 0.01 по модулю значением перед обучением модели, значимого результата не было получено)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Перекодируем категориальные признаки\n",
    "df = pd.get_dummies(df, drop_first=True)\n",
    "df.head(3)\n",
    "df0 = df.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    9091.000000\n",
       "mean        4.997690\n",
       "std         2.894723\n",
       "min         0.000000\n",
       "25%         2.000000\n",
       "50%         5.000000\n",
       "75%         7.000000\n",
       "max        10.000000\n",
       "Name: Tenure, dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Tenure.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "RS = 123\n",
    "\n",
    "#выделяем часть данных, где имеем пропуски, сохраняем и потом удаляем их из общих данных\n",
    "df_tenure_pred = df[df['Tenure'].isna()]\n",
    "df = df.dropna()\n",
    "\n",
    "#Выделяем фичи и целевой признак для обучения\n",
    "X = df.drop(['Tenure'], axis=1)\n",
    "y = df['Tenure']\n",
    "\n",
    "#берем фичи, где хотим предсказывать целевой признак\n",
    "X_pred = df_tenure_pred.drop(['Tenure'], axis=1)\n",
    "\n",
    "#разбиваем выборку на обучающую и тестовую\n",
    "(X_train, X_test, y_train, y_test) = train_test_split(X, y, test_size=1/4, random_state=RS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9091, 12)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(909, 12)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tenure_pred.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 69, 'max_depth': 5}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# подтягиваем код который уже писали ниже. На этом этапе уже стоило бы написать функцию, \n",
    "# но так как все проект правлю на коленке, просто скопипастю\n",
    "\n",
    "params = {'n_estimators': range(5, 100) , 'max_depth': range(1, 20)} \n",
    "rfc = RandomForestClassifier(random_state=RS)\n",
    "clf = RandomizedSearchCV(rfc, params, n_jobs=-1, cv=2)\n",
    "clf.fit(X_train, y_train)\n",
    "clf.best_params_  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10910690717113947"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.score(X_test, y_test) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 2., 2., 1., 3., 8., 2., 2., 1., 1., 2., 7., 8., 1., 2., 7., 7.,\n",
       "       1., 8., 8., 2., 8., 1., 9., 2., 1., 5., 8., 4., 9., 8., 7., 7., 2.,\n",
       "       4., 1., 5., 1., 7., 7., 6., 8., 1., 8., 1., 1., 3., 1., 1., 1., 5.,\n",
       "       1., 5., 1., 7., 7., 1., 5., 9., 1., 2., 5., 5., 2., 8., 2., 5., 7.,\n",
       "       5., 1., 8., 7., 9., 1., 1., 2., 4., 1., 2., 6., 8., 8., 2., 8., 5.,\n",
       "       1., 2., 2., 8., 5., 5., 4., 2., 7., 8., 8., 5., 2., 7., 1., 2., 1.,\n",
       "       1., 1., 1., 7., 2., 7., 5., 1., 2., 2., 1., 5., 1., 1., 3., 5., 1.,\n",
       "       1., 1., 1., 9., 1., 2., 8., 1., 5., 1., 1., 5., 1., 1., 7., 2., 1.,\n",
       "       4., 9., 1., 1., 1., 8., 6., 5., 2., 1., 2., 1., 5., 7., 5., 1., 2.,\n",
       "       1., 8., 2., 2., 1., 2., 2., 4., 7., 2., 2., 8., 1., 5., 1., 2., 1.,\n",
       "       9., 6., 7., 7., 2., 8., 3., 5., 7., 8., 7., 1., 7., 3., 7., 1., 2.,\n",
       "       5., 1., 5., 1., 8., 5., 5., 2., 7., 7., 1., 2., 1.])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = clf.predict(X_pred)\n",
    "y[100:300]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Похоже, что модель предсказывает адекватные данные."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Age</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "      <th>Geography_Germany</th>\n",
       "      <th>Geography_Spain</th>\n",
       "      <th>Gender_Male</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>591</td>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>140469</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>550</td>\n",
       "      <td>38</td>\n",
       "      <td>103391</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>90878</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    CreditScore  Age  Balance  NumOfProducts  HasCrCard  IsActiveMember  \\\n",
       "30          591   39        0              3          1               0   \n",
       "48          550   38   103391              1          0               1   \n",
       "\n",
       "    EstimatedSalary  Exited  Geography_Germany  Geography_Spain  Gender_Male  \\\n",
       "30           140469       1                  0                1            0   \n",
       "48            90878       0                  1                0            1   \n",
       "\n",
       "      0  \n",
       "30  5.0  \n",
       "48  1.0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tenure = pd.concat([X_pred, pd.Series(y, index=X_pred.index)], axis=1)\n",
    "df_tenure.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Age</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "      <th>Geography_Germany</th>\n",
       "      <th>Geography_Spain</th>\n",
       "      <th>Gender_Male</th>\n",
       "      <th>Tenure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>591</td>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>140469</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>550</td>\n",
       "      <td>38</td>\n",
       "      <td>103391</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>90878</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    CreditScore  Age  Balance  NumOfProducts  HasCrCard  IsActiveMember  \\\n",
       "30          591   39        0              3          1               0   \n",
       "48          550   38   103391              1          0               1   \n",
       "\n",
       "    EstimatedSalary  Exited  Geography_Germany  Geography_Spain  Gender_Male  \\\n",
       "30           140469       1                  0                1            0   \n",
       "48            90878       0                  1                0            1   \n",
       "\n",
       "    Tenure  \n",
       "30     5.0  \n",
       "48     1.0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tenure.rename(columns = {0:'Tenure'}, inplace = True)\n",
    "df_tenure.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_tenure['Tenure'].hist(bins=50);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 12)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.concat([df, df_tenure], axis=0)\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    10000.000000\n",
       "mean         4.826100\n",
       "std          2.915529\n",
       "min          0.000000\n",
       "25%          2.000000\n",
       "50%          5.000000\n",
       "75%          7.000000\n",
       "max         10.000000\n",
       "Name: Tenure, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Tenure.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Показатель точности модели оставляет желать лучшего. Но улучшить показатели мне не удалось. Возможно тут стоило использовать другой подход к классификации или возможно, использовать регрессию. Но в любом случае характер распределения почти не изменился, данный подход, полагаю, будет лучше, тем заполнение значений случайными значениями."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 10000 entries, 0 to 9999\n",
      "Data columns (total 12 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   CreditScore        10000 non-null  int64  \n",
      " 1   Age                10000 non-null  int64  \n",
      " 2   Tenure             10000 non-null  float64\n",
      " 3   Balance            10000 non-null  int64  \n",
      " 4   NumOfProducts      10000 non-null  int64  \n",
      " 5   HasCrCard          10000 non-null  int64  \n",
      " 6   IsActiveMember     10000 non-null  int64  \n",
      " 7   EstimatedSalary    10000 non-null  int64  \n",
      " 8   Exited             10000 non-null  int64  \n",
      " 9   Geography_Germany  10000 non-null  uint8  \n",
      " 10  Geography_Spain    10000 non-null  uint8  \n",
      " 11  Gender_Male        10000 non-null  uint8  \n",
      "dtypes: float64(1), int64(8), uint8(3)\n",
      "memory usage: 810.5 KB\n"
     ]
    }
   ],
   "source": [
    "df.info() # глянем, все ли в норме после преобразований датафрейма"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как я и говорил, возвращаем все на место. Эксперимент провалился)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 12)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df0.copy()\n",
    "df['Tenure'] = df['Tenure'].fillna(0).astype(int)\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Борьба с дисбалансом"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.7963\n",
       "1    0.2037\n",
       "Name: Exited, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Exited.value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Виден дисбаланс классов 0/1 : 4/1 соотвественно. Для борьбы с дисбалансом будем использовать методы upsamling и downsampling, а также воспользуемся встроенным параметров модели.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop(['Exited'], axis=1)\n",
    "y= df['Exited']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "RS = 12345\n",
    "(X_train, X_test, y_train, y_test) = train_test_split(X, y, test_size=1/4, random_state=RS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5998, 12)\n",
      "(1502, 12)\n"
     ]
    }
   ],
   "source": [
    "train = pd.concat([X_train, y_train], axis=1)\n",
    "\n",
    "df_exited_yes = train[train.Exited==0]\n",
    "df_exited_no = train[train.Exited==1]\n",
    "print(df_exited_yes.shape)\n",
    "print(df_exited_no.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11996, 11)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RS=123\n",
    "\n",
    "df_exited_no_up = resample(df_exited_no, replace=True, \n",
    "                           n_samples=len(df_exited_yes), random_state=RS)\n",
    "train_up = pd.concat([df_exited_yes, df_exited_no_up])\n",
    "X_train_up = train_up.drop(['Exited'], axis=1)\n",
    "y_train_up = train_up['Exited']\n",
    "\n",
    "\n",
    "df_exited_yes_down = resample(df_exited_yes, replace=False, \n",
    "                              n_samples=len(df_exited_no), random_state=RS)\n",
    "train_down = pd.concat([df_exited_no, df_exited_yes_down])\n",
    "X_train_down = train_down.drop(['Exited'], axis=1)\n",
    "y_train_down = train_down['Exited']\n",
    "X_train_up.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#код ревьюера\n",
    "df_exited_yes_down.duplicated().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Тестирование модели"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "В качестве модели будем использовать Random Forest (RF). Для данной модели условия масштабирования признаков не является обязательным.  \n",
    "Поэтому этот этап обработки можно пропустить. Будем использовать RandomizedSearchCV, поэтому валидационная выборка не обязательна."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Фукнция, которая выводит результаты обучения и метрики качества обучения:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_resume(clf, x, y, scv=True): #scv - SearchCV\n",
    "    print('Результаты обучения модели:')\n",
    "    if scv: # если это серч\n",
    "        print('лучшие параметры:', clf.best_params_)\n",
    "    vld_pred = clf.predict(x)\n",
    "    print('f1-мера:', f1_score(y, vld_pred))\n",
    "    print('roc-auc-мера:', roc_auc_score(y, clf.predict_proba(x)[:, 1]))\n",
    "    \n",
    "    cm = confusion_matrix(y_test, vld_pred, labels=clf.classes_)\n",
    "    disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=clf.classes_)\n",
    "    disp.plot()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Результаты обучения модели:\n",
      "лучшие параметры: {'n_estimators': 169, 'max_depth': 16}\n",
      "f1-мера: 0.5697399527186761\n",
      "roc-auc-мера: 0.8508225725904258\n"
     ]
    },
    {
     "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": [
    "params = {'n_estimators': range(5, 200) , 'max_depth': range(1, 20)} \n",
    "rfc = RandomForestClassifier(random_state=RS)\n",
    "clf = RandomizedSearchCV(rfc, params, n_jobs=-1, random_state=RS)\n",
    "clf.fit(X_train, y_train)\n",
    "get_resume(clf, X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "На исходных данных f1-мера ниже требуемой. Попробуем обучить модель на сэмплированных данных:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Результаты обучения модели:\n",
      "f1-мера: 0.6275720164609054\n",
      "roc-auc-мера: 0.8549485148985756\n"
     ]
    },
    {
     "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": [
    "rfc.set_params(**clf.best_params_) # установим оптимальные гиперпараметры RF, чтобы не повторять их поиск.\n",
    "rfc.fit(X_train_up, y_train_up) # обучим модель на upsample выборке\n",
    "get_resume(rfc, X_test, y_test, scv=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>f1 - мера получилась выше требуемой. Требуемый результат получен.  \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "  \n",
    "Обучим модель на downsample выборке. Предположительно, результат будет не лучше того, что получился на общей выборке."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Результаты обучения модели:\n",
      "f1-мера: 0.6115581565471836\n",
      "roc-auc-мера: 0.8549580271574992\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT8AAAEGCAYAAAAT05LOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAAchUlEQVR4nO3de7xVZb3v8c8XEAEvXEQRAYMSMVLxgkpZ5l20dmgnTeUkx9yRadlRe5XWPrm3HcsuZnnauWUjeam8VSbtbZp5SdtHUDRFQA3S5CKICChekbV++4/xLJjc1ppjrTnXnGuO7/v1Gi/GeMYzx3gGvvzxXMZ4HkUEZmZF063WBTAzqwUHPzMrJAc/MyskBz8zKyQHPzMrpB61LkCpgQO6x/Bh29S6GJbDs88PrHURLIe3317Fu2vfUEeucdwR28UrK5vKyvvY7HfujojxHblftdRV8Bs+bBseuXtYrYthORx5xlm1LoLl8NiMn3T4Gq+sbOKRu3cvK2/3wfPr9l/Hugp+Zlb/AmimudbF6DAHPzPLJQjejfKavfXMwc/McnPNz8wKJwiaGuCzWAc/M8utGQc/MyuYAJoc/MysiFzzM7PCCeBd9/mZWdEE4WavmRVQQFPXj30OfmaWT/aFR9fn4GdmOYkmOjQ3Ql1w8DOzXLIBDwc/MyuY7D0/Bz8zK6Bm1/zMrGhc8zOzQgpEUwOsgOHgZ2a5NUKzt+uHbzPrVIFYG93L2toiaZqk5ZLmbOHchZJC0sB0LElXSVogabakA0ryTpI0P22TynkOBz8zyyV7yblbWVsZrgM2W+BI0jDgWGBhSfLxwMi0TQauTnkHAJcAhwAHA5dI6t/WjR38zCy3pvSic1tbWyLiQWDlFk5dCXwVNvqIeAJwQ2RmAP0kDQaOA+6JiJURsQq4hy0E1E25z8/McokQTVF2vWmgpFklx1MiYkprP5A0AVgSEU9KGwXQIcCikuPFKW1r6a1y8DOz3JrLf9VlRUSMLTezpD7A18mavFXl4GdmuWQDHlULHe8DRgAttb6hwOOSDgaWAKULew9NaUuAwzdJf6CtG7nPz8xyqfCAx8bXjngqInaJiOERMZysCXtARCwDpgNnpFHfccCrEbEUuBs4VlL/NNBxbEprlWt+ZpZbU4Xe85N0E1mtbaCkxcAlEXHtVrLfCZwALADeBM4EiIiVkr4FPJryXRoRWxpE2YiDn5nlUskvPCLitDbODy/ZD+DcreSbBkzLc28HPzPLrbn80d665eBnZrlkExs4+JlZwQTi3TI+Xat3Dn5mlksEeV5yrlsOfmaWk/K85Fy3HPzMLJfANT8zKygPeJhZ4QRqiMlMHfzMLJds6cquHzq6/hOYWSfzouVmVkCBv/Aws4Jyzc/MCidCrvmZWfFkAx7+vM3MCifXGh51y8HPzHLJBjzc52dmBeQvPMyscPyFh5kVVnsWJ6o3Dn5mlksEvNvs4GdmBZM1e7t+8Ov6T2Bmna4pfd/b1tYWSdMkLZc0pyTt+5KekTRb0u2S+pWcu1jSAknPSjquJH18Slsg6aJynsHBrx2uOH8Yp+zzASYfMWp92o0/2JXTDxjNF44exReOHsUj9+4AwLJFPfmH9+67Pv3HXxu6/jc/u3xXJh44mgl77NPpz2DQTc1c863fctkF9wBw4tHzuPH7t3HfDdPYcfu3N8s/asTL3POzn3HYQc93dlHrSsurLuVsZbgOGL9J2j3A3hGxL/BX4GIASaOBU4EPpN/8VFJ3Sd2BfwWOB0YDp6W8rapqs1fSeODHQHdgakRcXs37dZZjP72ST5y5gu9/efeN0k/63Muc/IWXN8s/+D3vcPUfn90sfdwxr/GJM1fw2UPfX7Wy2tZ98rh5LHyxH316vwvAnPmDePiJYVx58e83y9tNzUz+9CxmzRnS2cWsQ5Vr9kbEg5KGb5L2h5LDGcCn0v4E4OaIeAd4XtIC4OB0bkFEPAcg6eaUd15r965aza+90bgr2GfcG+zQv6nD13n/gW+y06B1FSiR5TWw/xuMG7OIOx/Yc33aghd24qUVO2wx/0nHPs2Dj76HVa/16qwi1rXmtI5HW1sFfBZo+ddoCLCo5NzilLa19FZVs9l7MCkaR8RaoCUaN6zf/Wxnzj5qFFecP4w1qzd8+7hsYU/OOWZPvvLJPXhq5nY1LKG1OHfiTK655aCymmYD+7/Bhw98gen3uYYOLaO93cvagIGSZpVsk8u9j6RvAOuAX1TjOarZ7N1SND5k00zpL2MywO5Duu7g88cnreD085chwfXf25Up/7IbF165iAG7vMvPH53HjgOamD+7N/985gimPPAM2+3QXOsiF9a4/Rayek0v5v99IGP2Wtpm/nMnzmTKLWOJBnixtxJyvuS8IiLG5r2HpP8FfBw4KiIiJS8BhpVkG5rSaCV9q2oebSJiCjAFYOyYXtFG9rrVf+cNzdfjJ67km2eMAKDntkHPbbMm8sh932K34WtZ8ty27DnmrZqU02Dvkcv50P4LOWTfxfTcpok+vddy8ef/xHeu+egW8+85YgX/55wHAOi7w9scMmYxTU3d+K/H39OJpa4v1Vy6Mo0VfBX4aES8WXJqOvBLST8EdgNGAo8AAkZKGkEW9E4FTm/rPtUMfq1F6Ybzyks91vff/f/f92X4qGy0cPUr3dmhXxPdu8PSF3qy5Pme7Lr72loWtfCm3jaWqbdllZExey3llBPmbDXwAUy88JT1+1/93IPMeGJYoQNfJSc2kHQTcDhZ83gxcAnZ6O62wD2SAGZExNkRMVfSrWQDGeuAcyOiKV3ni8DdZIOr0yJiblv3rmbwe5R2ROOu4DtfeA+zH96eV1f2YOKBo/nMhcuY/fD2/G1ubyQYNHQt530va/E/NWN7bvj+rvToAd26Beddvpgd02DJ1G8N5v7f9uedt7ox8cDRjD9tJZ/5yrJaPlqhnXTMXE792FMM6PsWUy+7nZlPDuOKaR+udbHqUgVHe0/bQvK1reS/DLhsC+l3Anfmubc2NKcrT9IJwI/YEI03K3SpsWN6xSN3D2sti9WZI884q9ZFsBwem/ET1ry2uEPVtv577RJHTvtU2xmB3xx69WPt6fPrDFXt82tPNDaz+udZXcyscDyZqZkVloOfmRWOJzM1s8Kq5nt+ncXBz8xyiYB1nszUzIrIzV4zKxz3+ZlZYTXCJA8OfmaWmwc8zKxwItznZ2aFJJo82mtmReQ+PzMrHH/ba2bFFFm/X1fn4GdmuXm018wKJzzgYWZF5WavmRWSR3vNrHAiHPzMrKAa4VWXrt9raWadLqK8rS2SpklaLmlOSdoASfdImp/+7J/SJekqSQskzZZ0QMlvJqX88yVNKucZHPzMLJdANDd3K2srw3XA+E3SLgLujYiRwL3pGOB4YGTaJgNXQxYsyRY7PwQ4GLikJWC2xsHPzHKLMrc2rxPxILByk+QJwPVp/3rgxJL0GyIzA+gnaTBwHHBPRKyMiFXAPWweUDfjPj8zyyffgMdASbNKjqdExJQ2fjMoIpam/WXAoLQ/BFhUkm9xSttaeqsc/Mwsv/Lf81sREWPbfZuIkFSVtwrd7DWz3CJU1tZOL6XmLOnP5Sl9CTCsJN/QlLa19FZtteYn6f/RSnyPiPPauriZNZ4Ampur+qrLdGAScHn6846S9C9KuplscOPViFgq6W7g2yWDHMcCF7d1k9aavbNaOWdmRRVAhd7zk3QTcDhZ3+BislHby4FbJZ0FvACckrLfCZwALADeBM4EiIiVkr4FPJryXRoRmw6ibGarwS8iri89ltQnIt7M8Vxm1qAq9W1vRJy2lVNHbSFvAOdu5TrTgGl57t1mn5+kD0qaBzyTjsdI+mmem5hZg6nUuy41VM6Ax4/I3qN5BSAingQOq2KZzKyulTfYUe/f/5b1qktELJI2epCm6hTHzLqEOq/VlaOc4LdI0oeAkLQN8GXg6eoWy8zqVkBUd7S3U5TT7D2brJNxCPAisB9b6XQ0s6JQmVv9arPmFxErgImdUBYz6yoaoNlbzmjveyX9TtLLaeqZOyS9tzMKZ2Z1qiCjvb8EbgUGA7sBtwE3VbNQZlbHWl5yLmerY+UEvz4RcWNErEvbz4Fe1S6YmdWvSk1mWkutfds7IO3+XtJFwM1kMf/TZJ+ZmFlRNcBob2sDHo+RBbuWp/x8ybmgjA+HzawxVWeSqc7V2re9IzqzIGbWRXSBwYxylPWFh6S9gdGU9PVFxA3VKpSZ1bP6H8woR5vBT9IlZFPOjCbr6zse+DPg4GdWVA1Q8ytntPdTZNPLLIuIM4ExQN+qlsrM6ltzmVsdK6fZ+1ZENEtaJ2lHsimlh7X1IzNrUBWczLSWygl+syT1A/6dbAT4deDhahbKzOpbQ4/2toiIc9Luv0m6C9gxImZXt1hmVtcaOfhJOqC1cxHxeHWKZGZWfa3V/K5o5VwAR1a4LPx1dh+O222/Sl/WqunoWhfAaqGhm70RcURnFsTMuoigIT5v86LlZpZfhaa0knS+pLmS5ki6SVIvSSMkzZS0QNItknqmvNum4wXp/PCOPIKDn5nlpihva/Ua0hDgPGBsROwNdAdOBb4LXBkRewCrgLPST84CVqX0K1O+dnPwM7P8KjeZaQ+gt6QeQB9gKdl4wq/S+euBE9P+hHRMOn+UNllZLY9yZnKWpP8p6ZvpeHdJB7f3hmbWAMoPfgMlzSrZJq+/RMQS4AfAQrKg9yrZu8SrI2JdyraYbP0g0p+L0m/Xpfw7tfcRynnJ+adkH6ocCVwKrAF+DRzU3puaWddVTpO2xIqIGLvF60j9yWpzI4DVZLPEj69AEctSTvA7JCIOkPQXgIhY1dIBaWYFVZnR3qOB5yPiZQBJvwEOBfpJ6pFqd0OBJSn/ErJPaxenZnJf4JX23rycPr93JXUnVWIl7Uzdf7JsZtVUiQEPsubuOEl9Ut/dUcA84H6yCVUAJgF3pP3p6Zh0/r6I9k+WX07wuwq4HdhF0mVk01l9u703NLMGUIEBj4iYSTZw8TjwFFk8mgJ8DbhA0gKyPr1r00+uBXZK6RcAF3XkEcr5tvcXkh4ji8oCToyIpztyUzPrwvL1+bV+qYhLgEs2SX4O2GxQNSLeBk6uzJ3Lm8x0d+BN4HelaRGxsFKFMLMuppE/byvxn2xYyKgX2cjMs8AHqlguM6tjaoBe/3KavfuUHqfZXs7ZSnYzsy6hrAWMSkXE45IOqUZhzKyLKEKzV9IFJYfdgAOAF6tWIjOrbxUc8Kilcmp+O5TsryPrA/x1dYpjZl1Cowe/9HLzDhHxlU4qj5l1BY0c/Fo+L5F0aGcWyMzqm2j80d5HyPr3npA0neyj4zdaTkbEb6pcNjOrRwXq8+tF9vHwkWx43y8ABz+zomrw4LdLGumdw4ag16IBHt3M2q0BIkBrwa87sD0bB70WDfDoZtZejd7sXRoRl3ZaScys62jw4Nf116Yzs8qLxh/tParTSmFmXUsj1/wiYmVnFsTMuo5G7/MzM9syBz8zK5zy1+Staw5+ZpaLcLPXzArKwc/MiqkBgl85S1eamW2sAktXAkjqJ+lXkp6R9LSkD0oaIOkeSfPTn/1TXkm6StICSbPTkhrt5uBnZvmUuWB5mU3jHwN3RcRewBjgabL1eO+NiJHAvWxYn/d4YGTaJgNXd+QxHPzMLL8K1Pwk9QUOIy1KHhFrI2I1MAG4PmW7Hjgx7U8AbojMDKCfpMHtfQQHPzPLTc3lbW0YAbwM/EzSXyRNlbQdMCgilqY8y4BBaX8IsKjk94tTWrs4+JlZbjmavQMlzSrZJpdcpgfZhMlXR8T+ZJMlX1R6n4io2luFHu01s3zyhaMVETF2K+cWA4sjYmY6/hVZ8HtJ0uCIWJqatcvT+SXAsJLfD01p7eKan5nlV4E+v4hYBiySNColHQXMA6YDk1LaJOCOtD8dOCON+o4DXi1pHufmmp+Z5VLhLzy+BPxCUk/gOeBMskrZrZLOAl4ATkl57wROABYAb6a87ebgZ2a5qbky0S8ingC21CzebEq91P93bkVujIOfmeXliQ3MrKj8ba+ZFZODn5kVkWt+ZlZMDn5mVjgFWL3NzGwznsnZzIorun70c/Azs9xc8zMu+OFCDjl6DatX9ODzR2afKH7k46v5zIXLGDbyHc47YSTzZ/cB4IiTVnHyOcvX/3bE+9/m3OP25Lm5vWtS9qLrpmauvnQ6K1Ztxzd+eAwnHj2P/3HcXIYMWsOJ55zOa6/3AmC73mv5+tl/YpedXqd7t+DW3+/NXQ/tWePS11CDvORctYkNJE2TtFzSnGrdox784ZYBfGPiiI3S/v5MLy79x+E8NWO7jdLvv70/5xwzinOOGcX3vrQ7yxb2dOCroU8eN4+FL/Zbfzxn/iC+8t3xLHt5+43yTTj6af6+pB+f+6eTOP87x3P2aY/Qo3tTJ5e2vlRoPr+aquasLtcB46t4/bowZ+b2rFm1cQV60YJeLP5br1Z/d8SJq/nTHf2qWDJrzcD+bzBuzCLufGBDDW7BCzvx0oodNssbAX16vwsEvbddx5o3tqWpudgTIjn4tSIiHgRWVuv6Xd1hn1jN/b/tV+tiFNa5E2dyzS0H0RxqM+9v/zia3Qev5rarbubab9/OT34+jijjdw0ryP5FKGerYzX/50vS5JZZXt/lnVoXp1OM2v8N3nmrGy886yZvLYzbbyGr1/Ri/t8HlpX/oH0W87eFO3HyeafyuX86kfPOeJg+vdZWuZT1rYILGNVMzQc8ImIKMAVgRw2o87+uyjh8wmoecK2vZvYeuZwP7b+QQ/ZdTM9tmujTey0Xf/5PfOeaj24x//iPzOem/9gXEC8u35FlL+/A7ru9yjPP7dy5Ba8nDfB/as2DX9FIwWH/sJoLT9qj1kUprKm3jWXqbdkUcmP2WsopJ8zZauADWP7K9hzwgRd56q+70n/Htxi266u8uHzzvsGi8EvOBsBFP32BfT/4On0HrOPns+Zx4xWDWLOqB+f83yX03Wkd37rxef42txffOP19AOwz7g1efrEnyxZuW+OS26ZOOmYup37sKQb0fYupl93OzCeHccW0D3PjHfvxtc89yNTLbkcKptw6dv1rMIUUUbHJTGtJUaVOSUk3AYcDA4GXgEsi4trWfrOjBsQh2mwCV6tj7x59YK2LYDk8NuMnrHltcYdGa3boNzT2P+zLZeV96HdffayVBYxqqmo1v4g4rVrXNrPacrPXzIongAZo9jr4mVl+XT/2OfiZWX6N0Oyt+UvOZtb1qDnK2sq6ltRd0l8k/Uc6HiFppqQFkm5Ja/oiadt0vCCdH96RZ3DwM7N8IsdWni8DT5ccfxe4MiL2AFYBZ6X0s4BVKf3KlK/dHPzMLJfsJecoa2vzWtJQ4GPA1HQs4EjgVynL9cCJaX9COiadPyrlbxcHPzPLr7nMDQa2fLuftsmbXOlHwFfX54adgNURsS4dLwaGpP0hwCKAdP7VlL9dPOBhZrmVU6tLVmztJWdJHweWR8Rjkg6vUNHK5uBnZvlUbibnQ4FPSDoB6AXsCPwY6CepR6rdDQWWpPxLgGHAYkk9gL7AK+29uZu9ZpZTeSO9bY32RsTFETE0IoYDpwL3RcRE4H7gUynbJOCOtD89HZPO3xcd+D7Xwc/M8qvuZKZfAy6QtICsT69lToBrgZ1S+gXARR15BDd7zSyfKixaHhEPAA+k/eeAg7eQ523g5Erd08HPzPKr8ynqy+HgZ2b5df3Y5+BnZvmpuc6XZiuDg5+Z5RNseCW5C3PwM7NcRHmfrtU7Bz8zy8/Bz8wKycHPzArHfX5mVlQe7TWzAurQp2t1w8HPzPIJHPzMrKC6fqvXwc/M8vN7fmZWTA5+ZlY4EdDU9du9Dn5mlp9rfmZWSA5+ZlY4AbSxPkdX4OBnZjkFhPv8zKxoAg94mFlBNUCfn5euNLP8KrB0paRhku6XNE/SXElfTukDJN0jaX76s39Kl6SrJC2QNFvSAR15BAc/M8upzMDXdu1wHXBhRIwGxgHnShpNth7vvRExEriXDevzHg+MTNtk4OqOPIWDn5nlE0Bzc3lba5eJWBoRj6f9NcDTwBBgAnB9ynY9cGLanwDcEJkZQD9Jg9v7GO7zM7P8yu/zGyhpVsnxlIiYsmkmScOB/YGZwKCIWJpOLQMGpf0hwKKSny1OaUtpBwc/M8sp1+dtKyJibGsZJG0P/Br43xHxmqQNd4oISVUZXXHwM7N8AqJC7/lJ2oYs8P0iIn6Tkl+SNDgilqZm7fKUvgQYVvLzoSmtXdznZ2b5NUd5WyuUVfGuBZ6OiB+WnJoOTEr7k4A7StLPSKO+44BXS5rHubnmZ2b5VeY9v0OBzwBPSXoipX0duBy4VdJZwAvAKencncAJwALgTeDMjtzcwc/M8olocyS3vMvEnwFt5fRRW8gfwLkdvnHi4Gdm+TXAFx4OfmaWUxBNTbUuRIc5+JlZPp7SyswKy1NamVnRBBCu+ZlZ4YQnMzWzgmqEAQ9FHQ1ZS3qZ7KXGRjMQWFHrQlgujfrf7D0RsXNHLiDpLrK/n3KsiIjxHblftdRV8GtUkma19XG31Rf/N2t8/rbXzArJwc/MCsnBr3NsNnmj1T3/N2tw7vMzs0Jyzc/MCsnBz8wKycGviiSNl/RsWmf0orZ/YbUmaZqk5ZLm1LosVl0OflUiqTvwr2RrjY4GTktrklp9uw6oy5dyrbIc/KrnYGBBRDwXEWuBm8nWHbU6FhEPAitrXQ6rPge/6tnaGqNmVgcc/MyskBz8qqeia4yaWWU5+FXPo8BISSMk9QROJVt31MzqgINflUTEOuCLwN3A08CtETG3tqWytki6CXgYGCVpcVo71hqQP28zs0Jyzc/MCsnBz8wKycHPzArJwc/MCsnBz8wKycGvC5HUJOkJSXMk3SapTweudZ2kT6X9qa1NuiDpcEkfasc9/i5ps1W+tpa+SZ7Xc97rnyV9JW8Zrbgc/LqWtyJiv4jYG1gLnF16UlK71mGOiH+MiHmtZDkcyB38zOqZg1/X9RCwR6qVPSRpOjBPUndJ35f0qKTZkj4PoMxP0vyCfwR2abmQpAckjU374yU9LulJSfdKGk4WZM9Ptc6PSNpZ0q/TPR6VdGj67U6S/iBprqSpgNp6CEm/lfRY+s3kTc5dmdLvlbRzSnufpLvSbx6StFdF/jatcNpVU7DaSjW844G7UtIBwN4R8XwKIK9GxEGStgX+S9IfgP2BUWRzCw4C5gHTNrnuzsC/A4elaw2IiJWS/g14PSJ+kPL9ErgyIv4saXeyr1jeD1wC/DkiLpX0MaCcryM+m+7RG3hU0q8j4hVgO2BWRJwv6Zvp2l8kW1jo7IiYL+kQ4KfAke34a7SCc/DrWnpLeiLtPwRcS9YcfSQink/pxwL7tvTnAX2BkcBhwE0R0QS8KOm+LVx/HPBgy7UiYmvz2h0NjJbWV+x2lLR9uscn02//U9KqMp7pPEknpf1hqayvAM3ALSn958Bv0j0+BNxWcu9ty7iH2WYc/LqWtyJiv9KEFATeKE0CvhQRd2+S74QKlqMbMC4i3t5CWcom6XCyQPrBiHhT0gNAr61kj3Tf1Zv+HZi1h/v8Gs/dwBckbQMgaU9J2wEPAp9OfYKDgSO28NsZwGGSRqTfDkjpa4AdSvL9AfhSy4Gk/dLug8DpKe14oH8bZe0LrEqBby+ymmeLbkBL7fV0sub0a8Dzkk5O95CkMW3cw2yLHPwaz1Sy/rzH0yI815DV8G8H5qdzN5DNXLKRiHgZmEzWxHySDc3O3wEntQx4AOcBY9OAyjw2jDr/C1nwnEvW/F3YRlnvAnpIehq4nCz4tngDODg9w5HApSl9InBWKt9cvDSAtZNndTGzQnLNz8wKycHPzArJwc/MCsnBz8wKycHPzArJwc/MCsnBz8wK6b8BInMN+bx+77UAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rfc.fit(X_train_down, y_train_down)\n",
    "get_resume(rfc, X_test, y_test, scv=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Выяснилось, что даже на downsample выборке результат f1-меры оказался лучше, но roc-auc оказался немного меньше, т.к. значительно увеличился показатель FP, что в свою очередь связано с уменьшением выборки класса 0 - поэтому модель стала хуже предсказывать этот класс."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Теперь воспользуемся встроенными в модель методами балансировки классов (https://www.analyticsvidhya.com/blog/2020/10/improve-class-imbalance-class-weights/#free-courses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8336112037345782 3.3288948069241013\n"
     ]
    }
   ],
   "source": [
    "len_ex0 = train[train.Exited==0].shape[0]\n",
    "len_ex1 = train[train.Exited==1].shape[0]\n",
    "w0 = len(df)/(2 * len_ex0)\n",
    "w1 = len(df)/(2 * len_ex1)\n",
    "print(w0, w1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Результаты обучения модели:\n",
      "лучшие параметры: {'n_estimators': 130, 'max_depth': 18}\n",
      "f1-мера: 0.5666666666666667\n",
      "roc-auc-мера: 0.8526884021783073\n"
     ]
    },
    {
     "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": [
    "params = {'n_estimators': range(5, 200) , 'max_depth': range(1, 20)} \n",
    "rfc = RandomForestClassifier(random_state=RS, class_weight={0: w0, 1: w1})\n",
    "clf = RandomizedSearchCV(rfc, params, n_jobs=-1)\n",
    "clf.fit(X_train, y_train)\n",
    "get_resume(clf, X_test, y_test, scv=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Результат чуть лучше обучения на исходных данных, но хуже чем на сэмплированных. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Вывод:\n",
    "Проведена предобработка данных, проведен краткий исследовательский анализ данных.  \n",
    "Категориальные переменные преобразованы методом OneHotEncoding.  \n",
    "Выявлен дисбаланс классов, для их устранения были применены методы upsampling, downsampling, весовая балансировка классов при обучении модели.  \n",
    "Изначально была выбрана модель RF, которая наиболее эффективна на задачах такого рода, кроме того не пришлось масштабировать признаки. Подобраны гиперпараметры с помощью класса RandomizedSearchCV, произведено тестирование модели и сравнение результатов f1-меры. После обучения модели выявлено, что наилучшие результаты дает upsampling, downsampling, весовая балансировка классов, исходные признаки -  в порядке убывания.\n",
    "Лучшие гиперпараметры модели: 'n_estimators': 169, 'max_depth': 16, метрики качества: f1-мера: 0.6275720164609054\n",
    "roc-auc-мера: 0.8549485148985756.\n"
   ]
  }
 ],
 "metadata": {
  "ExecuteTimeLog": [
   {
    "duration": 452,
    "start_time": "2022-05-04T17:27:50.332Z"
   },
   {
    "duration": 64,
    "start_time": "2022-05-04T17:28:20.366Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-04T17:34:28.343Z"
   },
   {
    "duration": 94,
    "start_time": "2022-05-04T18:09:24.161Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-04T18:09:26.880Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-04T18:11:49.535Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-04T18:11:54.900Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-04T18:11:59.035Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-04T18:15:29.476Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-04T18:15:35.087Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-04T18:15:38.582Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-04T18:19:37.543Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-04T18:24:52.843Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-04T18:25:34.085Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-04T18:49:17.957Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-04T18:49:27.972Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-04T18:49:47.988Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-04T18:49:58.317Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-04T18:51:23.117Z"
   },
   {
    "duration": 1357,
    "start_time": "2022-05-04T18:51:30.224Z"
   },
   {
    "duration": 571,
    "start_time": "2022-05-04T18:51:35.434Z"
   },
   {
    "duration": 543,
    "start_time": "2022-05-04T18:53:18.897Z"
   },
   {
    "duration": 456,
    "start_time": "2022-05-04T18:53:25.502Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-04T18:53:50.670Z"
   },
   {
    "duration": 477,
    "start_time": "2022-05-04T18:53:55.620Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-04T19:15:04.466Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-04T19:16:17.318Z"
   },
   {
    "duration": 195,
    "start_time": "2022-05-04T19:19:36.739Z"
   },
   {
    "duration": 1349,
    "start_time": "2022-05-05T05:39:30.956Z"
   },
   {
    "duration": 45,
    "start_time": "2022-05-05T05:39:32.308Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T05:39:32.355Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T05:39:32.371Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T05:39:32.379Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-05T05:39:32.385Z"
   },
   {
    "duration": 899,
    "start_time": "2022-05-05T05:39:32.409Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T05:39:33.310Z"
   },
   {
    "duration": 852,
    "start_time": "2022-05-05T05:39:45.414Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T05:44:28.717Z"
   },
   {
    "duration": 115,
    "start_time": "2022-05-05T05:44:43.093Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T05:44:46.813Z"
   },
   {
    "duration": 1221,
    "start_time": "2022-05-05T05:45:00.851Z"
   },
   {
    "duration": 44,
    "start_time": "2022-05-05T05:45:02.074Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T05:45:02.120Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T05:45:02.135Z"
   },
   {
    "duration": 32,
    "start_time": "2022-05-05T05:45:02.154Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-05T05:45:02.188Z"
   },
   {
    "duration": 818,
    "start_time": "2022-05-05T05:45:02.213Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T05:45:03.034Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T05:53:02.705Z"
   },
   {
    "duration": 113,
    "start_time": "2022-05-05T05:53:22.827Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T05:53:36.123Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T05:53:42.451Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T05:54:06.908Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-05T05:56:10.907Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T05:56:27.051Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T05:56:44.922Z"
   },
   {
    "duration": 2409,
    "start_time": "2022-05-05T05:56:52.155Z"
   },
   {
    "duration": 109,
    "start_time": "2022-05-05T05:57:08.042Z"
   },
   {
    "duration": 109,
    "start_time": "2022-05-05T05:57:31.051Z"
   },
   {
    "duration": 208,
    "start_time": "2022-05-05T05:57:59.475Z"
   },
   {
    "duration": 213,
    "start_time": "2022-05-05T10:44:36.868Z"
   },
   {
    "duration": 1370,
    "start_time": "2022-05-05T10:44:46.683Z"
   },
   {
    "duration": 39,
    "start_time": "2022-05-05T10:44:48.055Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-05T10:44:48.096Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T10:44:48.112Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T10:44:48.118Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T10:44:48.125Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T10:44:48.141Z"
   },
   {
    "duration": 851,
    "start_time": "2022-05-05T10:44:48.152Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T10:44:49.005Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-05T10:44:49.013Z"
   },
   {
    "duration": 120,
    "start_time": "2022-05-05T11:08:21.441Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T11:08:32.356Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-05T11:08:32.361Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T11:08:32.411Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T11:08:32.426Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T11:08:32.432Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-05T11:08:32.438Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T11:08:32.459Z"
   },
   {
    "duration": 833,
    "start_time": "2022-05-05T11:08:32.471Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:08:33.306Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T11:08:33.315Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T11:14:25.794Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T11:14:52.154Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T11:19:13.091Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:19:29.834Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T11:19:36.665Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T11:19:42.905Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T11:20:21.841Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T11:23:14.344Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:23:27.889Z"
   },
   {
    "duration": 325,
    "start_time": "2022-05-05T11:25:02.233Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T11:25:13.129Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T11:25:48.330Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:26:16.953Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T11:26:27.297Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T11:27:23.330Z"
   },
   {
    "duration": 542,
    "start_time": "2022-05-05T11:29:57.983Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T11:35:31.001Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T11:35:45.585Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T11:35:54.352Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T11:45:58.192Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-05T11:46:08.752Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T11:46:14.663Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-05T11:46:49.633Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T11:46:55.888Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T11:51:00.536Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T11:51:35.584Z"
   },
   {
    "duration": 68,
    "start_time": "2022-05-05T11:51:43.318Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:51:48.927Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T11:57:59.199Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-05T11:58:14.983Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T11:58:45.566Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T11:59:17.295Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T11:59:51.318Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T11:59:53.919Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T12:00:01.207Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T12:19:57.893Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T12:20:33.382Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:20:58.270Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T12:31:04.453Z"
   },
   {
    "duration": 1458,
    "start_time": "2022-05-05T12:37:55.016Z"
   },
   {
    "duration": 38,
    "start_time": "2022-05-05T12:37:56.476Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T12:37:56.516Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:37:56.535Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T12:37:56.541Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T12:37:56.549Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-05T12:37:56.567Z"
   },
   {
    "duration": 818,
    "start_time": "2022-05-05T12:37:56.611Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T12:37:57.431Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T12:37:57.452Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:37:57.460Z"
   },
   {
    "duration": 138,
    "start_time": "2022-05-05T12:37:57.466Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-05T12:37:57.606Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-05T12:37:57.607Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-05T12:37:57.609Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T12:38:11.946Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T12:38:12.770Z"
   },
   {
    "duration": 1543,
    "start_time": "2022-05-05T12:38:19.009Z"
   },
   {
    "duration": 42,
    "start_time": "2022-05-05T12:38:20.554Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-05T12:38:20.597Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:38:20.619Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T12:38:20.625Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T12:38:20.631Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T12:38:20.650Z"
   },
   {
    "duration": 863,
    "start_time": "2022-05-05T12:38:20.661Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-05T12:38:21.526Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T12:38:21.547Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T12:38:21.556Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-05T12:38:21.572Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-05T12:38:21.584Z"
   },
   {
    "duration": 137,
    "start_time": "2022-05-05T12:38:21.618Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-05T12:38:21.757Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T12:39:00.303Z"
   },
   {
    "duration": 4521,
    "start_time": "2022-05-05T12:39:02.493Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T12:40:07.101Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:40:09.925Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-05T12:40:10.205Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:41:09.478Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-05T12:41:09.901Z"
   },
   {
    "duration": 4053,
    "start_time": "2022-05-05T12:41:14.886Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:41:21.325Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-05T12:41:22.710Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:41:29.805Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-05T12:41:30.301Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:41:40.622Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-05T12:41:41.102Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:43:09.421Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-05T12:43:09.997Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T12:43:15.813Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-05T12:43:21.805Z"
   },
   {
    "duration": 3832,
    "start_time": "2022-05-05T12:44:19.957Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:44:25.636Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-05T12:44:31.581Z"
   },
   {
    "duration": 18521,
    "start_time": "2022-05-05T12:44:44.949Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:45:07.508Z"
   },
   {
    "duration": 58,
    "start_time": "2022-05-05T12:45:09.109Z"
   },
   {
    "duration": 12997,
    "start_time": "2022-05-05T12:47:37.732Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T12:47:53.804Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-05T12:47:54.188Z"
   },
   {
    "duration": 363739,
    "start_time": "2022-05-05T12:48:24.805Z"
   },
   {
    "duration": 10666,
    "start_time": "2022-05-05T12:54:35.074Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T12:54:49.964Z"
   },
   {
    "duration": 58,
    "start_time": "2022-05-05T12:54:51.907Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T12:57:40.580Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T12:57:41.027Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T12:57:50.916Z"
   },
   {
    "duration": 57,
    "start_time": "2022-05-05T12:57:51.331Z"
   },
   {
    "duration": 17648,
    "start_time": "2022-05-05T12:58:38.388Z"
   },
   {
    "duration": 60,
    "start_time": "2022-05-05T12:59:12.892Z"
   },
   {
    "duration": 29645,
    "start_time": "2022-05-05T13:00:32.676Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:01:05.075Z"
   },
   {
    "duration": 58,
    "start_time": "2022-05-05T13:01:08.388Z"
   },
   {
    "duration": 35124,
    "start_time": "2022-05-05T13:01:22.420Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T13:02:43.235Z"
   },
   {
    "duration": 124,
    "start_time": "2022-05-05T13:03:05.731Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T13:03:33.627Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T13:03:39.811Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T13:07:38.883Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:08:24.059Z"
   },
   {
    "duration": 979,
    "start_time": "2022-05-05T13:08:47.731Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T13:08:58.659Z"
   },
   {
    "duration": 79,
    "start_time": "2022-05-05T13:09:35.187Z"
   },
   {
    "duration": 29682,
    "start_time": "2022-05-05T13:10:10.683Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:10:45.738Z"
   },
   {
    "duration": 99,
    "start_time": "2022-05-05T13:10:46.666Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T13:11:09.287Z"
   },
   {
    "duration": 1329,
    "start_time": "2022-05-05T13:11:10.346Z"
   },
   {
    "duration": 103,
    "start_time": "2022-05-05T13:11:13.843Z"
   },
   {
    "duration": 72,
    "start_time": "2022-05-05T13:12:43.554Z"
   },
   {
    "duration": 92,
    "start_time": "2022-05-05T13:12:45.828Z"
   },
   {
    "duration": 520,
    "start_time": "2022-05-05T13:16:09.516Z"
   },
   {
    "duration": 465,
    "start_time": "2022-05-05T13:17:31.090Z"
   },
   {
    "duration": 1333,
    "start_time": "2022-05-05T13:17:45.362Z"
   },
   {
    "duration": 580,
    "start_time": "2022-05-05T13:17:55.530Z"
   },
   {
    "duration": 547,
    "start_time": "2022-05-05T13:19:44.298Z"
   },
   {
    "duration": 3370,
    "start_time": "2022-05-05T13:20:33.220Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-05T13:20:36.592Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-05T13:20:36.640Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:20:36.657Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:20:36.663Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T13:20:36.669Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T13:20:36.687Z"
   },
   {
    "duration": 828,
    "start_time": "2022-05-05T13:20:36.699Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T13:20:37.528Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T13:20:37.547Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T13:20:37.555Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T13:20:37.562Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T13:20:37.570Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T13:20:37.579Z"
   },
   {
    "duration": 30985,
    "start_time": "2022-05-05T13:20:37.594Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:21:08.581Z"
   },
   {
    "duration": 90,
    "start_time": "2022-05-05T13:21:08.587Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T13:21:08.679Z"
   },
   {
    "duration": 1082,
    "start_time": "2022-05-05T13:21:08.686Z"
   },
   {
    "duration": 458,
    "start_time": "2022-05-05T13:21:09.770Z"
   },
   {
    "duration": 68,
    "start_time": "2022-05-05T13:42:31.269Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:44:54.012Z"
   },
   {
    "duration": 198,
    "start_time": "2022-05-05T13:45:14.435Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T13:45:25.405Z"
   },
   {
    "duration": 175,
    "start_time": "2022-05-05T13:45:31.684Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T13:45:37.517Z"
   },
   {
    "duration": 175,
    "start_time": "2022-05-05T13:45:37.861Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:48:27.252Z"
   },
   {
    "duration": 152,
    "start_time": "2022-05-05T13:48:28.239Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T13:48:37.843Z"
   },
   {
    "duration": 1124,
    "start_time": "2022-05-05T13:48:39.603Z"
   },
   {
    "duration": 74,
    "start_time": "2022-05-05T13:49:08.043Z"
   },
   {
    "duration": 151,
    "start_time": "2022-05-05T13:49:08.364Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:49:19.316Z"
   },
   {
    "duration": 153,
    "start_time": "2022-05-05T13:49:20.082Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T13:49:25.635Z"
   },
   {
    "duration": 1113,
    "start_time": "2022-05-05T13:49:26.307Z"
   },
   {
    "duration": 501,
    "start_time": "2022-05-05T13:49:32.962Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T13:58:39.097Z"
   },
   {
    "duration": 276,
    "start_time": "2022-05-05T13:59:29.193Z"
   },
   {
    "duration": 494,
    "start_time": "2022-05-05T13:59:44.649Z"
   },
   {
    "duration": 204,
    "start_time": "2022-05-05T13:59:46.633Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T14:00:07.385Z"
   },
   {
    "duration": 222,
    "start_time": "2022-05-05T14:00:12.882Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T14:01:33.209Z"
   },
   {
    "duration": 285,
    "start_time": "2022-05-05T14:01:34.281Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T14:01:40.369Z"
   },
   {
    "duration": 1256,
    "start_time": "2022-05-05T14:01:40.818Z"
   },
   {
    "duration": 646,
    "start_time": "2022-05-05T14:01:52.880Z"
   },
   {
    "duration": 233,
    "start_time": "2022-05-05T14:01:53.528Z"
   },
   {
    "duration": 1551,
    "start_time": "2022-05-05T14:14:11.513Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-05T14:14:13.067Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T14:14:13.115Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T14:14:13.133Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T14:14:13.140Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-05T14:14:13.151Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T14:14:13.178Z"
   },
   {
    "duration": 842,
    "start_time": "2022-05-05T14:14:13.208Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-05T14:14:14.051Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T14:14:14.072Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-05T14:14:14.083Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T14:14:14.110Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T14:14:14.121Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-05T14:14:14.134Z"
   },
   {
    "duration": 28944,
    "start_time": "2022-05-05T14:14:14.150Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T14:14:43.095Z"
   },
   {
    "duration": 331,
    "start_time": "2022-05-05T14:14:43.101Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T14:14:43.434Z"
   },
   {
    "duration": 873,
    "start_time": "2022-05-05T14:14:43.441Z"
   },
   {
    "duration": 463,
    "start_time": "2022-05-05T14:14:44.315Z"
   },
   {
    "duration": 90,
    "start_time": "2022-05-05T14:28:00.378Z"
   },
   {
    "duration": 1535,
    "start_time": "2022-05-05T18:03:29.163Z"
   },
   {
    "duration": 35,
    "start_time": "2022-05-05T18:03:30.700Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T18:03:30.737Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:03:30.756Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T18:03:30.761Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T18:03:30.772Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T18:03:30.791Z"
   },
   {
    "duration": 782,
    "start_time": "2022-05-05T18:03:30.801Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T18:03:31.584Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T18:03:31.603Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T18:03:31.610Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T18:03:31.618Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-05T18:03:31.632Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-05T18:03:31.655Z"
   },
   {
    "duration": 27668,
    "start_time": "2022-05-05T18:03:31.668Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T18:03:59.338Z"
   },
   {
    "duration": 225,
    "start_time": "2022-05-05T18:03:59.350Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:03:59.577Z"
   },
   {
    "duration": 983,
    "start_time": "2022-05-05T18:03:59.583Z"
   },
   {
    "duration": 468,
    "start_time": "2022-05-05T18:04:00.568Z"
   },
   {
    "duration": 28221,
    "start_time": "2022-05-05T18:33:14.341Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:33:42.564Z"
   },
   {
    "duration": 329,
    "start_time": "2022-05-05T18:33:42.570Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T18:33:42.901Z"
   },
   {
    "duration": 1609,
    "start_time": "2022-05-05T18:33:42.908Z"
   },
   {
    "duration": 784,
    "start_time": "2022-05-05T18:33:44.519Z"
   },
   {
    "duration": 9814,
    "start_time": "2022-05-05T18:34:46.975Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:35:00.114Z"
   },
   {
    "duration": 292,
    "start_time": "2022-05-05T18:35:01.394Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:35:07.536Z"
   },
   {
    "duration": 1610,
    "start_time": "2022-05-05T18:35:08.265Z"
   },
   {
    "duration": 25417,
    "start_time": "2022-05-05T18:35:23.356Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:35:51.248Z"
   },
   {
    "duration": 211,
    "start_time": "2022-05-05T18:35:51.778Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T18:35:54.110Z"
   },
   {
    "duration": 971,
    "start_time": "2022-05-05T18:35:55.029Z"
   },
   {
    "duration": 36717,
    "start_time": "2022-05-05T18:50:15.193Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:50:56.469Z"
   },
   {
    "duration": 338,
    "start_time": "2022-05-05T18:51:05.155Z"
   },
   {
    "duration": 25333,
    "start_time": "2022-05-05T18:52:17.155Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T18:52:45.561Z"
   },
   {
    "duration": 266,
    "start_time": "2022-05-05T18:52:47.689Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:03:43.324Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T19:04:55.817Z"
   },
   {
    "duration": 28894,
    "start_time": "2022-05-05T19:06:16.877Z"
   },
   {
    "duration": 114,
    "start_time": "2022-05-05T19:08:41.334Z"
   },
   {
    "duration": 325,
    "start_time": "2022-05-05T19:08:59.853Z"
   },
   {
    "duration": 32968,
    "start_time": "2022-05-05T19:11:12.882Z"
   },
   {
    "duration": 345,
    "start_time": "2022-05-05T19:11:53.105Z"
   },
   {
    "duration": 1391,
    "start_time": "2022-05-05T19:12:40.020Z"
   },
   {
    "duration": 35,
    "start_time": "2022-05-05T19:12:41.413Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T19:12:41.450Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:12:41.465Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:12:41.470Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:12:41.477Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:12:41.493Z"
   },
   {
    "duration": 709,
    "start_time": "2022-05-05T19:12:41.503Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T19:12:42.213Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-05T19:12:42.231Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:12:42.254Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:12:42.260Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T19:12:42.269Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:12:42.280Z"
   },
   {
    "duration": 27500,
    "start_time": "2022-05-05T19:12:42.295Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:13:09.797Z"
   },
   {
    "duration": 328,
    "start_time": "2022-05-05T19:13:09.802Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:13:10.132Z"
   },
   {
    "duration": 1202,
    "start_time": "2022-05-05T19:13:10.148Z"
   },
   {
    "duration": 586,
    "start_time": "2022-05-05T19:13:11.352Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:13:11.947Z"
   },
   {
    "duration": 22374,
    "start_time": "2022-05-05T19:13:11.955Z"
   },
   {
    "duration": 228,
    "start_time": "2022-05-05T19:13:34.331Z"
   },
   {
    "duration": 1393,
    "start_time": "2022-05-05T19:17:01.617Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-05T19:17:03.012Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T19:17:03.054Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:17:03.068Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T19:17:03.073Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T19:17:03.081Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:17:03.098Z"
   },
   {
    "duration": 578,
    "start_time": "2022-05-05T19:17:03.107Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:17:03.687Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:17:03.703Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:17:03.711Z"
   },
   {
    "duration": 32,
    "start_time": "2022-05-05T19:17:03.716Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:17:03.749Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T19:17:03.759Z"
   },
   {
    "duration": 21769,
    "start_time": "2022-05-05T19:17:03.774Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-05T19:17:25.548Z"
   },
   {
    "duration": 226,
    "start_time": "2022-05-05T19:17:25.553Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:17:25.781Z"
   },
   {
    "duration": 892,
    "start_time": "2022-05-05T19:17:25.787Z"
   },
   {
    "duration": 521,
    "start_time": "2022-05-05T19:17:26.681Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:17:27.205Z"
   },
   {
    "duration": 23289,
    "start_time": "2022-05-05T19:17:27.213Z"
   },
   {
    "duration": 313,
    "start_time": "2022-05-05T19:17:50.503Z"
   },
   {
    "duration": 1370,
    "start_time": "2022-05-05T19:19:11.948Z"
   },
   {
    "duration": 37,
    "start_time": "2022-05-05T19:19:13.320Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:19:13.358Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:19:13.374Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:19:13.380Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-05T19:19:13.387Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:19:13.406Z"
   },
   {
    "duration": 782,
    "start_time": "2022-05-05T19:19:13.416Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-05T19:19:14.200Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T19:19:14.219Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-05T19:19:14.227Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:19:14.251Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-05T19:19:14.261Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-05T19:19:14.270Z"
   },
   {
    "duration": 31735,
    "start_time": "2022-05-05T19:19:14.284Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:19:46.021Z"
   },
   {
    "duration": 302,
    "start_time": "2022-05-05T19:19:46.026Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:19:46.329Z"
   },
   {
    "duration": 1544,
    "start_time": "2022-05-05T19:19:46.335Z"
   },
   {
    "duration": 745,
    "start_time": "2022-05-05T19:19:47.880Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:19:48.627Z"
   },
   {
    "duration": 26764,
    "start_time": "2022-05-05T19:19:48.634Z"
   },
   {
    "duration": 233,
    "start_time": "2022-05-05T19:20:15.400Z"
   },
   {
    "duration": 31674,
    "start_time": "2022-05-05T19:21:48.634Z"
   },
   {
    "duration": 306,
    "start_time": "2022-05-05T19:22:24.816Z"
   },
   {
    "duration": 29197,
    "start_time": "2022-05-05T19:25:40.990Z"
   },
   {
    "duration": 303,
    "start_time": "2022-05-05T19:26:14.794Z"
   },
   {
    "duration": 10339,
    "start_time": "2022-05-05T19:29:09.171Z"
   },
   {
    "duration": 193,
    "start_time": "2022-05-05T19:29:21.181Z"
   },
   {
    "duration": 1498,
    "start_time": "2022-05-05T19:29:47.763Z"
   },
   {
    "duration": 55,
    "start_time": "2022-05-05T19:29:49.263Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-05T19:29:49.320Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:29:49.346Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:29:49.353Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-05T19:29:49.360Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-05T19:29:49.382Z"
   },
   {
    "duration": 749,
    "start_time": "2022-05-05T19:29:49.395Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T19:29:50.147Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:29:50.168Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-05T19:29:50.176Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:29:50.188Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T19:29:50.204Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-05T19:29:50.219Z"
   },
   {
    "duration": 29056,
    "start_time": "2022-05-05T19:29:50.256Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:30:19.314Z"
   },
   {
    "duration": 353,
    "start_time": "2022-05-05T19:30:19.319Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:30:19.674Z"
   },
   {
    "duration": 1340,
    "start_time": "2022-05-05T19:30:19.680Z"
   },
   {
    "duration": 615,
    "start_time": "2022-05-05T19:30:21.022Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-05T19:30:21.647Z"
   },
   {
    "duration": 13980,
    "start_time": "2022-05-05T19:30:21.660Z"
   },
   {
    "duration": 222,
    "start_time": "2022-05-05T19:30:35.641Z"
   },
   {
    "duration": 12881,
    "start_time": "2022-05-05T19:31:48.377Z"
   },
   {
    "duration": 209,
    "start_time": "2022-05-05T19:32:08.047Z"
   },
   {
    "duration": 33009,
    "start_time": "2022-05-05T19:32:15.926Z"
   },
   {
    "duration": 1475,
    "start_time": "2022-05-05T19:34:13.168Z"
   },
   {
    "duration": 44,
    "start_time": "2022-05-05T19:34:14.645Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-05T19:34:14.690Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:34:14.705Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:34:14.711Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-05T19:34:14.716Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-05T19:34:14.732Z"
   },
   {
    "duration": 691,
    "start_time": "2022-05-05T19:34:14.762Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T19:34:15.455Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:34:15.476Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:34:15.484Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-05T19:34:15.491Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-05T19:34:15.500Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-05T19:34:15.547Z"
   },
   {
    "duration": 38393,
    "start_time": "2022-05-05T19:34:15.567Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-05T19:34:53.962Z"
   },
   {
    "duration": 214,
    "start_time": "2022-05-05T19:34:53.968Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-05T19:34:54.184Z"
   },
   {
    "duration": 1061,
    "start_time": "2022-05-05T19:34:54.189Z"
   },
   {
    "duration": 510,
    "start_time": "2022-05-05T19:34:55.252Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-05T19:34:55.763Z"
   },
   {
    "duration": 24049,
    "start_time": "2022-05-05T19:34:55.770Z"
   },
   {
    "duration": 220,
    "start_time": "2022-05-05T19:35:19.821Z"
   },
   {
    "duration": 1560,
    "start_time": "2022-05-06T06:44:59.241Z"
   },
   {
    "duration": 37,
    "start_time": "2022-05-06T06:45:00.803Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-06T06:45:00.841Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T06:45:00.862Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-06T06:45:00.867Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-06T06:45:00.886Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-06T06:45:00.912Z"
   },
   {
    "duration": 776,
    "start_time": "2022-05-06T06:45:00.922Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-06T06:45:01.700Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T06:45:01.719Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T06:45:01.748Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-06T06:45:01.755Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-06T06:45:01.766Z"
   },
   {
    "duration": 138,
    "start_time": "2022-05-06T06:45:01.777Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.916Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.917Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.919Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.920Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.920Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.921Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.922Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.923Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-06T06:45:01.924Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-06T08:20:46.790Z"
   },
   {
    "duration": 30782,
    "start_time": "2022-05-06T08:21:23.425Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T08:21:54.209Z"
   },
   {
    "duration": 407,
    "start_time": "2022-05-06T08:21:54.217Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T08:21:54.626Z"
   },
   {
    "duration": 2574,
    "start_time": "2022-05-06T08:21:54.647Z"
   },
   {
    "duration": 901,
    "start_time": "2022-05-06T08:21:57.223Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T08:21:58.126Z"
   },
   {
    "duration": 34367,
    "start_time": "2022-05-06T08:21:58.134Z"
   },
   {
    "duration": 231,
    "start_time": "2022-05-06T08:22:32.503Z"
   },
   {
    "duration": 43470,
    "start_time": "2022-05-06T08:23:15.310Z"
   },
   {
    "duration": 1543,
    "start_time": "2022-05-06T08:31:27.805Z"
   },
   {
    "duration": 62,
    "start_time": "2022-05-06T08:31:29.351Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-06T08:31:29.415Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T08:31:29.447Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-06T08:31:29.463Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-06T08:31:29.490Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-06T08:31:29.511Z"
   },
   {
    "duration": 1016,
    "start_time": "2022-05-06T08:31:29.525Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-06T08:31:30.547Z"
   },
   {
    "duration": 44,
    "start_time": "2022-05-06T08:31:30.571Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-06T08:31:30.617Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-06T08:31:30.634Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-06T08:31:30.671Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-06T08:31:30.700Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-06T08:31:30.716Z"
   },
   {
    "duration": 26759,
    "start_time": "2022-05-06T08:31:30.730Z"
   },
   {
    "duration": 1017,
    "start_time": "2022-05-06T08:31:57.490Z"
   },
   {
    "duration": 478,
    "start_time": "2022-05-06T08:31:58.508Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T08:31:58.987Z"
   },
   {
    "duration": 54607,
    "start_time": "2022-05-06T08:31:58.996Z"
   },
   {
    "duration": 1369,
    "start_time": "2022-05-06T08:36:36.166Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-06T08:36:37.536Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-06T08:36:37.580Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-06T08:36:37.596Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T08:36:37.601Z"
   },
   {
    "duration": 52,
    "start_time": "2022-05-06T08:36:37.607Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-06T08:36:37.661Z"
   },
   {
    "duration": 702,
    "start_time": "2022-05-06T08:36:37.676Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-06T08:36:38.379Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-06T08:36:38.396Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T08:36:38.422Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-06T08:36:38.428Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-06T08:36:38.441Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-06T08:36:38.470Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T08:36:38.491Z"
   },
   {
    "duration": 25937,
    "start_time": "2022-05-06T08:36:38.496Z"
   },
   {
    "duration": 808,
    "start_time": "2022-05-06T08:37:04.434Z"
   },
   {
    "duration": 411,
    "start_time": "2022-05-06T08:37:05.254Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T08:37:05.667Z"
   },
   {
    "duration": 36702,
    "start_time": "2022-05-06T08:37:05.675Z"
   },
   {
    "duration": 32900,
    "start_time": "2022-05-06T08:38:35.430Z"
   },
   {
    "duration": 1392,
    "start_time": "2022-05-06T08:41:54.700Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-06T08:41:56.094Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-06T08:41:56.135Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T08:41:56.158Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-06T08:41:56.168Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-06T08:41:56.188Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-06T08:41:56.205Z"
   },
   {
    "duration": 768,
    "start_time": "2022-05-06T08:41:56.216Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-06T08:41:56.986Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-06T08:41:57.006Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-06T08:41:57.028Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-06T08:41:57.057Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-06T08:41:57.074Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-06T08:41:57.086Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T08:41:57.100Z"
   },
   {
    "duration": 28900,
    "start_time": "2022-05-06T08:41:57.106Z"
   },
   {
    "duration": 2409,
    "start_time": "2022-05-06T08:42:26.008Z"
   },
   {
    "duration": 973,
    "start_time": "2022-05-06T08:42:28.419Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T08:42:29.394Z"
   },
   {
    "duration": 27950,
    "start_time": "2022-05-06T08:42:29.402Z"
   },
   {
    "duration": 105,
    "start_time": "2022-05-06T09:01:09.561Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-06T09:04:06.416Z"
   },
   {
    "duration": 1954,
    "start_time": "2022-05-06T09:07:59.206Z"
   },
   {
    "duration": 41,
    "start_time": "2022-05-06T09:08:01.162Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-06T09:08:01.205Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T09:08:01.220Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-06T09:08:01.229Z"
   },
   {
    "duration": 41,
    "start_time": "2022-05-06T09:08:01.251Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-06T09:08:01.294Z"
   },
   {
    "duration": 761,
    "start_time": "2022-05-06T09:08:01.311Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-06T09:08:02.073Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-06T09:08:02.092Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T09:08:02.115Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T09:08:02.121Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-06T09:08:02.130Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-06T09:08:02.163Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T09:08:02.186Z"
   },
   {
    "duration": 34792,
    "start_time": "2022-05-06T09:08:02.193Z"
   },
   {
    "duration": 1490,
    "start_time": "2022-05-06T09:08:36.987Z"
   },
   {
    "duration": 849,
    "start_time": "2022-05-06T09:08:38.478Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T09:08:39.329Z"
   },
   {
    "duration": 30821,
    "start_time": "2022-05-06T09:08:39.347Z"
   },
   {
    "duration": 1534,
    "start_time": "2022-05-06T09:11:06.562Z"
   },
   {
    "duration": 35,
    "start_time": "2022-05-06T09:11:08.097Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-06T09:11:08.134Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T09:11:08.159Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T09:11:08.172Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-06T09:11:08.180Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-06T09:11:08.198Z"
   },
   {
    "duration": 743,
    "start_time": "2022-05-06T09:11:08.210Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-06T09:11:08.955Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-06T09:11:08.976Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T09:11:09.003Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T09:11:09.009Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-06T09:11:09.019Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-06T09:11:09.034Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-06T09:11:09.062Z"
   },
   {
    "duration": 19645,
    "start_time": "2022-05-06T09:11:09.072Z"
   },
   {
    "duration": 1469,
    "start_time": "2022-05-06T09:11:28.718Z"
   },
   {
    "duration": 616,
    "start_time": "2022-05-06T09:11:30.189Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T09:11:30.807Z"
   },
   {
    "duration": 24087,
    "start_time": "2022-05-06T09:11:30.816Z"
   },
   {
    "duration": 107,
    "start_time": "2022-05-06T09:11:54.905Z"
   },
   {
    "duration": 36471,
    "start_time": "2022-05-06T09:13:14.463Z"
   },
   {
    "duration": 1469,
    "start_time": "2022-05-06T09:14:34.858Z"
   },
   {
    "duration": 50,
    "start_time": "2022-05-06T09:14:36.329Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-06T09:14:36.381Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-06T09:14:36.397Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T09:14:36.402Z"
   },
   {
    "duration": 45,
    "start_time": "2022-05-06T09:14:36.411Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-06T09:14:36.460Z"
   },
   {
    "duration": 979,
    "start_time": "2022-05-06T09:14:36.479Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-06T09:14:37.459Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-06T09:14:37.483Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T09:14:37.505Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-06T09:14:37.512Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-06T09:14:37.525Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-06T09:14:37.561Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T09:14:37.591Z"
   },
   {
    "duration": 38317,
    "start_time": "2022-05-06T09:14:37.597Z"
   },
   {
    "duration": 2244,
    "start_time": "2022-05-06T09:15:15.916Z"
   },
   {
    "duration": 1055,
    "start_time": "2022-05-06T09:15:18.162Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T09:15:19.219Z"
   },
   {
    "duration": 29895,
    "start_time": "2022-05-06T09:15:19.228Z"
   },
   {
    "duration": 1390,
    "start_time": "2022-05-06T18:52:12.315Z"
   },
   {
    "duration": 39,
    "start_time": "2022-05-06T18:52:13.707Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-06T18:52:13.750Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T18:52:13.774Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T18:52:13.783Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-06T18:52:13.793Z"
   },
   {
    "duration": 39,
    "start_time": "2022-05-06T18:52:13.821Z"
   },
   {
    "duration": 714,
    "start_time": "2022-05-06T18:52:13.862Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-06T18:52:14.578Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T18:52:14.597Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T18:52:14.605Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-06T18:52:14.612Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-06T18:52:14.657Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-06T18:52:14.670Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T18:52:14.692Z"
   },
   {
    "duration": 35323,
    "start_time": "2022-05-06T18:52:14.698Z"
   },
   {
    "duration": 3089,
    "start_time": "2022-05-06T18:52:50.023Z"
   },
   {
    "duration": 1319,
    "start_time": "2022-05-06T18:52:53.114Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-06T18:52:54.435Z"
   },
   {
    "duration": 33168,
    "start_time": "2022-05-06T18:52:54.455Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-06T19:22:02.113Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-06T19:22:15.096Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-06T19:22:36.670Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-06T19:23:14.433Z"
   },
   {
    "duration": 165,
    "start_time": "2022-05-07T16:31:53.009Z"
   },
   {
    "duration": 163,
    "start_time": "2022-05-07T16:31:56.005Z"
   },
   {
    "duration": 61,
    "start_time": "2022-05-07T16:32:11.444Z"
   },
   {
    "duration": 1781,
    "start_time": "2022-05-07T16:32:45.464Z"
   },
   {
    "duration": 58,
    "start_time": "2022-05-07T16:32:47.248Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-07T16:32:47.308Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T16:32:47.332Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T16:32:47.337Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T16:32:47.351Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T16:32:47.370Z"
   },
   {
    "duration": 927,
    "start_time": "2022-05-07T16:32:47.384Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-07T16:32:48.313Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T16:32:48.345Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T16:32:48.356Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:32:48.363Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T16:32:48.373Z"
   },
   {
    "duration": 117,
    "start_time": "2022-05-07T16:32:48.384Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-07T16:32:48.503Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T16:33:26.054Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-07T16:33:31.467Z"
   },
   {
    "duration": 56,
    "start_time": "2022-05-07T16:33:31.475Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-07T16:33:31.533Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T16:33:31.551Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T16:33:31.559Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-07T16:33:31.577Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-07T16:33:31.605Z"
   },
   {
    "duration": 836,
    "start_time": "2022-05-07T16:33:31.636Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T16:33:32.474Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:33:32.494Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-07T16:33:32.504Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:33:32.537Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T16:33:32.546Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T16:33:32.560Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T16:34:05.083Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-07T16:35:30.593Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T16:35:31.120Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:36:16.998Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T16:36:27.151Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-07T16:36:42.639Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:36:47.942Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T16:37:07.896Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:37:21.928Z"
   },
   {
    "duration": 63,
    "start_time": "2022-05-07T16:37:21.934Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T16:37:21.998Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-07T16:37:24.333Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:38:12.206Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:38:41.662Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-07T16:38:55.689Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-07T16:39:05.498Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-07T16:40:10.839Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-07T16:52:23.987Z"
   },
   {
    "duration": 318,
    "start_time": "2022-05-07T16:52:41.089Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T16:53:41.314Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:53:51.327Z"
   },
   {
    "duration": 49,
    "start_time": "2022-05-07T16:53:51.333Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T16:53:51.384Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T16:53:51.400Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-07T16:53:51.404Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-07T16:53:51.413Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T16:53:51.433Z"
   },
   {
    "duration": 262,
    "start_time": "2022-05-07T16:53:51.444Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:54:45.714Z"
   },
   {
    "duration": 46,
    "start_time": "2022-05-07T16:54:45.720Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T16:54:45.768Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T16:54:45.783Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-07T16:54:45.787Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-07T16:54:45.796Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-07T16:54:45.815Z"
   },
   {
    "duration": 842,
    "start_time": "2022-05-07T16:54:45.843Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T16:54:46.687Z"
   },
   {
    "duration": 45506,
    "start_time": "2022-05-07T16:54:49.728Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T16:58:16.600Z"
   },
   {
    "duration": 62,
    "start_time": "2022-05-07T16:58:16.606Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T16:58:16.670Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T16:58:16.692Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T16:58:16.697Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-07T16:58:16.707Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T16:58:16.752Z"
   },
   {
    "duration": 949,
    "start_time": "2022-05-07T16:58:16.766Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-07T16:58:17.717Z"
   },
   {
    "duration": 93349,
    "start_time": "2022-05-07T16:58:21.075Z"
   },
   {
    "duration": 19894,
    "start_time": "2022-05-07T17:02:20.048Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-07T17:03:28.021Z"
   },
   {
    "duration": 18983,
    "start_time": "2022-05-07T17:03:43.732Z"
   },
   {
    "duration": 19394,
    "start_time": "2022-05-07T17:04:52.668Z"
   },
   {
    "duration": 19940,
    "start_time": "2022-05-07T17:05:22.596Z"
   },
   {
    "duration": 19565,
    "start_time": "2022-05-07T17:05:58.081Z"
   },
   {
    "duration": 20024,
    "start_time": "2022-05-07T17:08:48.185Z"
   },
   {
    "duration": 38104,
    "start_time": "2022-05-07T17:11:23.701Z"
   },
   {
    "duration": 9453,
    "start_time": "2022-05-07T17:12:24.809Z"
   },
   {
    "duration": 9392,
    "start_time": "2022-05-07T17:13:35.035Z"
   },
   {
    "duration": 9436,
    "start_time": "2022-05-07T17:26:32.429Z"
   },
   {
    "duration": 9525,
    "start_time": "2022-05-07T17:28:26.356Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T17:28:44.262Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-07T17:28:44.278Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:28:44.321Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T17:28:44.337Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T17:28:44.341Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-07T17:28:44.352Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:28:44.371Z"
   },
   {
    "duration": 928,
    "start_time": "2022-05-07T17:28:44.383Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-07T17:28:45.313Z"
   },
   {
    "duration": 9405,
    "start_time": "2022-05-07T17:28:51.363Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T17:30:00.635Z"
   },
   {
    "duration": 67,
    "start_time": "2022-05-07T17:30:00.642Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-07T17:30:00.711Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-07T17:30:00.729Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T17:30:00.738Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T17:30:00.749Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T17:30:00.772Z"
   },
   {
    "duration": 924,
    "start_time": "2022-05-07T17:30:00.785Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-07T17:30:01.712Z"
   },
   {
    "duration": 39165,
    "start_time": "2022-05-07T17:30:01.745Z"
   },
   {
    "duration": 19590,
    "start_time": "2022-05-07T17:31:36.175Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:32:28.763Z"
   },
   {
    "duration": 65,
    "start_time": "2022-05-07T17:32:28.769Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:32:28.837Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T17:32:28.853Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-07T17:32:28.859Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T17:32:28.868Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T17:32:28.888Z"
   },
   {
    "duration": 863,
    "start_time": "2022-05-07T17:32:28.938Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-07T17:32:29.803Z"
   },
   {
    "duration": 19460,
    "start_time": "2022-05-07T17:32:32.798Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:34:21.794Z"
   },
   {
    "duration": 19256,
    "start_time": "2022-05-07T17:35:30.784Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:36:00.733Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-07T17:36:00.739Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T17:36:00.784Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T17:36:00.800Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T17:36:00.804Z"
   },
   {
    "duration": 36,
    "start_time": "2022-05-07T17:36:00.814Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T17:36:00.853Z"
   },
   {
    "duration": 956,
    "start_time": "2022-05-07T17:36:00.868Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-07T17:36:01.827Z"
   },
   {
    "duration": 19213,
    "start_time": "2022-05-07T17:36:01.854Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-07T17:36:32.662Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T17:37:08.791Z"
   },
   {
    "duration": 38,
    "start_time": "2022-05-07T17:39:30.665Z"
   },
   {
    "duration": 61,
    "start_time": "2022-05-07T17:39:46.092Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-07T17:39:51.505Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T17:40:15.255Z"
   },
   {
    "duration": 78,
    "start_time": "2022-05-07T17:42:50.864Z"
   },
   {
    "duration": 59,
    "start_time": "2022-05-07T17:47:40.505Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T17:48:09.471Z"
   },
   {
    "duration": 45,
    "start_time": "2022-05-07T17:48:14.664Z"
   },
   {
    "duration": 56,
    "start_time": "2022-05-07T17:48:21.131Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-07T17:48:30.834Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:48:47.517Z"
   },
   {
    "duration": 74,
    "start_time": "2022-05-07T17:48:47.524Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-07T17:48:47.601Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T17:48:47.623Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-07T17:48:47.628Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T17:48:47.660Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T17:48:47.682Z"
   },
   {
    "duration": 935,
    "start_time": "2022-05-07T17:48:47.695Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-07T17:48:48.633Z"
   },
   {
    "duration": 20495,
    "start_time": "2022-05-07T17:48:48.662Z"
   },
   {
    "duration": 58,
    "start_time": "2022-05-07T17:49:09.159Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-07T17:49:09.219Z"
   },
   {
    "duration": 42,
    "start_time": "2022-05-07T17:49:17.788Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T17:49:34.229Z"
   },
   {
    "duration": 45,
    "start_time": "2022-05-07T17:49:45.470Z"
   },
   {
    "duration": 60,
    "start_time": "2022-05-07T17:49:53.134Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T17:49:59.953Z"
   },
   {
    "duration": 64,
    "start_time": "2022-05-07T17:49:59.961Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-07T17:50:00.027Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T17:50:00.054Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T17:50:00.069Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-07T17:50:00.088Z"
   },
   {
    "duration": 29,
    "start_time": "2022-05-07T17:50:00.117Z"
   },
   {
    "duration": 937,
    "start_time": "2022-05-07T17:50:00.150Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-07T17:50:01.090Z"
   },
   {
    "duration": 20529,
    "start_time": "2022-05-07T17:50:01.134Z"
   },
   {
    "duration": 57,
    "start_time": "2022-05-07T17:50:21.665Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-07T17:50:21.738Z"
   },
   {
    "duration": 50,
    "start_time": "2022-05-07T17:50:24.325Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T17:51:35.475Z"
   },
   {
    "duration": 57,
    "start_time": "2022-05-07T17:51:57.468Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T17:52:06.150Z"
   },
   {
    "duration": 38,
    "start_time": "2022-05-07T17:52:11.115Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:53:13.666Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:53:36.158Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T17:54:01.587Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:54:13.893Z"
   },
   {
    "duration": 69,
    "start_time": "2022-05-07T17:54:13.900Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:54:13.971Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-07T17:54:13.987Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T17:54:14.013Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-07T17:54:14.035Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-07T17:54:14.084Z"
   },
   {
    "duration": 870,
    "start_time": "2022-05-07T17:54:14.109Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-07T17:54:14.983Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-07T17:54:15.005Z"
   },
   {
    "duration": 18788,
    "start_time": "2022-05-07T17:54:15.037Z"
   },
   {
    "duration": 59,
    "start_time": "2022-05-07T17:54:33.827Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T17:54:33.889Z"
   },
   {
    "duration": 45,
    "start_time": "2022-05-07T17:54:33.939Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:54:33.986Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T17:54:43.606Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:56:01.823Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T17:56:01.830Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:56:01.880Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T17:56:01.897Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T17:56:01.902Z"
   },
   {
    "duration": 32,
    "start_time": "2022-05-07T17:56:01.915Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:56:01.949Z"
   },
   {
    "duration": 851,
    "start_time": "2022-05-07T17:56:01.961Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-07T17:56:02.814Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:56:02.850Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-07T17:56:02.861Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:56:13.846Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T17:56:48.481Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:56:53.806Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T17:58:30.690Z"
   },
   {
    "duration": 53,
    "start_time": "2022-05-07T17:58:30.697Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:58:30.752Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T17:58:30.769Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T17:58:30.774Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-07T17:58:30.787Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T17:58:30.834Z"
   },
   {
    "duration": 836,
    "start_time": "2022-05-07T17:58:30.850Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T17:58:31.688Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-07T17:58:31.709Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T17:58:31.735Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T17:58:31.750Z"
   },
   {
    "duration": 18405,
    "start_time": "2022-05-07T17:58:31.761Z"
   },
   {
    "duration": 55,
    "start_time": "2022-05-07T17:58:50.168Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-07T17:58:50.233Z"
   },
   {
    "duration": 39,
    "start_time": "2022-05-07T17:58:50.268Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T18:00:01.679Z"
   },
   {
    "duration": 105,
    "start_time": "2022-05-07T18:00:01.686Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-07T18:00:01.793Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-07T18:00:01.811Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T18:00:01.840Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-07T18:00:01.852Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:00:01.880Z"
   },
   {
    "duration": 875,
    "start_time": "2022-05-07T18:00:01.894Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T18:00:02.771Z"
   },
   {
    "duration": 67,
    "start_time": "2022-05-07T18:00:02.792Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-07T18:00:02.862Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T18:00:17.891Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:00:42.559Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:01:41.792Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T18:02:17.669Z"
   },
   {
    "duration": 42,
    "start_time": "2022-05-07T18:02:17.674Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T18:02:17.718Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T18:02:17.733Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T18:02:17.739Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-07T18:02:17.749Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T18:02:17.777Z"
   },
   {
    "duration": 849,
    "start_time": "2022-05-07T18:02:17.793Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-07T18:02:18.645Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T18:02:18.665Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:02:18.675Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T18:02:18.689Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-07T18:02:18.695Z"
   },
   {
    "duration": 18800,
    "start_time": "2022-05-07T18:02:18.737Z"
   },
   {
    "duration": 55,
    "start_time": "2022-05-07T18:02:37.539Z"
   },
   {
    "duration": 48,
    "start_time": "2022-05-07T18:02:37.596Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T18:02:37.646Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T18:02:37.661Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:02:37.673Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T18:03:17.993Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-07T18:04:37.228Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-07T18:04:42.762Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-07T18:04:54.076Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-07T18:04:55.836Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:05:12.151Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T18:05:21.556Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-07T18:05:27.919Z"
   },
   {
    "duration": 55,
    "start_time": "2022-05-07T18:05:27.926Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-07T18:05:27.983Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-07T18:05:27.998Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T18:05:28.002Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-07T18:05:28.018Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T18:05:28.047Z"
   },
   {
    "duration": 822,
    "start_time": "2022-05-07T18:05:28.062Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-07T18:05:28.887Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-07T18:05:28.910Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T18:05:28.940Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T18:05:28.955Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-07T18:05:28.961Z"
   },
   {
    "duration": 18446,
    "start_time": "2022-05-07T18:05:28.970Z"
   },
   {
    "duration": 64,
    "start_time": "2022-05-07T18:05:47.418Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-07T18:05:47.484Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-07T18:05:47.519Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-07T18:05:47.542Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T18:05:47.557Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-07T18:05:52.522Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-07T18:05:55.704Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-07T18:05:59.427Z"
   },
   {
    "duration": 1596,
    "start_time": "2022-05-08T08:03:57.443Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-08T08:03:59.041Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:03:59.085Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:03:59.101Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:03:59.104Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-08T08:03:59.115Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:03:59.133Z"
   },
   {
    "duration": 927,
    "start_time": "2022-05-08T08:03:59.143Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T08:04:00.072Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:04:00.092Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T08:04:00.103Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:04:00.120Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-08T08:04:00.125Z"
   },
   {
    "duration": 15850,
    "start_time": "2022-05-08T08:04:00.141Z"
   },
   {
    "duration": 49,
    "start_time": "2022-05-08T08:04:15.992Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-08T08:04:16.043Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-08T08:04:38.932Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:04:38.964Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-08T08:04:38.976Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:04:39.004Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T08:04:39.013Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-08T08:04:39.029Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:04:39.056Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-08T08:04:39.061Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T08:04:39.083Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T08:04:39.100Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:04:39.116Z"
   },
   {
    "duration": 22,
    "start_time": "2022-05-08T08:04:39.127Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:04:39.151Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:04:39.159Z"
   },
   {
    "duration": 33605,
    "start_time": "2022-05-08T08:04:39.170Z"
   },
   {
    "duration": 1964,
    "start_time": "2022-05-08T08:05:12.777Z"
   },
   {
    "duration": 889,
    "start_time": "2022-05-08T08:05:14.742Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:05:15.632Z"
   },
   {
    "duration": 17625,
    "start_time": "2022-05-08T08:05:15.640Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:05:33.272Z"
   },
   {
    "duration": 36,
    "start_time": "2022-05-08T08:05:33.277Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:05:33.315Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:05:33.328Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:05:33.331Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-08T08:05:33.342Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:05:33.374Z"
   },
   {
    "duration": 724,
    "start_time": "2022-05-08T08:05:33.384Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-08T08:05:34.110Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:05:34.129Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T08:05:34.144Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:05:34.172Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:05:34.178Z"
   },
   {
    "duration": 32362,
    "start_time": "2022-05-08T08:05:34.190Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:06:06.554Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-08T08:06:06.564Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:07:05.866Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:07:14.960Z"
   },
   {
    "duration": 115,
    "start_time": "2022-05-08T08:13:26.615Z"
   },
   {
    "duration": 2610,
    "start_time": "2022-05-08T08:13:33.408Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-08T08:13:36.020Z"
   },
   {
    "duration": 35,
    "start_time": "2022-05-08T08:13:36.050Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:14:04.632Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:14:10.063Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:14:44.512Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:14:55.833Z"
   },
   {
    "duration": 36,
    "start_time": "2022-05-08T08:14:55.838Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:14:55.876Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:14:55.889Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:14:55.893Z"
   },
   {
    "duration": 21,
    "start_time": "2022-05-08T08:14:55.900Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:14:55.923Z"
   },
   {
    "duration": 745,
    "start_time": "2022-05-08T08:14:55.933Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T08:14:56.680Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:14:56.698Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:14:56.706Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:14:56.718Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T08:14:56.723Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:14:56.729Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T08:14:56.742Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:17:06.364Z"
   },
   {
    "duration": 2676,
    "start_time": "2022-05-08T08:17:25.684Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-08T08:17:35.193Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T08:20:43.577Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:26:52.850Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:27:07.146Z"
   },
   {
    "duration": 4011,
    "start_time": "2022-05-08T08:27:20.492Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-08T08:27:28.067Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T08:27:33.266Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-08T08:27:40.890Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:27:57.474Z"
   },
   {
    "duration": 47,
    "start_time": "2022-05-08T08:27:57.481Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-08T08:27:57.530Z"
   },
   {
    "duration": 23,
    "start_time": "2022-05-08T08:27:57.550Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:27:57.576Z"
   },
   {
    "duration": 41,
    "start_time": "2022-05-08T08:27:57.581Z"
   },
   {
    "duration": 35,
    "start_time": "2022-05-08T08:27:57.623Z"
   },
   {
    "duration": 765,
    "start_time": "2022-05-08T08:27:57.660Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T08:27:58.427Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:27:58.445Z"
   },
   {
    "duration": 28,
    "start_time": "2022-05-08T08:27:58.454Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:27:58.483Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T08:27:58.490Z"
   },
   {
    "duration": 4341,
    "start_time": "2022-05-08T08:27:58.506Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-08T08:28:02.849Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-08T08:28:02.877Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T08:28:02.903Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-08T08:28:02.922Z"
   },
   {
    "duration": 37,
    "start_time": "2022-05-08T08:28:02.950Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:28:02.989Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:28:03.002Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:29:09.398Z"
   },
   {
    "duration": 38,
    "start_time": "2022-05-08T08:29:09.403Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:29:09.443Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:29:09.456Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:29:09.461Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-08T08:29:09.471Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:29:09.488Z"
   },
   {
    "duration": 701,
    "start_time": "2022-05-08T08:29:09.499Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-08T08:29:10.202Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:29:10.221Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:29:10.231Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-08T08:29:10.244Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T08:29:10.277Z"
   },
   {
    "duration": 6083,
    "start_time": "2022-05-08T08:29:10.283Z"
   },
   {
    "duration": 34,
    "start_time": "2022-05-08T08:29:16.368Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-08T08:29:16.404Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:29:16.430Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:29:16.472Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:29:16.485Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:29:16.497Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:29:16.507Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:29:23.037Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:29:23.831Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:33:05.985Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:33:06.666Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:33:13.204Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:33:13.211Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:33:13.224Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:33:13.232Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:33:13.241Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:33:13.255Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:33:13.263Z"
   },
   {
    "duration": 33107,
    "start_time": "2022-05-08T08:33:13.268Z"
   },
   {
    "duration": 1963,
    "start_time": "2022-05-08T08:33:46.377Z"
   },
   {
    "duration": 883,
    "start_time": "2022-05-08T08:33:48.343Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:33:49.228Z"
   },
   {
    "duration": 31705,
    "start_time": "2022-05-08T08:33:49.236Z"
   },
   {
    "duration": 1486,
    "start_time": "2022-05-08T08:36:47.646Z"
   },
   {
    "duration": 42,
    "start_time": "2022-05-08T08:36:49.134Z"
   },
   {
    "duration": 15,
    "start_time": "2022-05-08T08:36:49.177Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:36:49.194Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:36:49.198Z"
   },
   {
    "duration": 17,
    "start_time": "2022-05-08T08:36:49.205Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:36:49.224Z"
   },
   {
    "duration": 787,
    "start_time": "2022-05-08T08:36:49.235Z"
   },
   {
    "duration": 19,
    "start_time": "2022-05-08T08:36:50.024Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-08T08:36:50.045Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:36:50.077Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:36:50.092Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:36:50.097Z"
   },
   {
    "duration": 4171,
    "start_time": "2022-05-08T08:36:50.106Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:36:54.279Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:36:54.291Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:36:54.302Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:36:54.315Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:36:54.326Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:36:54.335Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:36:54.344Z"
   },
   {
    "duration": 20,
    "start_time": "2022-05-08T08:36:54.354Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:36:54.375Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T08:36:54.385Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:36:54.392Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:36:54.400Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:36:54.412Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:36:54.425Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:36:54.433Z"
   },
   {
    "duration": 33349,
    "start_time": "2022-05-08T08:36:54.438Z"
   },
   {
    "duration": 1730,
    "start_time": "2022-05-08T08:37:27.788Z"
   },
   {
    "duration": 792,
    "start_time": "2022-05-08T08:37:29.519Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:37:30.313Z"
   },
   {
    "duration": 28513,
    "start_time": "2022-05-08T08:37:30.322Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:38:52.563Z"
   },
   {
    "duration": 1626,
    "start_time": "2022-05-08T08:39:44.857Z"
   },
   {
    "duration": 40,
    "start_time": "2022-05-08T08:39:46.485Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T08:39:46.526Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:39:46.544Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T08:39:46.548Z"
   },
   {
    "duration": 25,
    "start_time": "2022-05-08T08:39:46.568Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:39:46.594Z"
   },
   {
    "duration": 859,
    "start_time": "2022-05-08T08:39:46.606Z"
   },
   {
    "duration": 31,
    "start_time": "2022-05-08T08:39:47.467Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:39:47.501Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:39:47.515Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T08:39:47.530Z"
   },
   {
    "duration": 38,
    "start_time": "2022-05-08T08:39:47.537Z"
   },
   {
    "duration": 5318,
    "start_time": "2022-05-08T08:39:47.579Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-08T08:39:52.899Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-08T08:39:52.934Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:39:52.974Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T08:39:52.986Z"
   },
   {
    "duration": 10,
    "start_time": "2022-05-08T08:39:53.002Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:39:53.014Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:39:53.027Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:39:53.042Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-08T08:39:53.047Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:39:53.081Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T08:39:53.089Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T08:39:53.096Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T08:39:53.107Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T08:39:53.122Z"
   },
   {
    "duration": 43,
    "start_time": "2022-05-08T08:39:53.132Z"
   },
   {
    "duration": 34368,
    "start_time": "2022-05-08T08:39:53.177Z"
   },
   {
    "duration": 1727,
    "start_time": "2022-05-08T08:40:27.546Z"
   },
   {
    "duration": 773,
    "start_time": "2022-05-08T08:40:29.274Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:40:30.049Z"
   },
   {
    "duration": 37345,
    "start_time": "2022-05-08T08:40:30.058Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T08:51:16.192Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:51:49.339Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T08:51:57.013Z"
   },
   {
    "duration": 108,
    "start_time": "2022-05-08T08:53:17.540Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T08:53:53.361Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T08:54:20.645Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T08:54:32.422Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T08:54:39.992Z"
   },
   {
    "duration": 1364,
    "start_time": "2022-05-08T09:08:12.102Z"
   },
   {
    "duration": 41,
    "start_time": "2022-05-08T09:08:13.468Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T09:08:13.511Z"
   },
   {
    "duration": 3,
    "start_time": "2022-05-08T09:08:13.525Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:08:13.529Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T09:08:13.539Z"
   },
   {
    "duration": 24,
    "start_time": "2022-05-08T09:08:13.557Z"
   },
   {
    "duration": 726,
    "start_time": "2022-05-08T09:08:13.583Z"
   },
   {
    "duration": 112,
    "start_time": "2022-05-08T09:08:14.311Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.426Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.427Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.429Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.430Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.432Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.433Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.435Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.436Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.437Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.439Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.440Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.442Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.443Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.472Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.474Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.476Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.477Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.478Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.480Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.481Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.482Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.484Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.485Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.486Z"
   },
   {
    "duration": 0,
    "start_time": "2022-05-08T09:08:14.488Z"
   },
   {
    "duration": 1386,
    "start_time": "2022-05-08T09:09:03.116Z"
   },
   {
    "duration": 37,
    "start_time": "2022-05-08T09:09:04.504Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T09:09:04.542Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T09:09:04.558Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T09:09:04.562Z"
   },
   {
    "duration": 18,
    "start_time": "2022-05-08T09:09:04.575Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T09:09:04.594Z"
   },
   {
    "duration": 711,
    "start_time": "2022-05-08T09:09:04.605Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T09:09:05.318Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:05.332Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T09:09:05.341Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T09:09:05.374Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:05.379Z"
   },
   {
    "duration": 3933,
    "start_time": "2022-05-08T09:09:05.388Z"
   },
   {
    "duration": 33,
    "start_time": "2022-05-08T09:09:09.323Z"
   },
   {
    "duration": 32,
    "start_time": "2022-05-08T09:09:09.358Z"
   },
   {
    "duration": 11,
    "start_time": "2022-05-08T09:09:09.392Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T09:09:09.405Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:09.421Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:09.431Z"
   },
   {
    "duration": 42,
    "start_time": "2022-05-08T09:09:09.441Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T09:09:09.485Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T09:09:09.493Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T09:09:09.502Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T09:09:09.509Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:09.518Z"
   },
   {
    "duration": 46,
    "start_time": "2022-05-08T09:09:09.528Z"
   },
   {
    "duration": 8,
    "start_time": "2022-05-08T09:09:09.578Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T09:09:09.587Z"
   },
   {
    "duration": 33752,
    "start_time": "2022-05-08T09:09:09.594Z"
   },
   {
    "duration": 2217,
    "start_time": "2022-05-08T09:09:43.348Z"
   },
   {
    "duration": 1065,
    "start_time": "2022-05-08T09:09:45.566Z"
   },
   {
    "duration": 7,
    "start_time": "2022-05-08T09:09:46.633Z"
   },
   {
    "duration": 21339,
    "start_time": "2022-05-08T09:09:46.642Z"
   },
   {
    "duration": 105,
    "start_time": "2022-05-08T09:20:53.960Z"
   },
   {
    "duration": 1817,
    "start_time": "2022-05-08T10:26:33.543Z"
   },
   {
    "duration": 37,
    "start_time": "2022-05-08T10:26:35.362Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T10:26:35.401Z"
   },
   {
    "duration": 2,
    "start_time": "2022-05-08T10:26:35.417Z"
   },
   {
    "duration": 6,
    "start_time": "2022-05-08T10:26:35.422Z"
   },
   {
    "duration": 32,
    "start_time": "2022-05-08T10:26:35.430Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T10:26:35.463Z"
   },
   {
    "duration": 760,
    "start_time": "2022-05-08T10:26:35.474Z"
   },
   {
    "duration": 16,
    "start_time": "2022-05-08T10:26:36.236Z"
   },
   {
    "duration": 9,
    "start_time": "2022-05-08T10:26:36.254Z"
   },
   {
    "duration": 12,
    "start_time": "2022-05-08T10:26:36.264Z"
   },
   {
    "duration": 4,
    "start_time": "2022-05-08T10:26:36.278Z"
   },
   {
    "duration": 5,
    "start_time": "2022-05-08T10:26:36.283Z"
   },
   {
    "duration": 4993,
    "start_time": "2022-05-08T10:26:36.291Z"
   },
   {
    "duration": 26,
    "start_time": "2022-05-08T10:26:41.286Z"
   },
   {
    "duration": 36,
    "start_time": "2022-05-08T10:26:41.313Z"
   },
   {
    "duration": 13,
    "start_time": "2022-05-08T10:26:41.351Z"
   },
   {
    "duration": 14,
    "start_time": "2022-05-08T10:26:41.366Z"
   },
   {
    "duration": 796,
    "start_time": "2022-05-08T10:27:27.219Z"
   },
   {
    "duration": 113,
    "start_time": "2022-05-08T10:27:55.699Z"
   },
   {
    "duration": 111,
    "start_time": "2022-05-08T10:27:58.208Z"
   },
   {
    "duration": 155,
    "start_time": "2022-05-08T10:32:37.970Z"
   },
   {
    "duration": 123,
    "start_time": "2022-05-08T10:35:23.243Z"
   },
   {
    "duration": 30,
    "start_time": "2022-05-08T10:35:58.189Z"
   },
   {
    "duration": 27,
    "start_time": "2022-05-08T10:36:03.767Z"
   },
   {
    "duration": 1540,
    "start_time": "2022-05-08T10:48:48.298Z"
   },
   {
    "duration": 177,
    "start_time": "2022-05-08T10:49:18.358Z"
   }
  ],
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Содержание",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "319.463px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}