2022-fall/seminars/sem02-sklearn-linregr.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Машинное обучение 1, ПМИ ФКН ВШЭ\n",
    "\n",
    "## Семинар 2\n",
    "\n",
    "## `sklearn` и линейная регрессия"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "На этом семинаре мы:\n",
    "* Узнаем о том, что такое переобучение, что такое линейная регрессия и как не допустить переобучения в линейных моделях;\n",
    "* Обучим линейную регрессию для предсказания цены дома;\n",
    "* Научимся работать с разными типами признаков;\n",
    "* Поймем, в чем отличие между разными регуляризаторами;\n",
    "* Научимся пользоваться основными инструментами в `sklearn`: моделями, трансформерами и pipeline;\n",
    "* Обсудим преобразования признаков и целевой переменной, которые могут помочь в обучении линейных моделей."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import warnings\n",
    "warnings.simplefilter(\"ignore\")\n",
    "%matplotlib inline\n",
    "\n",
    "sns.set(style=\"darkgrid\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Часть 0. Введение в линейные модели"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Напомним, что линейная регрессия &mdash; это модель следующего вида:\n",
    "$$a(x) = \\langle w, x \\rangle + w_0$$\n",
    "где $w \\in \\mathbb{R}^d$, $w_0 \\in \\mathbb{R}$. Обучить линейную регрессию &mdash; значит найти $w$ и $w_0$.\n",
    "\n",
    "В машинном обучении часто говорят об *обобщающей способности модели*, то есть о способности модели работать на новых тестовых данных хорошо. Если модель будет идеально предсказывать выборку, на которой она обучалась, но при этом просто ее запомнит, не \"вытащив\" из данных никакой закономерности, от нее будет мало толку. Такую модель называют *переобученной*: она слишком подстроилась под обучающие примеры, не выявив никакой полезной закономерности, которая позволила бы ей совершать хорошие предсказания на данных, которые она ранее не видела.\n",
    "\n",
    "Рассмотрим следующий пример, на котором будет хорошо видно, что значит переобучение модели. Для этого нам понадобится сгенерировать синтетические данные. Рассмотрим зависимость $y(x) = \\cos(1.5\\pi x)$, $y$ &mdash; целевая переменная, а $x$ - объект (просто число от 0 до 1). В жизни мы наблюдаем какое-то конечное количество пар объект-таргет, поэтому смоделируем это, взяв 30 случайных точек $x_i$ в отрезке $[0;1]$. Более того, в реальной жизни целевая переменная может быть зашумленной (измерения в жизни не всегда точны), смоделируем это, зашумив значение функции нормальным шумом: $\\tilde{y}_i = y(x_i) + \\mathcal{N}(0, 0.01)$.\n",
    "\n",
    "Попытаемся обучить три разных линейных модели: признаки для первой &mdash;  $\\{x\\}$, для второй &mdash; $\\{x, x^2, x^3, x^4\\}$, для третьей &mdash; $\\{x, \\dots, x^{20}\\}$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "np.random.seed(36)\n",
    "x = np.linspace(0, 1, 100)\n",
    "y = np.cos(1.5 * np.pi * x)\n",
    "\n",
    "x_objects = np.random.uniform(0, 1, size=30)\n",
    "y_objects = np.cos(1.5 * np.pi * x_objects) + np.random.normal(scale=0.1, size=x_objects.shape)\n",
    "\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "fig, axs = plt.subplots(figsize=(16, 4), ncols=3)\n",
    "for i, degree in enumerate([1, 4, 20]):\n",
    "    X_objects = PolynomialFeatures(degree, include_bias=False).fit_transform(x_objects[:, None])\n",
    "    X = PolynomialFeatures(degree, include_bias=False).fit_transform(x[:, None])\n",
    "    regr = LinearRegression().fit(X_objects, y_objects)\n",
    "    y_pred = regr.predict(X)\n",
    "    axs[i].plot(x, y, label=\"Real function\")\n",
    "    axs[i].scatter(x_objects, y_objects, label=\"Data\")\n",
    "    axs[i].plot(x, y_pred, label=\"Prediction\")\n",
    "    if i == 0:\n",
    "        axs[i].legend()\n",
    "    axs[i].set_title(\"Degree = %d\" % degree)\n",
    "    axs[i].set_xlabel(\"$x$\")\n",
    "    axs[i].set_ylabel(\"$f(x)$\")\n",
    "    axs[i].set_ylim(-2, 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Вопрос 1**: Почему первая модель получилась плохой, а третья переобучилась?\n",
    "\n",
    "Чтобы избежать переобучения, модель регуляризуют. Обычно переобучения в линейных моделях связаны с большими весами, а поэтому модель часто штрафуют за большие значения весов, добавляя к функционалу качества, например, квадрат $\\ell^2$-нормы вектора $w$:\n",
    "\n",
    "$$Q_{reg}(X, y, a) = Q(X, y, a) + \\lambda \\|w\\|_2^2$$\n",
    "\n",
    "Это слагаемое называют $\\ell_2$-регуляризатором, а коэффициент $\\lambda$ &mdash; коэффициентом регуляризации. Более подробно о регуляризации и переобучении вы узнаете из лекций.\n",
    "\n",
    "**Вопрос 2**: Почему большие веса в линейной модели &mdash; плохо?\n",
    "\n",
    "**Вопрос 3**: Почему регуляризовать $w_0$ &mdash; плохая идея?\n",
    "\n",
    "**Вопрос 4**: На что влияет коэффициент $\\lambda$? Что будет происходить с моделью, если $\\lambda$ начать уменьшать? Что будет, если $\\lambda$ сделать слишком большим?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Часть 1. Загружаем данные"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы будем работать с данными из соревнования [House Prices: Advanced Regression Techniques](https://www.kaggle.com/c/house-prices-advanced-regression-techniques/overview), в котором требовалось предсказать стоимость жилья. Давайте сначала загрузим и немного изучим данные (можно взять `train.csv` со страницы соревнования и переименовать в `train_sem2.csv`)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "!wget  -O 'train_sem2.csv' -q 'https://www.dropbox.com/s/6dxq90t0prn2vaw/_train_sem2.csv?dl=0'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>Id</th>\n",
       "      <th>MSSubClass</th>\n",
       "      <th>MSZoning</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>LotArea</th>\n",
       "      <th>Street</th>\n",
       "      <th>Alley</th>\n",
       "      <th>LotShape</th>\n",
       "      <th>LandContour</th>\n",
       "      <th>Utilities</th>\n",
       "      <th>...</th>\n",
       "      <th>PoolArea</th>\n",
       "      <th>PoolQC</th>\n",
       "      <th>Fence</th>\n",
       "      <th>MiscFeature</th>\n",
       "      <th>MiscVal</th>\n",
       "      <th>MoSold</th>\n",
       "      <th>YrSold</th>\n",
       "      <th>SaleType</th>\n",
       "      <th>SaleCondition</th>\n",
       "      <th>SalePrice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>60</td>\n",
       "      <td>RL</td>\n",
       "      <td>65.0</td>\n",
       "      <td>8450</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2008</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>208500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>RL</td>\n",
       "      <td>80.0</td>\n",
       "      <td>9600</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2007</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>181500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>60</td>\n",
       "      <td>RL</td>\n",
       "      <td>68.0</td>\n",
       "      <td>11250</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>2008</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>223500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>70</td>\n",
       "      <td>RL</td>\n",
       "      <td>60.0</td>\n",
       "      <td>9550</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2006</td>\n",
       "      <td>WD</td>\n",
       "      <td>Abnorml</td>\n",
       "      <td>140000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>60</td>\n",
       "      <td>RL</td>\n",
       "      <td>84.0</td>\n",
       "      <td>14260</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2008</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "      <td>250000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 81 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Id  MSSubClass MSZoning  LotFrontage  LotArea Street Alley LotShape  \\\n",
       "0   1          60       RL         65.0     8450   Pave   NaN      Reg   \n",
       "1   2          20       RL         80.0     9600   Pave   NaN      Reg   \n",
       "2   3          60       RL         68.0    11250   Pave   NaN      IR1   \n",
       "3   4          70       RL         60.0     9550   Pave   NaN      IR1   \n",
       "4   5          60       RL         84.0    14260   Pave   NaN      IR1   \n",
       "\n",
       "  LandContour Utilities  ... PoolArea PoolQC Fence MiscFeature MiscVal MoSold  \\\n",
       "0         Lvl    AllPub  ...        0    NaN   NaN         NaN       0      2   \n",
       "1         Lvl    AllPub  ...        0    NaN   NaN         NaN       0      5   \n",
       "2         Lvl    AllPub  ...        0    NaN   NaN         NaN       0      9   \n",
       "3         Lvl    AllPub  ...        0    NaN   NaN         NaN       0      2   \n",
       "4         Lvl    AllPub  ...        0    NaN   NaN         NaN       0     12   \n",
       "\n",
       "  YrSold  SaleType  SaleCondition  SalePrice  \n",
       "0   2008        WD         Normal     208500  \n",
       "1   2007        WD         Normal     181500  \n",
       "2   2008        WD         Normal     223500  \n",
       "3   2006        WD        Abnorml     140000  \n",
       "4   2008        WD         Normal     250000  \n",
       "\n",
       "[5 rows x 81 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"train_sem2.csv\")\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1460, 81)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Id', 'MSSubClass', 'MSZoning', 'LotFrontage', 'LotArea', 'Street',\n",
       "       'Alley', 'LotShape', 'LandContour', 'Utilities', 'LotConfig',\n",
       "       'LandSlope', 'Neighborhood', 'Condition1', 'Condition2', 'BldgType',\n",
       "       'HouseStyle', 'OverallQual', 'OverallCond', 'YearBuilt', 'YearRemodAdd',\n",
       "       'RoofStyle', 'RoofMatl', 'Exterior1st', 'Exterior2nd', 'MasVnrType',\n",
       "       'MasVnrArea', 'ExterQual', 'ExterCond', 'Foundation', 'BsmtQual',\n",
       "       'BsmtCond', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinSF1',\n",
       "       'BsmtFinType2', 'BsmtFinSF2', 'BsmtUnfSF', 'TotalBsmtSF', 'Heating',\n",
       "       'HeatingQC', 'CentralAir', 'Electrical', '1stFlrSF', '2ndFlrSF',\n",
       "       'LowQualFinSF', 'GrLivArea', 'BsmtFullBath', 'BsmtHalfBath', 'FullBath',\n",
       "       'HalfBath', 'BedroomAbvGr', 'KitchenAbvGr', 'KitchenQual',\n",
       "       'TotRmsAbvGrd', 'Functional', 'Fireplaces', 'FireplaceQu', 'GarageType',\n",
       "       'GarageYrBlt', 'GarageFinish', 'GarageCars', 'GarageArea', 'GarageQual',\n",
       "       'GarageCond', 'PavedDrive', 'WoodDeckSF', 'OpenPorchSF',\n",
       "       'EnclosedPorch', '3SsnPorch', 'ScreenPorch', 'PoolArea', 'PoolQC',\n",
       "       'Fence', 'MiscFeature', 'MiscVal', 'MoSold', 'YrSold', 'SaleType',\n",
       "       'SaleCondition', 'SalePrice'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Первое, что стоит заметить &mdash; у нас в данных есть уникальное для каждого объекта поле id. Обычно такие поля только мешают и способствуют переобучению. Удалим это поле из данных.\n",
    "\n",
    "Разделим данные на обучающую и тестовую выборки. Для простоты не будем выделять дополнительно валидационную выборку (хотя это обычно стоит делать, она нужна для подбора гиперпараметров модели, то есть параметров, которые нельзя подбирать по обучающей выборке). Дополнительно нам придется отделить значения целевой переменной от данных.\n",
    "\n",
    "**Вопрос 1**: Почему поля типа id могут вызвать переобучение модели (не обязательно линейной)?\n",
    "\n",
    "**Вопрос 2**: Почему стоит дополнительно отделять валидационную выборку?\n",
    "\n",
    "**Вопрос 3**: Обратите внимание на фиксацию `random_state` при сплите данных. Почему это важно?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "data = data.drop(columns=[\"Id\"])\n",
    "\n",
    "y = data[\"SalePrice\"]\n",
    "X = data.drop(columns=[\"SalePrice\"])\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим сначала на значения целевой переменной."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='SalePrice', ylabel='Density'>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Судя по гистограмме, у нас есть примеры с нетипично большой стоимостью, что может помешать нам, если наша функция потерь слишком чувствительна к выбросам. В дальнейшем мы рассмотрим способы, как минимизировать ущерб от этого."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Так как для решения нашей задачи мы бы хотели обучить линейную регрессию, было бы хорошо найти признаки, \"наиболее линейно\" связанные с целевой переменной, иначе говоря, посмотреть на коэффициент корреляции Пирсона между признаками и целевой переменной. Корреляция Пирсона записывается следующим образом:\n",
    "\n",
    "$$\n",
    "\\text{corr}(X, Y)=\\frac{\\operatorname{cov}_{X Y}}{\\sigma_X \\sigma_Y}=\\frac{\\sum_{i=1}^n(X_i - \\bar{X})(Y_i - \\bar{Y})}{\\sqrt{\\sum_{i=1}^n(X_i - \\bar{X})^2 \\sum_{i=1}^n(Y_i - \\bar{Y})^2}}\n",
    "$$\n",
    "\n",
    "**Вопрос**: Что означает, что коэффициент корреляции Пирсона между двумя случайными величинами равен 1? -1? 0?\n",
    "\n",
    "Заметим, что не все признаки являются числовыми, пока что мы не будем рассматривать такие признаки."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "numeric_data = X_train.select_dtypes([np.number])\n",
    "numeric_data_mean = numeric_data.mean()\n",
    "numeric_features = numeric_data.columns\n",
    "\n",
    "X_train = X_train.fillna(numeric_data_mean)\n",
    "X_test = X_test.fillna(numeric_data_mean)\n",
    "\n",
    "correlations = X_train[numeric_features].corrwith(y_train).sort_values(ascending=False)\n",
    "\n",
    "\n",
    "plot = sns.barplot(y=correlations.index, x=correlations)\n",
    "plot.figure.set_size_inches(15, 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на признаки из начала списка. Для этого нарисуем график зависимости целевой переменной от каждого из признаков. На этом графике каждая точка соответствует паре признак-таргет (такие графики называются scatter-plot)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHQAAAFcCAYAAACpy+lKAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOz9eYys2XnY93/f/a231q7ebt++M/fOnRmOhsNRRuLEFuOFgSRbksFIBpRABI1YiQMrlrLQSWRYMB1SkC0DIgVvcgDDMOAgiAPbgCAxGksiYNHOT6IkmzI10izUcObut2+vVV3ruy+/P96umu6+va9V1c8HGJC3q7vrvNVVzznvc855jpJlWYYQQgghhBBCCCGEGBvqZTdACCGEEEIIIYQQQhyPJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMftkNGEebm33SNLvsZjA9XaLR6F12M87UJF4TyHWNm3G+LlVVmJoqXnYzJsJJYv2ovnekXcc3qm2Tdh3PqLYLTt82ifdnZ5Li/X6kvedv3Nos7T1/Z9Hmw2K9JHROIE2zkUjoACPTjrM0idcEcl3jZlKvSxzdSWP9qL53pF3HN6ptk3Ydz6i2C0a7bVfJpMX7/Uh7z9+4tVnae/7Ou82y5UoIIYQQQgghhBBizEhCRwghhBBCCCGEEGLMSEJHCCGEEEIIIYQQYsxIQkcIIYQQQgghhBBizEhCRwghhBBCCCGEEGLMSEJHCCGEEEIIIYQQYsxIQkcIIYQQQgghhBBizOiX3QAhxNXU8yJWmy5uEONYOvN1h1LBuOxmCSHEWJFYKoQQ+5MYKSadrNARQly4nhdx90mbOE0pFXTiNOXukzY9L7rspgkhxNiQWCqEEPuTGCmuAknoCCEu3GrTxTI1LENDURQsQ8MyNVab7mU3TQghxobEUiGE2J/ESHEVSEJHCHHh3CDG1HeGH1NXcYP4klokhBDjR2KpEELsT2KkuAqkho4Q4sI5lk4Yp1iGNvxaGKc41s6QJPuehRBip64bcmepjRvEbHYCkjSlVrKGj+8VS4UQ4qrpeRGbnYDH6z3KjsF0xaawNf6UGCkmiazQEUJcuPm6QxAmBFFClmUEUUIQJszXneH3yL5nIYTYqedFvPdgcxgXK0WDBytdWr1g31gqhBBXzWAMWSkaaCr4QcKjtR6tXiAxUkwcSegIIS5cqWBw+3oVXVXpeTG6qnL7enXH6hvZ9yyEEDutNl3sbXGxVra4OV+m04v2jaVCCHHVDMaQtbLFjdkytqmRJBmdXiQxUkwcWW8mhLgUpYJBabG67+NuEFMq7AxRpp4ngIQQ4ipyg5jpuoa37WvVkommqbx6e/rS2iWEEKNk+xjSsXUcu0SWZfS8WJI5YuLICh0hxEga1NnZTvY9CyGuMsfSCaJkx9ckLgohxE4yhhRXiSR0hBAj6Sh1doQQ4iqZrzv4EheFEOJAMoYUV4kkdIQQI+kodXaEEOIqKRUMXro5JXFRCCEOIGNIcZXIujMhxMg6rM6OEEJcNWXH5HmJi0IIcSAZQ4qrQlboCCGEEEIIIYQQQowZSegIIYQQQgghhBBCjBlJ6AghhBBCCCGEEEKMGUnoCCGEEEIIIYQQQowZSegIIYQQQgghhBBCjBlJ6AghhBBCCCGEEEKMGUnoCCGEEEIIIYQQQowZSegIIYQQQgghhBBCjBlJ6AghhBBCCCGEEEKMGf0inuTx48f8D//D/zD8d7fbpdfr8R/+w3/g3r17/NRP/RStVotarcbP/dzPcevWLYALf0wIIcTpSLwXQojJJ7FeCCFGw4Ws0Llx4wZf/vKXh/99z/d8D5/61KcA+MIXvsBnPvMZvvKVr/CZz3yGz3/+88Ofu+jHhBBCnI7EeyGEmHwS64UQYjRc+JarMAz5lV/5FX74h3+YRqPBu+++O+wAPvWpT/Huu+/SbDYv/DEhhBBnS+K9EEJMPon1QghxeS5ky9V2X/3qV5mfn+eVV17h7bffZn5+Hk3TANA0jbm5OZaXl8my7EIfq9frR76G6enSWb4kpzI7W77sJpy5SbwmkOsaN5N6XRdp3OP9SWP9qL53pF3HN6ptk3Ydz6i2C0a7bUc17rEeJi/e70fae/7Grc3S3vN33m2+8ITOL/7iL/LDP/zDF/20Z6rR6JGm2WU3g9nZMuvr3ctuxpmaxGsCua5xM87XparKyCSdxz3enyTWj+p7R9p1fKPaNmnX8Yxqu+D0bRuVeD/usR4mK97vR9p7/satzdLe83cWbT4s1l9oQmd1dZWvf/3rfPGLXwRgYWGB1dVVkiRB0zSSJGFtbY2FhQWyLLvQx4QQQpwdifdCCDH5JNYLIcTlutAaOr/0S7/EJz/5SaampgCYnp7m5Zdf5o033gDgjTfe4OWXX6Zer1/4Y0IIIc6OxHshhJh8EuuFEOJyKVmWXdjeoe/7vu/jc5/7HH/6T//p4dfu3LnDT/3UT9HpdKhUKvzcz/0ct2/fvpTHjkq2XJ2fSbwmkOsaN+N8XaOyBH8S4v0kLcGXdh3fqLZN2nU8o9oumIwtV5MQ62Gy4v1+pL3nb9zaLO09fxex5epCEzqTQhI652cSrwnkusbNOF/XKAzwJ8UkDfClXcc3qm2Tdh3PqLYLJiOhMykmKd7vR9p7/satzdLe83cRCZ0LP7ZcCCGEEEIIIYQQQpyOJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMftkNEEKcv54Xsdp0cYMYx9KZrzuUCsZlN0sIIcQx9LyI5oMmSysdieVCiImx1zh19rIbJcSYkBU6Qky4nhdx90mbOE0pFXTiNOXukzY9L7rspgkhhDiiQSyPYonlQojJsd84teuGl900IcaCJHSEmHCrTRfL1LAMDUVRsAwNy9RYbbqX3TQhhBBHNIjltqlLLBdCTIz9xqlP1nuX3TQhxoIkdISYcG4QY+o7P+qmruIG8SW1SAghxHFJLBdCTKL9YpusPhTiaCShI8SEcyydME53fC2MUxxLSmgJIcS4kFguhJhE+8U2qQ8mxNFIQkeICTdfdwjChCBKyLKMIEoIwoT5unPZTRNCCHFEg1juh7HEciHExNhvnHp9tnTZTRNiLEhCR4gJVyoY3L5eRVdVel6Mrqrcvl6VmQ8hhBgjg1hu6BLLhRCTY79xatkxL7tpQowFWacrxBVQKhiUFquX3QwhhBCnUCoYzM6WqTuSxBFCTA4ZpwpxcrJCRwghhBBCCCGEEGLMXFhCJwgCvvCFL/Bn/+yf5b/4L/4L/vf//X8H4N69e/zIj/wI3/d938eP/MiPcP/+/eHPXPRjQgghTk/ivRBCTD6J9UIIcfkuLKHzpS99Ccuy+MpXvsKv/Mqv8NnPfhaAL3zhC3zmM5/hK1/5Cp/5zGf4/Oc/P/yZi35MCCHE6Um8F0KIySexXgghLt+FJHT6/T6//Mu/zGc/+1kURQFgZmaGRqPBu+++y6c+9SkAPvWpT/Huu+/SbDYv/DEhhBCnJ/FeCCEmn8R6IYQYDRdSFPnRo0fUajX+0T/6R/z7f//vKRaLfPazn8W2bebn59E0DQBN05ibm2N5eZksyy70sXq9fhEvhRBCTDSJ90IIMfkk1gshxGi4kIROkiQ8evSIj370o/z1v/7X+YM/+AP+yl/5K/yDf/APLuLpz9z0dOmymzA0O1u+7CacuUm8JpDrGjeTel3nbZLi/Ulj/ai+d6RdxzeqbZN2Hc+otgtGu20HmaRYD5MX7/cj7T1/49Zmae/5O+82X0hCZ2FhAV3Xh8sh/5P/5D9hamoK27ZZXV0lSRI0TSNJEtbW1lhYWCDLsgt97DgajR5pmp3HS3Uss7Nl1te7l92MMzWJ1wRyXeNmnK9LVZVLTTpPUrw/Sawf1feOtOv4RrVt0q7jGdV2wenbdpnxfpJiPUxWvN+PtPf8jVubpb3n7yzafFisv5AaOvV6nT/+x/84X/va14C8Gn2j0eDWrVu8/PLLvPHGGwC88cYbvPzyy9Trdaanpy/0MSGEEKcn8V4IISafxHohhBgNSpZlF7LU5NGjR/yNv/E3aLVa6LrOX/2rf5VPfvKT3Llzh5/6qZ+i0+lQqVT4uZ/7OW7fvg1w4Y8dlazQOT/jdE09L2K16eIGMY6lM193KBWMPb93nK7rOOS6Rs9lr9CByYn3kzRjK+06vlFt26i36zh940W2axSN8wodmJxYD5MV7/dzme09SVwYt9cXxq/N0t7zdxErdC4soTNJJKFzfsblmnpexN0nbSxTw9RVwjglCBNuX6/u2UGNy3Udl1zX6LnsAf4kmaQBvrTr+Ea1baPcrnsPm8fqGy+qXaP4esH4J3QmySTF+/1cVnuPO2YeGLfXF8avzdLe8zcxW66EmDSrTRfL1LAMDUVRsAwNy9RYbbqX3TQhhBDiUkjfKITYTeKCEOdLEjpCnIAbxJj6zo+Pqau4QXxJLRJCCCEul/SNQojdJC4Icb4koSPECTiWThinO74WximOdSEHxwkhhBAjR/pGIcRuEheEOF+S0BHiBObrDkGYEEQJWZYRRAlBmDBfdy67aUIIIcSlkL5RCLGbxAUhzpckdIQ4gVLB4Pb1Krqq0vNidFW91KKPQgghxGWTvlEIsZvEBSHOl6x1E+KESgWD0mL1spshhBBCjAzpG4UQu0lcEOL8yAodIYQQQgghhBBCiDEjCR0hhBBCCCGEEEKIMSNbroQ4pZ4Xsdp0cYMYx9KZrzuyL1gIIcQO0lcIIcThBrHyYcMl8iOJlUIcQlboCHEKPS/i7pM2cZpSKujEacrdJ216XnTZTRNCCDEipK8QQojDbY+VFceUWCnEEUhCR4hTWG26WKaGZWgoioJlaFimxmrTveymCSGEGBHSVwghxOEkVgpxfJLQEeIU3CDG1Hd+jExdxQ3iS2qREEKIUSN9hRBCHE5ipRDHJwkdIU7BsXTCON3xtTBOcSwpTyWEECInfYUQQhxOYqUQxycJHSFOYb7uEIQJQZSQZRlBlBCECfN157KbJoQQYkRIXyGEEIeTWCnE8UlCR4hTKBUMbl+voqsqPS9GV1VuX69KNX4hhBBD0lcIIcThtsfKjhtKrBTiCGT9mhCnVCoYlBarZ/575YhbIYSYHOfVVwxInyGEGBWniUeDWDk7W2Z9vXvOLRVi/MkKHSFGkBxxK4QQ4qikzxBCjAqJR0JcLEnoCDGC5NhGIYQQRyV9hhBiVEg8EuJiSUJHiBEkxzYKIYQ4KukzhBCjQuKREBdLEjpCjCA5tlEIIcRRSZ8hhBgVEo+EuFiS0BFiBMmxjUIIIY5K+gwhxKiQeCTExZKEjhAjSI64FUIIcVTSZwghRoXEIyEulqx9E2JEnfcRt0IIISaH9BlCiFEh8UiIiyMrdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxowkdIQQQgghhBBCCCHGjCR0hBBCCCGEEEIIIcaMJHSEEEIIIYQQQgghxsyFJXS++7u/m+///u/nh37oh/ihH/ohfvM3fxOAN998kx/8wR/k+77v+/hLf+kv0Wg0hj9z0Y+J8dLzIu4stXnrboM7S216XnTZTRJCIPFeCHFxZCxweSTWi8skn30hche6Qucf/sN/yJe//GW+/OUv86f+1J8iTVP+2l/7a3z+85/nK1/5Cq+//jo///M/D3Dhj4nx0vMi7j5pE6cppYJOnKbcfSLBXIhRIfFeCHHeZCxw+STWi8sgn30hPnSpW67efvttLMvi9ddfB+DTn/40v/7rv34pj4nxstp0sUwNy9BQFAXL0LBMjdWme9lNE0LsQeK9EOKsyVhg9EisFxdBPvtCfEi/yCf7yZ/8SbIs4+Mf/zj/6//6v7K8vMz169eHj9frddI0pdVqXfhjtVrtfC9enCk3iCkVdr59TV2l58WX1CIhxHYS74UQ503GApdPYr24DPLZF+JDF5bQ+ef//J+zsLBAGIb87M/+LD/zMz/Dn/kzf+ainv5MTU+XLrsJQ7Oz5ctuwpk7yjUtuhFRnGKbH76F/TCmWlVH9jUZ1XadllyX2G1S4v1JY/2ovnekXcc3qm2TduWOOhYY1dcLRrtth5mUWA+TF+/3MyntHeX7gMt+/uOS9p6/827zhSV0FhYWADBNk8985jP8+I//OH/xL/5Fnjx5MvyeZrOJqqrUajUWFhYu9LHjaDR6pGl23JfgzM3Olllf7152M87UUa/JVODxehfL1DB1lTBOCcKE29erI/maTOLfCuS6RpGqKpeedJ6UeH+SWD+q7x1p1/GNatukXR86ylhgVF8vOH3bLjveT0qsh8mK9/uZpPaO6n3AJL3Go2jc2gtn0+bDYv2F1NBxXZduN7+QLMv41V/9VV5++WU+9rGP4fs+v/d7vwfAv/gX/4Lv//7vB7jwx8R4KRUMbl+voqv58kpdVbl9vUqpYFx2045FKvSLSSPxXoiLIf3H5IwFxpHEenGWjhvP5LMvxIcuZIVOo9Hgf/qf/ieSJCFNU55//nm+8IUvoKoqX/ziF/nCF75AEAQsLi7ypS99CeDCHxPjp1QwKC1WL7sZJzao0G+ZGqWCThjnFfqlQxLjTOK9EOdP+o8PjftYYFxJrBdnZb94NjNz8Ooz+ewLkVOyLLv8vUNjRrZcnZ9JvCbY+7ruLOXHLVqGNvxaECXoqsrzY9JBXaW/17i47CX4k2SSluBLu45vVNs2O1vmd998PHL9xyi/XqPYLhj/LVeTZJLi/X5Gtb37jYdnp0vUnfFKUI/qa7wfae/5m5gtV0KIp7lBjKnv/AiauoobSIV+IYQQ+5P+QwgxKfaLZ1dxG6kQJ3Ghx5YLIT7kWPmy0u0zEmGc4lhn+7HseRGrTRc3iHEsnfm6c+WW5AshxCTZq/9o90M6vYi37jYk1gshDjRKY8P9xsOzVYlfQhyFrNAR4pLM1x2CMCGIErIsI4gSgjBhvu6c2XMM9iXHaUqpoBOn+b5kmfUQQojxtbv/aPUCHqx0qRQNifVCiAON2thwv/Hw9VnZTijEUUhCR4hLchEV+lebLpapYRkaiqJgGRqWqbHadM/sOYQQQlys3f1Hpxdxc75MrWxJrBdCHGjUxob7jYfLjnkp7RFi3MiWKyEu0XlX6HeDmFJh58c835csdRaEEGKcbe8/3rrbkFgvhDiSURwbyolVQpycrNARYoIN9iVvdx51eoQQQlweifVCiKOSeCHEZJGEjhAT7CLq9AghhLhcEuuFEEcl8UKIySIJHSEm2EXU6RFCCHG5JNYLIY5K4oUQk+XYa+uWl5dZXV3ltddeO4fmCCHOmuxLFicl8V6I8SGxXpyUxPqrR+KFEJPjyCt0njx5wqc//Wl+4Ad+gP/2v/1vAfj1X/91Pve5z51b44QQQlw8ifdCCDH5JNYLIcT4O3JC5/Of/zz/+X/+n/ONb3wDXc8X9vyJP/En+O3f/u1za5wQR9XzIu4stXnrboM7S216XnTZTRJibEm8F2L8SD8ojktivbgIEpuEOF9HTui89dZb/NiP/RiqqqIoCgDlcplut3tujRPiKHpexN0nbeI0pVTQidOUu0+kwxDipCTeCzFepB8UJyGxXpw3iU1CnL8jJ3Smp6d58ODBjq998MEHLCwsnHmjhDiO1aaLZWpYhoaiKFiGhmVqrDbdM/n9MrMgrhqJ90KMl939YJpmNDo+v/3WsvRbYl8S68VJHGdcfN5jdCHEMRI6f+kv/SX+yl/5K/ziL/4icRzzxhtv8L/8L/8Lf/kv/+XzbJ8Qh3KDGFPf+VY2dRU3iE/9u2VmQVxFEu+FGC/b+0EviHm81kNRMhQF6bfEviTWi+M67rj4PMfoQojckU+5+i//y/+SWq3Gv/yX/5KFhQV++Zd/mc9+9rN87/d+73m2T1wBPS9iteniBjGLboSpcKyjEx1LJ4xTLEMbfi2MUxzr2Ie4PWX7zAIw/N/VpiunA4iJJfFeiPGyvR9sdHwMQ0VBoWApB/Zb2/tfx9KZrztydPEVIrFeHNdxx8XHHaOf9p5AiKvoWHe83/u93ytBXpypQabfMjVKBZ0oTnm83uX29eqRA/h83eHukzaQZ/3DOCUIExavl07dPjeIKRV2fkxMXaXn7T+zsNcAefbULRHiYkm8F+Jpo5oA2d4PekGMoamEScLiTN4P7tVv7e5/wzifaT9O/yvGn8R6cVSDrVaKAgVLp16xcWz9wHHxccboZ3FPIMRVdOQtV3/7b/9tvvGNb+z42je+8Q1+9md/9swbJa6O3XtrbVM/9t7aUsHg9vUqupp3KLqqnlnwH8wsuH6+hP2Dx23uL3dQ9vn+/Zaidt3w1G0R4qJIvBfiaSfdgnsRddi294NZClmWsThTwrHzCYm9ZsSltoWQWC+OahD/DEPFMBSSNGNpo4frxweuuDlojL47Nt5f7pz6nkCIq+jICZ033niDj33sYzu+9rGPfYw33njjzBslro6z2ltbKhg8v1jl1dvTPL94dpn8+bpDqxtyf6VDnKToOnhhghtEew7K9xsgP1nvnUl7hLgIEu+FeNpJEiAXWYdt0A/+Z68uMF0toGkKWZYRRAlBmDBfd3Z8v9S2EBLrxVEN4t+1ukMUZWRkGJrKarO/Z3zZbq8x+n6xMY7THT8rMUmIwx15y5Wi5AOD7ZIkIU3TfX5CiMOdZ/2bs1AqGBRtHTfQiZMU29J5bqGIqip77hfef4tWRN05PMk0qsv5xdUi8V5MmrOIrSfZgnsZddgGM+KrTZeel1/v4vXSU9c76v2vOH8S68VRDeKfoijcmCvR6Ph4QUyWcqJV8XvFxnLRZG3T5bnCh7FxkmKSjPHFeTnyCp3XX3+dv//3//4wyKdpyi/8wi/w+uuvn1vjxOSbrzsEYUIQJWRZhh/Gh2b6L1oG3LpW5oUbNW7MlihY+r4zBoMB8nZhnB4pYMuJWmJUSLwXk+SsYut+8f2gm43LWgVzlFWru/vf/VbyiMklsV4c1fb4V7B0bsyWeHa+fOJV8XvFxvmpAl0vHul7gpOSMb44T0dOeX7uc5/jv//v/3v+5J/8k1y/fp3l5WVmZ2f5x//4H59n+8SE2z2TWK2erv7NeWS/jzOLuV/xt+uzJfx+cODzyIlaYlRIvBeT5Kxi60kK8F/0Kpjj9IFHXckjJpfEenFUJ4l/B8WjvWKjpqncXqgM6+2c9p5glMgYX5ynI48orl27xi/90i/xB3/wB6ysrLCwsMC3f/u3o6pHXuQjxJ5KBWMYzGZny6yvd4ePHWdwel4ndhynE9tvgFx2zEMTOidZzi/EeZB4LybJWcXWkyRAzvMUxt1O0gdu739HnWxXOHsS68VRHTf+HRaP9ouN2+PV4J5gEj77MsYX5+lYU0SqqvId3/Ed59UWIXY47uD0vLLfx+3ETjpAlnoGYpRIvBeT4ixj63Hj+0WugpnkGWA5Yv38SKwXR3Wc+HdYPDpqbJyUz76M8cV5OvBd9AM/8AP82q/9GgCf/OQnUZS9D2v+d//u3515w4Q47uD0PLPfFzGLeZEzuULsJvFeTKrLjq0XtQpmkmeAJzlZddEk1ouLcJR4dJTYOCmf/cvuh8RkOzCh87f+1t8a/v8vfelL594YIbY77uB03LPfUs9AXCaJ92JSXZXYOu594EEmOVl10STWi4twVvFoUj77V6UfEpfjwE/VoMp9kiT84i/+In/rb/0tTNO8kIYJcdzO4Cyz35e1X3ec6hmIySLxXkyyUYqt59W/TPIM8CQnqy6axHpxHnbHtZJjsNp0gdPFo0n67I9SPyQmy5Gqnmmaxte+9rV9l2UKcR6Oe6TqIPs9qI6vqyerjn/ZRwv2vIg7S23eutvgzpIcaSgulsR7IY7nODH7PPuXs+oDR5EcsX72JNaLs7JXXFttuszXnVPFo64b4gUx7z3Y5N5yB9eP5LMvxB6OXMb+R3/0R/mFX/gFokhuLsXFOMngtFQweH6xyqu3p3l+8WQD2e37dRVFwTI0LFMbzjScp8tOJgkBEu+FOKrjxuzz7l/Oog8cRZOcrLpMEuvFWdgvrvXc6MTxqOdFvPdgE8NQeX6xClnGB0sdoiiVz74Quxx5vdr//X//32xsbPDP/tk/o16v78joS+E0cRZ6XkTzQZOllc6OZegXvTzxsP2657kda1KKv4nxJvFeiKO5jOL9XTfkzlL7VH3QOB4DLNsVzp7EenESu+NHs+MzU7N3fM9p69zcX+7gxymtjodt6VybLnJdVdBVdeRjlRAX7cgJnbMqnPaP/tE/4hd+4Rf4lV/5FT7ykY/w5ptv8vnPf54gCFhcXORLX/oS09PTABf+mLg8g1nO+dnypR9LeNB+3fM+PnFSir+J8SbxXoijueji/T0v4sGGO1wRdJI+aFKOARanJ7FeHNde8aPZDTAMlVrJGn7faerc9LyIu8sdFufKFCydKEl5vNZjcbZIGKVndSlCTIwjbbn6rd/6Ld555x3SNOWP/bE/9tR/R/XOO+/w5ptvsri4CECapvy1v/bX+PznP89XvvIVXn/9dX7+53/+Uh4Tl2swy2mb+oVvc9rtoL36571cfjDY325ci7+J8STxXoijO27MPm0tmNWmi33KPugytxWL0SGxXpzEXvFjfqrAatM9sxpXq02XciG/H1AUBVPXMAyV1U1PxsNC7OHQhM4/+Sf/hP/xf/wf+df/+l/z4z/+4/zzf/7PT/REYRjyMz/zM/z0T//08Gtvv/02lmUNK+5/+tOf5td//dcv5TFxeQYFJR+udHmw0sb185lNU1dxg4tfmXLQXn03iDH1nR+bs2ynFH4Ul0nivRDHc9HF+90g3rG6B47fBzW6AWtNlw8et3m81sP140vrb8XlkFgvTmqvcXC1ZDJVKZyqxtXu4vKVokkQJ0RxCmSQQbcfynhYiD0cmub8l//yX/J//p//J6+99hr/8T/+Rz7/+c/zF/7CXzj2E/2Df/AP+MEf/EFu3Lgx/Nry8jLXr18f/rter5OmKa1W68Ifq9Vqx74mcbx9+Hsdabi+6XH3SRs/SqgWTeI4Y7XVZ3GmhKYpl5aJ32+v/nkfnzgY7K82XXpe/jotXi/JMnhxISTei3ExKjVgThKzj1ILZr/rcyydIEp2fG+7F9LpR7x1t3Gkfniz46GpKo6tEcUZSxs9ZqsFyo4cXX1VSKwXJ7XfOHi6bPH8YnUYu+4td458X9DoBqw0epi6hqYptPoBbhDx6otzLK/FuH6CpiLbQoXYx6F3oZubm7z22msAfPzjH2djY+PYT/L7v//7vP322/zkT/7ksX92FE1Ply67CUOzs+VLe+6uG/Jgw6VUtpmuawRRQqMfMTNTempguPt7N7s+b93fxDZUFucrBFHCo7UupVJMvergRgmzpSIv3ZwaqUGmXbR478HmcMl7ECUoYXKkdh71bzULPPds/QxaezEu8z14nib1ug4i8X6nk8b6UX3vTEq7jtP3XETbzjpmH3R9r3xkjvcebFJwLCxDY7MbsNELeX6xylTZPvS1aD5o8pFb06w0XCxDo6yr9L2IfpjyXa/Nnfr1m5T32EW6jLZJrH/apMX7/Zy2vQeNg4ET3RdsdH38KCVKFZ6ZL1F0LO49abO80eNjL84RRAn+Ecfao+CqvScu2ri1F86/zUdaVpBl2Z7/DajqwTu3vv71r3Pnzh2+53u+B4CVlRX+u//uv+O//q//a548eTL8vmaziaqq1Go1FhYWLvSx42g0eqRpdvg3nrPZ2TLr691Le/47S/lRrVms4W19LYgS3vnWWn7E4AHfu7TegyTlcaPPs9dKgELNMej2I1wlIEvh227U8PsBfj+46Es70HTRYLXpsrZt5vSwdl723+q8yHWNHlVVTpV0lnj/oZPE+lF970xSu47T91x0287CYdf30s0p3vnWGmtBzGYnYKZsoqQprbb71PfutrTSoVTQqTk6jY6PH8RYpo5taKfubyfpPXZRTtu208R7ifU7TVK8389ZtXe/cfBJ7wvuLbUxNBXShCdrXeZqBWYqJhttn0fL7SOPtUfBVX1PXJRxay+cTZsPi/WHJnRc1+WjH/3o8N9Zlg3/nWUZiqLwzW9+88Df8WM/9mP82I/92PDf3/3d380//sf/mBdeeIF/9a/+Fb/3e7/H66+/zr/4F/+C7//+7wfgYx/7GL7vX9hj4viOc7pHs+PjRwlBGGNbOu1eQK2UZ9mjOMPQFUoFA83QmSoa6FsDidMczXpeS/Ll6FQxqSTei3Ew6acBHnZ9Zccc3hy9dbdx7FO22v2QnhfhB3l/XCoYlAuHz3qPyja3kxr39p8lifXiNPYbB5/0vqDVDaiWTBzbIAjz77UMnXLZ5NXb0x9u4zrmZ1c+8+KqODSh8xu/8Rvn9uSqqvLFL36RL3zhCzuOGryMx8TxHbWeTM+LaHYDNFUZHj/Y6UeAwmytQBjn9QAyUizLIAgTanXrVMeqyrGsQhyfxHsxDs67ltllO871Hfe1KDkGb99rULQNCpaGFyQ02j7f9cq1A9s07n3quLf/rEmsF+fhpPcFlaLJWstjupJRLJhEcYobxLxws37iz6585sVVomTb11ceQZqmbGxsMDc3d15tGnmy5Sq3PViaukoYpwRh8lSwvLPUpuuGrLe9/OhBXWGzG7K26fEdL85g6Cprmy5dL+Y7X56nVsiXcsbpzk4hiBJ0VT3SkvrBUs6T/Px5ZPTP62912bMPl/0ePC/jfF2n3XK13VWP95O0BH+S2nXUvucy2ra7nSeJz4dd3/Z2Hfe1GPTHPS/CDxNsU8tX6Gxb9bOXo/Sps7Nl7j1sjtyM+Oxsmd998/GpxhTn2bbL2nK13VWP9TBe8f6kseW823vS+4IkTbmz1CaIUp67VsaxdRzb5JP/6bO88621HZ9d149ZbfaJ4oznF6v7Xvtp7gNOY1T72v1Ie8/fRWy5OvTY8oFOp8P/9r/9b3z7t387f/bP/lkgz/D/vb/3907VQDG+9jt+Fdhx9GCz41MtmfnJVaqC6ydUHIPnF6uUHZM0g2fnK/zxj87j2Ab3ljt5II7THc93nGNVT3q8+KAzitOUUkEnTvOMfs+LjvHKXIxxaqsYLxLvxSg77dHfF+E48XlwXO/Xv7nK/+/NJ7z7YBNVUYii9NDrO+5r4QYx1ZLJjbkSL9yocmOuRLVkHto3HqVP7brhyPZJJx0TTDqJ9eNnlMd+B8Wj3ceSm4Y6vC9IU3juWoWb82WevVZhtuZQtHW+9XBzx/2A68csbfRQFAVF5cBrl8+8uEqOnND5whe+QKlU4qtf/SqGkQ8UvuM7voNf+7VfO7fGidFXKuSJmVdvTw8z3oOORlXg4WqHd+41ee/hJsBwEDlXd1icKQ5/dr7usNp0afcCWr2AtZbLN761TrPjD5/rOEvqB8s+tzvKz682Xaytyv2KomAZGpapsdp0j/OyXIhxaqsYLxLvxajb3fcMEhi7bxou6ybnqPF5cHPW9UI2ewFhnNDqesRJSpplPLdQ2XF9e9nvtdj+HIPXZLMT0O6FOx4/St94lD71yXpvZPukk44JJp3E+vEz6mO/veLR9iSUqkCrF/K776yyvNGnXrF54UaVZ66V+eitOs8tVEizDMNQqTgmhqHyYLWL68csb/Rp90KebPTouiFpmu177fKZF1fJkd/Vv/M7v8Nv/uZvYhgGiqIAUK/XaTQa59Y4MX4GHU2SZDxp9DF1jYUZh7WWR5rBzfkyuq4ShAmL10s7fi7NMlYaLkmacn26yKO1Hu/ca/KdH5nd82cOMl93uPukDbBj2edhPz9OxTbHqa1ivEi8F+NolGomHDU+D/rM9ZaHaaiYukYUp/S8iNmpAqtN91RF+He/Jkma8mAlX/pdLZlH7huP0qf2vGjPGfFR6JNOOiaYdBLrx884jv123xvUyiZhHLPe9kiylLmag6oqLF4vPZWwulZ3uLfc4eFKl9VNF9tSURWVom3weK3H4myRMEqfek75zIur5MgrdMrlMpubmzu+9uTJE2ZnZ8+8UWJ8DZY4DrLoq81+3vnYBgVLZ2m9v+eScDeI6boRaZbR6oZstH1sUyNKsn1/5iAnXZI/Thn9cWqrGC8S78Uo228VzijNXB81Pg/6TD+I82N7AUNX8MPkTLYHDF6TNM1Y2uiz0fKwDI31Tf9YfeNR+tRSwRjZPmkctuldBon142ccx3677w1aXR9DV4niFC9IuPO4jR+lrDZdGt1gR2K4YOnculam2fXRNAVN0ZirO5S3Vu+sbnp7Xrt85sVVcuRP/3/1X/1X/M//8//MX/2rf5U0Tfn93/99/u7f/bt8+tOfPs/2iTExKNC2vNEfDhxrZQPL1PHDmDTNmJ+qkaTsWYzMsXTuLrVBV0nTDNvU8oKNhkqtYp+ogNlJjhcfp4z+OLVVjBeJ92JUHbQKZ5Rmro8anwc3Z/bWCZD5Cp28DzyLmzQ3iNFUWFrvYxjq1kmTCZ1+zHMLlWPd3BzWp16fLfF4ef9rvuwi/icZE0w6ifXjZ9zGfj0vYrMTcOdJh0erXWaqFo5t4IcxWZahqgpOwWC2ahHGKZsdD1NTqJWt4e/QNJWZaoHF2SJPGn00VQEyyKDbD5mvO3s+t3zmxVVx5BU6f/kv/2V+4Ad+gJ/5mZ8hjmP+xt/4G3zP93wPP/qjP3qe7RNjYPve2BtzRTbaLm4QEoZpXsgsg3rF3jeLDnkH1fMjkjhD0xSiJCXLMqolC9+/uMH4OGX0x6mtYrxIvBej6qBVOKM0c33U+DxfdwjCJF/dEqW4fkwQxZQKBkGY7HujclSOpbO66WFsbedSFAUFlXJBP/OVS2XHPLAg6qgWcr3KJNaPn3Ea+w0+95WSgeuF6GrGZjfIVx5moOsqfS+i7BjDeD5fd1jd9AiihCzLCKKEIEy4PlNE19UdB6xkWTay1y7ERTryKEdRFH70R39Ugrx4yvYBNobGVKWAoihsdHwW6kVmpwqoikLnkCz6s3MlGr2QvhtSsA2mKjYABUvb82fOyl6zhpd5jOlxyOyDOA8S78WoOmgVznMLlUuZud5v5clR4vPg5my16RJF+fYD27YoO+aZrGCZrzu8c69BpWgCGVGcEcYJ16eL53Lay37XvGOcAMP/PW2NIHE6EuvH03mN/c56Fd32z321ZKGpCqubLr1+xPM3Kiw3XII4pr413geoFk2iKEVXVTpuiK6qwxg+WJ25OFscxvdbC5VTX7cQ4+7AhM7v/M7vHOmXfOITnziTxojxtHuAXS2aFG2dWtmmWjTxwwRFfTqLvrvjmCrb3FiosrTawQ9ibEunVDAoF8xza/soFdG87OXo4mqTeC/O214x7riVOgarcAYJAfhwFc725EjPy59j8XrpXOPoUfqQw2L74ObsPCYSBq/JesvF9RNsU8tnuDUFXT3yIu1TG6XtcFedxPqrY3fssYvWgd971uPh7Z/7atGkVDCYqRVY3/RIUzA0jemKjaLA4/UefhCjaSqz1QLPL1aZnS2zvt4d/r6Lju9CjIsDEzqf+9znDv0FiqLwG7/xG2fWIDF+dg+w6xWb+ysdHFvfN4u+V8fR92N002C2Vtgxu3raJecHGZVZw1FKLImrSeK9OE/7xbiZmeOtnjmsfsRFr1o8rA8Zhdh+a+sYYMvULq3mxkGJOHGxJNZfDXvFnvcebDJdNPaMPecxHt7+ua9XbJY2eigoLM6VmJsqUHZMgijm3nIHx9bQtbwQfN+P85P+dv0+WZUuxN4O7Em/+tWvXlQ7xBjbPcDWNIXpSoGire+bRd+r46iVTYqOTuBl55Z93z1b0ez4zNTsHd9zGbOGo5JYEleXxHtxnvaLcU/We9Sd4xXmHaVZ2sNWnoxKbFcVhQfL+Uz39ZnihU8WjFsh10kmsf5q2Cv22Fv1xgaxZ/uYeHmjz425ImxLup52PLz9c1+wNGarBVY3PWxDQ1dVXnmuzv3lDkmSEsdgmyq3rlXQNIXVpstzz9ZP8QoIcXXI1Ig4tb0G2K88Vz9wsLjfIDjN9j4Fa7eTbE/aa7ai2Q0wDJVa6cNlqJcxayjL0YUQk2z/GBcdK6EDozVLe9jKk4Ni+0Vss93e7734THWYSLloo5aIE2LS7RV7LENjbat21u4xsaEr3F/p8txChcJW/NpvPHzU2LX7c192TF64UdvxvRn5KkJFUT78WpbJ+FeIYzjyXWuv1+MXfuEX+PrXv87m5iZZlg0f+3f/7t+dR9vEGDnuAHu/QfBs9fDB3UmXsO81WzE/VWC16VKw9EudNZTl6GKUSLwXZ+00MX+UHbbyZL/rVuBCtmKNygohGK1EnMhJrJ9ce8WeIEqG48rdsWG+XuT+SoeVpsuta+V9x8PHHYMf9rmX8a8Qp3fking//dM/zbvvvstP/MRP0Gq1+Jt/82+ysLDAf/Pf/Dfn2DwxqQZHtT51LOHs4YmUg46tPYgbxJj6zrd8tWQyVSnse/xjz4u4s9TmrbsN7iyd3xGr+70e51k/SIj9SLwXZ+00Mf+0zjOOH3aE8H7XDZyoHzuuvfo9U1fP5YQrMX4k1k+uvWKPv21cuTs2OLbOzfkyUZQeeBz6ScfgR2mn60fcW+7w3oNNvCCm64YnfwGEuEKOnND52te+xj/8h/+Q7/3e70XTNL73e7+Xv//3/z5f/vKXz7N9YkLtNwguO4efaHXSAepgFmC7ME6ZLlvM1x0cS8cN4q2lodFwFiJOU0oFnTjNZyHOI6lz2E2BEBdJ4r04a6eJ+adxEXG8VDB4frHKq7eneX5xZ9ze77ozOPNEy16Jq/36PZn9FiCxfpLtFXteujk1jE97xQZdV7k+U3xqPLzdWSeJB+2MopQPljqQZTy/WMUwVN57sHluE6lCTJIj9+hpmlIulwFwHIdut8vs7CwPHjw4t8aJyXbYMsz99uiedHmmqir83rtrBGFCrWhyfbaIYxvU6taey0dVRbnQpeqyHF2MCon34jxcRow7zZajrhtyZ6l96ho3e133WW8zWN10efP9ddIsf74kTek/iZivO8OZ81EqRnwR9YPE4STWT7btsafnRTxZ77G00sGxdEqO8VRsaPcCskzBMFRUBR6udnjnfpPbCxVuLVQOHYOf9HNdKhgULJ2Xnq3t+L27izgLIfZ25JHDt33bt/H1r3+dT3ziE3z84x/np3/6pykWi9y6descmyfGyVkO0A7ao3uU0zJ2t0VVFf7wzga1kkUQJfS8iD96sMknv+MGPTfac8D/YLnLi8/s7ESkULG4CiTei0lx0oLzPS/iwYY7XNlz1jVujnrq01796u6jfHtexJvvb6CpKiVLI0pSHq/10DWFlYbL9ZkiUZQSRulIFCMehaPcRU5i/dUw+MzNz5aHn7nVpst83aHnRsNC5Y5lECcpj1d7LG30cWyNcsFkve2RZtmBY/D9JkcHz3HYvcFhRZyFEPs78parv/23/zaLi4sA/M2/+TexbZtut8sXv/jFc2ucGB/HWdZ+lHoGB+3RPWx70l5t+e23n6CrKlNli2t1h2fmStiWzu+8vcydpTbxrmWng+WkslRdXEUS78Wk2G/LkQIH9kOrTRd7Wx+UJBmNtsdvv7V8JnV4jrLNdr9+dXddidWmS5pmOHbe1jSFdj+g60YoKhiGSpplPLdQeWpL2GU46xoc4uQk1l8Ng8+cbeo7PnM9NxpuF52vO9xb7vAHH6xzd6WDoSuoispmz8f140PH4NsnRwfPkaYZb76/fui9Qc+L2OwE/NGDFo/Xerh+nsTZXsRZCLG/Qz8lb7/9NqZp8pGPfASAZrPJ3/k7f4dvfetbfMd3fAcLCwvn3kgx+o66rP2oM3OHzaoetHR/r7YEUUoQxUC+Qmet6aJr4EYphqHyYLXLtbqD68f4YYKmQr1iDYtXjtJSdSHOi8R7MWn2mk1udUMUJcMw1H37ITeIma5reIDrxyxt9DA0FUXNhjclp11RctgWtPvLHRodnyRJsS2d6YqNZWo8We/tOOp90F9GcYahK3T6Ibah40cJBUu/1JOt9nLSVVPi7Eisv1oO+8wNxuZeEGGbOl7Xp9vP0KsaqqIShDFxnPJgvbvvSpt7ezxH1w1JMw68Nxg8d6Vk4Ef5GPzxepe5moOi63I4iBBHcOgKnb/zd/4OGxsbw39/7nOf4/79+3z605/m/fff50tf+tK5NlCMBzeISZKUx+s9Pnjc4vF6jyRJnyqSdtSZudMUctyrYFutaA47rnY/RNdV0lShVDC4tlVh/517TeIkRdfBCxPSLGO+7kihYnFlSLwX42yv1Z97zSYXbZ1qyTqwH3IsnSDKE/rNjo+pa6AwTJCc94qSnhdxd7mDomQULD3fBrGW96u7Z7cdS6fsmIRxQhSn+c1XmqAqMF2xgdE62UoKNV8+ifVXy2GfucHYvFjIi9TrhkamZLR7PlmWoSoKD1a7wyT4Xitt9nqOnhdjaOqB9waD566VrK3V8xpJCp1+tKOIsxBif4cmdO7cucPrr78OQKfT4Td/8zf5+Z//ef7CX/gL/N2/+3f5t//23557I8XoU4D7K13iJB0OPu+vdFF2fd9Rq+Mf9Rjvo57qcX22SJpm+QqcICKO8w7l5rUKBUunXDBQFPKEjqbx3EKFasnasRx1FJaqC3GeJN6LcbXf9qTVTfepGjRHOWFqvu7gb/VBXhCTkRJF6YUlSFabLuWCjoKKoiiYuoZhqKxuek/1Q/N1B1VVmK0WUBWI4pQoznjp2SkKWzdso5QwOWr/Ls6PxPqrZfCZ88N4z8/cYGxeLZpMlW0qjkkU5uPkOElZ2nBpdHyqRXPfJPhen+soSel64YH3BtvvCwqWzo3ZEt/2bI2pinXupyAKMSkO7d2TJMEw8sHDm2++yczMDM899xwACwsLdDqd822hGGmDgo33Vrq0uj665mDqKsquVM7g+5Y3+hiGyrW6c+BAczCrmh+ZmA/Ea3Urf66tgfmgQv9eBdhWmy6eH9N1Q3pejKoq/LGPzrPR8ul5+WqiuSkHL4zxgpgoybh5rcJM1abR8Vla72GZOrah8fyuLWNyMoeYVBLvxbjaa6utF8S8+f4GtxbK7D7B8LATpkoFg5mZEu98a40sgyxTuDFX3NFvDerwHKU/OG7f0ez4JEnGo7UOuqaiaxAlEEUpf+o7n4Es29HW29er3F/u4AX5NquuF/FotUffi4ardxzL4K27jUvvu/bq3y+7UPNVI7H+ahl85sKMp8bU73YDHi130HWVUsGg50WoKFiWRqPjY+kRJcdgrlag0fYpWDoFSydJUh4s79yCNV93eO9Bi1Y/oGBoxHFCqxeQJBn1ioWmfphIP+59gRBif4eu0HnhhRf4tV/7NQB+9Vd/lU984hPDx1ZXV4fHHYqrZ/uMqG2qw2Df6oZoqsLN+TLZru9bnC3iBTH3lju4fnTgzFypYOwo1rbadHfMvr75/jppmj21bL7nbh3VuunS82NKjsF8vUCcpLx0s8art2d4Zr7MVNkiThLuLXeIkhRDV3i81hvOJARhQrMbDJeUHqfwsxDjSOK9GFd7rf7sutGefQRwpBUiZcfk+cUq/9mrC0xXbFRVGX5/uxfQ9+MjHwRwnL6j50U0uwEoCjPVAq1uwHLTBzKu1R2W1vt7/myaZcxNFSgXDWZrBfpBTLMb8HCtixckB26XuGjb+3dZ/XrxJNZfPaWCwUs36zvG1F03pNX1sG2dzV5Axw1pdDzcIMYPEq5Pl1icLXNjpoht6hiGSqPj4wUx91e6GLoyjCnv3GvyYKXD/HSBFxYrRGlKqxdyre4AGY/X+sRxys35Ml4Qn+i+QAixt0PTnz/5kz/Jj//4j/PTP/3TqKrK//P//D/Dx371V3+V7/zO7zzXBorRtX1GVFUUWm5AluVF0CxD4+Fqj6JtEIRNum5IkoJtalyrO7R7AY/X+jy/WD3SzNxes6/p1nPVytbw+7YXebu1UNkxA9vqBfzWm09wCgaWrpEkGRlQMDXKjkmj46OpKqauEsUZGRnzU4Vh8bajFn4WYlxJvBfjarDVdnvMz2voPF0INIzS4QqRR+t91hsuKNDqhrx0s8b81M4bib1WlDiWgWGoR+oPjnNowGrT5d0Hm3S6Pn6UkAJTWzPbSZby7LUy9uC0mW0/O3iO9U0Py9ApFVSKBR1dU4nilDRNpe8SQxLrr7a94oWhqzxa62KZOiXbZLpqUyuZxElGHKf0vBDXz1e1P9noY+oqt65Vhsly1w9BUZibcnjc8nBsnYpj0uqFOLZOlsXcX+nR8yJcP2aqbKHr6onvC4QQHzo0ofP666/zb//tv+X+/fvcunWLUunDE34++clP8uf+3J871waK0TWomu8FMe1ewPqmhxfFtDoBlZLJ3FSRm9c0vvF+k5tzZQxDZaXp4voR16eL1Cr2ju1M+xnUyVHUfH/tdMWmYOn50lB35wzjYJnm7or+XhCztunR6PoomsJmL0EBXnq2xlTZoufFTFUKBEGE6yfYpsbiTImCpQ0TRHIyh5h0Eu/FuNrrNCtV4akaDIM+olQw6DsGm/d8qmWLgqXiBQm/+84K3/XKNeanHLpuyNt3GzzZ6ANwfabIcwsVSoV869JedXj26g+O0ncMVvGkaUaz42IbOgbQ6gb4QcJczabiWDh2XpT5QTeAbdu9Gt2A2aqFHyaoKmy2AoIgIkOh7Bh5UedD2ipbiq8OifVX2yAmdfohQZwSRTGGoaGg8Pz1Ml6QYlsaUZLmscJPyTKFJE0xdJUwSiiYOn4Y0+z4+GHC+qZLpZTHWz+IKVg6RUfn3t0OjqnR6vn4Maxvqqiqih9F3Jyv4AYxj9Z6lIsmpqZK3BHiBI60QbFUKvGxj33sqa/fvn37zBskxkPPi9jsBDxe69HpB3hhTKlosrHkoWrq1vb+jOWNPgVTZ6Xp4ocxaQaGrvKk2SfJsuEpJAc9z90nbQxdIYxSljf63F1qszhboljQUVWFIEqeOlZ8tenumK1tdHyiOCVJM6I4oWjnW6ree9ji5ZtTlB0Tx9LpagqKF+GHCc2OT6lgDG8I9poBln2+YtJIvBfjaK9VNK+9OMtq092zjwB470GLom3g2HkMd2x1x9f/6HGHh2tdHFuDTOHhWpe+H/PKc/Uj9Qc7akToCvP14vC5dtfg2ewEVIoGjbZPGGZ4vo+ua9iGSq2cb/eqbt0sbXYDNjse1aIxrA202fEwtbx63XLDxTY1NE0jyzKa3YDZSmHH6zVo66CNzY5PsxswP1WgWjL3PMpdTBaJ9VeXY+m0eyHtfl4iwdpKzgRRTKsXUS2aTFfsvAxBnOH7MdWSiWUUuDFX4slGn/vLHd6+16BWsqiXLaIkjzWNtk+nH7K26Q1/vxclBFG6LbGc0e5H/NHDTbKt+4IsTVmYLUncEeIEDq2hI8RugyRLpWSgaQrrLY++H+H6EaquUS2aqApstDx6Xr56572HLTr9CE1TSJOMjZZPtWQeeuzrYFlopWix3HTJMnBsjfWWx2rT4yPP1PY8Vnx3tf2uG9Hu+yzOliCDOE6xjHwp+uqml88IOAYPVrv4YULBUvHDhAerXUpO3qnIyRxCCDG6dtdlmZ9ynjqyfPuNQqsfULB2DoMKlkqrH+SJITfAsXVMXcc0NIq2geuHrDbdQ/uD7XVzbswV8cKE+ysd+l60Zw2efhDxaL3Lt5baRElCux/S6gVs9gKWG/nNk+/HtLoBT9Z7zNedHbWB8rpxHlGckmUZSZKRJCnVokW5YBAl6VNtLTnGsI1+lKCpCuttDy9ILuRodiHE5RjEi4pjkmZ5opsMrtWLrG+6lAoGtqkxO1UgSVMgwzI1bszlyfB2L2Cl6UGWYeoKG22fnhehAO8+aFK0DTJSmh0PTYUwSvPtVZaOFyZ0+hHrLY+Hq11UBchgZdOjWjQl7ghxArK0QBzb9noAlqFxd6lNnCR03CifIVQUDF3DDxI2Wh5JmlFxdAxdpdUNqDgm1+oOUZwdeuzrcFtXGLMwnRdOC8IYMrg5XyZNsz23be2erS1aBmGYMF2xCAoGGy2XtVaAAoRRAkDPjbh5rUzPi/CCGNvSmanZeZHlKTmZQwghxk2pYOxbJ6ZWtPCCZLgyB8ALUmpFKz+uN80wtA8fM3SFvpdt9UsH9wc76uYYGs8tVFhpuiyt5zUidtfgKTsG95Y8Ov2Aetlmplqg0fHpuhGmqTE/VWBl0+XxRo9rM2WmijsnEqpFkyhK2ewG1EoWYZRgmfnqo8XZIq4fDxNbg7Zub2MQxlvHCmc0Oz6OXZItxUJMqFLBoF628KMEFAjCGMs0qDgG07UCZcek58WUCyYLLxZ570ErT9goeUH5NIWSY6CQkWZgmSqOZaACMaBpCtfqJTa7EXGc0umHTFftYQ0zP8q3kbb6EV03pFK0uFZ3cIN4WAZBCHF0ktARx7a9HkDB0nnxRo1Ha1022l1MQyXoJ2i6im2qaIrKesfn+nQRy9BJs4w0zZivO/S8vJbOQQbL2v2t5yw7BlGcom0tPz8o6G8fyPe8iK+99QQvyOsLRElGpWgwUylgGip3n7QJwoSZmk2t9GGR5SzLdjzHQTcHQgghxsdLN2v87jsrAFs1dFL6fsSrr0zTcyOCOMVPPtwmEMX5ypfNTnDo8d+76+YULJ1b18r0vJjnF6tP1eCZrth84711LCPfepCoCpqqMD/lECcZUxU7Xy2aQS9IuL/S5bmFyo5jfusVm3rFJk53bgULooT6HjXr7m1ro23pw3oZrp8Mf6dsKRZiMg1ixTNzH9ZPCqIEXVWHsWK4Ir9o4EcRfpDweL0HWV5cvVo0Kdr6cOVfz4t58Zna8HcGYcJbdxtoqoLrR7iArqsYaYofpszXbBzbQFEU5usOfhBL3BHiBGTLlTi2QZJlYKqSZ/kNXUVTFAxDId1a8m1aOs/MlKiVLbwgRlGUreAPqsKh25UGy9o1TSVKEqI4JYzzwen2GgRv3W1wZ2n/Y1hLBYPXXpwlSbNhdf7ZqoOq5XUNLFOj1Q24v9Llg8ctHq/38KRjEUKIiTEosD/oL4q2wXe9cg1T12h0QkxdGxZEzrfhWrh+TBjHhFFCsxPQ9UMqJePQ479395OwM0Gy/XHXj/PaOVFCmoBuaOiaiq5p2JaGYSiUtgobm4ZKaauu20rTfWq713G2Bm9vw3TFJopSXD/BMlTZUizEhNseK1w/4t5yh/cebOIF8TCmDVbx1coWN2bzE/Z0VSGIUmarNp1+yJONPmstjyBMscx84hXyuJaRDet/aapK348wdIXpio2hK9SrNijKVs0v0DRV4o4QJ3BhCZ2f+Imf4Ad/8Af583/+z/OZz3yGb37zmwDcu3ePH/mRH+H7vu/7+JEf+RHu378//JmLfkwcze4BY7sXMF93eH6xwvx0kbmpIs/MlqhXbJ6dK/PMfBkFhZKjk2Yp4VZx4tdenD10u9JgWftstUCnH5OmKdeni2iaQqsb4gbRsAbBQYNrgPkphz/x6gJzNYdaKT8tZHGmhGPrxHFKyw3xghhdU4mThHvLneG17b4R2O85hBAS78Xo2V7TZnt/AbA4m59etThbpGjnfVKpYPDaS7M8O1cmDDPCKKVk5ytSayVrWLtmv3oPhyVW5usOrW7IH93f5OvfXOHxeo9ayaJSMrEMlZvzZRZnipi6RrVoDbd+RXFGvWJxc75MFKVP1QYa9Jn71Q3ar43b62VYlnHgzwkxILF+fA1iRRSlfLDUgSwvYWBsrVrveRFuEA9XEjq2zo25Et/xkVnKjkGrH2DoCm6Y8KThst5y+bZna6hKfljJk/UeXTfC0FRKBYPF2SLXporUSjbPzJd5dq7MrYUKzy2UKVgGnX7MbLUgcUeIE1CyLD+P6Lx1u13K5TIA/+bf/Bv+j//j/+CXfumX+It/8S/ywz/8w/zQD/0QX/7yl/nFX/xF/q//6/8CuPDHjqrR6JGmF/KyHWh2tsz6evdSnnv78abLG31uzBVRFIXHaz0MQ8XQFDZaAWGcoGsqjq3R6Ud03YiZaoE//so881NPZ+APuqbdR6p6QbyjBgE8vVx0L3eW2k8tSb/3pA1Kvrx9cASjpsJszeHWQoW7T9pYprbjpJTjdDqX+bc6T3Jdo0dVFaanS4d/4zmalHh/klg/qu+dq96uveJ+qxuwuulya6GyZ2zf3ba37jYoFXQURRl+bbAt99Xb008950HHgPe8iHfvN7i/3CFKMkxdRVPz3xvGKVkGU2WLrhtSdkxsS0VBJYwTXnlhjm7XO7SvO4qzPKr8qr/HTuK0bbvseD8psR4mK97vZ6/27hUbB2NpYM/HHi538cKIBys9NF1humxRcix0TeW1F2dY3/T4nXdXqBUNSo5Js+2DAvWyxXLD49q0w1ytQNeL6PZDbl+vcmuh8lTsGbfXF8avzdLe83cWbT4s1l/YCp1BwAfo9XooikKj0eDdd9/lU5/6FACf+tSnePfdd2k2mxf+mDie3aeJaFuzh6qqsNJwubfcxbEMnluoUCuZaKrGwnSR73rlGi/drNFzj7/CZfdzZrCjBgFb/z6s0PJeM6ddL2Z+qjCcgXjhRt65ZOwsbnnYrKwQQuK9GD3bZ5oHum6YF/Q8Ymw/bBvVbrv7rO03K6tNl2rJol4p8NxCmcXZEvWqTbVksThbolgwePFGjT/92iI3Zks7V6iqynC1z2lXjx7URiEOI7F+/O0VGwdj6f1WGtbKFlOVAq++MM1rL8zyzHyFqbJJmmb03IiCpfPC9QpzU3n9TMPQaPcj7i53mas5PDtXxjJ1np0r890ff4aP3Z6W2CPEKVxocZDPfe5zfO1rXyPLMv7pP/2nLC8vMz8/j6blmV9N05ibm2N5eZksyy70sXq9fuTruOzZ7+1mZ8uHf9M56bohT9Z7xIrC0oZHnKZMVwrUKgV6frS1tFzhhZvTT81odtxw37Yf9ZoW3YgoTrHND9/GfhhTraoH/o5ZYGamxJP1Hj0vYrZqUCxaGLq25+/qeRHTjnmsazjNdY0buS6xl0mI9yeN9aP63rnK7dqrv3i43uf6bJmp2ofF+XfH9u1ts4sW7z3YxB6cDBUlKGHCSzenKG/VtTmqhw2XacckSDLiOMM0NLIso+9HPHe9iqGrvHTzw/fp627IB482ebjao73U4tn5MrZjsrTep1S2ma7n7Wn0I2ZmSsduz1m5yu+xkxrlth3FJMR6mLx4v5/d7T1oLP3cs/WnxsvXZ/N/v323wXSlMBwbh1GSj6VtA9ePKTgWHzxuEQQJM7UCtxZMWv2IF25N8dpH5o4co8bt9YXxa7O09/ydd5svNKHzsz/7swD88i//Ml/84hf57Gc/e5FPf2Zky9WH9QjSNKPrhtxf7uD5Ee5MkZlagdmKjZKmrDY8lDTZcynnXm0/zjWZCjxe75JmGV03oudFqAq89uLsob+j50XcedDk/cctojhlumJTLppcm3aeWnof+RGrbnDka9jLOC4RPAq5rtFz2UvwByYh3k/SEvxJb9dhW4cG/cX2rbOeF1Ep6Gy2+sPv2x7b92rbdNFgtemytu15/H7AxkbvSFuXBu28s9TG0BWqJYtGxydLod33CeKMdsd7qh/reRHNTZfpksH8TJnVjS6/8cEG8/UCtZKFR16EdLXZ5727Gzy/WD3V9qmTmPT32HkY9y1XMBmxHiYr3u9nr/a22x6//fYKPS9EVxQqJZNK0eK1F2e497C5I66ZjoHfDzAV8NyA0I9wbI0ozgjjJK932fZY2ugTJyn9fkCrG7LR6nNtusiNmRJKmvLOt9aOtF103F5fGL82S3vP30VsubqU43v+/J//83z+85/n2rVrrK6ukiQJmqaRJAlra2ssLCyQZdmFPiaOZvtgNM0y4jSlXDAp2hpFS8OLYuoVG2frGENUuL/cIc3ypd1lx0BVFBavn34Akv8+k99+ezlfAlo0uT5bZLXpUrSNfQeyPS/iG99a4/3HLUoFHVPXaXR9Wm5I0daxTB3H0lm8Xsp/R90ZFs/cnuzZfQ1nWYtAiEkh8V6ch+3xVgHcIKJasigV8m1Rd5+0d9Q5GxQAXW269Lw8Rr/24gyrTZcgSg6M7duVCgalXTciq5sub76/PuznkjSl/yR6qs7aYCLEMjVuzBW5v9LFa7pUiwb3nnQI44xn50rM1ApP9WN7bf3NJ1QiaqX8NK67y208Px7Wl2t0PD56a/+tDNJnibMksX70dd2QO0vt4WdeVRX+8M4GjqnR6af044x+EDM7VeDBSocsU6iVzT3j6msvzvLv311lpdlHU1UMDZbW+8RJhqYq+GFMmmbMTxfwg4ROP2TqtoWp5wXbhRBn50Jq6PT7fZaXl4f//upXv0q1WmV6epqXX36ZN954A4A33niDl19+mXq9fuGPicMNijg+XOuy0uzzcLXLRssnSTMsU0fTFLJUodnxAWj3Qzw/Zn7KoWTr9NyI1aZ36KDxqDUBel7Etx61mJ9yeOnZGvWaTdeNSLOM+8udfX/HatNlpelSKhgUTAPT0CgXTHQVXD95qpbAUU4N2e8EFTkNS1w1Eu/Fedsdb9dbLhttnzTNDqyFs7tezPyUc+QTofZrx9t3G/za7z5gsxtg6SquH/HN+03uLXf4xnvrT/U9g6SMY+c15gqWzoOVHouzJb7rlWu88EyNWtl6qv171bkoFfTh71/e6NPuhaRZRrlooiiw0fb55v3mnn2h9FnitCTWj49BrPrXv3WXh6sdVCUvdvzbb6+gaypRkmLqOgVLwzJUVhp9+n6M64f71hgr2gbzUw6LM0VsQ+XxuovrR/hhxFqzz0bbI0lT/CDBMjUqjonrxwfWHRNCnMyFnHK1sbHBT/zET+B5HqqqUq1W+et//a/zyiuvcOfOHX7qp36KTqdDpVLh537u57h9+zbAhT92VFd1y9Xbd/NkjmPrbHYCnqz3UVSYqdrMTjk8We9hGiq1ks2z18rcX+4wP+VQK1vD33HYKVR20eI//OHSkU6UurPU5luPWlRLBjA4HSQhiVP6QcJLz9b2/B1v3W3w+99ao1I0UbfVxXH9CKdg8oN/4rljvzYHnRLw/GJ1LJcIHoVc1+i57CX4kxTvJ2kJ/iS1a3e8/eBxG10HXdO4MZu/9w86feos2jZIiDTaHk82XGxTGxbkNw0VQ9comDqLs8Udfc9ep2S9dbfJq7frB56etf2ap2pFNlv9Had0feO9NVQVyBTm6g6WodHthyw3XD7x6rWn+sLVpntgn3XWr9dlGtV2wXhvuZqkWA+TFe+3G8aqjk+lZNPr5yfQLs6U+N13VijaGhvtgIKloakKUZLS7Ue8+GwNS9d48Zna8Hdtj0vbY9I33lvDDWI6bkgQpiRJQgYkacbiTAnTUDF1Fds0dsTEw4zD67vbuLVZ2nv+LmLL1YUdWz5JrkpCZ7Acu9EN8P2Y9x5tUiuZTFcLALz/qEWapqiKwku3pui5EbqmoioKzy9WaXQDZqsWiqLg+jHNjo8XxGQZ/GevLuw4vnWw7DsGlDSlVto7CbT7uHRVAcPQMLZmLrMs495Kl8Vph+euV/f8HXeW2vzHb62RZRmFrSJwcZIRxjHXp8v86deuH/u1Ouw423EMQEch1zV6LjuhM0kmaYA/Se3aHW8fr/WIk5Q4SXnhRg04enLioG1HB7VtcDPzaLVH1w3JMvLVqUrGbM2h78c8O1didqqwo+/ZK4my2vColAx6XoQfxNiWnm8pLpjD9m/frjWooTM86cqN+K0/fIKuKxi6BlleZLnVDen7ES89O4VtatQrNpqmoKv5CTaHHcF+3C1Zk/QeuyjjnNCZNJMU77f7MFZ1mZ8ps7HZZ7nRp9ML6bkRQZxQr9hYhoauKfhRAkCtZDFbtfcdS791t4GmQrMb8LtvrxBGCaqqoGkqWQq6BkGUcnOhjIKCY+tMlWy+86VZSgXjSPFlHF7f3catzdLe8zexNXTE6Nte9LjV9VAVlZ4bEkYRa5ses7UCC9MOj9a6hEmKpqrcmCuhKsqHmfelNmGckiQZSxs9TF3DMBSyTBnuwwWGg9RSQefRhovvR6RphhvE+EGMZeqQZXhBzN3lDuWCztyUg2GotHsBepRSdgwMXcELEuIoZW7K2XE92/fsztcdrtUd3r3fpEVIkqYkaUq9UuClm7UTvV6D42y3D9RlWakQQuy0exBvF61j/4wCO+JtvWJzf6VDwcprt+1VC2f3BEXB0ihYOn0/3rdGxEEGCRHb1EhTg82eT5SkaCoEYYKy1a7dfc9e9diemS/xh3c2KNoGBUvDCxIabZ/veuXa8Pm21wDquCG6qg7rvM1P5cmkP7y7gaHlW7q6bsS95TbPLVSGRUuXNnpcny4SZumhfdb2BNJxXxshxOgYxipLp+OGPF7r0fVClAymqhbvPWwRJymGpmLoKrqu8fximTBKcWxzzxpjPS9iZcPl0XoX29RI4hTXj0hRMHUVy9DwoxTL0lBQePnW1I77g8uOL1I/TEwaudsUexrs9X+82qPTj/DDBD9KcIOM2SmbjhviBgq6plEqaHh+jL01MPztt/I91QVLY7Mb0tk6IWqwxWm6mp/qsdJwsU2dStEYDiprZYtHvYD3Hm6yMFNEVRXuLXdYb3nM1wvMVm2iOOMP725gGzp9P6JcUFAVaPciVFXhlefq6LtqDWwfqJYKBi89O8XapsfqpkuawFTF5pm5EkX7ZAF9v4H6WRR/FkKISbDXIP69B5tMF3cWsd9d8Hh30qXvxwTdmCTNiJMUXVMp2gbTFXtY8HhY1J69JyjabkC7FxBGCTfnK9TKFl4Y03UjWt2Q73xpltkDrmWQEKlXbLywx1TJpt0LCaK8XS89W8OxdYIo2dH3bC/MrACqovD2vSa9fkS7F2LqKrO1Ajfny/TciPmpD59zUJB5r9m+gqVTLxdI0gQ/TOi6IYah0XVD1lo+tqFt1bFzeXauwtzNwrAWxl591vZ6P8Dwf1eb7lNFoYUQF+e4yYhBrHIsnT+40+TJepc0yTANlVJmMlW2CMJ8i5SuaRQsDVPXuL1Q49ZCZUch+UF8uPukTZKlmIZKqxcQxCkdNz9pVikYpGl+aMoL1ytMV2x0VWWz6/P//tZdojjF1FWeW6xQKeZHl19kfLnsZJIQ50ESOmJPbhCjKrC00adY0IjihIpj0OgEBEFCp++TKVC2TT7+0ixxkvLmB/kpH/WyjWWqPFrvUi6YJElGqCQ0OwE3r5VotH0MQ0FRoB9E+FGEaWg4ts5MzeHt99eJk4w0zVhuuPTckLlagW4/JEsy2CrOBjBbs1lueCzMlrg+U2K+nq/MOSy50nMjXn1+mteNueHXgig5cWey1wkq228ohBDiqtsrSWBvFdks7bG1qFTQ85OggpiyY6AYW4U5DZXVZkjZ0SEDsgzbMri1UNkz5g6ed33TwzJ0kjSj0w/ouBFzVZvVVp9H6z0WpotUHJ1OP2/DzMz+CflBEt8yNa5PF3m02kNBwTYNpkoWlqERRMlTfc8gKbM9ydTsuNiGjoZCvWKTkmHo6rAmz1FkwEeeqbLZDWi7EU8aLhXHoOtGLK11CaOUmVqBNM2obB29PtiutVefNZjV305OpxHicp0kGTFfd3j3foONto+uKwRRTJZmWKZFmmX5ykAzX/l+a6GE6yckaTqMp7vHxHeW8ufXNYXpis3Sep+EDMc26HkBjY6PY2uUCjb9IOETz9R4tNbdcbLso/U+PT/G1POtoHBx8UWS1WISSUJH7MmxdB6udnBsDQWVOMnyGcCKgutHqJqCZagkWcq795vEcUYcp+i6hqrC8nqfomOgqlB2LKarFhkZK02XesVGQaFgKQRhwnrbpdkOmK0VKJctul6Epqg8Xuvj2BqVkk3Z1vDCGDeI0VSVatHEDxNs0+bmvMZ02dpRK+Gw5Mp5DFb36viEEELk9oq7lqGxti1xsXuwnSQpmqrwRw9aVIoGtqnhbx3Lvbu2w34D8sHz+mGCY2tsbvpYho6mRCiKQrcfM1W28IIY29QoOwaWqfFkvUfd2fsmaXsSvx1EKCr8px+dw9RVVjc97izlN1n73WhtTzIVrfxUKpS8rVMli7VNl2fnK0/9XM+LaD5osrTS2TE771j5SVU35koEj9tYRl7Pbqaq0d86yrzrhbx0Y4pa2SKIEnputG+NodNuI5YtDUKcvZMkI/L4YFAwI5Iko1o0sQ0DP8prUZqGimUYaFsnvVqGimVZ+35et2/havVcZqoF1ts+QZDg2CZZmpGREUQJzY7H23cbeGG8dbJsHj8qjkEYJdxf6QwTOhdVpkCS1WISSUJH7Gm+7vCND9aJwoRm18WPMkxdwdBUdE3FNJR8Waah5XtpGy62pWEbGlMlk5Q8OBuaSqGoEcYJhqbScyPqZZswSZiu2LT6Qb6CJ05YbvZp9AIKhsZ0rUCj7VEv23TciCBKqJUs1pouiZbR96O8gGQYc326yFKjDzBcog/5jOV+A8mDBqsyEBVCXDUXEff2irvbtyTB04NtRVFYb7moisq16XxL1bv3m9QrNuWCQb1i49j6gQPywfPaZl5LJowS4jQFRWFpo48fJEyVDVr9BC+IqRZNsiwDTaPu7F+8c5jEX2pT3bZ1+LkFY1g89LCbIj9MqFct1jc9NE0lCCKykkHXi4crTgcGs/Pzs+WnZue3b/td23QpOyaNjo9t6PmRxIZKEKQszBSBw29gTrONWLY0CHE+TpqMyIBbCxVUQydNEh6t9fCCmCBKCBMVP0jY7JrMVAsYuorvx7x1t7FnX7B9C9dyow9ktLs+QRRhGwZoACpTVYsszfiD9zewra2i7IqCuVWCYaXRp9UL9617dhQn6bek5qWYRPLuFfuydBXfzwfUURTTdWM0BSxTR1Mh3Fp9s77poSoKSZyR6bDZDdBUFc+PqTgGVScfdK82++iaSpblxxgub/Txg5g4yWj1IuoVk1LBpGTrGJoKisIHSx1MU6PTC3jueoV6xaLZCVhp5MvF5+sOfpDweL1FxTEwdZX3HrXouRFTFZuipbHR9nnlufqOIL/fYLVWt2QgKoS4Ui7qBnyvuKtsndQ0sNdgO0kybFvFDxPWN100VUFT8yNxlzZ6LM6U0DRl3wH54Hl1TeHhWoe1po8bxFybdihYGn094sFKH11Xma87dL2QjhsSpQqmAt961CJNM0oFnSQx6fvRcLvS4MTFG3NF2Nbmw26ytieZkjRjdsqh0fbIyA8OuL3H9rHB7Lxt6niusmN2/vnF6nDFUBin6JpCxTFJswxNUUi3DjT1gvzEya4bUbTzZNVef+PTbCOWLQ1CnI+jTkbuntgcFJKfKltomkqaZigomIZOmubxIs1SnjR6eEHCR56t0er6PF6L+WCpzWsvzjC/ddjI9i1cZcdgtdHHixK8ICHN8tOsipaOkimESYqug6JAEMVsdrfaoKpUSzZFSz9xmYKT9ltS81JMIvXwbxFX0WrTZW6qQJJmzNRsbi+WKdp5h9HzIuI4pdULebDaxY8SLEMjU8A0NbIMwjghilM0VdvqQBQc2+TFGzW6XszDlS53lzv5QN3UqJfzmgMzUwVMXWW6apNlGV03oGip3LpWptUNAIWSY/KRZ6a4tVBG11Q6bsjslEPPi7i/0mV5o0+nH7LWdInijEbH4/5yZ8f1DQarupoPunVV5fb1Kj03Gg5Ela1aPdZWjQchhJhE22/AzzPu7RV3X7o59VSyPQgTgigZztzWqxZTZYu1TQ/T0HjhRpU4zpf1G5rKarP/4RHeXsSdpTZv3W3w9tZ/95Y7+fbelo9jW9imhqGp+EHMTLXA89drDE7vNnUFTVVJ0pR+EPH/vbmEpipUSwZpButtD9ePePP9deI0pVTQMXRlWOtn4LAZ38F1lgpG3l9t9mn3QuIkY6PjA/kNy3ZuEGPuKvhvbqu1UyoYPL9Y5eMvzZFu/XuuVmBxtoRpalSKBu89auKHCZqmUCkZ3H3Sfup5tv+9nl+s8urtaZ5fPHpy77B2CiFOZnd8HNTpKjn5ZzlOU1QFHq51ebjWRVMhTlPcIC/2XnYs2t0Q29SZmbKZrdnMlG2qJYtuP6LvxcxUbZY3XLwgwQ9jVpsu/+b3HrG6mfcHgy1cUZyw3vLoeDG1okmtZBGEKX0vJoxTum6+ir5SMLHNPOESpwmb3YCuF2JbOn/ytevHji8DJ+239hv/y6StGGeyQkfsyQ1iojhjYdrhyUafbz1o4UUxhqahqgoZKpqS4vkRpq4SZxmLMw6WobPUyJewv3q7zjNzRdIMuv2Q1c0+hq7hWDorzT79IMKxdRami3T6IWGUsL7pU3V03CBmcabEjZkSlpXXTCgXTWarBdwwIQgiXD/BNjWqRZNqyaDVC3j/UTtfQhrHJM38Op6dK/Fko8/Hbk/vuMa9at7ck721QogxdpIl6BdZU2B33C07JhsbvR0zy0GYsNLIB+W6orAwW6ZWzuuwFSydOMmYreVHdffcCF1T+ehzeXwfzNiqCjxYzU+CunWtTM+P0DSF569VKNn5DcBm16fTC7mxdcJh1wtZbXoULJ2FmSJpptDuB7x4o4ofpnT6Aa6f8GC5x+yUjWVouH5MmsJKo0/Pi/jYc/V869QhM76Dm4r7yx2CKF+pigK6pmAbKnHy9GzzYHZ+u70SR7cWKnzz4SZkKV4Qo+sKz8yWCcMYN0ywLY3pik3B0k91GMB+ZEuDEOdjv5Vzq02XNMtYb3k8Xu2hayrlokGzG3BjtkS1ZNHphbR6Pv7WltOCopOSUqvZqKpCEKXUShYAYZSw2fMxNI2CqbK66fNrv3uf11+a59ZCBS+IaXVDQKVSMOh4MV03IMsgzcD1Y1AyFupFKiWbNMuwDZUHa11cP+AjN6r8p982N1z1cxKn6bek5qWYNNK7ij05ls7j9R5BGPNko0+mZpRsjQyVdjcgshMUVcH3Y4xyfuxgqxuiajFFS+fmXInbi1VURWFuqsC/f3cV149RiPOOI0y4MVNkvZXPRGZZRqPlwabLK89O0fdjCrbGjdkyjq0Pv6fnxUyXLeJt9Qoer/Vo9UIer/ZYa7lkaUbBNjAthb4fcX+5s6N45mHXfdhAVGrsCCFG0UFL0I9yBPdl3IB33XDfJIymqbR7AUGUEkQJlqnns8ZRDBnUKzb1cr6ac7XpoioKlqmRJBnvPWzT9QKCKKHR9jD0fLVos+NjWzpxknJ9pojrJ0xXbfp+XiDZ0BVSMhptD03XUBWF+ysdWt0Ay8xrP3S9EKeg02j7NLs+pq5xc77MctPl3fubVB2TasXOZ4oP6B9KBYOCpfPq8zOstzziJMXUNaI4zbdCOQbfeG+dqYqFY+mUnPx0Kj+MD6w7USoYvPzs1FbfGBDHGbM1k66mMjuV/42X1nvYlk7B1Gi0gzPtz2RLgxDnZ69kxDc7Ppu9ANNQUZR8i9NmJyBOUm7Mluj2Q9651+T2s1Ms1Ausbnp03RBDV2h0A5I4o1w0KBdNGm2fJMkwNI00TVlv+2ga9P2YN99fZ2mjT98LaXYDFCUjU6DnhpApqGpGHKV4SUS1ZBOECe1+gGVotHshz8yVmakWeGa+xHrLpe9FB9a7PIgkjoX4kGy5uqK2L0u/s/T0kuv5uoOqwPuPWpi6hq6qZIqKZWooap6B11TIVIjiFMfWafQCkjSjUjSpVwusb3qkWcYfvt9go+VhaCq2mQ+Qe17E0oZLwdJJ04xHqz1a/YiCabDZC1hq9Gj3Appdf7iMfRCody851TWFpfUeXpRX509R8IIYw8iX1Xe9EHtbwD/IfstZBzUeBjdMg6X2cZoeuGRdCCEuykmXoB8W987Tk/XesM2b3YCibeDYOs1ufhNQLVkUbR1dVbGNvN6MrqqUHB0FhShJma8X81OpNvrEccrSRo+uG+AHeeKj3Y1I0pQnG33uPunQ7oUsb7hsdkIsQ+XBShdQMHQVXdVQgVY3oNHyhjciuqaioLLW9NG1/KTFBysdTF3D0FU0XWVhuohj6+i6QtHSeLja4avfeMzbdxv79hGD7Ul+EOe14wBDV2j3Q9Y2Pfp+NOxrBkeNG/rhWwVmpwoEccJMrcAz80UyYH2zz0bbJ05SCpZO34t484ONYX2gs+rPZEvD/naPvbpueNlNEhPACxJURcHUNSxTR1EUFDUvOu8FMX/0oJkfXGJqxCkoqoKpqzQ6AatNlyRLWZguEsVb/QAZmgobbZ92P6LR9gnCfHWgpircX+2gKpCmGY22nxeRV8DQ8xpfpqkRJwlJmhHFKY2uR5ylJEmKH8b0vYiNts96yz0w9nTdcN97lcvst4QYNZLGvIJ2z+K2+yEfvNViqlJgumwNs+SvvTjL7769SsFWKZgaQZIHTNvU8IIEUzcoWQZRnPJgpYuqKsO9sqWCQRgndN2IpUaPsmPkHczWDKhp5B1JrWSwupX4ybIMN4jxg4j5qQJRnOEHeTX+uakCqqIMZ/hUReHBcj6Tq6Dw7c9P8+b7G/Q0DcdS8wJwQYxt6ziWQXXrWMTtr8F+p5YcVAhSij0KIUbVSZegn6YA7mn1vGhYb2VwrHiWKcNEvqmrhFHK84tVnl+s0vMifvutZaIoo2ApLM7kfcJa0+XJ1qkptZJJkkKSZYRBShgnKD3oexEVx+D5xXL+M5se5aJJsxPwkWcrtHsRYZTms9O6gh9lzE0VaPdDkjQjjFJUFRZmSmRZyqP1HlGcoqoKBUunYOk4tobrJdxb6eD7CUEc8+6DJm4Q8dFb0/ueuGhbW6dR6flJXEEYYxkWjqMPk3MAPTfiuz4yPzxOfZAgaHZ8vCDBtnWmt45gv3mtTKPt82gtP4lG11T6foyCgqJAt59vVzN0dcdz3F/uULD0U63akS0NT9trBd17DzaZLhqS7BKnYts6fjciilMqjsFy0yXLMkoFg5WmSxhnTFdN/uh+k422RxAldN1wmECP04yVRp/5KYfpisXj9T4dt4vrR+iagqKpxGnK0rqLoauQKlQrJk8a+Tg6TVNUVQUUNE3F0PJC7G6QkKYBYZIyVVa5PltCUdg6rtwijtkRe7aPpXtexIMNdziBurvo8WX2W0KMGknoXEHbkxJeEOfHpar5calx0fjwGNQph488U6XVC8BR6LoRrh8RZwAJfpRAfrIrqpIPENc2XRzbYHaruHHHjTB1jYpj0uoHgJZX00/ANFQ6vXhrKxZbRZdDqo5Oq+czP1XENjW6bkSnF/GdL+WbBgYDohefqeYDooct6rpJrWTS7AaYloqmKYRRRsHUWZwpMV22htd/WGX8gwaiF1lrQgghjuM0S9Av6wa8VDBYd/PVOINjxTMy7K02727/oFBvnKbDGjZLGz0UFG7Mlni41iWMY6I4od0NUVWoly0avRBDzU918YKEaslkcbZIuWDiBzGmqeQFQ/shYRijKCp+nPDijRrvPdzMjxcv21QcAy9K6bshUZSxupnX3NFUBVVVcFSdTj8iyVJsQ6NoG7h+Pht9f7nzVC23wfakUsFgbdMjjjPSLEVTVdIspV4pDr93d18z6MvSLGOzF6AqCn43wtQUHq71WJxxSNOMa9MOhqZyf6WD5yfEcUoUk2/HmMlX7wzEW/1hXqxap90L+WCpTb1sUa/YssX4FPaaELK3VtBJ8kucxnTZwtTy1e9RDNemnGGyuedG6KrCNx9s0gtiiqaGQr5l09RTqkUTRclPp3rSdLk1X+ZPvHqN33l7GTKFKMlXRaoKlEsGrV6IYWj0/TivY+mYdN2QOM2wDAXLUOl5EaqqUjbBtlS8bky3F7Le8pip2vTciHrZxjY/3CiyO76tNl1KZZss3n8CVRLHQuQkoXPFDGbzFJW8IGKYYBhqfgKFn+yYoQNAUVjddJmu2MxULR76EWGSMlMp0HHzIpOaoqBp+VLMME5YbfSolU2mKza6pvCRZ6qstTymSnmdgp4XkykwWy1gGTpZlpFk+fLO2E/IMohjSLOMG3OlYe2cUsHgzlL7qQGRqam8+2CTuZrDfBCzsekRJCm1ssVMJS/21tw6NWS+7pxqlY3s2RVCjKpRqV1ynDpj12dLPF7O2zxVtnbU0Bksod/d/u3X2Wh7KCj4YUzB0jE0lZ4b0epFVIompqGRJCmGqjA35VAs6LxwowZ8WJft+kyRh2tdHFtntmoTJSmuH3OjWkbXVV6+VefxWg/DUCGDVtun54fM1Gx6bkiaZnTcvA4NFZsoTXBMHUNXiZIUp2Di2Nq+xfkHs8wdN2K94YKSFyUFeLLRz0+CrNhPHc0+6MvWWx6moe6ov1Mu6Nx50mFuqoCpb/V1uo7m5HWGpqs2HTdgaSPf8uAF+eu3tulSLprDZNl620NTlWEh1fM4zv6q2GtCyDI01uT0L3FK83WHvh8NJ1MHsX++7vDm++ugKgRRQpykbLQjMjLUDDRVxQ9TqiWTOM5QyEhSKFgGlVKBNPNYWu9RKVoYik7Xz0+5fW6hwlKjT5Lkqwu9MKGgq+iaQrcfkioK0xWTLAMvyFceenHKo9XucFWmG8TculYZXsPusbQbxEzXNbxt1ykTqELsTe5Cr5DBbJ6h51uf4iTl8VqPZ+ZKROTHhwMkScofPdzcOoHKwdAUvvW4zWrTY3bK4dueten7MZ1HLQwtH7BqijKcMVVVhUbbI00yfuATNynaBknaxPVDSgWD2tbM6FrbgzRDUSCOE5JEyQtO9gNKjk7fj3nnXpO+F1KwDBxLp9ENmK1auH5Ms+Pjhwmtro/nR9imxnPXKluDUg9T07BNg2fmS1SL5nAljh+lzFatHa/NUTuJ49wwbb+pWXQjTAUZBAshzs0oLEE/bAXkbmXHHLY5zFKencu3QyUpWIa6Z/u3X+dmL6RoaYRRgh8mKAoEcULXC4iTBFPXcGydmWoBRQHbzIc9XhCz0nSJorw4cqlgkKT5tl9dU5mp2nznR6/x7gfrWKbG4mwxLyTaD/HCGMtUKRfy7Qqdfki7F1G0YbpSoNnJa0pESUocp9QrNvkymIy9lAoGbN2QzU8VSJKUbz1qsdHx8poXSV6YebpS4JXn6sOfGyQI/K1kDOT1d1w/YWHa4YOlNnO1ApARxRm2rZFlGo22jxvEWIZOJwuxDH24tbnrxbywmN9kNTt5wWddy7fAyRbj09lrQiiIEpkQEvvaKzkO7Jkw3+/0q/m6w6O1HkmSUTA0Aj8myVKMrdP0vCBiYcZhfTOvdYkCSxs9HEvFMYv0/Zi+F6MoUDB1Co5Bxw1RMoXF2RKNto9paGy0PcjyhHHB0rfq6ai0OgGqqlIrGkRpRrPtcXuhynSlgKYp+xZ4d7ZO4Rtw/ZjVZp8ozg59LYS4aqQXuUIGs3nz9SJLG72tga7BWstjumoPaxGsbnqkKRRtA0PP97yGcUqzG6Cr+fLLOE4pFQzSNCOLMnRTxVJ14ihF01SKtk7ZMSjaxtbpI6CpGkVHH9bpub/c4ZsPmqx3fJI4xdBVgjil70ZoqkKpkLLZ9fMimJrK+49brDRcpioWuqZQ2pr1DNMMy9CJ45QMmK85VB2TO086dNyARkvHMrThgLfVDQm3nZIFx9uWcJQbpt03NVGc8ni9KzObQohzddlL0LevgGy0fR6sdGj1Qt5/2OJPvnZ9z2Nq92rz4Ebm3nLnwME7wPuPW3hBjG3lsT6KU4I4xTIzdF0hjBOSRENTNXpuyNt3G7S6PsWCxbPzRdZbLhvdgFrBorZta9HCTAnfDfNkU5Qnm+brDv/vb91FU/MiygZg1wq0DY2eF+Z9naLQ6PhMVwvMbtV/6/sRz86V9129tP11e9zyqJUtHMug54VoWl6jR1Xy62+6EZEf5bXi9qi/Y5sauq5ya76CH8asNEPIYLZWoOjoLK32SZKMatFkYdrB9ePh1ubbCxU0bWdNo8FzgMyQn8ZeE0KKFHEV+9grOf7u/QZZplArm3uXDNiqM3Z/ucNvv7XMw7Uelqay1sxPlCoWTaarNlGc0u6F+FGMaWg8Xu8Txwk3r5UhA1PXqFcKLDd6WEb+/H6YEsUpmqrQ9IKt0gkaLz5T48lGH0NT8YKYkmqia8qOQvdBkuXbVisGtxcq2KbOK8/VDxxLz9cdGv0oX1kUpztWb8bp4a+FEFeJJHSukMFsnmLkhSSbHR/H0lhvR8xWCxQsLS+U1s+PZDV0BcgHdastDy+ImCrZOJZOkmbUSib9MKFg6qBkeWfhGNy+XkFTFFY3Pb76Hx/nW6qSPNFjhzo35/OgXSwYbHYDbswUabZ9ul5IlKSUCzpBnBdfe2bOwbEMPlhq5wPQkknjkcfMVIGCpZPFCoamUqmYWFvL0u8tt9no+Oi6gm1orGz28cKY5xer2KaWX2eYZ/1Psi3hKDdMu7d12aY+PG1GZjaFEJNq0M802j5v3W3gWDr1iknHjfndd1b4rleu7ZnU2e44NzLzdYeVZp+CmW+3Wu/6ZCksTDl03ZCCpaNrCpmi0HFDKiWTteUOuq5i6BoPVnqkWUoYJmwmKZWSOUyydLeSObuTL2XHpN0PMJL8lEXXj2l0fCqOyUzNxjDqfOtRm5JtkCQZCinTlTy5s9/qpe3bcQYrbkxHRVUVXrhRxfUjPljqUCmZTDsmq26QFzgO8jY8XHWJkxRdVbh9vUIQJrzwTI0/vLPBTLVAwVLxgjRfaVt3eGauiKLkfXy98uH2s1sLlWHSwTLyrdhZlg3/ZrLF+OT2mhB66eYUfj+47KaJEbRXeYC+H0OWMV8vDL82+N5BMuede00aHY84Tlhv9um4MQoZBUsjTjPSOMHQVaolk/VNDy+MSeKEqYrFg+UuxYIBpBQLJtWiRRCktPsBUZJyY6ZEzwuJvRBTz5PYQZSAktfKMXSTF5+pstrwaPUDFPJ1ibahYhcNTF3j3kqHWik/qOT5A8bDpYLBzEyJd761xoP1LgVL51rdGU7OHvZaCHGVSK98hWxf7uvYOo5dYnaqwM1r+RGmgwHG7etV1lsuUZyf9LHR8uj1IqI0xdO3tjpFCV6QYCgKcZbS7ofUihY35ktYukbXCwnjhFYvwDQUVEWlHwR03JBf+50HvP5tcyxv9JmtFtjshaSAaeloikKQpHz01hTLGx4PVnp03BBNVbFtDcjoeBH1aiE/ocQxiZOU9x5uYhkat66V6Xp5Vf6ZagFFVSiYBn6Q0Oj4zNYKw9nX89yWcFjx5OPUmBBCiMty3Fg16GcerOQra2xTI0pSKkWTom3w3oPWoQmd49zI9NyI6aqDH0T4YUKSpFimShDFWJaObaqsNvM6EIqiUK/aOJZGSTO5+6SDpipMV23iJCXJMgxdp2jr3Fqo7HvCynMLFd5/3CZJE7wgo+vlp8XcnC+jKAq1ksVHnqnS6UVMVazh63ZQ/bbt/fNgxY2CMtwK/XC1l5/6uNojSDJ0oFY26fRC1jc9um5Eu+ejqQpRnPKnX1skTTNuzuergrwgP6FypmrT6Uf71oLbnnSwLAM38JivO9imtm9NI3F0uyeEyo4pCR2xp73GkXGSPrVzM45THqx3cYOYzU6Qn0ylqjxq9EABhQw/yg8yKdg63X6EbWmYmkLJMbZOtIt40vBYmHbQVNjshqxueizUHWplkyjJ61uuNPoYhka1aGPoKo5lECcZj9d6qIqCYajcXeqQ5bliojhF11SKBR1VUVEVBRUFXVGOtJqm7Jg8v/hhwnuQhN7vtZAVhOKqkoTOFbJf/ZfdAbXnRfT9mCcbfdo9n/cft/DDGF3VaCYpza6PqoCua8xWba7PlHjnQZNyUSdNM9IsQ1UUirbOWrNPyTHRNBU/iLAtHcc2eLze5Q/vNNA1lWrRolbKa9yggOvFJCn0vbxDyeseZAQBdN0Ix9KI4phayWS95eFYOtfqeZHmd+43mK4UeG6hgqIqrDVddD2vF9R1IyqOOUzenGcG/6DiycetMSGEEJfhJLFq0M+0eiH1irmjjoypKzQ6h9+8HvVGZjB4f+5amfcft0iS/ISszV6+bXe+bpOkCuvtvKymYai4fsxmx+f6DLR6AUGYoGoK5YKBpev0/ZB7W8d2l8o2bj9jaaOPH8RoWn5DMjtV4N5yh76XoKkKuqoyUyuwMJOfSOUF+RamVj9gcbY4TILdOyDR/9y2lTH1ssW3HrXoehFTJWv4d3j+ehXH1ojjjNVWn3rZ4t37zTyRlaY8M1fGNBR6XszbdzeYny4xW823kQ1kWZZvSTtgleqgf3yeDxN6ciywEBdrr3GkrqmQfRgIXT/mwWq+eqVU0Hm83mNtM1+xGCUpPT8mjBPcIEbJMiDDtvItlf0gwdAUKkWLsqLiRzFxnIJpoGkKVcskBRQUltZ73FooEycZoND3Ywq2RqPj0XVDOv2ALANDV+m7+Wp9sgzb1qlXbFw/wjRA1TSqtsVUxTrWqvWjvBYgKwjF1SXv+itkMPP2Rw+afOtRmzBOWJwuPTXjWioY3Foo83ity5NGvi9WtUzcIIIUVBRSFSxFwQsT1loeJVun78WYus71mSJdN+TBao+eF6Oo+ZHnXhijqSr1Sn7EoaIoJEledb/RCQijBMc20DWFZtvLTwwhQ1VVQCHNMhqd/HjbzV5IqxeyMF1E0xSCCOplmyhO2OyG3JjLKFs6c3WHRtsnTjKKlnFhSZPdyTM/jIcD5tOcsiWEEBflJLFq0M+8/7BFx42pFE3qFXt4alKtaO35cwM9L2KzE/B4rUfZMahXbBxbP3DwXnIMvvmgSZJmOLaOkmVEUYIKvPdokzhNUVSVNM2LBisoPFrvoWsaigKtXkinF3BtuoRpqHTdEDeIsZ384IAM2Oz6NFo+X/cjnr9e3apDkbDR9mn1AixDpdnx8QKdZsdHURSmSuaOk6EOSvTvOO2ql69Kna0USNKUhytdPD+m0fUxTY1KWSNL4d37m/T8CFPXUNV8ZeyUblO0ddpuRLUY71kvrl6xKTn5aqlWP6BWtHjpZm3PvvGyazIJcVXtNQlbtHWyLD+xytRVVpt9gjDB0jXuLHXw/AhFVdno+Lh+hOuFRHGe6E4z6Lkxhq5QdExU8vqUXTdE1xR0LU/U9LyYl27W8IOEe8sdgiglTjPWNj1KRQvbUFioO/S8iKUNF5WMLGW4ytHUVYIwIU3y02+jOCFJM+I4o1I3qZXyVYvHWU1zlNfisk51FGIUSELniun7EUsbfeamPtxTv72uwWA27s5SmyRNuT5bpmhp3FvuEsQpKhlJmkEGXpSQAY6lMV93WFrPa9XcX+nQaPt0eiGqruIHMR03JMsybBOSOKPZzk+yWmkGdN0YVVUIovx/ywWdIEpRVYWCbZAlGYqiEMYpcZSSpVCw8qNpN3SPWcUBJc/Wz08VuLfc4/F6lxszJSwz3172zFyZV56rX9jM4u698tWqOkwmHTRLK4QQo+KwraP7KRUM/uRr1/ndd1Yo2gamnteZ6fsRr74yve/PDVaiVEoGfhTjhwmP17vM1Zw9B+/tXoBjGdxZalN2TAxdxQ8T+l7EphvxYK2H58dUijpxCt1+iKZAmuXFPbN8+hk9TCja2lZCJuLWQp58WV7Pkzl5HA/RdRVFUXi41mWtna8ODaOEgqmxuulRLds8XOtQdWxMU2W6WtiRBDvslMTtyZNKySRJMpY2epQck1JRZ7MTkKUZxaI1rGlRMHUabQ9NU9FUFYWAqbKFprJvvbha3crbM13gmfkiYZyy2nQp2oasvhFiROxVc+mjt/L4Ofhaz0/QdWWrJpiCF2hstLp0+iFdN8QPE6IEVBVQgRTSLB/DRymoSUaSJaRpXq9LU/NixnGc8WC1i6VrlB2dNNVo9SKKto6fZviBy0rTRyWjUrRQlDwxBODHCUqYkZEnq10vQdHyE928IGGqDPWKfazVNEd5LWQFobjKJKFzxbz3oEXRzo9aBdC0DC+I+Y3fe8S3Pz+DG0RUSxaKClGS4fsRtq5SdgzcIKLV8UmBomUQRTG+H5GmGc1uiB/EqAo8CmJAIUlTjFRlox+iqmDoedFKRVUoWhrNbpAH3izf36sqKqauEUYZcRJRr9hkWb583Q9j/G5CRoauwq2FKqvNPn6YstrsUyvZ2GZGhsKta2WqJYtGx2emWuDZuTK3FioXHuS3D85nZ8usr+cV+g+apRVCiItwlNo4p4lV81MO3/XKNd570KLRyVeBvPrK9IH1cwYrgtI0Q1MV1lsecZKiZCqfePXa8Ht6XowfxKw2PXQtoNX1qVdsvCAGBW5fr3HnSQtL03C9kCDKyABFyWeg0yzLj9Q1tGGCSFcV4jglUqHvhjzZ6PHuwxa+H5GkecFPFIbHlHt+QjPzsSwNXVUJk5SeG+KHKZae8Nz1yrCfHSTB9jslEeDOUnv4t2h2fGZqNkub/eGx60mSUi3lp8qsbLgEUcK1qQIb7fwkSE0DsoyNlo+hq1yrO/vWi9tv5dX9ra1mUttNiNGw3wmAA2EQY9s6hq6y3vJ4/3GbVi+g6wZ03TyZqymgKvnx4Gq+2wrL1EjSLC+TkGQohoKq5jV1yrbB2mafJE6JAC/Mt2nGacqjNZepkkW1bJKkKbqp4hR0gjjBtjT6XkTsK9gWGJpKux9imhq2qtMLIx6vdfnYc1P5yvpjrqbZb7WgrCAUQhI6V06rHzBdMYE8W77WdNE1cKOU9ZaLFyaUHTM/ZtWLCOKEx+s9SgUDS1MxDRXLNDCNfKlnGGdkKqRJhqqoJEmCvXUKlqLkxdoGR6CmaYofplRL+VYtFIWibVB28mO9W70A09RJ4oR2PyIIU+anbW7MFfnmg00yUsqOybfdnMKxDTZaHuttF0PPl80PClN+/KVZpsoWPS/m1dv7zwZflsNmaYUQ4jwdtTbOaWPV/JRzaAHk7dwgRlPh7pMOfpigkGEZOm03r7uz/Vjer37jEa6fTyL0/JggdjH1fFKgPmtjaBq9JKJetlna6BHFGVs7AFAVKA62cZHmyQ1LxyJDUVT8MMULE+oVi/c3XVIgyzQWZkq4fkjfi4jifHlPXVdI02x4hPiLN2r03GiYzIGdSbDdNyV7/S2a3QDD+P+z92exlmZpeh72rOEf93jmmDIjcqjKqq4udqlZZLdMiRRIXchG3+mGIHhnAb4z4AsJBAw3AVG+aIiQAQMEZAO+FKRrkwLUlCDAtkQ1qWaxumuunCIiI+LEmff4j2vwxdrnxImIE2NGREZm/g9QWZln/Pc+e39r/d96v/eVZ7Hhw17M7tGCNNJc3sywQpHocIO2s55jnQ/NLw+RDj4568OEfh5d2LS7SCVqVq+Bj66vdd5uHR1vKY/WC71q5FSN4ePbE1rnWJYNzUqZB2A9COdJYkVrLJGSaCm5vJHTGEvTOsrasNaPibViPEjYm5RY57CNJ46Cybz3Hi1DAydPo2AmbywHJxVSCawNjSMnAQSTRUMSKaLVz9Ra4IGjac24n54pBT/vGsgdHV8a+VVfQMebZdxLKOuwq50ug4TcOUE/i7AO8lRx73BJWRsiLYmUQCvJsmrZn5QoqUhiyWTRYF0o7mVpwAuEBGM9ZWMw1tFaj9YhJlwrQbwq7GVjuX+0xDSWRdVw92DJ7vESYxyHkwqP4N2dPmvjmMkizP8O0ojL6z1+5/o6a/2EybzCec8oT+mnEVXjaFrL1jh9YSnnm+b0lFbLcGqrpew2zR0dHW+M8wqNqrEcTEruHi75yW8OHjr9fdO1Kk80t/cW3Dtc8sX+glv359zeCyO8N3dnZ193c3fG4bRCK0kSKQZ5xHTZsHu05HBWcnN3ysm8YXOcsTXOgAfJKGqlsknjcPCwMcrIEk2eKIa9hJ21DOcdcSS5cXnMeJCQxiGqvKgapJCUjUGIcMq9rAz7k4qTWcXn92dY51YjxCHu+zQZamf94sbW+b+FEGHcYWctY++4QEloWoeUMOolrA0SZkvDII/40Xc2aVpLEimuXxpwbbtPlkRsr/fYHGVcvzRk7/hBStepl8+ibM+UV+fZPykY9OKHruPUtLSjo+Pt4NF6sTFKGfVifn3rhGnRsH9SslgajD1f9UB48HiEhyRRRFqyNkj57jtrrA1Tttcy3tkZ8N3ra6wPUzYGCZ6g2hFCsDZI0So0ZASC9WGKFgLrQEcr/52mpbEWKQRta/B4okgxXzYY65BCsLMyjz9VDl5Unzo6Ol6ct/OOt+O18dH1MX/2i/sAVHWLIKhlfvj+BmUdGjEHkyWXNnr0sojdo4LjaYVxjjyNyBLFYtkiEWwMM4oqxKFa56hKi0AgfZjF9Su/nVNJZ21DbGqQfgbztEEWMS9aFkuLjaCfRggRpPFX1nsczSqyWPHXvrfD8TxcR2sck3lw1B/0YjbHKZN5jVaCWdF8LaJVO6PJjo6Or4pTb5yyNtzZXxBFkmGumS3bx1QZb7JW7azn/PP/9RZH05pepkkizbJuWVYF/7+/vEe2OsW9d7gkjSTTRYPzjihS9DPFzd2SJNJkiWY8TJgsanaPCpJYsj6ImRYGrcIIsRQwzCO0loz7CR9cHXL/uMQ5Ry+NiFbK0mEv5nA6o2lb+nnExjAcIhjnkEowXTQoIWitR1SWf/3rA/69f+vqWRPsWb4OF/kUjfoxjfXkseKze1MGvZj3rwxRKpiN/vUfXqFa1rx/ZcTBpMA5eHd7wI8+3EKpkLq1KNonGlpfpLyal4YPrw4fuo7O262j4+3ifL0oa8N02fD57pTpsgHnKSuLA6QPDZ1TG/nWQ09L+oOUQa7ZWeuzMUqpW8OHV0ekkcLhibXC44i0IooEWRRhrEVJUFKRZ5r1cYa1lvEwpWcsJ8uGtrGs9+KVWXFE3VhyEZT5TRtU95fWU7IkIk80N3dnHE1LisZR15Y0lmRJuD/43bdQWd/R8bbTNXS+ZZz3NVgUBuscW2s5ZWPIEs3eSUFZW45ndUj6qFqE8DRN8K+ZFg1mZSbZ2OBiL0VYNLwDrcMSYq1HyuCjU7eWREu0UiipqGrDsJfinOdgUlHULVVtcEv48J2YUS9FCpgVDXmiWdaG7wwSZqvkkdmypDKOfHUym8aKrbWMk1nJ4UlF27pO8dLR0dHxBE4VGkeziigKY0qtcQzy6IWiZL8Mpx4+t48K2qo9U7BMV35sZW3CgcAq2appLLf3Znx6d8ovbh4xXdQIJdkYJEgR/HYGacSVrRxrwVlLVVnK2hBHkrJxRErSWse8aJACvvvuGO8leaL4ze0TTuYNeaq5cWnAomxp5iFB6qN3xtSt42RRM1u2bI0yiiaY/WeRom0dFshSDVLwL3+1x3/4tz7gg5eM422MY2OQnD0n9w6X3NlfcmWzx/tXRgzymGpZc+PyEOfDKMWjI3Gf786eaGh9kZfP+5eHZ+PR56/jbVW6dnR8GzmtF4ui5bdfTDicFkyXLdNFjXVwqrs7r79TIgQEaqV5/+oYYyx/5YMNRv2YX9+eoCRc3uwBrIzhPWmseWdzwNG8YrkwCCXoZRotJVI4itoRaYkUgjxyNCI0kmcnBUoaBGAaT2PD1y3Lmn42JF2lEv7i8yOSSDFf1kghmBUGpRI+uzf9SjwvOzq+7nQr9beQnbWcXho27kezkNRhrGXvpMFaz3zZcjipQISTwjv7cyCMV1V1iN821iMFuNP2/6qDL6UAEWTip5tMKQXe+XDKeHXIzXszpouGqjE454kiQRopytowmbfkSUTTGu4eFuSJppdqPt+dEmnNO1t9WuvJYk3dWvpZOEkVePp5zPVLQT7fLQYdHR0dF3Oq0JgXLcOVh1ljLFc3+29ElXHeB2Ijj9kran7x+TFChERDj6c1nqo1eLc6ao4jjuc1n92b0rTB1Lg1llv3FySRpDKOtV7M/nGJkhKPYFG1WGtBS5x3CCRKQFm1JJFma5iRpxE///yYcT/mg8sD5lXL/eMCrSTroxznPeuDFCHhu++M+PTelHEv5XheMi9awJMmCoRgY5BgnOVwUvI//Plt/v0fv/tMD6En+RSN15Oz5+g774zOPn6eJ5ks97PomYbWT/LyefQ63mala0fHt42d9ZxffH7Mb++cgIeDacXRSYn35/bjj+A89BLJ+1cG/K3ff4ej4wWzZcvt/QVtY9lYSzmeVVRNUNEP8ohykGBdTBRJslizqFp6iaaXaRaFYbKokVJyZaNHHElO5hV1Y/HOUTbhEFgqyTCL0TrUnsNpyb//nS0WRcugF3M0qYh1sGQIjfaWjVH6Rg4UOjq+aXQeOt9S9o4LkkiSaMX9o4L7hyVt6zieVXxwbcSgF5PGivtHSxZVS1EZamNpnQchwuKx+lk+fAgPWOexxqOkZJQnDPKYLA6LQCQFi6INLvjWAoI8jYijiCjW9HsJxjnu7s+5cxgMLtNYEUeK3ZV5c5povn9jjR99ZxOhBEIE2alznlEv4d2dPkXdScQ7Ojo6nsRpI6CXRMyWBiUFVzf75Kl+IVXGomz59O6Un312xKd3n9//4CLfmKJqWFaGcT9erS8eF7zziZRCipAEhYdF2YSbByRCgLWeWErq1jErahZ1y3RRkSaKcS+laizWe6wPTQov4PJmzrwMNyaXN3I2RhnvXBry/pUx437MbNmgdDicOJ4F5efNezOWZVDmjPspm8MUIQTOw9YoZVm1nCxaBHAyb/iXv9x75nPyJJ+i8yNTT/O06WcRH1wd8cP3N/jg6gNl6s56Tt3Y5/by6bzdOjrefvpZRC/VCCGYLWsWRYtUwetGiou/Ryvo5RF5EvH5vQnOe3Y2Mn74/jpXt3v89vaE6bIhS4IR+69vnbC9lvLRu2PyRJNoSR5rrPc4L2haw7ifMupp7h0u+cXNY46nFVVriJTEmpUiSAbvzKoOB655rFmWLUVt2FnLWFYt3vvV/2BZtuysZd0evqPjJegUOt9wnhRNezSv2TtaMi9a5ssa6z3mxKO14MP+iP0TyWxRs6ws1joa6ygag1idAngfZnQ9p91/HUaxAC0hjkMiFgi2xmGEarpsOZpV9DNN3awc87UIDvuNoZ+GhepoWhEpydogYXOcczQtaI3ns7szPnxXcY2gHLq8ltPPI6x1pIlmY5gipSCJuj5lR0dHx9PoZxG//9HWmQok1vKF/MeeNynrIi7yjTHWg/Bsr+fc2puTaEmrQ5JUHAVz/cNJyc5ahljF8B7OCto2jP2uDYKn2yCPqCqDUoKqtuxsZESxZFm0lI0liiSjfkIcaW7tzcHB9lrCyTwkaQ17Me9s95mXNiSxZCGed/eo4N7BkmGuaVrHncWC9XH4nY11VG3L4aShtZYs0fhpyaJoyCLNv/3DS099Ti7yKfr8nM/R0ayiqg1JrEnPKW4u+ps8ut4viva5vHyedB0dHR1vFx64vjPg54saJQSNFxjrLlToCFiZ3zv2jpeMhglXN8Znyj1jPVe3+iyKhvmygdXh7KJoubTRY2OUMshjPrs3Y142LIqGNFEcTgoiLfGAMTYc5taQrvbfxnna0pBEhjzTfOfqkMb4szVCKcm1rT4n85q6tQghuLbdRynZ7eE7Ol6CrqHzDebRDfd02fDJzyasDTM+/SI44s/LhumipWnDSd6oF/NhZVaRrJ5BqlkUq9nclSRHyKDOUQTfHAE4H9zs8R4vQMggh8/SmDTRNI0jT8P8rcdTVBaLYVYYskiQxQqtJMYE8+XL6xnDfkJVG6SQJDEsK8NsUVOsNuvvPcU/oKOjo6Pj6TxtZOdZnFfZwMOmu89qClw0DqSVACEY9mLyLGJRNHgHHkcSa/AeJQWNCWvR3ckSu1p/lITJsiSNI8rKsKha1voJvTRCacXaIMV7wagHSRyFUS48aSw5nFTIGWyvZStft4J+GtNPwrV5PNNlg3Oe1jmMg7VBTJYorPP8/vd2+Mlv99k9KpH4YJhsPcZ6IgV3DhcvFf+dJ2HNPjgpw9hDoilrS1EZ5kXz2Ndf1GDbOy46lU1HxzeMPNHBn9I4Wmtx/uJmTqRCc0ZJycYwoW4tv755zOYg5urWACCMWUWSo9Zyab1HUTcczw0//fQw+KzNK+7sLVhW7VnEuVQC7z2bw4w4UiilMCYk/zXGkcYaJR0egZDBUsHYsEYMejEAdWPZGKVY75BC4rwPJs3dHr6j46V4Iw2dk5MT/pP/5D/h9u3bxHHM9evX+U//0/+U9fV1fvrTn/LHf/zH1HXN1atX+c//8/+cjY3gcP6mP/dN43TD7Zzns3sz7h4sQvEVgnnZ8tmdKVVrSeIQ/WoMTBcNn96dUjWWeRlONNvVxtQIaG140UjCQqFV+I+qtihpcR7iSCGj4HFgreNoWmGNwziHEpI4UayNUqJlw6I0GGuJtKSXaQa9BOk9vTzm/uESFUmssSxKQxZp1ocJe8dLNkYZ718JNw03d2fc2p0DcGVl7PY28yTV1Jv6/o6O18m3td7vnRT85tYEQ6iRH10fP9M/5U3zpNrxMqqMi1Q2z+u/088jfvrxAc7DlcLgnSVPY+o2jEBlsWJZCpI43AgsqxaBY3OUs6zasKasbmA8YBzYBoxpyZKIUS8m0pLGWIrS8L3rY+bLQ5z3LMrQDGlbe3aqfTSraVtLmmiUENw7WrI+SCirlrt7c07mNVpLdsYZVRMOPOJIMivCONW/+1cu88//1S0QIaUlikOjSUqoW/NSRtM76zmf/GyCkpJYS1rj8Xh21jLuHSxYzx+u+V+mwdbR8TJ8W2v9m+bRut3PI5ZVy6gfEylJUVkkjxshWweRhKo1lFVEFAkiDf/zX97n3/9rEevDlDRW3DmYo1XwwdEqWDHcmy/5Nx/v00sj7p8sKMvgiyklmMZRGYdp5kRxjLEe58PvNA6MczgXxmWNtRhruXlvxtWtHjcuDfBwdpDQWE9VGbJEMcjibj/b0fGSvBFdmxCC/+g/+o/40z/9U/7pP/2nvPPOO/zjf/yPcc7xH//H/zF//Md/zJ/+6Z/y4x//mH/8j/8xwBv/3DeRojZY6/j07pTffnHC/eOCL/bm/PSTQ6bLmtp6lAC8oKgNWiuk8OwdL6nqMBKVaEUWKaSUREqEF4wAqcJiIYQg1hLvQyEPkeSe2bLlcFKyf7xkMq/p9TTDPKasDWUVHPAb41AqSEWVkmyMMn7vu1tc2erTGsuiapHegRdEWrI2TGiNozX+oVNH5z3XLw/4zjsjokjy2b3n93J40yzKll/ePOL2/pz7x0tu78/55c2j577e01NY4xz9TGOce6sfb8e3j29jvd87KfizX9ynMZbNUUpjLH/2i/vsnRTP/uY3xLxoXmntOFXZnOd5/HdOb0521nL6qWa6rNk7LrlxecD6IGXci+nnMUoIskQzyGP6mSbPEtb6SVDyXIAnKHfK2pDEkqa1DPKg9oy1QoiwJs6KhnnZcn9SsndS0Ms0N3b6IASf3Z1SNYZequhlEZN5jbEeax1l1bKsDKNBglKCZWXop5r1QcKon7K93qdpDEezkpNpRdM6WuvppZpYywt9IZ7mQdTPItaGGUkUbthOfY5G/fjCv1lRG2L98JbuSb+3o+NV8G2s9aecvnf/9a/3Xsg/7EV//r/69T7/88/uMS+bs7q9d1yws56zOczI0zjYIDzy/XZlj+AAJT3LukFJweZ6RtkYfnnzmKJq6WcRx7OaybLmeFaye1Tw2b0Z4JkuGr44WFA3DikBEbx6tFZ4B0UDWgSFpFrdEwBY68iz0EBqW8+0aDHeoSPBF/tLBA+8v/7697b5mz+6wl/7/s5DHmAdHR0vxhtp6IzHY/7gD/7g7L9/9KMfce/ePX7+85+TJAk//vGPAfi7f/fv8t/9d/8dwBv/3DeRPNHc3luwe1Swe1hSNUEueTKvmMwbWmMxzmOcw7SeRdFgrOPwpKKXavI05t2dPlvrOZGWRJEiSwTCB3l8nkiyLMI6QRwJlAxGlk3raI1jVhgcgp21lLbxVCZI5yMNh5OSxljiWHF1q08v1Yz7SfDh8WIVsagAxeY448cf7fD9G+v0Uv1Q0b/IXPMi48i3hZu7Mw6nFUKEv48QcDituLk7e67v/7o93o5vH9/Gev+bWxN6aUSeaqSU5Kmml0b85tbktfy+l+HeweKV1o4XNd095bSGjQcJ7+wM+PCdNWIt+IuPD7l3uGR7LSePNVe3+4z7MaNezFo/4dpmzq39OdN5iBy/qK0TVKMCD/zwg02GeULdhhugYS9CK4lWErynKg2m9SghcAKubvf5/vV1WgOXN/s47zmZ1wgJvSxitmy5c7jg3sGSyaLGWodWgpN5zc3dGbNVjHoaK4QMxs1Na9gYZjTGIeCh5s3eSfHMBtvGIGF7PefDayOubT8wrb7opudlG2wdHS/Lt7HWw8MHa8M8fuUHa+d/fl23KCk5OAl7+CRSOO/5za0JjbG01tHLFEn8uCmyArwArRVZrJmXLZ/fm5NGKviA7S/RSrK9luNsaADdP1owL2uWtaUxDmMd1oAlNGpaEzwvT712okgx7sdkiUYpQawh0opxP6WfxfTyiI1hyjCP0VJxPC8puyZzR8cr542v9M45/uv/+r/mb//tv83u7i5Xrlw5+9z6+jrOOSaTyRv/3Hg8fu7HsLHx9sx3bq3mYC8i7SX8f/7yHncP59TGUjYej6BtLVkSkaxO7yyEdr4SSC2R3nMwLUliTVFbxoOEyaKmaix5FiNViB5JVqeBcSyxFkxlEQKQoIWgbj3GWU4WLc550lQzGiUcHBX0e8Fb5+C4RKQRqVJMFjW7xwVb44zWeT66PqaxjvevjBnmMU1rmSxqfvDdbQZ5mMO9fVSwkccI8WAl894zK5qnPjdfBVtbA2a/2mNnox88IVbUjWFWm+e63rfx8b5tz/Or4pv6uN4kX/d6/7y13gDbGz2kDDVxOMjo9xyH0+qteR3d/vUeO5uDV1Y7toDNzT73DhYsypatUcSVrf5ZbX7idZyrYcuq5db9GcNBRmsdWgo+35+ze1wQRZJeHoPzzIqWyDr6uUZryaJs8azGmgRn41exhkEvYr5sSVLNhsppWsvnuzOOJhWNdQz7wUuisRXgKRqLLA1VbUJK1azh9z7a4mjWcHmzx6xoWdYlw168+nwNCNJY4YXgb3ywxZ//ep+isSgpiVIVUhx9SHEcDhO8lOhY0R8kbKwr6tbyyZ0JVzb7rA/Ts+emagyNf1B70l7Cb26dkK4acXVrEY196HmeFw33DhYYITicN1zZ7LE2SM++9qPra8/8m7xK3pbX+6O8rdcFb/e1PS9f91oPz1/vj28ds7M1IF3t4y5tDR97734Zzv/8k2XLeBTRGkfjBKM0pjguQQl0pOhnEdY5wGCMfchLR8gwctVaSCIfkmitJcsz+r2Y69fGbI0zZKSZV7uhmVMZ2tZTN23YzyOwgLDghMM7gbUhlUoCRWOCAhKxakwLEi1CM8h4skwQRYo4jlgb5VzdGdAiXuh5+jq+P75u19xd7+vndV/zG2/o/KN/9I/I85y///f/Pv/9f//fv+lf/0o4OlrgLnIge8NsbQ04OJg/9vFTSfvxrOKL+zMWixalJW3jcAKct9StQGuJthJrQzpVIiTWeJzwHM+qkGglBf1EE0uJ0x6sJ5KS1lmcE2SxxDWCog4+OAhorcN5j5LQmuBZYIxjWTZMZhXL2lA0LVoILIK6MaSR5N5hg9YSZxxbaxl4j20cX9ybsDnOUEqyM8qoljXVMiSStFXLXlE/ZK5ZtxYt5YXPzZOeq6IOY2Blbc7STq5s9rhxefhKJKCnf6vFvKaJJPW5623aoJx6nuv9so/3VfOk1+DXna/z45JSvDVN5697vX/eWq+B/aMleaoZDjJm85KiChvdt+V11M8i9g7nX7p2XOTD8+5GUOWcr81Poq1abp4sOZpWfPzFBKkk1lr6WcTGMOPXN4+w1oUbsnnDfFmv0gsVo17MMFXsyvCcO/egmaNlGPk1xiOF4F/+bBcpBWmimc1rFssGqQVVbUgThZaSojQc1CVHskRKyCLFeJhyb3/GJ19MWBvl1HWDQmBduIlBgrOOuhV46/ntzWN2D+YkSgST0jb4SYx6CqUExbLFZIYokpRFTbl6HmbzCtcahHuwyfPesyjNQ/44G72IveOC/XPP9yCPw3pyzgg51pJeJPntzWPWBwnrw5Sd9fy5/iavire1br6t1wVf/trelnr/da/18Pz1/u79WUigKwRr4x4nk+WF792X5fzPXy5qPrszYb6saA1sjVP6eUwaKT6/N6VpDFVjqSrLeYHeadveWKC11DKMS+VpzMm0pKlalIft9YyysQxSza+mFXVtMDaMUHnE2RiHX/3DOI9zDzw0i6qlEi3OgdKCPNKkK1VgWbfUrSWWkq1RSi+RNG1L0/jnfs2/ze/dJ/F1u+buel8/r+Kan1Xr32hD50/+5E+4desW/+V/+V8ipeTy5cvcu3fv7PPHx8dIKRmPx2/8c98Uzm/wqtYSKYnUEq0EST9iXrY4E2LH1/oZ6wPHvaMC5zxJrGlaQ1G7s4KdaM+ybtFKEmtNP9MoHeZiq8ZStY61fhwKdxMaQ6f/kxKEO53jFZjGYr1FK6hKA1IwzIOkflo0pLHmnZ0+i1V04rI0tMYyLxw76zl5EnHj8vChx7uznvPZvSnACyddnX+ulITffjHhaFZxbbNPEitu789ZVoYfvLf+yuZ6r2z2uL0/RwhBpAWt8RS14d3t5+vcfpnH29HxJvk21fuPro/5s1/cB6DfcxRVUHv88Advjynnla0+d3a/XO34MlHlp/TziJ/8dp+iCvW9qYM6JomCStM5qFvHbNmEwwHrsM6zVIJICdaGKevDhMm0Rung26BViDbXSjKd18jVYUIchfE3KWRoyBhPI1cRu84HdSrhFNs4WFaWKDb8/LMj0lhzeFKwrFqMCR5zsVYM8ghroZeGf7+9NyeJFZN5CARYHwYPpaaxXN7I+eDq6EJ/m34WsSgeHtG4aETqaabVjxohjwcJ2SpN8oPOCLnjDfFtqvVwcUrfqxxvFITx/KJx3D9cMl1UZIkmiQX3jgrUSUkcSY4nFdZ5jPFn/jWnnJrFQ9iLezxaayKt0MJTtQ4L3N5fMu7HaCVYG2YsK4MXDi0V4FdhJuEHCgm+feDX4z1ID1aCsxApQRJrqsoSxYpeGmOspawNh9OSrXGGdf6597sdHR3Pzxvx0AH4L/6L/4Kf//zn/JN/8k+I4yD//d3f/V2qquLP//zPAfhv/pv/hv/gP/gPvpLPfVM4v8GbLZsgsXSOk3nDojI455FC0Us1o55G6yDZXO8nqxMGi3UPCnZtoKo9ZWUR3tNax6IwWGeJtcI7z7AXnPItwYDt9HudC+bJAoGzLpwUwCrKUCIArSXzoiXRoaninCeNNZGSZ3O2xnkOphVf7M24uTt7zDjy/SsjtAzpKlrK5765OP9cHc9rWuMYZDFlY4kjRS+NKKrmlfrT3Lg8ZGOY4ZxjWRqcc2wMs8caVU/iyzzebzJPMxftePN82+r9zlrOH/7gErFWHE4rYq34wx9ceqtSrgZ5fGHtAJ77vfMqPLwOTkrq1nI0q7i9v+BgUtEay+5Rwa9vnbB7UnA0LQGPc8FY2FqIlGRetRxMSoZpRD+PWRukjPsxHo8x4Wt1JNFa0dqQ1DiZNyzKBuc9Ze2YLhpmiwprQzKMlmGt8h50JPHeUbWOtnWrUQaP84L5ssVaTy8LNz9VE+6WjPEMsxglBc57jLHgPXXrSGO1Oox43N9mkEdIKV7Yg+g8L2qE3NXJjlfNt63Ww8v7h13Eo+/JvZOCog4Js0XZYKxFSInzkCUR3sHtgwWf3ZvhBEyXNa2xuKf8Du8gizWRCgrFybLFWEtdG3pZxNG05PbekqYxGOvAhbAR532okSoodmKlUSrcOApCM914wEEcCXSssM6D8AhgOIgY9BIG/ZiiNkyXDZuj9Ln3u98Uurrb8SYQ3vvXPjv08ccf80d/9EfcuHGDNA3z4teuXeOf/JN/wk9+8hP+4T/8hw9FDW5ubgK88c89L2/zyNXPPjsKUs3a8q9/s8cXe0uqpmVRGlprkYRRq3dXqR77x2HjXDWGycI85pR/Hilg2NMM85TGGMraYK0nSzXzoqZqHv+eOIIs0sSr+X9jHMZ5+smqAeSgl2h+8P4ax/MapSPWepqysZSVIU81SgrevTQAHyTpp5HlX7aJcfpcCSH45M6Eo2lFGivq1nFtuw94lqXh0kaPH77/5U7az/+tvkmx42+D9PHRsYNT5cGXeY28DY/rZfmqJfjfpHr/MrX+bX3tXHRdL/reOV8zTzkdNXieGrkoW/7Hn9zh7v6ce4dLFmWD9QIhPM56kkjTOkOzapYoJXAWjPX4VZpKWMNCfG6INffgPUKGmxXnPFJK6ib4SWgV4nOzWLGs7Fm8LhKcWY0WiFVDR0m0FERasTnKSFPNsqipW0fThkb/IIuII4XzsDFKSLSkMcEvrqwsy7IFAZfXcz66scFf/972E5/nnfWcRdG+8Fpw+rf89G4wTr1ojO5Rhc7rqJNPuq63jbf1uuDrPXL1Tar18GL1/nQfF6URbdW+1D7uovfkzd0ZO2s5Saz4i08OOZpWxJFk1E8YZDE/+e0ex/M61CohOJxVtPbJv+O0UidxSKod9RK0Cv+uI8W7231u7s5ZVC3OhYZ3u5L7aCVYBeFiWo9YKXHsI6laSoVxVYEgSRSxFqz1Uwb9mH4Wsywbqtry0btr/G9+ePmFnqe3+b37JB7d77/uuvtl+bo9x1+364U3M3L1Rho63zTe5obOp3enq3jaOXcOZxRlmK9dlm1o6EjB5jAjSzW37s+JtWTcTzie1xzPL+jIPIIkNHWSSNNaG1z2W8fpgeBKmYkgFHmAjVFK3Rjq1hNHgkiF012tQArJKI/YXu+RRJLRMGexrLh7sCCLNWmizjwDwFNUls1xymzRsjZMvlRD5Pxm+M7Bgt3DJd4H2fz2WkZrgo/DuzvDLy1f/zoWoOfhbXhcL3JT87y8DY/rZfmqGzrfJL7pDZ0Xfe982ffap3enfPzFhP/pZ3eZl4a2tTQmeM5EMhjsp7GmaS1lY4NnzSp+F86tLXI10isg1posVpwsa6wJX6PUahxABENQAfRSDcKjpKCfRkyXdUhwceGk2fuVybKCUT/h0nqKQGG9I4kkZWWYLVu0ViRacGWjz8Y4ZTRI+GJvzrifMO7HtDaoe7bWMgZZfPa8LMqWm7sz7h0ugWd7tD2t8X/myfYCNwuvo04+ytfptf+28HVu6HzTeNP1/uefHXEwLbHWkSaajWHKzftz+qnmnZ0Bd/YX3D8qQATT6UVp2D8pqGpD0ViKylDUT+nmrEh18AADyHsRV9ZzRnlCax3TZR2a2VpwNKkpa0PZWKQITaA00pzMG+Io+PGYJ0iBskTSTyKSWDEvG0Z5wpWtHoNejHOejWHCd66tPbPWPFr3fvDd7TfmAfaqOP+aeBN198vyNtfHi/i6XS98Az10Ol4/Ugp++vEBJ4uGprUY65gVDQJIohCNOlk0FJVBiDDyJKSgtc9eFCCMVM3LoORpWkfTPlzdPWGzfXos4B3ksUZ4T9M2VBXoTFBbTy/RpKkkzyKqxjDq54wHMTjLqB9zab2HdQ8iWlsTZJz7JyXWet7Z6b2Uh8Mp5/1o1gcJh5MyeOj0+jStpahD5OzLyGg73hxFbehnD5eyWIexko6Ojgc8ulk+mtdsjZKHvuZp753TmlnWhnnRsihbpIAffWfrwp//aLP9aF5z92DB4bTCrH7F6QrSuvCPcU/jPSzLi8cI5Mrv5uyT3lA25mykVyuCob/nzFdCrTwkvPP0s4hLGznTog0+ZvWqccSqYeQJsb6NwwlPJDyTyjDsxVzrhXGraWHYGKf89d/ZYWctZ++k4KcfHzArWvpZxHgtQQrx2NrhvOf65cFZ8+VJa9fzehWdjuDuHRcsyvCcX73Sv3AtLGqDFJzFH6exYm2Q0PinDWt0dHS8DhZly2e7M4Y9TZZoWuu4s78gUg/q7/ow5WBScOdgiZIwWTQ4Fzwpgedq5pziPSSJItJBXdi6MHJalC1RpChrx7JqsPZBQ6s1DudblAIhJMY9uVY0jcPFjjxLibRkNIhoW8fxrGJzmLE2ePZe+qK695tbJ2z0ordGzfKidPvTjjdF19D5BrF3UvAvfn6fqrFMFg2TeY11jlhJamdpjENKgYg8i6olTzSCMINblM+/MFhH8NF5wkGGc6B1aOb0c83mKGVZhQj0WK8S0qVAKMmwF+O9Z9xP6CXhZFYIwd/60TXmRcPRrKK1FoGkMWFkTApBnuszDwcI3g5PMo58Euc3w03r+PDqmKubIeWqaR3vbg9eWcpVx+vjdRsUdnR8EwjKzYc3y/ePFhxPS7QSZyfEUoonvnf6WcTOes5PPz7EOU8/0wzy+MxD59Rj56ImxKJsuX+4ZPekQAoubNY4F1Q1jXHBp+FcJPkpj/YfHrOLcSA12PbB7zAObGWJ9CpV0FjyWIdob2w4iBAQaUEv1WglKWobRr4II19h5DghiRTvX064ttNnUbTsrAUPpb/xwytPbWY9amD8tLXrRb72aabJ5xHArb05vTQiTxWt8dzam3cGpR0dXwF7xwWDTCOQCCGIdXiPV3U4iP383pSiMpzMG5QQOOepakfVtrTWURQv0BCQAuU9TWsZZHH4XT6k8nkP87JlsWxo7IMRLS9CamBTh1Gr1rozBf5FuNUY1tXNHhvDlCRWHEwKjAsHps8TLnJR3UtXHm0vur9/W+j2px1viu4V9Q1hUbb8y1/ssX9SUFYtdd2iJbStZW5sKNKrE8tYh+IrVw0RpcQTmzMXIXh8k32esDkW5L2I7bUMYz3DfswPBxsczyomi4Zr2z0GeczJrGY8SDDWMysa/vA728znJc553r8yQgrBZ7szBpnmykaP2/sLlIT1Ye/s932ZbvfzboY73l665K+Ojmdz72Dx0GbZrZKeDiYl72z3Mdby+e6MzVHK79x4sh/Oomi5cXnwmIT8N7cm7GxkD/38o1nF/aOCD66OKGtDHAm8FzzpoDeKoLGOujVYwlrz6E3Es5Yq5wH74OvUuaaQtUHpuXtYMsgjqtaQp4osjcCF5Je6tYiq5Xff38A4x+d3ZywrQ68ymL5DCsHaMMEYx62D+UMNnKdJ6J/3pPbUQFNIyFZNtizRr+xU1+PxXuCf+Ux2dHS8LorasL2Wc+8ojGBGWsCqubKzFtKgZkVLpCV1KxgkEVXj2L+7xNqn78EfRXoQK/N0KeHqZsa8tBxM5xSNoWmD8fz5WmsdZ+NnitU46gUNdkFIv5IC8iRiPIi5cXlInmo+uDo681h7VjPnSXUviRT7TzB5/zrQ7U873hRdQ+cbwKJs+V9+tsvHd04oa7tymQfrPI5gYgaQRhJwoZPuYFm1FI1BvKCN0rO+WhCk5XmiWRukpJFiumyItGB7LSdLojOjtINpRZoqRr2YqrHc2p0xyhRNG0atfvf9DW5cHp6dfPaSiGEvIk8fvHTPd7u/SYbDHc/Hi4wddHR8W1mU7UOJSEezirVBTKJD3HfVWLJYIYVg77jg8yfU0CeN7kyWNe/shEZ7WZswPhAJhIB52fDTjw+wxhMrj5TAI6JQKcJaNS+CtEaeHhWvDItPeeb6I8L3GlY3ImL1/6ubDiVhrR/jV+0irVVIuXLhRLjGoyRkiWJeedYGKVAxLwyjyrC+lXDvsMBYy6ifnKmRfnnziDyJ8HDh8/Y8J7WnIweRFgghMKsxjGvb/acqp54HD9y4NOB4Hnwy0kRz49Lgsbjjji9Ptw/55nP6N759VLyUKXKeaOZFg0Swu/LVGvcTxnnMznpOEimKquXO3oLDSYFHkKeaLNZMX0SdA1jviKSkl8Rkacwvbk5wLih28B4hPOIC+c2pZ5ld1eOL7IXkqt5qJehnEZFST9yfP4mn1b2stV9rNUu3P+14U7yx2PKO18NpIfziYAFAVRtO5hWVCTJyKQQrX0lqE5o5ZtXdNxa89Y9L1r8kaRQ2w1LCp3emHM4r8lSRxhoEXNvuhQ09krVBgvdwNK1x3nMyq/jJx4eczOqzaL9+FvHB1RE/fH+D3/9o64lRr6fPhXGOfhb8gj6710UEfhs4/xr54Orbkx7Q0fG20M+ih6Kzq9qAF8EbZrvPh9dG4cT4cPnEGno6NvUvf3mf+8cFUoaDg1t7c7JInf38o1mF83A4CQqdX908wVrP9FTWLx7fejgfInDTWCHVg5Ng8dhXPh0pQwNICYhi8eBGJYRh4RDEkSRNJNvjFEloMKWJIlo1vJSQ3L4/Z7FswXs2RzH9XLM+TDHWsyjD+nRpPUesxiEOpxUHk+KJa8/zRB2fjhzsrPdWIw6CKBLcPy5eOhb5lDzRKCW5ttXnw2tjrm31UUp+rW+W3ka6fcg3n/N/42Eev9TfuJ9H3Nqb44F3dnpsjjNqY0nSoMYr62DCPlk0SBn28fNl81DC4LNYhflhPbStQymBkp6mcfTzKBzq1pa6cSvV/uM/4/T+4dFkq1OcBy0FWRIhJHyxt2CyqF8ozv1pda/6knXvbaDbn3a8CbqV/GvEo6c+aS9h77jAuRCvnSWaMg5x4q21OAfG+zMZ5ZsK5hJS4j3UrSNNFHmskFIwXTa8s92nqg1lbTmeVfQyzWReI6XAWcewn2KNZdiLXtgE8tO70+f2Hejo6Oj4NnFlq8+d3QfSb6UkRW24tJ5zZ39B1ViOZzXDXvRYDb25OwPgs90Zs2WDFALrHPvHBeujDIDxIKFuguxmMq+ZLGpa42gaS2ksy6JFylPfs4tlId7BxjBhWYbExRcZK4DVabKFOIJISIx1DyJ9XTBMxnnq1uGBcS8mSxV17SlbQxZreqliURrSSLGxlnH7/oyq9VzeyIkjwWwZzJTHg4Rs1Qw5mlUoCXsnJdZ50kTTz6KH1p7nOak9HcsSkeDqZp/jWUVZe7z3XzrmtpP+vxlexP+o4+vJ+b/xy3o5LoqW65cGLMr2TDHXzzV395Ysy5aybunn4f1e1g4hYFm2NOb5/S4Fwc9SCkkcC6bLGiUFVdOye9iCeNDwdvBs+eMFeCBPIzbHCcM8mMbf3ltweT0nSzRSCD7fnT1Vqfa0uvfR9bWvXcpVR8dXQdfQ+ZrwJPf3o3nNbF6hleB4VtEYS9UarPNoJYiUwJknT8u3LyG3Pj1bvTCBBFBKIISkagxbo5TeaiO7KFsWRcPxoqGXaMb9GIA4V0wWNfUqdeva1iDcHLT2hUwgn+RRcDipzj7fyZ87vg50kv2OV80gjx9qKPTTiKNpxU8/PmSQa4Z5wqJsyFJFWZuzZoVZmRsPejFJJJgXTTh5XY0PLcqG7727hnXw3mo8drZszg4Q5mVLVQcT/RgYj2I+u/fAk0HIoBYVQKLCyFXzlHuWpxlzRhKUBq0UzrmHfo4kmC577/Eerl8eUFQG5yEdCNaI6OUx+yclW4kmjhVFbYi0ItaOPI1YG2RsjgStsWGWa8V0UTOZ1yTRg8Sa/ZOStnUP+eo8y7Pt/FhWnmrytH8Wcftl3/+d9P/N8DKpNl29/3rxKpKLitow6sWM+yFlsKwNX+wv0FqiJCxKQ6wscSQ5OawRziOkwDwpN/wRNKAjgfMe6xz1qicyL1uqxiCERK8Uic/fIgoIQnPcWlCKoPRLY3pZxNY4J44kWaJx3pPE6pmpfk+re4M87ho6Hd8oXle97xo6XxOe5P4+nVV8cTCnlyn2TjyLyp7NvbZtcKd/VUiCnF0pgURQXtANckDdWHqpRCqFkJJlaahbSy/VWOfZGcFokKCUYP+4QOtwkhpHkrVhymg1f3u6QD7vi/8ij4LpsuF4XjPsx09MXuk2Uh1vE88bWdzR8aKcNhROX2P9LMJ7R2vgeFFxaT0njhRHs4prW0G58cX+nKq1HNydMi8bposmjFatlCiDPCZPIjaG6VktPU2+qhqDUgLjfFBhEnT764OUo1kVTrdjRVmZ0GyRgsmyeWLT5vRGon3CHYjzkGjNINHMa0OsTFgLV18faQECvvPOiB+8t8FffHJEL3PcO1pS1ZasCuvA9Z0Baaw5WrakkaI1djUuZdge50yXntmy5le3TuhnESfzGus8l0fpWWKNMZ5yFSv8vOvM61bRdCEAr58XTbW5KH2uq/dvN68iuejRn3E0q5BCsDFKWR+mLArL/qTgaFazMUiYFw1VbZ7a7AZINQgVmkJCSOom1FYpQ00ta0NtQEmHeYqBVnDcfJzThyxOkwodlFVQX26MMvqZpqgN9w6XXD9nnv80FVOnHuz4tvA69/ddQ+ct49GNXz+PWBQtv7h5zNogIU80ZW2YLhuM9/zy00MWlQmbzcZgHRjnEae74XNzr0872XwaitCFjyOFdS6cqgqBWp2qSinCjK9z1Cb8jlE/RkvJvGhQUvDLz49X87uCjWFK1bQM8oSttYzjWY1zECnF9UtDmirI7RsTYhKf98V/0aKwd1yws5ZfuKiw+vpuI9XxNtFJ9ju+LBc1ELbOff70NaaV4MpmHyEErXEY4yialvvHS8raYK3ns3tzPrg65OCk4GBSIXyQ7DfGs6wMjXF8cnfC0SzlBzfWGfVjIi24d7QkVgIpJYiwfsRacjStUEqSxook1owGKftHC+alxTmPFP6J48GhoSNoz81iyZVpslzFjq/1E5xzaCWwSiKUgNaitWKQRyuTT09rHd+5NuK3d09II00/0cjVmrWsDO9dGfH9D3M+u33M0bTCGMd4kKGVJI01o17MvGhYFC1VY1nrJygZVtnWeJx3pGnyQhu4TkXz9edFb04fTZ/r6v3bz/m/8XmvmBdpQDz6OpkX7VmCa55qvnd9zP3jBWks2dnI+c3tKbV5sgJIEAzfo1jTNobKgCTEkksZTOK1hKYNtfNZZuiPfvrsbFhApCTOebT0JJEkTzVFZbh/tESKHnGkaJ17yIT/9HFepGLq6l7Ht4XXub/vGjpvEY9u/KaLhp9/fsT1SwPW+jGzRcOnd6dsDFMmi5rDWcXepELgmS8NjlDUIx2KsYSHHCVf1kLHiyCJd9iwQRcSa92Z830aKSIlMc6BsGSJZtRLAHE2AlbXlvEgYWujR55GlLXFmOBjcGk959pWj+kyNH+892ebILk6wX2eF/9Fi8LaMGO0Gu065XRR6W6cO95GXoWcu+Pby5MaCJub/bNGz+kBgRCC1jpirYi0YFFajPHEWoOH2bJFS8HRtOL+SYl1YZVpjUdrifCetnWriFvHomxJYoWxHikEkZb0sojGOmbLmmx1Kh18FzSthcm8wloXFKCwMv28eLVygHnEWOe0+eN8GNnFC/pZzKIyKC0wFrSSWBtGj0d5zHDlP7cxTFksw2FI6SCKwujAsmzZO15yZWfI1lrGsBefNWA+vTsliiRJpBgPwrhEmijq2qKkoKhC8tf2OD9TKj0aF38+zv1UrfNoE+69y8Mn3tDMi7AX6JSlbx8venP6aPocdPX+bef833hWNGgpX7gB8ejr5NEE1zzVaCnppRHWgHtGB8YTDl+TSNHUFiX8A2N5Ad6AcS93H3AaTa7k6aiswEtQPkSM97MYAdw7XDJbNlzd7DHqJUwXzVmNhKermDr1YMe3gde5v+8aOm8RpwbHp3GwR5OS2lp+fcswzGM+253RGsfdgwXOeWZFQ10bwJ910z3BWBIgWilqlPCYl6jipzGFqxR0rAWdKJQAh0c7iFYnksaFTXqiQ2rG5igLG5VIcLJouLTe44NrI9RqBtjFIVnkxuXhWfPmvSsjotUL+3QT9Pnu7IU2O48tCnenT5TGdjfOHW8jr0LO3fHm+bIxtq+KJzWqP/nihOOT8Lm1fkzdWKom1Lp+DnhYlg39POajd4fkqeaXnx+zqCSTZY0AYq04mlY458gSTZ4oQCCF597hkumiRauQnHV1q8edwyWtbcgSRS/RLCtDP9FBiSMjFlVLpBXGWCLlqG04IKjaJ6fF5KlmtjSP3Zh4oKgsSdwyWVbksSaKE6bL0DCKtcZbz7ifMlvU9POYW/fnLKoapRXSueBRYR3OSVrjz27WxuvJWZT77uGSq1u9h96fO2sZn9ydsbWWPaTK2FnP+Xx3drbOPBrnfpqOM8hjfvtFiBLuZxprY5ZVe6GKZ1G23DoszlKUXpWytBs/fnW8yM1pP4s4KOqu3n/NOP0bb20NODiYf6mfAQ8a8ZN5HZR/pcFYx+Y4o6hbmmc0dBINcRxR1Tb4khnOZDXOPJ/p8UXR5FqdHhRLLq/n2NW1LpYNvSw0c5aVYbKssdaTxYrv31intY5b98PzMurHTBcNeycl66sGT1dfOr6NvM79fbdivEUczWsm85Ik0kgBe5MCYy1V47DOU1aGJFZUtaVoGorKBd+BR37OqbWNqS1aQRxJbOPOvk6J50sPOS3sSqzknDpsQovG4i0ksSTRKvgFOE8cKdJEMMxT0kRyPLcsypbWOoa9ULgjLWgN3Lg04M7+8rETrK2tAev5gyL/ZV/8T5M/7x0X3Y1zx1tHN0/+9eO8KmYjj9kr6q9sfPNJjerbe3M2+jFJpOhlEXcOJrTGEWuBVpKmtWRJxM5axvGs4t6h5f5RQaQlUZ4wX7QczkqkFGgVVJnL2rLWj1kUhmXVsDHKaI1lsqiJIkU/jcgThQAmRcvVrT5XNnuUdcuvbk3IYsXaMEMKz7JsyRD084jJ8skNHWcfmPwnGurz/XcPVWPxzrFsDH0lcBbwgiiWSCUZ9CPWBylxrPg3v51wPK0oqjCurLWgn0a4UcoHV0f81e/t8Pnt44cUT1EkubU358al4dlpulKS9y8P0VI+tqadX8OOZhVRJBEIsiSk45S14V/8fJedtZxBrmmN52BasjXKLlSL7h0X9Acp3ry8snRRttzcnXHvcAnA+jDBec+on3Tjx2+YR9Pnunr/7aSfReys5/z04wOcJ6ghBfzlp4eMewnqKWnlIYxEYq3Fe6hWY1Uv6nZ8vpkjBcQ6xJEHj0tBnsVcWstwzvOb2ydEkcYLTy/VLCtPogVCBQuGcZbAJZgtWhrrOZmV7KznjHox02XDJz+bsDbM2BgkXXOn41vD69zfd3eubxFVZVgWhjuLgvvHS2bLlrJpAMm4H9MaR1kbIi2pavfMGVgIo1Leu9BlF6ClxHtP2T6/ZOe0+aN8kLt774ljhdaSSCvevTRgWRvuHy5pWgE0+H2/8s+RbPaDtP9gUrDWT8lTjVKSD66OzhJAFmXLzz87YvarPRbzmiubPW5cHl744p8uavIk4mefHT3zJPGp8ufuxrnjLaSbJ//6sXdc4LznYFJysmwxrXkstvpN8aQmOAR15J29BTfvz6mNxblwWJCnMX/wg0scnJR8cneCWXnALOuWemG5stFjbRBzMCtxzmO9Z9m0eOc5njm8lGSRop9FlJXgcFZxvL/AOs+4FzMeZgyzmOs7fZJYoSR89O4aJ9OS2nrSKMTbLivD4bR66uM7HS2+6GDCAsYELx7lg0H/1c2caVFT1hbnHbf3FixLw+9/d4uiNBxNKqQKhxVt5ZkXLca6s8jgU8WTc567h0uKquVkXiMRfHR9/JAaZ1E83og6v4aVtSFSksZarm6GdWZetNSNJVupnSItVh9vUOrxVIOiNmysK8pzH3uasvQiX75b92ccTivyVIEXwdw5jxjk8UvHMHe8HI+mz3X1/uvB61BkLoqWG5eHLIqWn312hJKCzVHCwbSmauwTfTCjSGDtyrvrS13B6uedqnIiiZAQCUHrPFXdksR95kWLjhSVMeSxWil7QrJuokPjen2QMl82TJYNY2BnLWfcTyhrw8FJiZKSum4xvahrHnd8a3id+/uuofMWURvLx3dPsE4wmVeUdUtZexCWpjF472ktxDo0ap4X58JGN9bgvKM+t+d8kpP9RVjnEe6BX4JWIaL1ZFYxKwzWOd7Z7rOoLLf3FwxyzdYoI000dWtYFC17RyXffWdMpBU/eG8dCAvjL28ecTit2Nno00SS2/tzlpXhB++tP/TiF4D3giiSz4xCPOVJ8ufuxrnjbaWbJ/96cTyrOFnUxJFkNIg4qpoLY6vfBE86AdocpXxy+5jJoqE2hkgpjPcMejHTZc3/9Bf3WJYtXxwsSCJFlmjqxmKM5WRWc/dwSaIVMhMUpWFZGiItaFobfG9cxN5JgQDuHy/JIkWWaoQQ7J8s+f71dZSStNaRJpo01gzziK2NPr/+7IBf3DzCOod/xlzAsB/T2grX+iesgyGW3FhH3Rrun5RUdYuzkOeaugmHIpN5zaxssATVz/k04GVluHV/xvVraxS1QUm4e7AkiiTjfoJWgt2jgvVxOGE+Hcl6kvHx6TrjHXjpubrZP1P3LMqWcS8+8zKCoGSdLlqubD7Yop3eQO4eLqmtI19F/MKTlaUX+Sn99ONDvPfkqT77fUpKjPEPpZt148dvjq7ef714XYrMU3Xlrfsz8kQzX7aUtWOQx6QqjL4WzcP1MZIQS8G8fhWtnEBrg1+O91CWLQiPQGJcMMJPI0k/00wOKkzr6GURWRahhGCQR0wXDZNFzaJo8Z6Vf5pi3Es5mJYhHGWUUreuax53fOt4XfW+a+i8RRyeFHgE86qhaixFHba2woMx/kw9WZuHvI6fyWmMefPI3uxFmjkQXPGl9TgHReOxrmbciyjbMNrl0eydVAgBwnuaxnL3cImUEmMcaRIMMhEgxINFae+4YFkZ8lSTxJo6UgghKKqGveNgHHn64j9vSAlf3si420h1dHR8WcrasigbJgcNn9ydIQWM+zFp9NUssVIIbu0G/4Irmz3evzJiUoZO/nReEWuJlFBZz6JsGfZi7h4ucc6HtCvrmC4bslhTNObsc1X7cLvFWI8WEMcC5x0n0xovPGv9BCklkZYoAVkaGv9JJEMDY5BQVobb+wu2jgs+vzvBWo/Ukr5WLArzxLHgunZowUMKlbPHffZPh7WwtI5IOZQUSBVuRrI4JEV+fGdKUbYh0ne1oJ6egLfGcjit+OSLE/JEc3t/vjpEUKuvk+SJYjKr2BgkHJyUD/nfpbF6SKF1us6cNtuUemD+LwVc2eoxX6l7IiUpa4uUgp31HHj4BvLado+DecP+0ZLrOwO0lmfK0kfVOGVtHvNTcs4zWdS8mz9QomaJoqgN1bn5tW78uKPjYs77lL2sou0iz6pTdeXJvEFKz+7RgmrVMK9rQ2X8WY3SgNDhP5avsJlzinPQNJYoUsEfU8mzx/vZ/QWjXsSltZzGBH+yPAveaJFWzJc1dRuSBi9v9rh/XHDr/oJ63Z3Fqe8eFVxaC/Wtax53dHx5utX6LeJwWjGZV8wWwXfmXPL4Y6OwUj47dvA85/fGp071Hh7q6ISEkQcydi3Df+ODL48gGC5rGb4t0oplaSnb4Ki/rCzeO6IoFPbpsj3zHMjTCOOCJPPO/oLdo4Jf3zzh0kaPug0mbmv9B274kRYsS09RP1zkX6WRcWcC2dHR8SqojeE3t05orV95GTj2jgSD78fP/uZXyPkb/++8MzpT50DwR9hZy/jFZ0e0ztNPIrQWOA8Hk5KyalkUhtZa8ixmkEergwXz1BsGpUJjp2kNRW0QCPJUcXVjyDuX+rTGcXCy5IvDBVmiGfcT6sZx52BJLw3/vawMZdUyHqQkkeR80JVWwZDfE8as+pliWYVPPjqC4AgRAVEkaVfX7HxQ4AhCysuibImUoGwMrQ0NFURIiPGEU+nWek7mNf/zX9zjB9fXOJqWbI4ywLMoDPeOllxez8+MjX99+5gk0gx7MXmqaI1nf1LQWM8H567vIlXoj76zxd5xQZpo5kXLrGiRAn70na2z9egho+tI8Z1xj7/8zR5/8ckRm6OUK5s9llV7oUrog6ujh8bv+pnmaFY+pAjqpSFlSyn5UMpkN37c0fE4X3Yf+qQkwp31nFv3Z1RNi/OeqrU0bWjutjZ4ZmoVDnmdCMbHijAi1doXO6B9FlKs0nIFJEkUNv8+NLvzROE9pInGesMg1zjvubSekyWa+8cF437C5jgFwlhpWRvuHC65sTPAOc/5yt01jzs6vjzdO+gr5rSpcDwLsbBlZTA+FO6n8SLNnMfwYVFofWjOeB8aRN4/MEWTrD4GICDVAuEgz4PCxhiHlALnPd555nUbNtxSgPfUbTBns86RopDAvGyJlcR5QVG1Z02Uw0nJrGgY9ROubnuU8CgZZoJPZvVDXjlP8ocQ8EIxrk9aULs53o6Ojhdl96AIJ5JaonW4Ka5bx+5B8Uav47yXT1Ub0kSfKUWSLOL2/oJ+FjFZ1LTWcbJoEMCsNAjvqNvgneNpKMoWrSWLxZMNiiEoRmMdTnTxgPS0jWeyqBguIpzzaCnZHGRsjjNmy4bjaUWiwwJzsqhQQhBrRWMsaSxDGqJ7vGmjFORpdBYGoGS4zzi/HHovcOduFlb3IWcfaa2jqA07azlZNGNeGfwjaiAp4GRWcWmzRxSFKPS7+0usDz4V22spaaJRMpzOO8/KFDqlaiyzZc1s2TJftPzO9bWH1pSLVKG9NPyNlJRc2eg9tn5ddAMppWBtkJw17n768SE769lDapxBL2b/pOC97MHvG+Qxw15MURlIPXiBcY53dwYhxr0bP+7oeCp5opkuGxZly/1JdeaZNsier4H/pCTCRdGSJxHXtgb8/LMjJJJeKllWBmPDvt3Zh+uZBfwrbuZA+AXeh/311prGGI8QwXR+0IuJI43EI0UFQpBEmt+5sc6Ny0P+xc92iSOJB/ZX6b3rg4R5aZgXLZGW7KxlOMK9Qtc87uj48nQNna+QRdnyk9/uc/+4YO+opG4My9LieQ3F+RzWhxhDIUJSVWsdAoGSHufDaNap8WQWa7wE2zp0HNI5nHUoIVabbksaKZZ1Cx6M8wjp8S6cGhjjabXD15ZeopmXhiyNUELSzyKmywZEkOsvq5aybpkvKrRSSAWXNvPHTjD2jsNN0nmT5FNfnedtzjxpQe3meDs6Ol6Uw1lJGgdPkzSJqeqGojIczi4aDHp9nPfyyRJNa92Zl89ICI5nFYNegvVQ1YbJosZ6vzoxhboJHmlN65AyNDbM8yQiurDZP/WYWVYNTWv5xaoZb5xnNIj5/N6U8SBhWjRIPLV19KoY6zx1a3DoMCIsQxT66bxwrE+jc4NvmxICQ1B8okLU+OkhR5YocJ6KYCKqVPgxbnXT453n2laf799Y5y8/PWJWPX6qHoyJPXEcYa1Ha0HVet7d6XH3oGC6rBFC8P6VIRCi1A8nFfOi4WRWByNRLdBaPtdBwbNGfx89yDg4KZFCkOf6bOTDuWDoPD6ndD2NU69be7ZeSin4g98JBtgh5crz7vaAG5eHXQOno+M56OcRP//8iF4asd1P2V/UHE0r/vAHly78+kfV4Efzmq1R8tDXnFf4/JUPN3De88ubx0wXDUo9UNY7HjeEfy33C2cKfY8xHr0aufLOEamIVEtG/Zh/67tbIflQyjO/uCubPW7vzykqg9ahWV/WDTcuDdhZz1cHwtCa0OzvmscdHV+erqHzFfKrm8f88uYxguAXA/5FUwZfitMTT++hatxqcfC0PiwUvVRSNw7noWotWaxpCdLyTGvSOAk+OlWLFyGiME80VW1wHkwbFDOthTiCfi8425cr6X/TWrQSDHsx+5OKRAukkGDDaYD3wXT5o+vrZ5vT8ycYj0rW8yR6YV+dVzm61dHR8e1GK0VrDLuHJY1bEkvJeKDRSj37m18hZW2RK7ULQKwVxnjK2tJOK65t9ilXaSl1ElQhVdkghcB6f9YUMQ4iAe1zlkPngqJTSUEvj9BKYKynmNdYTzD9bR37VcnhNDSdtBLkSYwg1OxlHZJSrPFEStC0EEchUt25cMSQaElZGiCMUDnnV6fED8aSI61QQlAbS9OGGyElBZX3sIomXxQtRWVo2otvhZzztNbTNIbdwyVZrJkXDfePCubLhjyLyFYNvLI2FJXBO8+d/QVKScaDZLU2aZJYfemDgkeNrqfLGucd68Pe2df0M82ifFhN9bQ49Z21nN99f+Olr6mj49vKomi5vjNgUbYUtSGNFZujlEXRsrP2yNdeoAY/mZXESpDEIRGqqg1KSbZGGdmqebu1lvGRXaNsDJ/cm1KWy7P6/CR/sVeKCEp9rRXGeSIN72z3UUpw/6Rka5xxdauHlOIxhc2Ny0OWleF4eoJWgixROB8xHiRoBUXluLrV6xTxHR2vkK6h8xXyy5vHtMbSS2OiOMzdvwlOf4tk5S9w7tdKFTbEIfXDB0WO98Ra4myIPQfopYrW2lUDxpPFGmMdQgiMsZzuk9NYIUUYy7LWkmUxCNga51jrWRQNUT+hl0aM+wlXt/sME8Wn92aMeg/LV08bLo+eZv7ssyNiLS/82ifxpNGtbo63o6PjRVkbxHx69wStFEkkqVvLvcOGH39v8EavI0011bylNY5IixBj6x1pmmC8J0kkg14MaxkAv7x1tPKBCarK80RKUjznbO/ZaG6sqBtLEinSWFC3ln4WsTZIOJiUzMuWSAU1SWMdVVMiCP+uhKSfRmytZXjv8YSmD0ja1oMUGG85mFU4vwoGEEHNg/UoIEkk33t3zGzZUjYG74O5smlXzRwFgyxBSMGf/3qfeVkTq5Aa+dDYliP4QaQxn9+foRX0s4TLGzlrg5bdoyLczFUtN+/PibUkW8s4mdVEWpDHGiFgfZi+koOCR713BnnEMHmQcgVhlKqozUNqnLqx3U1TR8crpqhDUt55Ii0f83yEi9XgO+s5t/cWoamdKrQK37usDL0s4ubdOVVtgmq9CdHgcSwo6jdzjwCQRwqEYHOU8Vc/2kKuDm8/uDriD34QhaZ4bUiixxU2/SziB++tUzeWZd0yyCPyRFPUhlnR0kujri51dLxiurvXr5Dpsj2bvy8rQ/0iWeQvyanngBerDfEj64O1QXkjPCSxxntPpCRKCXSmiCOJlJKismgpGAxSxv2YW3sh6raoWppVCpeWQY3jVhGpzimubffZHKXMy5ajWUkaS47nFVmiGPZjvPPsnxSMe8lzN1xepjnzpGjfbo63o6PjRWkai1RQt5aqDcbAURQ+/ibZGCTESqyaDSFtaXucM8hjkiziL3+7D6kgWsWHm9bR2gcGmKcIoH2WkdsKKSCJww/I0igkTYlwiptoSWss1jqalQlvOI0O17AsDYfzikEWM+pFCCHwHjbXc7yAqraksYY0Yl7WaKWwLsj/pfRIKfE+yPcF4fdPlg3GOJJYYY1DRQprLYgwVjzIIybzmvlKmeTxJInCWoezHutBSBlG1+KIk1mF1pLrl0YIIRjkMcZ6qspwZ39JFituXBoiBPzKHLMoDYvS8L3rY/JUU7f2obXoZcz4H/2e69fG/PKTg8dGqX70nS0WRdv54HR0vEYEcGtvTi+NGKURx1XLrb05724/3sC/SA0+6sUrSwOFMZDGkhuXhjSt5bdfnLCznjMvWhpj+WJ/wbJqaUxoWr+JFUUryLOIKxs5v/PeBh9cHeG9Z1Gas7GqR5VIp5yvVevDhKRSjAcxdpWc2LaW9a1un93R8arpGjpfIaN+zPGspGlblmV75mPwKjg1qTSP7MlPZ3ATLajb8PvOx5f7lYdOognNGyCOFW3r0DKYnG0MUy6tZ5wsmlX6R8LJrObEWJQSpEqiZZDggyBWmg+vDRBIxv2UqjXgW8rK4vDkmWZzlCEF7B4u8c7x7/7elce8cp7UcLmoOTOZN/RS/ZCh8qMnCI+ObnWb346Ojpdhf1qQqOAsH3SNkCjB/vTNmiLvrOcsq5attSyM5iwa9k5KWuO4nEX00gjnQ3qgVhKtBVTnGvzn8D6sI81TxCVKgNIQSYnBkyhFY+1KearAe/q9hEVlwvrmPVmqyVPNbFGTpxFxJEPD33vW+inee8b9hMm04qiqzkzvjXNsrfXQMngFzcsWJQVK6dAwMpYkkgzzCOs8xep3No2laR1pohn3M7bWc4Z5xOJeSxKHwwmxCgY4TdOKI8m4l4ao8khRt5a2dSSRpDWeSAu2Lg/JVmMUQoRn7/vX1/lif4G1nixRjxl+vowZ/0Xfc/dgyc56fmHz5kk3Ws9Dl/zY0fH8+NWJqH/0ZPQcTzpwTCLFjcvDs9oBcDQtcR7G/YQkCs3vXhZRr8ziq9qRSEH9mtT8kQr3A1kWsT3OSVfK+dNrflb4yEW1StSe2aJh97hgkGk+uDp6bm+xjo6O56dr6Lxhzm+YNgYpd/fnVI2jrM1jkvfn5VT5eb5541cO9Y9iXCjYxj1Ygh6NfYVghtY0ljTRZHFEP/GcLBre3emB9yyqILX8q9/b5t5BwZWtHmVjSFNN3dhw+ms8435MpAXOC/JUM+zHvDvoE2vJ//qrA8q6ZXs9p24MTWtJI0mWhvn+0+SPixoui7Ll5u5sZeoIa4MEKQTNyr9HCB8246sGz0WLx7OMKDs6Ojqeh6pyNNYHL7CVSVljHVX1Ou3tH6efReys5/zm1oT9aUndtFzd6rM/Kfnz3x6yf7xg2Iu5vjPgu++skacRdWOpW3c2xnSaoPI8glHvQ3RujWOtF/P+1RGH05KTZU3kQuxuXRtirRhkEcvSoEQYgfIOZCxQSmKMB+FRKpjtf3Z3xvG8Ckkp3p95A3nniJIIpTXrA0VVW/p5RNP6ME7VDybGQgqyWCJETNSXzJcVQkqctSyLhkEakacRWRIjMBjrMU4QaY/2sDFKuXF5wHCQ0VSGsjYsygYpxUOqJ+Chm7Us0WyvZcwWF6tkXsaM/6LvSWPFYl6dnZa/Crrkx46O58MTDMdv7c05mNYo4bm+M7iwrfMkNfiVzd5jjZ6jeYMzlk/uTJkta3pZULbUtTkbi61f4cHveSIt2B6nrA8ztBJoJbiy0aOoWvp5xGTenO2r+5lmumj45O6U9UHC+jA9Cyx5tFaN+gl7RyXv7vRZlC27R8uH0he7PXhHx6uha+i8QR7dMF3b6fGXnx2g9crtXYF9iVF7tzJ6PG+G8+hmXBBOU41fJX2c+/xFy0MSaRAiNGOcozJhQ91PY+TK5EwKwfG0oqhbrm31uX+85HhWI5VAeEgTFWIOW0PVGD68OmLQi8+KfS+PaIxhWbRc2crDyWei0asTiyc1XBZlyy8+P+ZoVgYpu/AcTEs2Rym/c2ODvePihU2SOzo6Ol4WpQWLwpypEsHjHIz76Ru9jtMDg52NDCE8s0Lxy89PmJcNsdbEkeJk3lC1U5wHrSRidSDw6DogLvjYRchVg6YxnpNFhXEOLQReSfpKEmkVPiYFCM/mKAMZ0qcWlSGJNb1UszVOOZhUTJc1s0WDlhIpBVVjEEIgFUwWzdkaYowhSSTfeXfEvcMCgWdjlPGdayNa4/mXi5pBJFkfJiwHMSfziqZxLCvDoBcOGpZlS6Ilxjpq4/DOo7VklMd478MhQ6JIEolWig+uDs9uyHbWcwB+efOIZWUoqpZlZZDA966vX5ga9TJm/Bd9TxIp9i/w6/gydMmPHR3PhwD2TkrWhykb4x5HkyV7J+WFI1fnm+yTZc24l/DR9TG9NHrY6HzRMJmVbK3l5Kli/8TRmHIVXmKp66fpgL4cSoYxsA+vjUkjxdY4o2otgzzmZF5zZbNPL9Vn++qiMhxMy2A231qMC83fqnUXpnftTwqqNr4wffFVNqU7Or7NdA2dN8jN3RlHswprg/x7Y5iyNU65u1/QyzVahU1n1Txf2T4dlfKEhpDggWxeyNDoOcWvPqbOfewip3wBjHqa7XFGaRzTeUNRGvp5zMYooWwt272MXhpxMqtxK4+dsrZoJYm0RCtJFinSOMI4xziP+PFHO0HKfs5IbpRHaAEn8/rM7+HyRh/3jGiVveOComropdEDY7pUsKzMmfqpS7Dq6Oh4U1jrEI/MLAkRPv4mOX9TXreO1hiKqqExjl4qKauwTlhj+eTOhONZRdWEZMHzo7dnj4GnN3UcoAjNmcY6pBCrqHABQnJpI2etnzBbtjSN5cff3+Gze3NOFhX9LA5rRaa4ttnDeU+aSIb5gF+XJzjrQEjSRNO0lrq1OOfZGQfT5Hnh6eURAri6ma+SvVYjVKuLK1vDwYnDGEesFDqRCBGSs/6d37vK9lrOT357QFEbRlqwNc6pGktvlV7V72d8cHVE3Vhmy8dVN4uyxXtBXVsOJxVJLMmy+OwG51F1y8v4vV30PY/68rwKunWzo+PFEKsdt3hsYPUB55vs7+wEVc7eccH7V0YPjf3Pli3fu7HGfNnSGEeWhH21dR4lBUJcrLr/skggTxTb45SmtbyzPUBKGMYxW+OMKxs9Prg6eih85HhWEWuFVoKyNme1aTJvaHrRY/VNeJ6YvtjR0fFq6Bo6b4hF2fLZ7oxhT591qO/sL1AyuNsb6yhrizH+wo31o6hzxX0V8kGkIIkkxjisf9w87bxqR0nQIvgGGBN+n2QVPaskRePAO7bHGa2x9POYcT+YQTbGIesWKcMytjnqoaRgbZhhXRDsR1qvbhIEUazZWstYFO1DG9P1YcqsaNjZyLlxaXBm7Li1Ovl8EuH58vTiB4totEoJOJ3t7RKsOjo63hSnHmPWO5wPMdpaStonxGK/Ls7flKdxOOV1zmONpTYW4zzWObSTWOFpVw2nR5v7p42cJ9+mBLQMX6QjSRor1gYJu0cW8EgpODguGaQxa8OYpnF8950x72wP2D1ccvP+nL3jJXVt+Xx3hgCiSOFEi7EOj8c0ButCwymOJGmqWRumDHsJ86rhO9fGjHox02XDrftzru8MiCMZkqeikKZVt5ayCf46kdJc3+nzN354mf7KU6hqLJ/tziiqlqqxXN3ssTXOGQ9idjYH7B3OqVvH+jB5rLm1d1wwHsS0xpKlQyItaYxlUbZsjbPH1C0vY8Z/0feIcwqhF+VJPjndutnR8Xx4YNSL+PWtGfXNExKt+N714YXN76cp3z64OqJ/dRRGkQ53qQ1IKbDWE0eKsnErE3dF1RhesSgPCAbIW+NwSNvaEEpSt5adtZxYS37nxgYQGsvTZbOaNDi9l4norRrWsZZkiWK6qFlW4Z5GKxnUl+s5VXNx+mJHR8eroVup3xB7xwWxkhxOapxzxLFGScH94xLjHVpLbG0eMzF+EtY/2GyfbsaNg8RDay9W35x+SK3iZVtrsZYQ/eqDgkcKyJOgrIlXp6c9NBujlDTWTBc13nlmyxCdOuonpCtjye9cHbIxSIhWhmdSSUa9iI1h2NiezthCKP5KCTaGGb1Un518fnR9jWpZP/Wx54lGK3FmTgnQrhaP0w3qqQz+/KJyujB1dHR0vFKEwFqHF6GGeg/WOR6T7bxmzt+Urw9TvjiYgxBIFeLEG2NJtFqNhoGSkuoJKqJIPuzLdpFaR8iwnkgEO+OMa9t9ELB3FMz+zWoM2BjPME/ObgimRcOsrFkWFuMtw16Q42sJtbF471gUltWBLkIIrIXNYcKljR55ounn0Zkp8CCL+cMfXGJRtHx6d4pEkCeK+8cFeapZ68fhAETAte0HnjYHk5Kbu3Oa1pDFiixRFLVley3DOc+saGhb90RPttMGWtVY5EptWjcG72Gtn9A8Yoz3Mmb8F33P86yTF/E0n5wu+bGj4/moasO/+e0RVWtWqa8N/+a3LX/wOzuPfe2zlG+n78lIi2CQLMIBwftXRtSt5dMvppzMgrWBMe6Vp1yliT4brbIesOH+ZLZsyc41c/t5xM8/P6KXRvQzTVFZJouWv/JB2FcHVZFmWZmViSchJMAL1gcJxkYXpi92dHS8GrqGzhviaF7TWkfdBi+aw0nJ/nHB8bJhe5RRtUEWrqXFuuf0LgCEArdqyngPZe1CY2bllaNWc1jn9+xChplcJKExskq7YpX0EUeScRZOVMvK8t7lIXmqaY1lXrVc3ezT70WYVQxtaxxprOhnEZNFzaJsWBumq4695vqlAVIKFkX72Mb0B++tP7SZHeTxMzeqO+s5h9PqIQ+dorJsjtKzU0vvxWOLSkdHR8frQAkRfGGkPGt8OOfCx98g52/Ks0Tx3qUhJ7MGYyqcDbcC1jmQiixW4B+sNecbNp7QmDrPo2uSJKRbbY4yjPekSVgL8B4hIYk118YZw17M3qSgbgx3D+cMezF3D5YUpQHlWcsSqtYhhYM4wnuH95IoOjW4FySxIo81Saz54fvhBmJRtiyK9ux6eisz/aN5zWReYpzn2laPWdEyXbT0s4jv31g7k/kvypZ/8fNdeplmez3D2LAG91LNF3sL/uaPrrC1NeDPfnrniZ5spw00AeweF2SxRiuJ9zwxxvhlzPgf/Z7nWScv4llqgS75saPj2XyxN2f/pEBrSRKHZvls0fDF3py/9v2HmzrPUr6dvid31nt8tjulqiy1MexNCkZ5zNZ6zr3jBfcOl89U7r8oknC4uywbrIc0UoxHGTcu9WmNZ1mF8JHffX+DRdFyfWfAomypTUTdOrbGGcvK0MtCmp8UIigb17Oz33GaECilOEtffNSHrKOj48vTNXTeEFVlyBJNpCW37s/xhM1/tNo0Z7GiXUW3lrXB2AejVPBAjeN5MG7lRZC8R1oFefpK3h9rSeMcwoYGjXtkFUh0iFOVHoQUCBl+iZIw6qeMewleQGVXhpCRCnHnDoZZTGsd/Sxi/6TErKST2+OcunUMewnTRYM+PW1YcXoi8SqSpfpZxA/eW+dXN4/5+M6E1jiubPS4fimYUH56d3rhotKZO3Z0dLwOoliuaqTjtDUiVh9/kzyq5tga5/yH/94H3L4/59d3J/zm8xOiWPHuzgClJDd3Z8DFo1XGBjm+EkGp49yDUWAtoZdqRv2YtWHC1Y0eJ4sG58Lo8PpqPDeOFEeziu1RxvG8Ymucc3t/wbSoSSNNaxzOC1pr8E4SR5bttZy7h0sGcQw+nAxLGZpHp02lpylNqsoghURLiY4Fw17MogzjvvMyKIROx47qxrIxTMLfSoW/Vd1arHvQLHnaCft7l4d8dm9Kaxzee1pr8Q7Whslq/Pjt4llqgS75saPj2dzcm5PECiklQkoiDUoKbu7NH/vaZynfzlR+zp51zZ31HM1qeklML1XBr0ZLGmtfmY+OBHp5UGu2FmK1qpVFDQyItCBPNPcOl/zu+xvBZ6wfMx4kXAOKynA0LYNp8kaPq1f6fL47e8gn8+wxrxRHXbO4o+P18UYaOn/yJ3/Cn/7pn3L37l3+6T/9p3z3u98F4PPPP+cf/IN/wGQyYTwe8yd/8ifcuHHjK/nc6yZLFFVrWBQtm6NV8on3OMKm93hehbhVAWmsmZdBtn0q35cyNFzwkGU6pIpYS6QUkVZh5CgOi0qkNbENdsllbVGRJMFTm7AjH/XikMYhPcZ6pFBkaUQaS/CC9VEaGkTWsT1KESJsoIUQ/JUPNvAeBllM2zom85q6DY71AsHWWoq6PMQ6f+YncDSr2Bpnr3wWP4kVf/V722eL5N5xQS+NOnPHjo6vkG9jvddCoqSgsR7vgkIlVgIt3mxDBy6+Kd9Zy/nf/c0P+flv9/jpx4c454lWEeGRCusGhAMEtzpFyBKFUAItJWJ10NC6IOlZGyS8f3WE8JBnms1xRhJr3t0ZhlHYSHJpPedoVmGsQyAQ83B6O13GzFfqkiiSCAG9OKJsLFmiUEqyOQzrjhCSJJYoKTHWszMKTfqnKU1O19p+HhSjeMusaMiS4Ou2MUzPElnGvZi6DQpTCKlf02XDjZ3h2XP3tBP20wba/aOCtUFC3VqSSNHLItYHCW/YE/uZdD45Ha+ab2O9rxuLlAK5skOWhFnbunl8IOpZY5an78mjWUW0StybHDc4D7NlzV98XJEmmp21lNv7S6QLabUvSqKD2lFpSRbpcJi7+v3rw3SldvSEVs8K8eAXPVo78lSjVM6Vzf5ZUtWzamXXLO7oeH28kd3m3/k7f4f/6r/6r7h69epDH/+H//Af8vf+3t/jT//0T/l7f+/v8cd//Mdf2edeN+vDlO21DOMc1jmUElzZ6rE+SFiUDazSojxQNYY0CaehbpU+IgVYC1pLemlElmokAu8d1jrq1tAaR7ZKfsoSTdU4tFp9jQuy8CxVWOcZD+JQ3FWQjGoVjNi0gsm8ZP+kJNbqLEUrTRTDXsS8DHO1H1wd8f0b6+xs5Hz33THfeWeE8Y79SfAsaIwNBmhKMC/aVy6vPL+hF0KQRIokVg/J4M/TbVo7Ot4M38Z633pL1Xj8yjbHO6gaT+vfrhSPnbWcv/HDy3z3nTE7Gz0ub+RnSYFKCuKVJ1kchQaVArz3OMLjinVIMmyNCzG7i4r9k5L9SUkaKYracGWzR55ESBkSUPDQGMvWOKO1jixRWO9xHmIVToilCh5E02XL4aRk1I8pa0MvU6wN4pUq1fP739sGwqm2tY47Bws+uTPhzsECax1FHZSwzjmsD4bIi6oljSV5GrE2SHj30oAkVlSV4cpWj6I2VI3Fe8+ybHHW8dH18YPnbD2nbkLKll/9zPPrWT+L+ODqiA+vjfn9727zg/c2uLbVR6083d4mnvVYOjpelG9jvR/3E46mJbsnS+7uz9k9WXI0LRn3Lzb5Pa0RP3x/Ixghn1OmnL4nj6YVJ7OKRdkyWzYo6dk7Kvjk7oTf3p5wOK1IY41cqSZfBC1BSEkvi7i83mN7LSTspqkmT4N/WT+PWFY27KNPCm7uTrmzv2RtkDx0nU+rHV196ej46ngjDZ0f//jHXL58+aGPHR0d8ctf/pI/+qM/AuCP/uiP+OUvf8nx8fEb/9ybYGc9RwrBzlrG9lrOeJCQxZp3tgdc2xqgpMTj6acxWaLRKCIZjCmlWo1XaUEWR8FjxzuyVNNYT9kYtJI4DyezCqUEm+OUNFEM8pgkUkSxYtSL6efRKtlEUDUWLQVZHKLGi9qgteRwVrM5TvmD39mmn0WczCt6SUgEKWvDsjJnkvXTpkrVWMq65f5xyc3dBRvDFCUFs6Whl0SPxbd+WYraXCjtLGrTLSodHV8h38Z6v1waEKEBb134f8Tq418xizIYBf/rX+/x6d0g/T+9ufgbP7zCxjAlWSlUTOuItWBznHF1u8/2Wo61DmMsSitAEEVhvThZVFgH1nkOTkoG/WCWGVQ3nrZ1eBcaMVc3+1zZ6tG2wXctjtSqkeQZ5DHeBcWldy4kNVrPqBdTFIbbe0usg//tH17nvctBOSOAm/fnGBuMOI113Lw/p6oNRd3iPLyz3eN3rq8Ta0UvS3j/8pD3Lo/IU32WyJKnEd97d4yWgqNZjQf+1r91jZ21B2vF6Qm7lkHlqaV8bD37uqw5z/NYOjpehG9jve9nmtpYnHMIEUZta2MfU4Y/388K70nTOpaV5XBaI4DjecOsbEBA3RrKxmJtUDEmWrDqvYe02tXPevSGThKaOUkk8TakBQZrSYESgnRlp4AXXN3ssz5MsdZTVC2RDr6Yzoc15HlqR1dfOjq+Or6y46Pd3V12dnZQKmwklVJsb2+zu7uL9/6Nfm59ff2Frn1j48VSH+ZFw3HREqcxaWs5mdVsjzMub/X59M6EG1nEYJAi8Hz8xRQvgslmHEmsdQghmCwahAgjWtY6GuOIdJitbYzF+TC6FZTzgqq2vLM9YNRLEErgnKcoW45mFb1MkcaKUT9hVja0jUMqyfZ6Bl7w7uUe33l3g9Z7pNaMegm1g8trOd8d52G0wEOURmzkMUVt2D2puLQ54GASfAkq67l6aYSUgo+ur72Qm/3W1uNGko9ytWhXZswPXsJVYxiNJO+9u87mZp97B4sQHzuKuLLV/8od9Z/ncX0d6R5Xx7P4utb75631jbMIwlisI2yknQ8ffxOvo3nRnNW7fvag3s2LhluHBf1BirGO/WXDzYMlH7075sN31njvnTX+WmnYnyyZzBv2j0KTPsjwM/ZPCpAC23hiBWjBxiANKYtS8P47Y4x1XN3sk2Up62vh+RqPDJGW/PiHV/jNrRPSVeM/zWLuHS75QSQ5OCk4mIQT6TQJY1tXtodsraV8emfKbFmzPkx47/KIfp7wo4+2zmr4+LjgaNkwyOIzxZBQDUbA+1fWuHY5NJnKuuXD6zDuxfzww+2z56tqDB+Mcq5s9fnkixOcVFzC8+7OgA/feXi92toasAW89+6TXzdb8MbXnJd9XT3rsXxZ3ta6+bZeF7zd1/YyfNPrfWEco17KsmhojCdWgl6eUBj3Un/LLeCD+3N+c+sYK4JS/v5RifOglMTXFq0FSimsBaEE28MM6x2zZQveI0VQogsf/MDSVKGVpKwN1jiSVGN9MDq2aN7Z7nN5o8+oH7M+ytgaZxxOSmbLGmM9earZPLfff546ePpYXnV9+Tq+P75u19xd7+vndV/z26UH/ppwdLTAPafh4d5JwU8/PsD50L2OtMC2ht2DBceTkqZ1tM6GY10h0BK8c+hIr24QBEVpUJJg9KgkRdPSto6mCaepEtBC0BpLLwsjV9aFLnvVGAZZxM5ahmktSSQZZJr7xyXXtnrkcUZtDMZ6jHEcTCtGvZjprKSXauaLahVlrhmmmqZq8N6fzQLvFTUHk/Is8SrTCttaprMKU1t+/6MtqmX93IkcW1sDDg4eN5Z7lFjAnYNgTPeoa/6f/fQORW3OIsz7WfRC1/A6eN7H9XWje1xvH1KKF246d1zM89Z6Z1k1OUIz3nkflDqW1/46Om8QHGvJQVFzZ3d65ttgnOPwqOXO0ZLlsiHSgp9XLccnBVXreG+nzziPmBYt1jokUFaWveMli9UIkpSQxWGMKo5k8I+QgvVeHA4XpOfgaM4oCzdURdVyZ3/J5c0eApgTRnfzRPN7761zMCm5vTvj0lrG5lrOv/7VPpNFhVaCu/tTpBAMM401js1h8Kb5xW/3z7waTiYlm/0QtVs1IQZ3c5Dw6b0Zl0YpVWOZLyqq2pBKwf2DBVuj9GytmC5q8iTik9snnMxKdtZzRr2E5bLmX/3l3bNT5RepA4uy5eSkoKgNbdUSC6he08n021qfuut6cb7stXX1/tXxvPX+8HjJoqhxLow/WedZFDWHx/Kl/paLsuXTm8eczCqOpxXLqsX74HO2rFqUFkgLxvigLowjqqYlPlXRGEvRWpxjZXLs8KVDrHx+8MEfLBKSD66NKMqGzUHKv/O7lx5Sz9y9P2OjH58Fmpzf76/nX43K5m1+7z6Jr9s1d9f7+nkV1/ysWv+VNXQuX77M3t4e1tpV19myv7/P5cuX8d6/0c+9LvZOCv6HP79DVRuEAOs982XDuJ+QJZp+HjEvGpQSjHoJJ/MqbDhrS2NC5K1dRW/3Uo11jsZYIqmovaG10Fq/8tdxwSitaMmTiCSS9FeJVFIIysYx6kVsjVOGvYQsnpGlmpNFjVaSNJYhDUtIxv14ZSwc0Us1RWWomgeeNKd+NKfu/fOiZZiHxBIh4XvX1/Hec/dgyee7s4caK6+Ki4zmxuvJ2RjYo8knneSzo+Or45te75NIUlTB58ysEq7k6uOvm6cZBBe1QQr47Z0T+nlKFEmmi4Y7B0sgHBiMLg+4tt2H/QXvXxryl58dgfcsK4t1YRORJzp46TjP4aQmTSTvXR5xbbvP0bSirC3pyi+mrA0378/JztXhurEP1eGf/OaAzVHKZNFwe29O01okgsmyppcEw86Tec0gix+kqSwaIIw35YnGOBeue0XdWsa9hOmy4eCkJIqCl1xZW5JYh0OQNnjJeR9Gx+p5hZKSg5OSJFJk56KEX8TA82mpW19m7TkdbX70gKKj423mm17vy9pR15Yo0mdJsnVtKOsXd0E/rR0OR9MaBr2YZdVQ1papNXgXFDdaKbwNX1M1oUGTxJpr6xmf3Z0jPGSJxlpHaxytASk8WnuUUjjnSHsx1nm0UhxMSn7ymwPWhslZbelM0zs6vr68+QiOFRsbG3z/+9/nn/2zfwbAP/tn/4zvf//7rK+vv/HPvQ4WZctPPz6gagzWWYx1HE8KiqrlcLpK/hAwXTZEKhhQLivD8ayiWY1HGefxziHxVK3D2mCc3BhLY1bhuH4VKbs6VGgsHM+D1DzSwUPnnZ0Bf/CDHa5uDxj2Ejye77+3TmuDWXISKUa9hCRWXN/pIUSIJARPnkQsq5bJouLjLyZ8fm/KZN6cbSzfvzKil0TMlgYlwxwuwK29OVEk6Wdh4/3ZvSmLsn2lz/GjRnOLon2iUXJHR8dXxze93ierDe/p2a5/5OOvk6f5ieWJZv+kwDuBFHAyq3E++NbMy5rpsmYyb6hbS1kbqtacxfE2xuHxqNW/t8YF+b4LXgwCwZ39BYui5WSlrvlif8G//u0BR7OKYS95Yh3en5ahAdNPuHF5RC/XtNZRluEmyblVBLyEu4cL6taxNkjO1pJ+Hl3oWfPR9TF7xwVCCGIdvHg8nne2+2SJ5ofvb5AlmvEgeMvVrSNPFVEkOZpVDz13L8LTTPpfltMbPePca11HOzpeNd/0eo/3OEFortSWpjW40w35C7J3XOCcZ7a0tNZzPCuZFoaiamlbT9uGtNplaXDWI4VAS0GahCb1ZNEw7MWMB3EIN/F+1cwW6Iig/Fcy1CYJx9Oa6aJm77jgZF49VFueVFffNi+wjo6OxxHev0QFekH+s//sP+Of//N/zuHhIWtra4zHY/7b//a/5dNPP+Uf/IN/wGw2Yzgc8id/8ie8//77AG/8cy/C88gyP7075bd3JuweLqnqFusE9w4XaC3JU83OWs73rq9x92DBojRc2+pxMC357M6Uk0VDayxaBW8AKSFPIySwrFpaB2Vt0QqMDY0dKR/Em6daIpRgZ63HznrG9jjnvctDPtudsdaP2Rhl5KnmeFbxb35zwGRZ897lITvrOXsnBSezGoHg0kZOEikOpxVChqharSS9VPM7NzbOTgoflfzf3J1RNpb3Lg/PTjzr1qKlPJPMP4kvI0v72WdH9DN9JhcFzuSiP3x/46V+5qvi6ygRfB66x/X28VVL8L9J9f55Jfj/5//H/8LuSfnYxy+vZfxf/w//9gv9zhfl07vhpv/8qeppvd1Zz/kff3KHummpDRRVE2LHhwlt67i82WNrlJElmk/vTvnVrROc9wgBi7LBOZjOa2oTNvVSCtrWsaxaLq3n/JUPN8M61brwO5VkMq9YH6ZIKc4UNIfTipN5zQ9urLOznvOnf3ZrZSIa0eul/ObWEcfTEhDcuDRgVjSkcfCAuLLZx3vPtVVT5vxju0i98q9+vU9dt2dx5OvDFO89n92bkUSK3aMl17Z6XNkK6qJwYh0SuT68Nn5orXreOvA61p6n/V3/8EfX3sr69LbWzbf1uuDrP3L1ba33k2WJ9WGcCQFKeMa9F6/3/+rX+0zmJZ/emXHvcMFkWVM1Yd8fK4l1YYTKOYg1RFoiRFDnSEL6ICLEiLfGU5YttXHEUfiEFtB66KcRVWMZ9mOkCMbOm+OM3/tgizx9dl2Fr0Yx+Da/d5/E1+2au+t9/byJkas30tD5pvE8Rf9nnx0xWdT85vYJ+8cFzsN0WWFsKKzvXx1y/dKQ2/dnfHJ3ys56xrxoOZpWeCCLFEgRNtU2nKgua0NRGxIlaI3HE9Q5SoBW0Jjw70kq0VLxwZUBa4OUK5s9NkYZUgSZ+fkN4mRes3dSsLOes39Ssixabh8s6KeaLNH00ogoks9szpwv9LuHS65t98jTB4X+eTe3X+ZF/7QN8LMaSa+br2MBeh66x/X28VVv8L9JPO8G///0f///Mi0eV3WMcs3/7f/4N1/HpZ3xaEP90RGnn392xJ2DBZ/uzoi1YNxL8D74PvzehxtYBz98f4NF2fL//H//gqox9FKFMZ5Z0bI/WRIpzbgfUxu7MuMXGOv5m793lTRWLMqWJJK8d2XEnf0FRWWYLZuzBMZhHjPsRWyv0qD2jwvuHC7JEsX6KOezOyfMy5a1fsr1SwME0BrH/ZOCD66O2BylZ+vPs9aS8+tAUYX16JO7U4y1fHh1zKyoKWrLxjDl8kaP43mFQJDEiu217KHn7nnrwOtYe57WJPrbf3DjraxPb2vdfFuvC77+DZ1vEs9b7/8v/68/43hSEMf6LJikaQzr45x/9L//w/9/e3ceJ0V5J378U9XV1Xf33BcQUFTkWGUYDG7ACw/ccEi84+q6YlwVg8fGbFBXXfFIWN2gUWIkYPZy9eUmOigSr4j3EVgJ6C+oOAHkGGaYe6bvrqrfH820MzAzzAw90934fb9e83rNTPXx9FHfeupbz/N9BvScb/9pD+3hGJ/WNNIeidPcHiEWS5Yv0NSv67EZ+2dzuZ02NFVBURRsCiiKkhyZY1qYJM8BTAt0u42igAMshdZgjI5wDMO0KCvyEI8aVBR7GFnsw+3UGFniPWRcPdRxZqhk877bm1xrs7R36B3RNXSOdG5HsuZNOBonblgkDDP5s39eazgS56u6Npo6kit5OHWNnXUdxOJGcqlxXUsuI66oxEigKCo+lx3DSCZy3G4V3WajqS2yfxnCZKBXVSWVtfd5HBimRWmBB5steWU1GjMIRxK0h2J0hBOoqsJxo/LZWddBKGoQiRscXeHHMAzaggl21LUz/a/KUp1pSA5J7wh3P3nxuuypmgOd9Q26Go55uJ01fTrb2HnAGVEhnR0hxNCJxXuundDb/9Opp3piIyq8qU72mHI/pmXh9ejU7msnnrBQVIvjRuVjs6mpOj9elx2/x040HidhgK7bKHVo7GsLY5pmMvniT9ao6QjFsdttuJ024gmLuuYghYHksHy3U6NmTytOu41QOI7Po9PQFqa8yJ1KeOh2jeNH57GnIUhLR5Q8r4M8n4OSPBdjyv2p2D2yxHvQRYhDHUs6jwPhSIL6lhBNbVEMyyDgcdASjBLwOIglkqtrhaIJigMu6prDOO02NFXt9t7111Ace6SehRDZyeeyE4s6iBnJkYmKouB1OfANIrHhctjY1RAlYZk4bAp2m42EamJakDAsbCpYXQ4j8ZiBYlfRNJVI3MCuaeT7HNjtKvXNYVRFwee243JolBV4aQtGME07KAo+t0ZhwEM4Ek1O3dIUIjEDOHRs6atW20DqjQkhhob0DIZIaYGb4J44NptKwjBRVZU8jwNQiMZMdu0L4vfo6JoNh55M1LicdqJxk5hhYEsoGJaJYVqoqkrAq6OpCjYV2sIJnPZkfRzDNGkLxbDbbLh0DbdTIxiOE/A5iMZMxlb4cTuTBS1jcZPSAvfXq2657fj2F2Z2OjUKFQvTm1wGFqA4z2JbbRttoTiFAVfqtfW3Qw3Dm1g51ImNEEIMBWN/j1tVkj+mlfwxrKFP6ED3hHpP246uCNASjtPUGqEooFGS70bT1IPicmm+G01VSJjJAsiqqhBwOwhGYgSjBu3hOOGogWWZFAVcgIJdU3DYNUL7k/yhSILyQg9twRhxw8Tl1Chy2glFEhT4k8cEp1PDaVeZeFQhpUU+6hraUytPdY3dQL+PJV1HiaqKQl1zCMMCsPC5HPjcdgwTIjGD8v3TCprbo1QUejhmZN5hHSeG4tgjFyiEyE6l+W5M06QjYmBhoaDgddoozR94rZkCvxNPox2HZsNQVNwOg1AkjmWCqYBiJcsqwP6ipwokTItoKIF9f4LF79HR7TZsikI0nqwZptsVVFXFqWu4dDsT8924nBqF+R7qGzuobeygvjnMiCJPqlZOX7EluVBK935/Txd3hRCZIQmdIdLZwXtvcy2FfieqqqJpClgWLe1RgpHkKlIWFqZpoigKHoeN1lBypZS4aqGhkLApOHUNn1vDpqo0tUfwuex4nTp5ASemYWFYJnbFhl1LDsP0ue1UFLqwqSoVxR7g6yRMRyjOmHL/QUPDW9pjBCMJAt6vO6Bxw6Qs3017MFk0s7+dykwmVvo6sRFCiKGga8l6Z1hgAuzvhB9YrDhTvC47R32rgDyXPZX06Gk0ylHlfgzTJJGwMC0TVVHJ89qJJZLF/dmfn7IscDptWJZF3DDxuL4uphmOJtDtCgGPTr7Pgc2WLFAcinx9JbjQ50jVamgLxdBUtVtdtq76cyzpaZWpcCzB2BEBmt12ahuCJAwLu6YSiRnYbA7KCt18q9Sftum46T72yAUKIbLT0RV+ahuDBNx2nE6NSCSBYVocXeEf8GOVFrj5cncLpfluOsIxfF6dYCxBKJRI1sUEsIENsNlUVFXBsixsdrDbbRTlOVEUhfj+kUL5fh2bTcHr0vF7dLyuQDJxU+xhT2OQWNzAqasU+V3UNoZwOOz9GpkoIwaFyG6yJw4hr8vOyFIfrR0R3A47mk1hX0sEn8fE73Hg0FUgOerGMCx8HgeBqIFpfn31szUYpcDnoMDvJBJN4NLtuF0qbt3OyBIvTptCeZEHEwtNVWgNxoknDPa1Rjn+W3k4dVu37Pu22rYes+wuh41wLEEoYuBy2IgbJvG4SXG+C01V0VR1QJ1KSawIIb4p8nxOTDOEhbL/em3ytzyfM9NN6+ZQcXlMuZ9gJEEoEiNhJIsF1zbaOaZCJ25YRGL7lzJXkqNHw9EETofGqBJv6jhhWcllwUeWJC8m7KrvIJGwcNjVbseizrYcam55f44lPU0H8Hl0dta3Y9dstAZjxBIGPpeOU1cJRuIU+l1Zv3qLHEeFyD75PidHVwRoDUZBUXHuXyk2fxDx3uuyM/nYYj76cx2hmIFiWfgcGoplgaXgcmpEEwZmwiRhgt+jY5om0YSJioLXacfrtuNxatS3RPC6dMaPzifg1VMXYCuKPGiayogiLwkgFDHQ7SpTjy9hUj+LtsuIQSGymyR0hthRZT627jIwDItwInkF0+O0U1bgobYxRCyRQMFG3DQJuHVK8wv3LxmYLAo5qsSDaVl4XDoVRV5Gl/vZtqedAp+DEUVe4rEEiYRJazBKWyiO0w66ZqMtnKC5Lcau+g4qirypDnRvWfYCv5Nvlfn409Z9tIVMvC47efkOVEVhTLlfrgoKIUQvJo8t4qPP6ojGjP1TlVQcuo3JY4sy3bTUVKSvGkPEI/E+VybxuuxMPKqg20ome5vCgIVuT646FYok2NcSwunQGDsikOrYjylPXp0ORxOpjn9pvovifFcy4eJwDLpGzaH0NB3A77bz8a5Wjir3c1S5j90NIeqbw4wp8/GtEp8c14QQg2IBE48qoLk9iqZrJGIJ8n0O+lFPuUel+W5mThnJ9to29jQEMSyL9lAsNRqnPRKnIxjDroBpmoRjBjZFweVQMU2DlnaTfJ+TMWW+VE3Mr+o7yPM4GFXqJRiO8/mOZnwenePGFOKwKd1idn/IiEEhspskdIbYmHI/oWicYCRBwjBR1OTw885pT3UtYVrbw6iKSlmhB49LY19zBK9LS10hVVWVQr+TcDRBOEpyiLgFDoeGW7fTYcZw6BrRmEFzRxxVsRhZ5KWs0EMomujWge8ry+512Zn+VxXDviyhEELksspxJWzb2057KJaca2WBz61TOa4ko+3qOhWp0K2zvTnIl5+0kO93paY9HRjfDxwVsntfkPZwsh5ONJbA5dQYWezFMumx3o1DtzF2RID65hB/3t5MwK33+Xzp0NOFirZQnFHFXpy6jUjM4OhyPyccXYjPrWd81UMhRO7qXPhjZImX/DwPzS3B1Kp2g+V12Zl0dCGTji6kZncr+1pC7Njbzp7GEG7NxOF3YZkWkbiB15WMc/k+Bx0RE7um0tIR5TuTyqlvDmOYBi7dRigSZ8PndRw7Mi8Vkz/b3kRpwDmo1alkxKAQ2UsSOkPM67IzYUxhKklSXuAhGElgsymUFbqJJgzyvDpjynzYbCpffNVMwgBVVfHoyeXJg5E4boeGQ7fh9+qpRIzToZHn04mbBuGogdNho0izkTBMdLsNl8NGW8hMVaHvvFIbjSVr5jidGoU+R7csuwRsIYQYGNO0mP5XZexpCGOQrHdQUeTq1xK4Q6nrVKRQNMG+5jA2VSUajZPw2PnLntZDduzHjc7jw/+3F6/TjivgIBw1CUbinHxCWbcioDW7W1PPZRgWpgltwRimCRXFHhKm2a/nG4yeLlS0B2OMHRHA0+W5OpfmFUKIweoabyzL6ldR4YE+fjASZ/JxxXx7fzyrawrx+Vct1DZ04HMnzwNCUQPNBuX5Lo4elU8wHKexLYzHacejK+xpCBKKJGhsjTB2RICjXAFcbgcd7ZFDxuCuRebl4q4Q2U8SOsPgwCRJZ6CMWSbfKvEBYJjsXz5Wwe1QaW6PEosbycSMbmPrrhaqji/pViPAqdtobo8yutRHJNpMQ2scn1snz+tMTtkyklOnQtFEtyu1RXnO1MgcCdJCCHF4OkdClhV6UldssyF5kFzxCfY1h2nqiBGLJijwO4jGzX4vO1ua7+bkiWV8vqOFxrYoeR4HfzWx8KAVXTqnPYUiCXY3dNDaHsPvsRNLmOxpDDKiyItDtw3JMrc9TQc4uiKAdkBRainiKYQ4XF3jTWdR93ROP+opnhX4nIws9hJw22kPJzAtC4dDw+ey4/c4KPQ5qNndik1RUucPLe1RvG47+1rCqVGJDruN+mjfx6WeiswPVTJeCJEe0rPJgJ5GwXQmefY2h4nFYhT63bidyWlX+1rDRGLmQSumJAyThtYI4WiCWNzCtCAWN0nYTeyajXjcJC/fgduh9Vg0Eg7dmRdCCNG3zhibHKHTkBqhU5yX2aK7CrCjrh2P0w6KgqJAbVOIsv3JmP4uO1ua7z7kkryd056a2iLomm3/KlkKbqeGrtloaoswotgzbEmuzto9IEU8hRC5LRQz8Lo0bDYFiwiapmJTFdqCcRKGSTiaYEd9R7fzB5tNpbEtTJ7362LN0bhxyKS2nC8IkXuyY03Vb6COcJya3a188pdGPv1LI3/e3kjCNPE6NRKmQnN7hEjcwLIsDMPCs7+z3CkUSfDlrha8LhvNHVFMy8TnshOKxNndGMTltFGc70JVFEoL3ISiiYMSQskhmzL8XAghDoeqKnz8xT7qmjsIhWPUNXfw8Rf7UFUl000DwMLCabeRMM1u/0/niJXSAjfRmEF7KI5mA0VJFt4MeHTsmkIkZgzZCJnOK8oJ00zWnzOTUxRKC9ypFRo1VZUrzEKIw9Y13vjdemo6aUc43ud9Ovv8NbsPfdsD41lzWxi7pqIokO93oqoqbaE4lmXh9+jY7epB5w8el0Y8YaEoX08Ni+wfmd8XOV8QIvfICJ0MOHA44/baNsIxA59bJ+DRSRgmoUic5rYIBX4nBQEneR6daMwAkoG1rimIatfwuHQcukY4mgAUPE6NPJ8TVVHwufTUlKreVreS4edCCHF4vtzZim63ARaoCnYtOX32y53JVZYyxQLGlPloao+CCqYJRQEn5hDUfeicJtDSHqMtFKdgfyF/RUmOHLWpDNkImd6uKHeE4lIAWQiRVl3jjaIohxzBMtApTD3Fs9ICN3VNYUrzXbSHYhiGhTvfTYHPgd+r47Dbejx/GFPmw67ZUlO3xo3OJxKM9vn65HxBiNwje2cGHBisDRPcThuNbRH8Hh1NUwlG4rQHY5QXefC67KnkTOec2njCYuJReWyp2Zfc7tYBi1DEYOwIPx3hRLeObF+rWwkhhBi83Y0d5Hl1dM2Gx+MkGIwQSxjsbuzIaLtSq7EUJ1djKQ042dsUIh43h2QJca/LzpRxNao79wAAGihJREFUxamTF8MwqWsO0x6McXRFYMiWCu9p2fL+TicTQoiBGGi8GegUpp4eP+DRicdNfG4dm02loihZqHhbbVtqNE1f5w+d5wM+t37IhI6cLwiReyShkwEHBmunnlyZKhJNUFHkJdzQgc+lUxRwUZzn6la8uDP4ux0amk3F6dCIGya6ZiOesHDqth4z6T0VWUt3Z14IIb6JdC25slPXI6phWOiarfc7DYMDV2NRVYVC/+CWrO2vrseaWDxZ+H+oi+/LFWUhxHAZaLwZaAKot8cv8DsPGnHY9bYFfmev5w8DIecLQuQe6e1kwIHBusDvZPveNlwODZfDRnHARV1zGKfd1utV1NICN43BOF6XnfrmMImEhWmZlOS5e82ky5LkQgiRfseNCrCppgFFUXC5TCIxg45InBPHFmW0XUO9Gktfzzucxxq5oiyEGC4DXbZ8oAmggcSzrrft7/lDf8j5ghC5RRI6GXBgsLbZFAr9LjxOjY5wAp9b55iReX0GYa/LTlGRl/8XjROPm4SjBk6nA59b79fV0M5VtULRZPZdli8XQojBOX50AcFIgrqmMM3tUUDhmBF5HD+6INNNS3XMi4t97NvX3uvtcvmYIFeUhRDDZaCJ8oEmnHuKZ3kFDuqaQmw7ID4feNv+nD8IIY48ktDJgJ6C9cSjCgYcgH3u5LzYgRZ9HGiBNiGEEL3zuuxUjSuhrimE3WknHonnVELkSDgmyBVlIcRw6W+ivPO2A004d41nh4rPEvuEEJLQyZBMBuCBFmgTQgjRt4F08LONHBOEEGLoHE6fX+KzEOJQ1Ew3QAy/UDSRqorfSddUQlFZEUQIIb5p5JgghBDZSeKzEOJQJKHzDdRZoK0rWRFECCG+meSYIIQQ2UnisxDiUCSh8w1UWpBcCSsaN7pV6B/o0oZCCCFynxwThBAiO0l8FkIciiR0voE6C7RpqkpHOIGmqjlV/FIIIUT6yDFBCCGyk8RnIcShyHi9byipii+EEKKTHBOEECI7SXwWQvRFRugIIYQQQgghhBBC5BhJ6AghhBBCCCGEEELkGEnoCCGEEEIIIYQQQuQYSegIIYQQQgghhBBC5BhJ6AghhBBCCCGEEELkGEnoCCGEEEIIIYQQQuQYSegIIYQQQgghhBBC5Bgt0w3IRaqqZLoJKdnUlnQ5El8TyOvKNbn6unK13dlosO9ltn4G0q6By9a2SbsGJlvbBYfXtmx+XbnmSIv3vZH2Dr1ca7O0d+gdbpsPdX/FsizrsJ5BCCGEEEIIIYQQQgwrmXIlhBBCCCGEEEIIkWMkoSOEEEIIIYQQQgiRYyShI4QQQgghhBBCCJFjJKEjhBBCCCGEEEIIkWMkoSOEEEIIIYQQQgiRYyShI4QQQgghhBBCCJFjJKEjhBBCCCGEEEIIkWMkoSOEEEIIIYQQQgiRYyShI4QQQgghhBBCCJFjJKGTJZYuXcrMmTMZN24cX3zxRer/27Zt45JLLmHWrFlccsklbN++/bC3DZfm5mauueYaZs2axdy5c/nhD39IU1MTAH/605+YN28es2bNYsGCBTQ2NqbuN9htw2nhwoXMmzeP+fPnc9lll7FlyxYgtz+vrh577LFu38Vc/7xmzpzJueeey3nnncd5553HO++8c1htz5bXJbJHJvfhoYi16ZSueJJO0WiUu+++m3POOYe5c+dy5513ApmP0+vWrWP+/Pmcd955zJs3j1dffTUj7crmPklPbetrH4Dh+c719p51OnA/GK52ifTLhj5bNu+jPcnVc4Jc7e/nSj8+F/vn6e4/pOU7YYmssH79emvPnj3WGWecYX3++eep/19xxRVWdXW1ZVmWVV1dbV1xxRWHvW24NDc3Wx9++GHq75/97GfWbbfdZhmGYZ111lnW+vXrLcuyrOXLl1uLFy+2LMsa9Lbh1tbWlvr9tddes+bPn29ZVm5/Xp0+/fRT6+qrr059F4+Ez+vA/cqyBt/2bHpdIntkch9Od6xNp3TFk3S79957rfvvv98yTdOyLMvat2+fZVmZjdOmaVpTp05NxaotW7ZYkydPtgzDGPZ2ZXOfpKe29bYPWNbwxfPe3jPLOng/GM52ifTLhj5bNu+jPcnVc4Jc7O/nUj8+F/vn6e4/pOM7IQmdLNP1i93Q0GBVVVVZiUTCsizLSiQSVlVVldXY2DjobZn08ssvW1deeaW1adMma/bs2an/NzY2WpMnT7Ysyxr0tkx6/vnnre9973tHxOcVjUatiy++2Nq5c2fqu3gkfF49HTCOhNclskM27cOWdfixNl3SGU/SqaOjw6qqqrI6Ojq6/T/Tcdo0Tevb3/62tWHDBsuyLOuPf/yjdc4552S0XdncJ+kprnfq3Acsa/jj+YHt6mk/yES7RHpkW7zP5n20L7l4TpAL/f1c68fnWv883f2HdH0ntMMbdCSGUm1tLaWlpdhsNgBsNhslJSXU1tZiWdagthUUFGTktZimydNPP83MmTOpra2loqIita2goADTNGlpaRn0try8vOF8OQDccccdvPfee1iWxcqVK4+Iz+uRRx5h3rx5jBw5MvW/I+XzuvXWW7Esi6qqKv7xH//xiHldIvP62veHex9OR6xN1/c4nfEknfvWzp07ycvL47HHHuOjjz7C4/Fw00034XQ6MxqnFUXh4YcfZuHChbjdboLBICtWrBj0sSXd371cOcZ13Qc6251t+0E2tEsMTjbF+4G0LVv30Vzoi+VSfz8X+/G51D9Pd/8hXd8JqaEjhsW9996L2+3m8ssvz3RT0ub+++/nzTff5JZbbuFf//VfM92cw7Zx40Y+/fRTLrvsskw3Je2eeuopXnjhBX73u99hWRZLlizJdJOEGBLZEmuzOZ4YhsHOnTuZMGECzz33HLfeeiuLFi0iFApltF2JRIInnniCX/7yl6xbt47HH3+cm2++OePtyjXZsg9Adu8HQmRKNu2j/ZEr/f1cjDe51j/P1v6DJHSyWHl5OXV1dRiGASS/RPX19ZSXlw96WyYsXbqUHTt28PDDD6OqKuXl5ezZsye1vampCVVVycvLG/S2TJo/fz4fffQRZWVlOf15rV+/npqaGs4880xmzpzJ3r17ufrqq9mxY0fOf16d76Wu61x22WV8/PHHR9z3UGROtuzD6Yq16ZDueJJO5eXlaJrGnDlzADjxxBPJz8/H6XRmNE5v2bKF+vp6qqqqAKiqqsLlcuFwOLLi+JELfZID94HOdmfbfvDuu+/KcSZHZUu870ku7qO51BfL9v5+Lvbjc61/nu7+Q7q+E5LQyWKFhYWMHz+eNWvWALBmzRrGjx9PQUHBoLcNt5///Od8+umnLF++HF3XAZg0aRKRSIQNGzYA8Mwzz3Duuece1rbhFAwGqa2tTf39xhtvEAgEcv7z+od/+Afeffdd3njjDd544w3KyspYtWoVP/jBD3L68wqFQrS3twNgWRZr165l/PjxOf89FNkjG/bhdMbadEh3PEmngoICpk2bxnvvvQckV5hobGxkzJgxGY3TZWVl7N27l7/85S8A1NTU0NjYyOjRo7Pi+JHtx7ie9gHIzv1gxowZcpzJUdkQ73uTi/toNvfFcq2/n2v9+Fzsn6e7/5Cu74RiWZaVxtcpBum+++7j1VdfpaGhgfz8fPLy8njppZeoqalh8eLFtLW14ff7Wbp0KUcffTTAoLcNl61btzJnzhzGjBmD0+kEYOTIkSxfvpyPP/6Yu+++m2g0yogRI3jwwQcpKioCGPS24dLQ0MDChQsJh8OoqkogEOAnP/kJEydOzOnP60AzZ87kV7/6Fccdd1xOf147d+5k0aJFGIaBaZqMHTuWf/7nf6akpCSnX5fILpnch4ci1qZbOuJJOu3cuZPbb7+dlpYWNE3j5ptv5rTTTst4nH7hhRf49a9/jaIoANx4442cddZZw96ubO6T9NS2hx9+uNd9AIYnnvf2nnXVdT8YrnaJ9MuGPls276M9ycVzglzv72d7Pz5X++fp7j+k4zshCR0hhBBCCCGEEEKIHCNTroQQQgghhBBCCCFyjCR0hBBCCCGEEEIIIXKMJHSEEEIIIYQQQgghcowkdIQQQgghhBBCCCFyjCR0hBBCCCGEEEIIIXKMJHSEGGKVlZXs3Lkz080QQgghhBBCDMK4cePYsWNHppshxEEkoSNED1566SUuuugiJk+ezF//9V9z0UUX8dRTT2FZVo+3X7x4McuWLetx28aNGxk1alS/n/u5555j3LhxrF27dlBtF0II0d1AY3qmPfroo4wbN45NmzZluilCCJHVKisrUz/HH388J5xwQurvF154ocf7fPTRR5x66qmDfs7FixczadKk1POcf/75/PGPfxz04/Wmp3a2tbVx2223MX36dCorK5k1axYrVqxIbR83bhyTJ09OtW3q1Klpb5fILlqmGyBEtnnyySdZuXIld911FzNmzMDj8bBlyxZWrVrFRRddhK7r3W5vGEZan//5558nLy+P6upqvvvd7/Z6u0QigabJLiyEEH0ZaEzvy3DEXcuyqK6uTh0HTjzxxIy2RwghstnGjRtTv8+cOZP77ruP73znO0P+vFdffTW33HILlmXx29/+lkWLFvH+++9js9mG9Hl/+tOfEgqFWLt2LT6fj23btrF169Zut1m9ejWjR48e0naI7CEjdIToor29nV/84hfcfffdnHvuuXi9XhRFYcKECfzbv/0buq6zePFi7r77bq655homT57MRx991Odjdg7R3LRpE9OnT++WAHrttdeYO3du6u/du3ezfv16lixZwrvvvsu+fftS2zqz9CtWrGD69OncdtttmKbJihUrOOuss5g2bRo33XQTLS0tqfvceOONTJ8+naqqKv72b//2oIAvhBBHsv7E9DfffJP58+czZcoUTjvtNB599NHU/Xft2sW4ceP43//9X04//XSuvPJKoO/Y2tzczHXXXceUKVO44IILWLZsGd///vdT22tqarjqqqv49re/zaxZsw4ajblhwwb27dvHHXfcwdq1a4nFYqltzz33HJdeeikPPPAA06ZN49FHHyUWi7F06VJOP/10vvOd73DXXXcRiUQAaG1t5dprr+Xkk0/mpJNO4tprr2Xv3r1D8l4LIUQ2icVi3H///cyYMYMZM2Zw//33E4vFCIVCXHPNNdTX16dGsdTV1bF582YuueQSpk6dyowZM1iyZEm3+NsbRVGYM2cOLS0tNDQ0ALBjxw4uv/xyqqqqmDZtGjfffHPq9uPGjeOpp57inHPOobKykocffpivvvqKSy+9lClTpnDTTTf12c5PPvmEuXPnEggEUFWVsWPHcu655w7V2yhygCR0hOhi48aNxGIxzjzzzD5vt2bNGq677jo+/vhjqqqq+vXYJ554Ii6Xiw8//DD1vxdffLFbQqe6uppJkyYxa9Ysxo4dy4svvtjtMRoaGmhtbWXdunXce++9/Nd//Revv/46//3f/80777xDIBBgyZIlqdufeuqpvPLKK3zwwQdMmDCBW2+9tV9tFUKII0F/YrrL5WLp0qVs2LCBJ554gqeffprXX3+9223Wr1/P2rVrWbVqFdB3bF2yZAkul4v33nuPpUuXUl1dndoWCoVYsGABc+bM4f3332fZsmXcc889fPnll6nbPP/885xxxhn8zd/8DQDr1q3r1pbNmzczatQo3nvvPa6//noeeughtm3bRnV1Na+++ir19fUsX74cANM0Of/881m3bh3r1q3D4XB0O0YIIcSR6vHHH2fTpk2sXr2aF154gU8++YRf/vKXuN1ufv3rX1NSUsLGjRvZuHEjpaWlqKrKbbfdxocffsgzzzzDBx98wP/8z/8c8nkMw6C6upqRI0dSVFQEwCOPPML06dNZv349b7/9Npdffnm3+7z77rs899xzPPvss6xcuZI777yTBx98kLfeeoutW7fy0ksv9drOE088kWXLlvG73/2O7du3D8VbJ3KMJHSE6KK5uZn8/PxuQ9gvvfRSpk6dygknnMD69esBOPPMM6mqqkJVVRwOR78ff/bs2axZswaAjo4O3n77bWbPnp3avnr1aubMmQPAnDlzup0IAKiqyo033oiu6zidTp555hluueUWysrK0HWdH/7wh7zyyiskEgkALrzwQrxeL7qus2jRIj777DPa29sH9d4IIUSu6U9MnzZtGuPGjUNVVY4//nhmz559UC2ERYsW4Xa7cTqdQO+x1TAMXn31VRYtWoTL5eKYY45h/vz5qcd58803GTFiBBdccAGapjFhwgRmzZrFyy+/DEA4HObll19m7ty52O12Zs2addBxoKSkhCuuuAJN03A4HDz77LPcfvvt5OXl4fV6ufbaa3nppZcAyM/PZ9asWbhcLrxeL9dff33qOCaEEEeyF198kRtuuIHCwkIKCgq44YYbeq2pAzBp0iQmT56MpmmMHDmSSy65pM94+eSTTzJ16lQqKyt54IEHuOmmm1LTrTRNY8+ePdTX1+NwOA6qY/ODH/wAr9fLsccey3HHHcf06dMZNWoUPp+PU089lT//+c+9Pu+dd97J3Llzeeqpp5g9ezZnn302b731VrfbfO9732Pq1KlMnTqV++67rz9vl8hhMvFaiC7y8vJobm7uVpfgmWeeAZJXZE3TBKC8vHxQjz937lwuvfRS7rnnHl577TUmTJjAiBEjAPi///s/du3alUrwzJkzh2XLlrFlyxbGjx8PJDvnXRNIe/bs4YYbbkBVv87NqqpKY2MjRUVFLFu2jJdffpmmpqbUbZqbm/H5fINqvxBC5JL+xPRNmzbx0EMPsXXrVuLxOLFY7KDh62VlZanfDcPoNbZGIhESiUS3Y0TX33fv3s3mzZu7de4Nw2DevHlAchqupmmpIphz587lqquuoqmpiYKCgoPa0tTURDgc5vzzz0/9z7Ks1LEqHA7z05/+lHfeeYfW1lYAgsEghmEMeZ0HIYTIpPr6eioqKlJ/V1RUUF9f3+vtt23bxs9+9jM+/fRTwuEwhmEwceLEXm+/YMGCVA2drVu3smDBAgKBAKeddho//vGPeeSRR7jwwgsJBAJcddVVXHjhhan7do7kAXA4HAf93Tl1qydOp5PrrruO6667jo6ODlasWMHNN9/MunXryMvLA5IjPaWGzjeHJHSE6KKyshJd1/nDH/7ArFmz0v74xxxzDBUVFbz99tusWbMmNRoHktOtLMvqdjUXkkG5M6GjKEq3bWVlZTzwwAM9Tvuqrq7mD3/4A7/5zW8YOXIk7e3tnHTSSVm7qosQQqRbf2L6j370Iy6//HJWrlyJw+Hg/vvvp7m5udttusbeF198sdfYWlBQgKZp7N27l6OOOgqA2tra1H3Ly8s56aST+M1vftNjW6qrqwmFQpxxxhlAMjkTj8d58cUXU/V7urYlPz8fp9PJSy+9RGlp6UGP9+STT7Jt2zaeffZZiouL2bJlC/Pnz5fjgBDiiFdSUsKePXs49thjgWQsLikpAQ7uTwP8y7/8S6q+mtfr5d///d955ZVXDvk8iqJw3HHHMWXKFN566y1OO+00iouLUyNjNmzYwFVXXcVJJ5004CRLT+3sqnNU5hNPPMGuXbtSCR3xzSJTroTowu/3c8MNN3DPPffw8ssv09HRgWmabNmyhXA43Od9TdMkGo2mfnorpDZnzhz+4z/+g/Xr16euAkejUX7/+9+zZMkSqqurUz933nkna9asSU2hOtD3v/99Hn74YXbv3g0kr9Z21n4IBoPouk5+fj7hcJif//zng31bhBAiJ/UnpgeDQQKBAA6Hg82bN6emxfamr9hqs9k4++yzeeyxxwiHw9TU1LB69erU9tNPP53t27dTXV1NPB4nHo+zefNmampqqKur44MPPuBXv/pV6hiwevVqrrnmmm6P0ZWqqlx00UU88MADNDY2AlBXV8c777yTaqvD4cDv99PS0sJjjz12WO+nEELkitmzZ/P444/T1NREU1MTy5cvT9WtLCwspKWlpVsZgmAwiMfjwePxUFNTw9NPP93v56qpqeHjjz/mmGOOAeD3v/99qgB9IBBAUZRuo+n7q6d2Ll++nM2bNxOLxYhGo/znf/4nfr8/dRFBfPNIQkeIA1xzzTUsXryYlStXMn369NSqIbfeeiuVlZW93m/FihWccMIJqZ/Oq6kHmjNnDuvXr+fkk09ODaF//fXXcTqdzJ8/n+Li4tTPBRdcgGEYqc75gf7u7/6OmTNnsmDBAiorK7n44ovZvHkzAPPnz6eiooJTTjmF2bNnM3ny5MN7Y4QQIgcdKqbffffd/OIXv6CyspLly5enihH35lCx9a677qK9vZ3p06fzT//0T8yePTu1NLrX62XVqlWsXbuWU045hRkzZvDQQw8Ri8VYvXo148ePZ8aMGd2OA1dccQWff/45X3zxRY/t+fGPf8zo0aO5+OKLmTJlCn//93/Ptm3bALjyyiuJRqOcfPLJXHLJJZxyyimH/4YKIUQOWLhwIZMmTWLevHnMmzePiRMnsnDhQgDGjh3L7NmzOeuss5g6dSp1dXX85Cc/Yc2aNUyZMoU777yT7373u30+/qpVq6isrGTy5MlcffXVnH/++Vx66aUAfPLJJ1x00UVUVlZy/fXXc8cddzBq1KgBv4ae2qkoCrfffjsnn3wyp5xyCu+//z5PPPEEHo9n4G+SOCIoloy7FUIIIYQYEg8++CANDQ0sXbo0000RQgghxBFGRugIIYQQQqRJTU0Nn332GZZlsXnzZn77299y9tlnZ7pZQgghhDgCSVFkIYQQQog0CQaD/OhHP6K+vp7CwkIWLFjAmWeemelmCSGEEOIIJFOuhBBCCCGEEEIIIXKMTLkSQgghhBBCCCGEyDGS0BFCCCGEEEIIIYTIMZLQEUIIIYQQQgghhMgxktARQgghhBBCCCGEyDGS0BFCCCGEEEIIIYTIMZLQEUIIIYQQQgghhMgx/x/5AXPUFSWTkgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1152x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(figsize=(16, 5), ncols=3)\n",
    "for i, feature in enumerate([\"GrLivArea\", \"GarageArea\", \"TotalBsmtSF\"]):\n",
    "    axs[i].scatter(X_train[feature], y_train, alpha=0.2)\n",
    "    axs[i].set_xlabel(feature)\n",
    "    axs[i].set_ylabel(\"SalePrice\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Видим, что между этими признаками и целевой переменной действительно наблюдается линейная зависимость."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Часть 2. Первая модель"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Немного об обучении моделей*. В арсенале дата-саентиста кроме `pandas` и `matplotlib` должны быть библиотеки, позволяющие обучать модели. Для простых моделей (линейные модели, решающее дерево, ...) отлично подходит `sklearn`: в нем очень понятный и простой интерфейс. Несмотря на то, что в `sklearn` есть реализация бустинга и простых нейронных сетей, ими все же не пользуются и предпочитают специализированные библиотеки: `XGBoost`, `LightGBM` и пр. для градиентного бустинга над деревьями, `PyTorch`, `Tensorflow` и пр. для нейронных сетей. Так как мы будем обучать линейную регрессию, нам подойдет реализация из `sklearn`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Попробуем обучить линейную регрессию на числовых признаках из нашего датасета. В `sklearn` есть несколько классов, реализующих линейную регрессию:\n",
    "* [`LinearRegression`](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html) &mdash; \"классическая\" линейная регрессия с оптимизацией MSE. Веса находятся как точное решение: $w^* = (X^TX)^{-1}X^Ty$\n",
    "* [`Ridge`](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Ridge.html) &mdash; линейная регрессия с оптимизацией MSE и $\\ell_2$-регуляризацией\n",
    "* [`Lasso`](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Lasso.html) &mdash; линейная регрессия с оптимизацией MSE и $\\ell_1$-регуляризацией\n",
    "\n",
    "У моделей из `sklearn` есть методы `fit` и `predict`. Первый принимает на вход обучающую выборку и вектор целевых переменных и обучает модель, второй, будучи вызванным после обучения модели, возвращает предсказание на выборке. Попробуем обучить нашу первую модель на числовых признаках, которые у нас сейчас есть:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 32085.7681\n",
      "Train RMSE = 35631.3131\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "model = Ridge()\n",
    "model.fit(X_train[numeric_features], y_train)\n",
    "y_pred = model.predict(X_test[numeric_features])\n",
    "y_train_pred = model.predict(X_train[numeric_features])\n",
    "\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))\n",
    "print(\"Train RMSE = %.4f\" % mean_squared_error(y_train, y_train_pred, squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы обучили первую модель и даже посчитали ее качество на отложенной выборке! Давайте теперь посмотрим на то, как можно оценить качество модели с помощью кросс-валидации. Принцип кросс-валидации изображен на рисунке\n",
    "\n",
    "<img src=\"https://docs.splunk.com/images/thumb/e/ee/Kfold_cv_diagram.png/1200px-Kfold_cv_diagram.png\" width=50%>\n",
    "\n",
    "При кросс-валидации мы делим обучающую выборку на $n$ частей (fold). Затем мы обучаем $n$ моделей: каждая модель обучается при отсутствии соответствующего фолда, то есть $i$-ая модель обучается на всей обучающей выборке, кроме объектов, которые попали в $i$-ый фолд (out-of-fold). Затем мы измеряем качество $i$-ой модели на $i$-ом фолде. Так как он не участвовал в обучении этой модели, мы получим \"честный результат\". После этого, для получения финального значения метрики качества, мы можем усреднить полученные нами $n$ значений."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross validation scores:\n",
      "\t -38825.3172\n",
      "\t-45278.9230\n",
      "\t-31405.3380\n",
      "\t-33890.1048\n",
      "\t-73848.3061\n",
      "\t-24142.7915\n",
      "\t-45523.7066\n",
      "\t-52861.2908\n",
      "\t-26094.3376\n",
      "\t-29693.5005\n",
      "Mean CV MSE = 40156.3616\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "cv_scores = cross_val_score(model, X_train[numeric_features], y_train, cv=10, scoring=\"neg_root_mean_squared_error\")\n",
    "print(\"Cross validation scores:\\n\\t\", \"\\n\\t\".join(\"%.4f\" % x for x in cv_scores))\n",
    "print(\"Mean CV MSE = %.4f\" % np.mean(-cv_scores))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Обратите внимание на то, что результаты `cv_scores` получились отрицательными. Это соглашение в `sklearn` (скоринговую функцию нужно максимизировать). Поэтому все стандартные скореры называются `neg_*`, например, `neg_root_mean_squared_error`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Обратите внимание, что по отложенной выборке и при кросс-валидации мы считаем RMSE (Root Mean Squared Error), хотя в функционале ошибки при обучении модели используется MSE.\n",
    "\n",
    "$$\\text{RMSE}(X, y, a) = \\sqrt{\\frac{1}{\\ell}\\sum_{i=1}^{\\ell} (y_i - a(x_i))^2}$$\n",
    "\n",
    "**Вопрос**: Почему оптимизация RMSE эквивалентна оптимизации MSE?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Если бы RMSE не входил в стандартные метрики `sklearn`, можно было бы его определить через функцию `sklearn.metrics.make_scorer`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import make_scorer\n",
    "\n",
    "def rmse(y_true, y_pred):\n",
    "    error = (y_true - y_pred) ** 2\n",
    "    return np.sqrt(np.mean(error))\n",
    "\n",
    "rmse_scorer = make_scorer(\n",
    "    rmse,\n",
    "    greater_is_better=False\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross validation scores:\n",
      "\t -38825.3172\n",
      "\t-45278.9230\n",
      "\t-31405.3380\n",
      "\t-33890.1048\n",
      "\t-73848.3061\n",
      "\t-24142.7915\n",
      "\t-45523.7066\n",
      "\t-52861.2908\n",
      "\t-26094.3376\n",
      "\t-29693.5005\n",
      "Mean CV RMSE = 40156.3616\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "cv_scores = cross_val_score(model, X_train[numeric_features], y_train, cv=10, scoring=rmse_scorer)\n",
    "print(\"Cross validation scores:\\n\\t\", \"\\n\\t\".join(\"%.4f\" % x for x in cv_scores))\n",
    "print(\"Mean CV RMSE = %.4f\" % np.mean(-cv_scores))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Для того, чтобы иметь некоторую точку отсчета, удобно посчитать оптимальное значение функции потерь при константном предсказании.\n",
    "\n",
    "**Вопрос**: Чему равна оптимальная константа для RMSE?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE with best constant = 76636.5188\n",
      "Train RMSE with best constant = 80588.6191\n"
     ]
    }
   ],
   "source": [
    "best_constant = y_train.mean()\n",
    "print(\"Test RMSE with best constant = %.4f\" % mean_squared_error(y_test, best_constant * np.ones(y_test.shape),\n",
    "                                                                 squared=False))\n",
    "print(\"Train RMSE with best constant = %.4f\" % mean_squared_error(y_train, best_constant * np.ones(y_train.shape),\n",
    "                                                                  squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Давайте посмотрим на то, какие же признаки оказались самыми \"сильными\". Для этого визуализируем веса, соответствующие признакам. Чем больше вес &mdash; тем более сильным является признак.\n",
    "\n",
    "**Вопрос**: Почему это не совсем правда?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_weights(features, weights, means, scales):\n",
    "    fig, axs = plt.subplots(figsize=(14, 10), ncols=3)\n",
    "    sorted_weights = sorted(zip(weights, features, means, scales), reverse=True)\n",
    "    weights, features, means, scales = zip(*sorted_weights)\n",
    "    features = list(features)\n",
    "    \n",
    "    sns.barplot(y=features, x=list(weights), ax=axs[0])\n",
    "    axs[0].set_xlabel(\"Weight\")\n",
    "    sns.barplot(y=features, x=list(means), ax=axs[1])\n",
    "    axs[1].set_xlabel(\"Mean\")\n",
    "    sns.barplot(y=features, x=list(scales), ax=axs[2])\n",
    "    axs[2].set_xlabel(\"Scale\")\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_weights(numeric_features, model.coef_, X_train[numeric_features].mean(), X_train[numeric_features].std())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Будем масштабировать наши признаки перед обучением модели. Это, среди, прочего, сделает нашу регуляризацию более честной: теперь все признаки будут регуляризоваться в равной степени. \n",
    "\n",
    "Для этого воспользуемся трансформером [`StandardScaler`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html). Трансформеры в `sklearn` имеют методы `fit` и `transform` (а еще `fit_transform`). Метод `fit` принимает на вход обучающую выборку и считает по ней необходимые значения (например статистики, как `StandardScaler`: среднее и стандартное отклонение каждого из признаков). `transform` применяет преобразование к переданной выборке."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 32085.3150\n",
      "Train RMSE = 35631.2160\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train[numeric_features])\n",
    "X_test_scaled = scaler.transform(X_test[numeric_features])\n",
    "\n",
    "model = Ridge()\n",
    "model.fit(X_train_scaled, y_train)\n",
    "y_pred = model.predict(X_test_scaled)\n",
    "y_train_pred = model.predict(X_train_scaled)\n",
    "\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))\n",
    "print(\"Train RMSE = %.4f\" % mean_squared_error(y_train, y_train_pred, squared=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "means = pd.Series(data=X_train_scaled.mean(axis=0), index=numeric_features)\n",
    "scales = pd.Series(data=X_train_scaled.std(axis=0), index=numeric_features)\n",
    "show_weights(numeric_features, model.coef_, means, scales)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Наряду с параметрами (веса $w$, $w_0$), которые модель оптимизирует на этапе обучения, у модели есть и гиперпараметры. У нашей модели это `alpha` &mdash; коэффициент регуляризации. Подбирают его обычно по сетке, измеряя качество на валидационной (не тестовой) выборке или с помощью кросс-валидации. Посмотрим, как это можно сделать (заметьте, что мы перебираем `alpha` по логарифмической сетке, чтобы узнать оптимальный порядок величины)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best alpha = 297.6351\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'CV score')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "alphas = np.logspace(-2, 3, 20)\n",
    "searcher = GridSearchCV(Ridge(), [{\"alpha\": alphas}], scoring=\"neg_root_mean_squared_error\", cv=10)\n",
    "searcher.fit(X_train_scaled, y_train)\n",
    "\n",
    "best_alpha = searcher.best_params_[\"alpha\"]\n",
    "print(\"Best alpha = %.4f\" % best_alpha)\n",
    "\n",
    "plt.plot(alphas, -searcher.cv_results_[\"mean_test_score\"])\n",
    "plt.xscale(\"log\")\n",
    "plt.xlabel(\"alpha\")\n",
    "plt.ylabel(\"CV score\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Вопрос**: Почему мы не подбираем коэффициент регуляризации по обучающей выборке? По тестовой выборке?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Попробуем обучить модель с подобранным коэффициентом регуляризации. Заодно воспользуемся очень удобным классом [`Pipeline`](https://scikit-learn.org/stable/modules/generated/sklearn.pipeline.Pipeline.html): обучение модели часто представляется как последовательность некоторых действий с обучающей и тестовой выборками (например, сначала нужно отмасштабировать выборку (причем для обучающей выборки нужно применить метод `fit`, а для тестовой &mdash; `transform`), а затем обучить/применить модель (для обучающей `fit`, а для тестовой &mdash; `predict`). `Pipeline` позволяет хранить эту последовательность шагов и корректно обрабатывает разные типы выборок: и обучающую, и тестовую."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 31703.3590\n"
     ]
    }
   ],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "simple_pipeline = Pipeline([\n",
    "    ('scaling', StandardScaler()),\n",
    "    ('regression', Ridge(best_alpha))\n",
    "])\n",
    "\n",
    "model = simple_pipeline.fit(X_train[numeric_features], y_train)\n",
    "y_pred = model.predict(X_test[numeric_features])\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Часть 3. Работаем с категориальными признаками"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Сейчас мы явно вытягиваем из данных не всю информацию, что у нас есть, просто потому, что мы не используем часть признаков. Эти признаки в датасете закодированы строками, каждый из них обозначает некоторую категорию. Такие признаки называются категориальными. Давайте выделим такие признаки и сразу заполним пропуски в них специальным значением (то, что у признака пропущено значение, само по себе может быть хорошим признаком)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "categorical = list(X_train.dtypes[X_train.dtypes == \"object\"].index)\n",
    "X_train[categorical] = X_train[categorical].fillna(\"NotGiven\")\n",
    "X_test[categorical] = X_test[categorical].fillna(\"NotGiven\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>MSZoning</th>\n",
       "      <th>Street</th>\n",
       "      <th>Alley</th>\n",
       "      <th>LotShape</th>\n",
       "      <th>LandContour</th>\n",
       "      <th>Utilities</th>\n",
       "      <th>LotConfig</th>\n",
       "      <th>LandSlope</th>\n",
       "      <th>Neighborhood</th>\n",
       "      <th>Condition1</th>\n",
       "      <th>...</th>\n",
       "      <th>GarageType</th>\n",
       "      <th>GarageFinish</th>\n",
       "      <th>GarageQual</th>\n",
       "      <th>GarageCond</th>\n",
       "      <th>PavedDrive</th>\n",
       "      <th>PoolQC</th>\n",
       "      <th>Fence</th>\n",
       "      <th>MiscFeature</th>\n",
       "      <th>SaleType</th>\n",
       "      <th>SaleCondition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>RL</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>IR1</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>BrkSide</td>\n",
       "      <td>RRAn</td>\n",
       "      <td>...</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>Unf</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>MnPrv</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>837</th>\n",
       "      <td>RM</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>BrDale</td>\n",
       "      <td>Norm</td>\n",
       "      <td>...</td>\n",
       "      <td>Detchd</td>\n",
       "      <td>Unf</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1094</th>\n",
       "      <td>RL</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>NAmes</td>\n",
       "      <td>Norm</td>\n",
       "      <td>...</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>Fin</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1291</th>\n",
       "      <td>RM</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>BrDale</td>\n",
       "      <td>Norm</td>\n",
       "      <td>...</td>\n",
       "      <td>Detchd</td>\n",
       "      <td>Unf</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>436</th>\n",
       "      <td>RM</td>\n",
       "      <td>Pave</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>OldTown</td>\n",
       "      <td>Norm</td>\n",
       "      <td>...</td>\n",
       "      <td>Detchd</td>\n",
       "      <td>Unf</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>NotGiven</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 43 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     MSZoning Street     Alley LotShape LandContour Utilities LotConfig  \\\n",
       "184        RL   Pave  NotGiven      IR1         Lvl    AllPub    Inside   \n",
       "837        RM   Pave  NotGiven      Reg         Lvl    AllPub    Inside   \n",
       "1094       RL   Pave  NotGiven      Reg         Lvl    AllPub    Inside   \n",
       "1291       RM   Pave  NotGiven      Reg         Lvl    AllPub    Inside   \n",
       "436        RM   Pave  NotGiven      Reg         Lvl    AllPub    Inside   \n",
       "\n",
       "     LandSlope Neighborhood Condition1  ... GarageType GarageFinish  \\\n",
       "184        Gtl      BrkSide       RRAn  ...     Attchd          Unf   \n",
       "837        Gtl       BrDale       Norm  ...     Detchd          Unf   \n",
       "1094       Gtl        NAmes       Norm  ...     Attchd          Fin   \n",
       "1291       Gtl       BrDale       Norm  ...     Detchd          Unf   \n",
       "436        Gtl      OldTown       Norm  ...     Detchd          Unf   \n",
       "\n",
       "     GarageQual GarageCond PavedDrive    PoolQC     Fence MiscFeature  \\\n",
       "184          TA         TA          Y  NotGiven     MnPrv    NotGiven   \n",
       "837          TA         TA          Y  NotGiven  NotGiven    NotGiven   \n",
       "1094         TA         TA          Y  NotGiven  NotGiven    NotGiven   \n",
       "1291         TA         TA          Y  NotGiven  NotGiven    NotGiven   \n",
       "436          TA         TA          Y  NotGiven  NotGiven    NotGiven   \n",
       "\n",
       "     SaleType SaleCondition  \n",
       "184        WD        Normal  \n",
       "837        WD        Normal  \n",
       "1094       WD        Normal  \n",
       "1291       WD        Normal  \n",
       "436        WD        Normal  \n",
       "\n",
       "[5 rows x 43 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[categorical].sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Сейчас нам нужно как-то закодировать эти категориальные признаки числами, ведь линейная модель не может работать с такими абстракциями. Два стандартных трансформера из `sklearn` для работы с категориальными признаками &mdash; [`OrdinalEncoder`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OrdinalEncoder.html) (просто перенумеровывает значения признака натуральными числами) и [`OneHotEncoder`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OneHotEncoder.html).\n",
    "\n",
    "`OneHotEncoder` ставит в соответствие каждому признаку целый вектор, состоящий из нулей и одной единицы (которая стоит на месте, соответствующем принимаемому значению, таким образом кодируя его).\n",
    "\n",
    "**Вопрос**: Проинтерпретируйте, что означают веса модели перед OneHot-кодированными признаками. Почему пользоваться `OrdinalEncoder` в случае линейной модели &mdash; скорее плохой вариант? Какие недостатки есть у OneHot-кодирования?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 26282.2264\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.compose import ColumnTransformer\n",
    "\n",
    "column_transformer = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer),\n",
    "    ('regression', Ridge())\n",
    "])\n",
    "\n",
    "model = pipeline.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Вопрос**: Как вы думаете, почему мы не производим скейлинг OneHot-кодированных признаков?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на размеры матрицы после OneHot-кодирования:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size before OneHot: (1022, 79)\n",
      "Size after OneHot: (1022, 294)\n"
     ]
    }
   ],
   "source": [
    "print(\"Size before OneHot:\", X_train.shape)\n",
    "print(\"Size after OneHot:\", column_transformer.transform(X_train).shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как видим, количество признаков увеличилось более, чем в 3 раза. Это может повысить риски переобучиться: соотношение количества объектов к количеству признаков сильно сократилось.\n",
    "\n",
    "Попытаемся обучить линейную регрессию с $\\ell_1$-регуляризатором. На лекциях вы узнаете, что $\\ell_1$-регуляризатор разреживает признаковое пространство, иными словами, такая модель зануляет часть весов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE = 26468.4015\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import Lasso\n",
    "\n",
    "column_transformer = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "lasso_pipeline = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer),\n",
    "    ('regression', Lasso())\n",
    "])\n",
    "\n",
    "model = lasso_pipeline.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "print(\"RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Zero weights in Ridge: 0\n",
      "Zero weights in Lasso: 29\n"
     ]
    }
   ],
   "source": [
    "ridge_zeros = np.sum(pipeline.steps[-1][-1].coef_ == 0)\n",
    "lasso_zeros = np.sum(lasso_pipeline.steps[-1][-1].coef_ == 0)\n",
    "print(\"Zero weights in Ridge:\", ridge_zeros)\n",
    "print(\"Zero weights in Lasso:\", lasso_zeros)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Подберем для нашей модели оптимальный коэффициент регуляризации. Обратите внимание, как перебираются параметры у `Pipeline`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best alpha = 263.6651\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'CV score')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "alphas = np.logspace(-2, 4, 20)\n",
    "searcher = GridSearchCV(lasso_pipeline, [{\"regression__alpha\": alphas}],\n",
    "                        scoring=\"neg_root_mean_squared_error\", cv=10, n_jobs=-1)\n",
    "searcher.fit(X_train, y_train)\n",
    "\n",
    "best_alpha = searcher.best_params_[\"regression__alpha\"]\n",
    "print(\"Best alpha = %.4f\" % best_alpha)\n",
    "\n",
    "plt.plot(alphas, -searcher.cv_results_[\"mean_test_score\"])\n",
    "plt.xscale(\"log\")\n",
    "plt.xlabel(\"alpha\")\n",
    "plt.ylabel(\"CV score\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 25642.6115\n"
     ]
    }
   ],
   "source": [
    "column_transformer = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer),\n",
    "    ('regression', Lasso(best_alpha))\n",
    "])\n",
    "\n",
    "model = pipeline.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Zero weights in Lasso: 218\n"
     ]
    }
   ],
   "source": [
    "lasso_zeros = np.sum(pipeline.steps[-1][-1].coef_ == 0)\n",
    "print(\"Zero weights in Lasso:\", lasso_zeros)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Иногда очень полезно посмотреть на распределение остатков. Нарисуем гистограмму распределения квадратичной ошибки на обучающих объектах:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='SalePrice', ylabel='Density'>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "error = (y_train - model.predict(X_train)) ** 2\n",
    "sns.distplot(error)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Как видно из гистограммы, есть примеры с очень большими остатками. Попробуем их выбросить из обучающей выборки. Например, выбросим примеры, остаток у которых больше 0.95-квантили."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "mask = (error < np.quantile(error, 0.95))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 24963.5922\n"
     ]
    }
   ],
   "source": [
    "column_transformer = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer),\n",
    "    ('regression', Lasso(best_alpha))\n",
    "])\n",
    "\n",
    "model = pipeline.fit(X_train[mask], y_train[mask])\n",
    "y_pred = model.predict(X_test)\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train[mask]\n",
    "y_train = y_train[mask]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='SalePrice', ylabel='Density'>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "error = (y_train[mask] - model.predict(X_train[mask])) ** 2\n",
    "sns.distplot(error)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Видим, что качество модели заметно улучшилось! Также бывает очень полезно посмотреть на примеры с большими остатками и попытаться понять, почему же модель на них так сильно ошибается: это может дать понимание, как модель можно улучшить."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Часть 4. Подготовка данных для линейных моделей"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Есть важное понятие, связанное с применением линейных моделей, — *спрямляющее пространство*. Под ним понимается такое признаковое пространство для наших объектов, в котором линейная модель хорошо описывает данные, даёт хорошее качество прогнозов.\n",
    "\n",
    "Не существует общих рекомендаций о том, как найти спрямляющее пространство для произвольной выборки. Есть лишь некоторые общие советы — например, если добавить в выборку полиномиальных признаков, то скорее всего модель станет работать лучше (если не переобучится). Пример про этот подход мы обсуждали в начале семинара. Есть и другие трюки."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "У линейных моделей есть огромное преимущество: они имеют мало параметров, а поэтому их можно обучить даже на небольшой выборке. Если выборка большая, то параметры модели получится оценить более надёжно — но в то же время есть риск, что данные будут слишком разнообразными, чтобы линейная модель могла уловить все закономерности в них. Иногда можно улучшить ситуацию путём разбиения признакового пространства на несколько областей и построения своей модели в каждой из них.\n",
    "\n",
    "Попробуем для примера в нашей задаче разделить выборку на две части по признаку OverallQual. Это один из самых сильных признаков, и, возможно, разбиение по нему даст нам две выборки с заведомо разными ценами на дома.\n",
    "\n",
    "Для начала вспомним, какое качество получается у обычной гребневой регрессии."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 23427.2492\n"
     ]
    }
   ],
   "source": [
    "column_transformer = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer),\n",
    "    ('regression', Ridge())\n",
    "])\n",
    "\n",
    "model = pipeline.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на связь OverallQual и целевой переменной."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'SalePrice')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdkAAAGvCAYAAAAXPK/eAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAAA6jUlEQVR4nO3dfVTUdb4H8PcMCD4ADhDGoJVFm86iLa5UhzWj0EQ8Ph3vdiI2183Syl3TDG6eMPD6kAt620otbdO27Vru3c2VJBXbi1JyuK1e9a6gZkvploDKAJeHBGR+v/sHyyjIwPwGfk9f3q9zOke+Xwa+n2aYz3yfLbIsyyAiIqI+Z9W7AURERKJikiUiIlIJkywREZFKmGSJiIhUwiRLRESkEiZZIiIilTDJEhERqcRf7waYUU1NIyTJ9+3F4eFBcDob+rBF+mEsxiRKLKLEATAWo+ptLFarBaGhQzzWM8n6QJLkXiXZ9p8hCsZiTKLEIkocAGMxKjVj4XAxERGRSphkiYiIVMIkS0REpBImWSIiIpUwyRIREamESZaIiEglTLJEREQq0TzJbtq0CaNGjcLZs2cBAKNGjcKMGTMwa9YszJo1C19++aX7ewsKCjB16lQ8/PDDWLp0Ka5cuaJqHRERUV/SNMmWlpbixIkTGD58eIfynTt3Ijc3F7m5uRg1ahQAoLGxES+//DK2bNmCTz/9FEOGDMG2bdtUqyMiIuprmiXZlpYWrFq1CitXrvTq+z/77DOMGTMGI0eOBACkpKRg3759qtURERH1Nc2OVXz99dcxc+ZMjBgx4oa6uXPnwuVy4YEHHsDixYsREBCAiooKREVFub8nKioKFRUVAKBKHRERUV/TJMkeP34cJSUlSEtLu6Hu0KFDsNvtaGhoQHp6OjZv3oznn39ei2b5LDw8qNc/IyIiuA9aYgyMxZhEiUWUOADzx3Lof77F7/edRlXNFdwUOgg/T3bgwfG36N2sXlPzedEkyR45cgRlZWWYNGkSAKCyshJPPvkk1q1bh/vvvx8AEBQUhEceeQTvvvsuAMBut+OLL75w/4zy8nLY7XbV6pRwOht6daB0REQwLl+u9/nxRsJYjEmUWESJAzB/LMWllXhv3xm0tEoAgMs1V7DxP0+grr4J8TGROrfOd719XqxWS7cdL03mZBcuXIjDhw+joKAABQUFiIyMxLZt2zB27Fg0NTUBAFpbW5Gfnw+HwwEAmDhxIk6ePIlz584BaFsclZycrFodERF5tquwzJ1g27W0SthVWKZTi8xB16vuvv76a2RmZsJisaC1tRXjxo3DkiVLALT1bFetWoWnn34akiTB4XAgIyNDtToiIvLMWdesqJzaWGRZFudSQI1wuPgaxmJMosQiShyA+WNJf7Ooy4QaHhKI9Ysm6NCiviHEcDEREZnbnIRoBPh3TBkB/lbMSYjWqUXmoOtwMRERmUP74qZdhWWormtGWEgg5iREm3bRU3FppSaxMMkSEZFX4mMiER8Tafqh784rpZ11zXhv3xkA6PNEyyRLRERe0ar3p7buVkozyRIRkea07P2pTcuV0lz4REREPRJpn2x4SKCi8t5gkiUioh6JtE9Wy5XSHC4mIqIehYcEetwnazZarpRmkiUioh7NSYjuMCcLmHufrFYrpZlkiYioR6Ltk9UKkywREXlFlH2yWuLCJyIiIpWwJ0tERP0Oj1UkIiJSgZYHa3C4mIiI+hUtD9ZgkiUion6FxyoSERGphMcqEhERqYTHKhIREamExyoSERGpSKuDNThcTEREpBImWSIiIpUwyRIREamESZaIiEglXPhERERe0eq8X5EwyRIRUY+0PO9XJEyyREQqEqX31915v2aMRytMskREKhGp96fleb8i4cInIiKVaHnbi9q0PO9XJEyyREQqEan3p+V5vyLhcDERkUrCQwK7TKhm7P1ped6vSJhkiYhUMichusOcLGDu3p9W5/2KhEmWiEgl7P0Zl1arvplkiYhUxN6f8Wi56lvzhU+bNm3CqFGjcPbsWQDAiRMnMHPmTCQlJWH+/PlwOp3u79W6joiIxKflqm9Nk2xpaSlOnDiB4cOHAwAkSUJ6ejoyMzORn5+PuLg4bNiwQZc6IiLqH7Rc9a1Zkm1pacGqVauwcuVKd1lJSQkCAwMRFxcHAEhJScH+/ft1qSMiov5Byz2/ms3Jvv7665g5cyZGjBjhLquoqEBUVJT767CwMEiShNraWs3rbDab17GEhwcpDf8GERHBvf4ZRsFYjEmUWESJA2AsRvGL6THY9Mf/RfNVl7sscIAffjE9ps/j0iTJHj9+HCUlJUhLS9Pi16nO6WyAJMk+P16kBRCMxZhEiUWUOADGYiQxt9rw86mjblhdHHOrTXFcVqul246XJkn2yJEjKCsrw6RJkwAAlZWVePLJJzF37lyUl5e7v6+6uhpWqxU2mw12u13TOiIiNYhyQQAgVixarfrWZE524cKFOHz4MAoKClBQUIDIyEhs27YNTz31FJqamnD06FEAwM6dOzF16lQAwJgxYzStIyLqa+1bRZx1zZBxbatIcWml3k1TTKRYtKTrPlmr1YqcnBxkZWWhubkZw4cPx/r163WpIyLqayJdDydSLFqyyLLs++RiP8U52WsYizGJEovZ45j/6wKPdduXJ2rYkt4TKZbr9fY11tOcLG/hISJSiUjXw4kUi5aYZImIVCLS9XAixaIlnl1MRKQSkS4IECkWLTHJEhGpSKQLAkSKhbfwEBGRoYiyT7a4tBLb807B9c/1q866ZmzPOwVAgFt4iIjIfETaJ/vBp1+6E2w7l9xW3teYZImIqEdaXg+ntsYml6Ly3mCSJSKiHml5PZxIOCdLRKQiUeYxw0MCu0yoZtwnGzTIHw1XWrss72vsyRIRqUSkecy7o8MVlRvZY5Pvgr+fpUOZv58Fj02+q89/F5MsEZFKRJrH/FuZU1G5kcXHROKJaQ6EhwTCgrbe+BPTHNzCQ0RkJiLNY4oUCyDYVXdERP0Rz/slJlkiIpXwvF/icDERkUp43i8xyRIRqUiU834DB/ih+eqNhzUEDvDToTXmweFiIiLqkb+HXOqpnNowyRIRUY+0PIpQJEyyRETUI6tFWTm1YZIlIqIeSbKycmrDJEtERD0aMrDryVdP5dSGSZaIiHpksXQ9LuypnNowyRIRUY+6urWmu3JqwyRLREQ94hGRvuFhFEREKhLlPtk5CdF4b9+ZDrcK8YjInjHJEhGppP0+2fbE1H6fLADTJVoeEekbDhcTEalEpPtkyTfsyRIRqUSkO1hF6pVriT1ZIiKViLRYiL1y3zDJEhGpRKT7ZEXqlWuJw8VERCoRabFQeEhglwnVjL1yLbEnS0REPRKpV64l9mSJiFQi0mIhkXrlWtIsyS5atAjfffcdrFYrBg8ejJdffhkOhwOJiYkICAhAYGDbkENaWhomTpwIADhx4gQyMzPR3NyM4cOHY/369QgPD1etjoioL3W3WMiMySk+JhLxMZGIiAjG5cv1ejfHFDQbLs7OzsbHH3+M3bt3Y/78+XjppZfcdW+88QZyc3ORm5vrTrCSJCE9PR2ZmZnIz89HXFwcNmzYoFodEVFfE22xUHFpJdLfLMLMF3KR/mYRiksr9W6S4WmWZIODg93/bmho6PHmhpKSEgQGBiIuLg4AkJKSgv3796tWR0TU10TawtM+9O2sa4aMa0PfTLTd03RONiMjA0VFRZBlGe+88467PC0tDbIsY/z48Vi2bBlCQkJQUVGBqKgo9/eEhYVBkiTU1taqUmez2dQNnoj6HZHO+xVt6FsrmibZtWvXAgB2796NnJwc/Pa3v8WOHTtgt9vR0tKCtWvXYtWqVYYfwg0PD+r1z4iICO75m0yCsRiTKLGYOY6ZDwYjJHggfr/vNKpqruCm0EH4ebIDD46/Re+mKVbtYYi7uq7Z1M8RoO5rTJfVxbNnz0ZmZiZqampgt9sBAAEBAUhNTcWzzz4LALDb7SgvL3c/prq6GlarFTabTZU6JZzOBkiS7EvoACDUogHGYkyixCJCHHX1TZBcbT1AySWhrr7JlDGFedgnGxYSaMp42vX2NWa1WrrteGkyJ9vY2IiKigr31wUFBRg6dCgCAwNRX98WnCzL2Lt3LxwOBwBgzJgxaGpqwtGjRwEAO3fuxNSpU1WrIyLqayLNY3KfrG806cleuXIFS5YswZUrV2C1WjF06FBs2bIFTqcTixcvhsvlgiRJiI6ORlZWFgDAarUiJycHWVlZHbbbqFVHRNTXRJrHjI+JxN+/q0XhiXJIMmC1ABPGRpouDq1ZZFn2fdyzn+Jw8TWMxZhEicXsccz/dYHHuu3LEzVsSe91PlgDaOvJzksebepEK8RwMRFRfyTSFh7ewuMbJlkiIpXMSYhG5yMBLBaYch5TtIM1tMIkS0Skkr9/V4vOE3Ky3FZuNiL1yrXEJEtEpJLCE+WKyo1MtNXFWh0RyVt4iIhU4ml9ZC/WTepGpFt4tLwdiUmWiAynuLRSiDdz0YhyC4+WW6uYZInIUES6g1U0onz40XIRF+dkichQRNoqItJiIZFOr9LyeWGSJSJDEWmriEiLhUT68KPl88LhYiIylHAPB9Gbsfcn0mIhkT78aPm8MMkSkaGIdAcrIM5iIZE+/ADaPS8cLiYiQ4mPicSEsZGw/vOkJB5EbwwiDX1riUmWiAyluLQSRScr3XtJJRkoOllpygU2IuGHH99wuJiIDEWk6+EAYP2Hx3D6fK37a8dtNqQ/9mP9GuSj4tJKFJ6o6PDhp/BEBe4cYTPl86LVdiT2ZInIUERaYNM5wQLA6fO1WP/hMX0a1Au/3/8lpE4HMUuyjN/v/1KnFvlOy+1ITLJEZCgi7S3tnGB7Kjey5qsuReVGpuV2JCZZIjIULrAhtWk5WsI5WSIyFJH2lpIxabkdiT1ZIiKVOG6zKSo3sofGRSkqNzItR0uYZInIUEQ6I/f+u7tOQJ7KSRvxMZGYlzwa4SGBsKCtBzsveTRPfCIi8Ym0hcfTytvf7//SdLEcPN71RfMHj5djbtJojVvTezzxiYj6JZG28Ii0Ipd8wyRLRIYi0hYeIiZZIjIUbuEhkXBOlogMhVt4SCTsyRIREamEPVkiMpT2LTztK4zbt/AAMF1v1mqB+0D9zuVmY7EAchexWEwYi5bYkyUiQ9HyXFm1dZVguys3sq4SbHfl1IZJlogMRaQtPJ56rGbsyXLVt2+YZInIUER6MxepJ8tV375hkiUiQxHpzTxoUNfLXjyVG5mWRxGKxHzPNBEJTaQtPC0eTnbyVG50Wh1FKBImWSIyHFHezFtaux4X9lRO4mGSJSLDKS6tFKInS6TZnOyiRYswc+ZMzJ49G6mpqTh9+jQA4JtvvsGjjz6KpKQkPProozh37pz7MVrXEZH+iksrsT3vVIer7rbnnTLlVXee9pByb2n/oVmSzc7Oxscff4zdu3dj/vz5eOmllwAAWVlZSE1NRX5+PlJTU5GZmel+jNZ1RKS/Dz79Eq5Oo6kuua3cbLi3lDRLssHBwe5/NzQ0wGKxwOl04tSpU5g+fToAYPr06Th16hSqq6s1ryMiY2hs6npRkKdyIxNpOxL5RtM52YyMDBQVFUGWZbzzzjuoqKjAzTffDD8/PwCAn58fhg0bhoqKCsiyrGldWFiYlv8riKgfmJMQjXf3nkbrdV1zfz+LKbcjkW80TbJr164FAOzevRs5OTlYsmSJlr++z4SHB/X6Z0REBPf8TSbBWIzJrLEEDx6A+u+vdllutphCgmtvGBqWZSAkeKDpYunM7O2/npqx6LK6ePbs2cjMzERkZCQuXrwIl8sFPz8/uFwuXLp0CXa7HbIsa1qnhNPZAKkXR7aYfVvC9RiLMZk5lpRJP+iy95cy6Qemi+l3eaVwdXqvcEkyfpdXiphbbfo0qg+Y+fXVWW9jsVot3Xa8NJmTbWxsREVFhfvrgoICDB06FOHh4XA4HMjLywMA5OXlweFwICwsTPM6IjKG+JhIPDHN0eFkoSemOUy5hUekc5jJNxZZVn+dW1VVFRYtWoQrV67AarVi6NChePHFFxETE4OysjIsX74cdXV1CAkJQXZ2Nu644w4A0LzOW+zJXsNYjEmUWMweR/qbRV0m1PCQQKxfNEGHFvUNsz8v11O7J6tJkhUNk+w1jMWYRInF7HG8n38GB4+X31D+0LgozE0arUOL+obZn5frqZ1keeITERmOKCc+/a3MqaicxMMkS0SG0n7iU/u6p/YTnwCYLtFyTpZ41R0RGYpIJz4RMckSkaGIdOITEZMsERGRSphkichQggZ1vVTEUzmRkTHJEpGh3DN6mKJyIiNjkiUiQ+G2FxIJkywRGQq3vZBImGSJyFBEuoPVcZtNUTmJh0mWiAxlWOggReVG9u2lBkXlJB4u1yMSiAjHEX75j1pF5UbWcKVVUbnRifD60hqTLJEgiksr8d6+M2hplQC0zWG+t+8MAHMdR+jp7o1e3MlBfUCU15fWOFxMJIhdhWXuN8B2La0SdhWW6dQiEglfX75hkiUSBFflkpr4+vINkyyRIKwWZeVGJdLqYpHwefENkyyRIESZy5yTEI0A/45vTQH+VsxJiNapRQTwefEVkyyRIETpacTHRGLC2Eh3D9xqASaMjeTiGp3Fx0RiXvJohIcEwoK219W85NF8XnqgeHVxRUUFLl68iNjYWBWaQ0S+mpMQ3WH1J2DOnkZxaSWKTla6e+CSDBSdrMSdI2x8Q9dZfEzbh52IiGBcvlyvd3NMweuebHl5OVJSUpCcnIwnnngCALB//35kZGSo1jgi8p4oPQ2uYiWReN2TzczMxIMPPogPPvgA9913HwBgwoQJyM7OVq1xRKSMCD0NrmIlkXjdkz158iQWLlwIq9UKi6VtsiQ4OBj19eb8QyYiIlKb1z3Z8PBwnD9/Hrfffru77O9//zvsdrsqDSMiImPhsYrKed2TnT9/Pp555hl89NFHaG1tRV5eHp5//nksWLBAzfYREZmWxcMeZU/lRtZ+rKKzrhkyrh2rWFxaqXfTDM3rnuxPf/pT2Gw2/OEPf4Ddbsfu3buxZMkSTJ48Wc32ERGZluxhj7KnciPrbkEae7OeKdrCM3nyZCZVIqJ+iAvSfOP1cPGaNWtw7NixDmXHjh3D2rVr+7xRROSb4tJKpL9ZhJkv5CL9zSIO5VGfEeWwE615nWTz8vIwZsyYDmVjxoxBXl5enzeKiJTjnBmpiccq+sbrJGuxWCB3mkhwuVyQJMnDI4hISzzEgdQkymEnWvN6TjYuLg6vvfYa0tPTYbVaIUkSNm7ciLi4ODXbR0Re4pyZ8USFD0K580qX5WYkwmEnWvM6yWZkZODpp5/G/fffj6ioKFRUVCAiIgJbtmxRs31ERKblrGtRVE7i8TrJRkZG4s9//jP+93//F5WVlbDb7bj77rthtfIiHyKirjRfdSkqJ/Eo2sJjtVoxbtw4tdpCRISgQf5ouNLaZTmR2XT7qk1OTsa+ffsAAAkJCe4zizs7dOhQnzeMiJQRJTl1FUN35URG1u1f3+rVq93/Xr9+vc+/pKamBv/6r/+Kf/zjHwgICMBtt92GVatWISwsDKNGjcJdd93lHnbOycnBqFGjAAAFBQXIycmBy+VCTEwM1q1bh0GDBqlWR2Rmj02+C+/uPY1W17VdAP5+Fjw2+S4dW6Wc1QL3XbKdy4nMptsJ1faVwy6XCx999BFiY2Nx77333vBfTywWC5566ink5+djz549uOWWW7BhwwZ3/c6dO5Gbm4vc3Fx3gm1sbMTLL7+MLVu24NNPP8WQIUOwbds21eqIzC4+JhJPTHN02GLxxDSH6bZYdJVguysnMjKvVi35+fmhqKjI43BxT2w2m/sOWgCIjY1FeXl5t4/57LPPMGbMGIwcORIAkJKS4h66VqOOSATxMZFYv2gCPv73WVi/aILpEiyRaLyerJk3bx42btyIxYsXY8CAAT7/QkmS8OGHHyIxMdFdNnfuXLhcLjzwwANYvHgxAgICUFFRgaioKPf3tG8bAqBKnRLh4UGKH9NZRERwr3+GUTAWYxIplnYixWT2WMze/uupGYvXSfY//uM/UFVVhXfffRdhYWEderVKFj6tXr0agwcPxuOPP+5+rN1uR0NDA9LT07F582Y8//zz3kegA6ezAVIvxq5E2sjNWIxJpFiuJ1JMZo5FpNdXb2OxWi3ddry8TrK9WfjULjs7G+fPn8eWLVvcC53aL30PCgrCI488gnfffddd/sUXX7gfW15e7v5eNeqIiIj6mldzsocPH0ZpaSkkSfJp4RMAvPrqqygpKcHmzZsREBAAAPi///s/NDU1AQBaW1uRn58Ph8MBAJg4cSJOnjyJc+fOAWhbHJWcnKxaHRERUV/rsSf79ttv480338Sdd96JN954A2lpafjZz36m6Jd89dVX2Lp1K0aOHImUlBQAwIgRI/DUU08hMzMTFosFra2tGDduHJYsWQKgrWe7atUqPP3005AkCQ6HAxkZGarVERER9TWL3PlqnU4mTZqEf//3f0dsbCz+53/+B5mZmfjkk0+0ap8hcU72GsZiTGaOZf6vCzzWbV+e6LHOiESK5Xpmfn11pvucbE1NDWJjYwEA48ePR1VVlc+NISJ1FZdWYldhGarrmhEWEog5CdHcxqMjfz9Lh8NBri+n/sGrOVlZliFJElwuF2RZdn/d/h8R6a+4tBLv7j3d4dL2d/ee5qXtOvrBiKGKykk8PfZkv//+e/zwhz90fy3LsvtrWZZhsVhw+vRp9VpIRF758C9nb+g1tbpkfPiXs+zN6uTMP2oVlZN4ekyy//Vf/6VFO4iol3iwvvF4WvHS/UoYEkmPSXb48OEdvpYkCVVVVRg2bJhqjSIiIhKB1zeu19XV4YUXXsDdd9+NKVOmAGjr5f7mN79RrXFERERm5nWSzcrKQlBQEAoKCtxnF48bN44H7BMZhKf1qlzHSqQfr49VLC4uxueff44BAwa4zy0OCwuD0+lUrXFE5D1P03yc/iPSj9c92eDgYNTU1HQoKy8vR0RERJ83ioiUCw8JVFROROrzuif7yCOP4LnnnsPSpUshSRKOHz+OV1991X1MIpFZiXKAw5yEaLy37wxaWq/tXQ/wt2JOQrSOrSLq37xOsgsWLEBgYCBWrVqF1tZWvPTSS3j00Ucxb948NdtHpKri0soOiclZ14z39p0BANMl2vb2ivCBgUgUXidZi8WCefPmMamSUHYVlnXo+QFAS6uEXYVlpkxO8TGRiI+JFOpsWSIz6zbJFhcXe/VD4uPj+6QxRFpz1jUrKiciUqLbJOvNNXAWi4WnQpFpDRnoh8YmV5flZiTK/DKRKLpNsgUFnq9pIhJB+3Y0b8uNrLi0EtvzTqH9+GJnXTO2550CYL75ZSJReL2Fh0hEIp33+8GnX6LzrWouua2c9OFpRMSsIyWknNcLnxoaGrBx40YcOXIENTU1uP6u90OHDqnRNiLVhYcEdjn/asa9pV0Ne3dXTupLfXhUh9EFAPCztJVT/+B1T3blypU4deoUFi1ahNraWqxYsQJ2ux2/+MUvVGwekbrmJEQjwL/jnwH3llJfiY+JxAOxUbD+c/bBagEeiI3i8H0/4nVPtqioCHv37kVoaCj8/PwwefJkjB07Fs888wwTLZkW95aSmopLK1F0shLSP3uykgwUnazEnSNsfI31E14nWUmSEBwcDAAYPHgw6uvrERERgfPnz6vWOCItcG8pqUW0fdiknNdJdvTo0Thy5Aji4+Mxfvx4rFy5EkOGDMHIkSNVbB4RkXlxHzZ5PSe7Zs0a9wXuK1aswMCBA1FfX4+cnBzVGkdE3rN62HXkqZzUx0sbqMeebElJCQICAnDXXXcBAKqrq/HKK6/g7NmzGDduHOx2u+qNJKKeSR7utPNUTuobFjqoy17rsNBBOrSG9NBjT/aVV15BVVWV++uMjAycO3cOKSkp+Oqrr7B+/XpVG0hE3vF0foYJz9UQxpnztYrKSTw9JtmysjLExcUBAOrq6vD5559jw4YN+NnPfoZXX30VBw8eVL2RRNQz2UOP1VM5qc/T/3o+Jf1Hj0nW5XJhwIABAIATJ07gpptuwu233w4AsNvtqKurU7eFREREJtVjkr3zzjuxb98+AMDevXs73Lhz8eJF97YeIiLqyM/DO6ynchJPjwuf0tLS8Oyzz2LlypWwWq344IMP3HV79+7Fj3/8Y1UbSKQ23lxDavH384NLuvFYS38/nl3cX/SYZOPi4nDw4EGcO3cOI0eORFBQkLsuISEB06ZNU7WBRGoqLq3Ee/vOuA8McNY14719ZwDw5hrqvearXZ8b7amcxOPVoEVQUBDGjBnTIcECwB133IGbb75ZlYYRaaG7E3mIiHqLMwPUr/FEHlITr7ojJlnq13giD6kp9eFR6LxN2QJeddefMMlSv8ar7kh1XWVZ6jc0SbI1NTVYsGABkpKSMGPGDPzqV79CdXU1gLa9tzNnzkRSUhLmz58Pp9PpfpzWddT/xMdEYsLYyA73fU4YG8lFT9QnPvj0yxsOA5HltnLqHzRJshaLBU899RTy8/OxZ88e3HLLLdiwYQMkSUJ6ejoyMzORn5+PuLg4bNiwAQA0r6P+ydN9n8Wllfo2zAc8VtF4Gpu6XkXsqZzEo0mStdlsuO+++9xfx8bGory8HCUlJQgMDHQf25iSkoL9+/cDgOZ11D+JtLqYxyoSGY/mc7KSJOHDDz9EYmIiKioqEBUV5a4LCwuDJEmora3VvI76J64uJiI1eX1pe19ZvXo1Bg8ejMcffxyffvqp1r++T4SHB/X8TT2IiBDnOEozxxI8eADqv7/aZbmZ4+pMlFjMFkfgACuar0pdlpstls7M3v7rqRmLpkk2Ozsb58+fx5YtW2C1WmG321FeXu6ur66uhtVqhc1m07xOCaezAVIvLumMiAjG5cv1Pj/eSMwei+xhLFWWZVPH1ZkosZgtDn8/C5pv/AwHfz+L6WK5ntn/7q/X21isVku3HS/NhotfffVVlJSUYPPmzQgICAAAjBkzBk1NTTh69CgAYOfOnZg6daouddQ/NVxpVVROpAQXPpEmPdmvvvoKW7duxciRI5GSkgIAGDFiBDZv3oycnBxkZWWhubkZw4cPd18Cb7VaNa2j/ik8JLDL+VceRkF9ga8vssiexsvIIw4XX2P2WDpfEAC0HUYxL3m06fbKzv91gce67csTNWxJ74gSB9D2+np372m0uq69X/j7WfDENIfpXl/XM/vf/fWEGS4mMqL4mEjcNLRjr+KmoYGmfgMkY5E7fSDv/DWJjUmW+rX1Hx5DufNKh7Jy5xWs//CYTi0ikewqLIOrU051yTDlPmzyDZMs9Wunz9cqKidSgvuwiUmWiEglvOWJmGSJiFTCW56ISZb6NT8PfwGeyomU4C1PxLcS6tce+FGUonIiJUS65Yl8wyRL/drfyrq+T9hTOZESIt3yRL5hkqV+jas/SU18fRGTLPVrQwb6KSonUoKri4lJlvq1VteN15B1V06kxN3R4YrKSTxMstSvNV/t+og7T+VESvz19EVF5SQezS9tJzEUl1ZiV2EZquuaERYSiDkJ0dyWQNQJr7ojJllSrPPNNc66Zry37wwAmC7RBg3y7/Lu2KBB/NMgot7jcDEpJtK2hJDBAxSVEynh6cMaP8T1H0yypJhI2xI638DTUzmREveMHqaonMTDJEuKcVsCkXd42AkxyZJiPPScyDsijfqQbzgxQIq1L27i6mKi7g0Z6NflSmIedtJ/sCdLRKQSi8WiqJzEw54sKSbSFh4iNXW1Pay7chIPe7KkmEhbeIjUZPXQYfVUTuJhkiXFuJiDyDuSh9M5PZWTeJhkSTFu4SHyDv9WiEmWFOMWHiLv8G+FmGRJsfiYSMxLHo3wkEBY0PapfF7yaC56IuokPiYSE8ZGuudgrRZgwthI/q30I1xdTD6Jj2l7o4iICMbly/V6N4fIkIpLK1F0stI9ByvJQNHJStw5wsZE20+wJ0tEpBKuxCcmWSIilXAlPjHJEhGphKuLiUmWiEglXF1MXPhERIbiZ7XA1cVpDX4mPCaJl2kQkywRGUpXCba7cqPjSvz+jcPFREREKtGsJ5udnY38/HxcuHABe/bswV133QUASExMREBAAAID2xYCpKWlYeLEiQCAEydOIDMzE83NzRg+fDjWr1+P8PBw1erIe+/nn0HhiXJIctsG+4TYKMxNGq13s0gAIg0XE2nWk500aRJ27NiB4cOH31D3xhtvIDc3F7m5ue4EK0kS0tPTkZmZifz8fMTFxWHDhg2q1ZH33s8/g4PHyztssD94vBzv55/Rt2EkBNGGi6l/0yzJxsXFwW63e/39JSUlCAwMRFxcHAAgJSUF+/fvV62OvFd4olxRORFRf2WIhU9paWmQZRnjx4/HsmXLEBISgoqKCkRFRbm/JywsDJIkoba2VpU6m83mdXvDw4N6FzCAiIjgXv8MvXR3fZeZ4+qMsRiP2eMwe/uvx1i8o3uS3bFjB+x2O1paWrB27VqsWrXK8EO4TmcDpF4MXYm8ylCkuBiL8Zg5DpH+7hnLNVarpduOl+6ri9uHkAMCApCamopjx465y8vLrw0/VldXw2q1wmazqVJHRMYQNKjrz/6eyomMTNck+/3336O+vu0ThCzL2Lt3LxwOBwBgzJgxaGpqwtGjRwEAO3fuxNSpU1WrI+/xqDhSU+OVVkXlREam2UfDNWvW4MCBA6iqqsITTzwBm82GLVu2YPHixXC5XJAkCdHR0cjKygIAWK1W5OTkICsrq8N2G7XqyHtzEqLx3r4zHW4X4VFx1Fc8TcRwbTGZkUWWZb52FeKcLLD+w2M4fb7W/bXjNhvSH/uxfg3y0fxfF3is2748UcOW9J4osYgSR2ci/N23YyzXGH5Olszn/fwzHRIsAJw+X8t9skREnTDJkmKHjne9H9ZTORFRf8UkS4pxzoyIyDtMskRERCphkiUiIlIJkywRGUpU+CBF5URGxiRLRIbSfFVSVE5kZEyyRGQozrpmReVERsYkS0SG4ulqdl7ZTmbEJEtEhsItYiQSJlkiIiKVMMkSERGphEmWiIhIJUyyRGQoD42LUlROZGSa3SdLROSNuUmjAQCFJ8ohyYDVAiTERrnLicyEPVkiMpw7R9gQGhwIC4DQ4EDcOcKmd5OIfMKeLBEZSnFpJd7bdwYtrW0nPDnrmvHevra7iuNjIvVsGpFiTLIaKi6txK7CMlTXNSMsJBBzEqL5pkHUya7CMneCbdfSKmFXYRn/Xsh0mGQ1wk/nRN7hsYokEs7JaqS7T+dEdE14SKCiciIjY5LVCD+dE3lnTkI0Avw7vjUF+FsxJyFapxYR+Y7DxRoJDwnsMqHy0zlRR+3TJ1y/QCJgktXInIToDnOyAD+dE3kSHxOJ+JhIREQE4/Ller2bQ+QzJlmN8NM5EVH/wySrIX46JyLqX7jwiYiISCVMskRERCphkiUiIlIJkywREZFKmGSJiIhUwiRLRESkEiZZIiIilTDJEhERqUSTJJudnY3ExESMGjUKZ8+edZd/8803ePTRR5GUlIRHH30U586d062OiIior2mSZCdNmoQdO3Zg+PDhHcqzsrKQmpqK/Px8pKamIjMzU7c66p+GDPRTVE5EpIQmSTYuLg52u71DmdPpxKlTpzB9+nQAwPTp03Hq1ClUV1drXkf9V1OLpKiciEgJ3c4urqiowM033ww/v7Yeg5+fH4YNG4aKigrIsqxpXVhYmA7/B8gIXJKsqJyISAleEOCD8PCgXv+MiIjgPmiJ8YgUF2MxBjO3vTPGYkxqxqJbkrXb7bh48SJcLhf8/Pzgcrlw6dIl2O12yLKsaZ1STmcDJB96OsWllcJfdSfS7UKMRX8i3VjFWIypt7FYrZZuO166beEJDw+Hw+FAXl4eACAvLw8OhwNhYWGa12mhuLQS2/NOwVnXDBmAs64Z2/NOobi0UpPfT0RE2rPIsqz65NOaNWtw4MABVFVVITQ0FDabDZ988gnKysqwfPly1NXVISQkBNnZ2bjjjjsAQPM6JXzpyS5+rRCNTa4byocM9MPGpQmK26Cn+b8u8Fi3fXmihi3pPZFiEek11o49JmNiLNf01JPVZLh4xYoVWLFixQ3l0dHR+OMf/9jlY7SuU1tXb37dlRMpZbFYFJUTkfp44hMp5uktm2/l+mq40qqonIjUxyRLinkaKOemF30FDep6YMpTORGpj0mWSBBXW7ueevBUTkTqY5IlEkTz1a7HEjyVE5H6mGQ1wqE8IqL+h0lWI552Smmwg4qIiHTCJKsRbuEhtXnaqcMdPET6YZLViNXDG52nciKlPA2KcLCESD9MshrxdEAUL3uhvsIPckTGwySrkfCQQEXlRErxgxyR8TDJamROQjT8OvUo/Cxt5UR9gR/kiIyHSVZDlk7jdp2/JuqNu6PDFZUTkfqYZDWyq7AMra6O43atLhm7Cst0ahGJ5m9lTkXlRKQ+JlmNOOuaFZUTKcXXGJHxMMlqRKSVn7YhAxSVG1nggK6fAE/lRsY5WSLjYZLViEgrP+u+v6qo3Mh+PtWhqNzI5iREI8C/4590gL+Vi+uIdMSDczUyZKBfl6c7DRnop0NrekekDwxA2yrv66fLO68CN4v4mEgAbfP/1XXNCAsJxJyEaHc5EWmPSVYjFg9n23kqNzKrpeuEasah712FZei0Hg0uua3cjMkpPiYS8TGRiIgIxuXL9Xo3h6jf43CxRhqutCoqN7KE2ChF5UbGxUJEpCYmWY2ItChlbtJoOG6zdShz3GbD3KTR+jSoF0R6XojIeJhkNTInIRr+nSb7/P0splyUUlxaibILdR3Kyi7Uobi0UqcW+W5OQjQ6j3JbwJO4iKhvMMlqSO40kdn5a7PYVViGllapQ1lLq2TKgzX+/l0tOj8L8j/LiYh6i0lWI90tsDEbkeYxC0+UKyonIlKCSVYjIiUmT9uOuB2JiKgjJlmNiLTARrTtSErKiYiUYJLViEin8XA7EhGRd5hkNRIfE4l5yaMRHhIIC9p6sPOSR5vywAOReuUibUciIuNhkiXFROqVi7QdiYiMh0lWI8WllXhv3xk465oho23B03v7zpjyzVykXrlI25GIyHh4drFGunszN2NyEuWMXJFWfROR8TDJakS0N/P388+g8EQ5JLltJW5CbJQp5zHDQwK7fA7MOL9MRMbD4WKNiLRY6P38Mzh4vNy9l1SSgYPHy/F+/hl9G+YDkeaXich4mGQ1ItKbuUinJIk0v0xExsPhYo3Ex0Ti79/VdhhinTA20pRv5qKdkiTK/DIRGY8herKJiYmYOnUqZs2ahVmzZuHzzz8HAJw4cQIzZ85EUlIS5s+fD6fT6X6MGnVqKi6tRNHJyg5DrEUnK025upinJBEReccQSRYA3njjDeTm5iI3NxcTJ06EJElIT09HZmYm8vPzERcXhw0bNgCAKnVqE2mrCE9JIiLyjmGSbGclJSUIDAxEXFwcACAlJQX79+9XrU5tIq0unps0Gg+Ni3L3XK0W4KFx5lxdTESkJsPMyaalpUGWZYwfPx7Lli1DRUUFoqKu9YzCwsIgSRJqa2tVqbPZbF63NTw8SHF8EaGDcLnmSpflERHBin+e3pY9fg+WPa53K/qeGZ8LT0SJRZQ4AMZiVGrGYogku2PHDtjtdrS0tGDt2rVYtWoVHn74Yb2b5ZHT2QBJ4Sqf2fffjvf2nekwZBzgb8Xs+2839WIbkRYLMRbjESUOgLEYVW9jsVot3Xa8DDFcbLfbAQABAQFITU3FsWPHYLfbUV5+bUtIdXU1rFYrbDabKnVq41YRIqL+R/ee7Pfffw+Xy4Xg4GDIsoy9e/fC4XBgzJgxaGpqwtGjRxEXF4edO3di6tSpAKBKnRa4VYSIqH/RPck6nU4sXrwYLpcLkiQhOjoaWVlZsFqtyMnJQVZWFpqbmzF8+HCsX78eAFSpIyIi6msWWZZNeoSAfnyZk72eSD1ZxmJMosQiShwAYzEqtedkde/JkjkVl1ZiV2EZquuaERYSiDkJ0ZxfJiLqhEmWFGu/G7d9pXT73bgAmGiJiK5jiNXFZC4inV5FRKQmJllSTKTTq4iI1MQkS4qJdDcuEZGamGRJMZHuxiUiUhMXPpFi7YubuLqYiKh7TLLkE55eRUTUMw4XExERqYRJloiISCVMskRERCphkiUiIlIJkywREZFKmGSJiIhUwiRLRESkEiZZIiIilTDJEhERqYRJloiISCVMskRERCrh2cU+sFothvgZRsFYjEmUWESJA2AsRtWbWHp6rEWWZdnnn05EREQecbiYiIhIJUyyREREKmGSJSIiUgmTLBERkUqYZImIiFTCJEtERKQSJlkiIiKVMMkSERGphEmWiIhIJTxWUWOLFi3Cd999B6vVisGDB+Pll1+Gw+HQu1k+27RpEzZu3Ig9e/bgrrvu0rs5PklMTERAQAACAwMBAGlpaZg4caLOrfJNc3MzXnnlFRQXFyMwMBCxsbFYvXq13s1S7LvvvsMvf/lL99f19fVoaGjAX//6Vx1b5ZuDBw/i9ddfhyzLkGUZv/rVrzBlyhS9m+WTQ4cO4fXXX0drayuGDh2KdevW4ZZbbtG7WV7Jzs5Gfn4+Lly40OH96ptvvsHy5ctRW1sLm82G7OxsjBw5su9+sUyaqqurc//7008/lWfPnq1ja3qnpKREfvLJJ+WHHnpI/vLLL/Vujs/M3v7rrV69Wl67dq0sSZIsy7J8+fJlnVvUN9asWSP/27/9m97NUEySJDkuLs79+jp9+rQcGxsru1wunVumXG1trXzvvffKX3/9tSzLsrx79255/vz5OrfKe0eOHJHLy8tv+HufO3euvHv3blmW22KaO3dun/5eDhdrLDg42P3vhoYGWCzmPGS7paUFq1atwsqVK/VuCv1TY2Mjdu/ejSVLlrhfVzfddJPOreq9lpYW7NmzB//yL/+id1N8YrVaUV9fD6CtRz5s2DBYreZ76z1//jxuuukm3H777QCAhIQEHD58GNXV1Tq3zDtxcXGw2+0dypxOJ06dOoXp06cDAKZPn45Tp071aUwcLtZBRkYGioqKIMsy3nnnHb2b45PXX38dM2fOxIgRI/RuSp9IS0uDLMsYP348li1bhpCQEL2bpNi3334Lm82GTZs24YsvvsCQIUOwZMkSxMXF6d20XikoKMDNN9+MmJgYvZuimMViwWuvvYZFixZh8ODBaGxsxNtvv613s3xy++23o6qqCn/7299w9913Y8+ePQCAiooKhIWF6dw631RUVODmm2+Gn58fAMDPzw/Dhg3r05jM93FKAGvXrsWhQ4fw/PPPIycnR+/mKHb8+HGUlJQgNTVV76b0iR07duDjjz/GRx99BFmWsWrVKr2b5BOXy4Vvv/0WP/zhD7Fr1y6kpaVh8eLFaGho0LtpvfLRRx+Zthfb2tqKrVu34s0338TBgwfx1ltvYenSpWhsbNS7aYoFBwfjN7/5DdatW4c5c+bA6XQiJCTEnaCoa0yyOpo9eza++OIL1NTU6N0URY4cOYKysjJMmjQJiYmJqKysxJNPPonDhw/r3TSftA8hBQQEIDU1FceOHdO5Rb6x2+3w9/d3D3396Ec/QmhoKL755hudW+a7ixcv4siRI5gxY4beTfHJ6dOncenSJYwfPx4AMH78eAwaNAhlZWU6t8w3P/nJT/Dhhx9i165dePzxx9HU1IRbb71V72b5zG634+LFi3C5XADaPqheunTphmHl3mCS1VBjYyMqKircXxcUFGDo0KGw2Wz6NcoHCxcuxOHDh1FQUICCggJERkZi27ZtuP/++/VummLff/+9e75MlmXs3bvXtKu9w8LCcN9996GoqAhA26pJp9OJ2267TeeW+e7Pf/4zEhISEBoaqndTfBIZGYnKykp8/fXXAICysjI4nU7TJqbLly8DACRJwquvvoqUlBQMHjxY51b5Ljw8HA6HA3l5eQCAvLw8OByOPh3+5qXtGqqqqsKiRYtw5coVWK1WDB06FC+++KIp55qul5iYiC1btphyC8+3336LxYsXw+VyQZIkREdHY8WKFRg2bJjeTfPJt99+i5deegm1tbXw9/fH0qVLkZCQoHezfJaUlISMjAw88MADejfFZx9//DF++9vfuhejPffcc5g8ebLOrfJNRkYGjh07hqtXr2LChAl46aWX3FvfjG7NmjU4cOAAqqqqEBoaCpvNhk8++QRlZWVYvnw56urqEBISguzsbNxxxx199nuZZImIiFTC4WIiIiKVMMkSERGphEmWiIhIJUyyREREKmGSJSIiUgmTLBERkUqYZInIZxs3bkRaWhqAtuvpRo0ahdbWVk1+9xdffGHq/bPUPzDJEpnQrl27MGPGDPzoRz/ChAkTkJWVhbq6Or2bdYODBw/ipz/9KWJjY3HfffchLS0NFy9e1LtZRJphkiUyme3bt2PDhg1IT0/H0aNH8Yc//AHl5eV44okn0NLS0me/p7c90v379+OFF17AvHnz8N///d/Iy8vDgAEDkJqaasgPBERqYJIlMpGGhgZs3LgRK1aswAMPPIABAwZgxIgReO2113DhwgVs374dd999N2pra92POXXqFO677z5cvXoVAPCnP/0JycnJuOeee/Dkk0/iwoUL7u8dNWoUduzYgSlTpmDKlCkA2o6jS0hIwI9//GPMmTMHR48e7bGdsiwjOzsbzz77LGbMmIGBAwciIiICa9euxaBBg/D73/8eQMfhZuDGIeePPvoIycnJGDduHCZNmoSdO3f2+v8hkZaYZIlM5NixY2hubnYnwHZDhgxBQkICzp49i9jYWBw4cMBdt2fPHiQlJWHAgAH4y1/+gq1bt2LTpk0oLi7G+PHj8cILL3T4WX/5y1/wn//5n9i7dy8AYOzYsdi9ezf++te/Yvr06ViyZAmam5u7befXX3+N8vJyTJ06tUO51WrFlClTvL6xKTw8HFu3bsWxY8ewbt06rFu3DqWlpV49lsgImGSJTKSmpgahoaHw9/e/oS4iIgI1NTWYMWOG+1aR9puF2q+K27lzJxYuXIjo6Gj4+/vjmWeewenTpzv0ZhcuXAibzYaBAwcCAGbNmuX+nfPnz0dLS0uP1+e1X9/Y1UUL7e30xoMPPohbb70VFosF9957LyZMmOBVT5rIKG78SyUiwwoNDUVNTQ1aW1tvSLSXL19GaGgopkyZgtWrV+PSpUs4d+4crFYr4uLiAADl5eV45ZVXkJ2d7X6cLMu4ePEihg8fDgA33KW5bds2/OlPf8KlS5dgsVjQ0NDQY5Jsv5ru0qVLuOWWW7pspzcKCwuxefNmnDt3DpIkoampyZS3PVH/xSRLZCLjxo1DQEAADhw4gGnTprnLGxsb8dlnn2HZsmUYOnQoJkyYgL179+Lrr7/GtGnT3Nes2e12PPPMM5g5c6bH39H+vQBw9OhRvPPOO/jd736HH/zgB7BarbjnnnvQ0+Vdd9xxByIjI7F//34sWLDAXS5JEg4cOIDExEQAwKBBg9DU1OSur6qqcv+7paUFzz33HLKzszFp0iQMGDAAixYt6vF3ExkJh4uJTCQ4OBi//OUvsWbNGnz22We4evUqvvvuOyxduhSRkZGYNWsWAGDGjBnIzc1Ffn6+e6gYAFJSUvD222/jq6++AgDU19dj3759Hn9fY2Mj/Pz8EBYWhtbWVmzatAkNDQ09ttNiseDFF1/EW2+9hT179qC5uRmXL19GRkYGampq8PjjjwMAHA4Hjhw5gvLyctTX12Pr1q3un9HS0oKWlhaEhYXB398fhYWF7gvpicyCPVkik1mwYAFsNhtycnLwj3/8A0FBQZg8eTI2bNiAgIAAAEBiYiIyMjIQFRWF0aNHux/78MMPo7GxEcuWLcOFCxcQHByMn/zkJ0hOTu7yd91///2YOHEikpKSMHjwYMybN++G4WRPpk2bhoCAALz11ltYsWKFe6j3/fffd8/VTpgwAdOmTcPMmTMRGhqKBQsWoKCgAAAQFBSEFStWYOnSpWhpacFDDz3k7gETmQUvbSciTRw+fBgvvPACfve738HhcOjdHCJNMMkSkWYKCgpw8eJFPPbYY3o3hUgTTLJEREQq4cInIiIilTDJEhERqYRJloiISCVMskRERCphkiUiIlIJkywREZFK/h8tZBH/CS4K/gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7, 7))\n",
    "plt.scatter(X_train.OverallQual, y_train)\n",
    "plt.xlabel('OverallQual')\n",
    "plt.ylabel('SalePrice')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "threshold = 5\n",
    "mask = (X_train.OverallQual <= threshold)\n",
    "X_train_1 = X_train[mask]\n",
    "y_train_1 = y_train[mask]\n",
    "X_train_2 = X_train[~mask]\n",
    "y_train_2 = y_train[~mask]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE = 22745.7569\n"
     ]
    }
   ],
   "source": [
    "column_transformer1 = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline1 = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer1),\n",
    "    ('regression', Ridge())\n",
    "])\n",
    "\n",
    "column_transformer2 = ColumnTransformer([\n",
    "    ('ohe', OneHotEncoder(handle_unknown=\"ignore\"), categorical),\n",
    "    ('scaling', StandardScaler(), numeric_features)\n",
    "])\n",
    "\n",
    "pipeline2 = Pipeline(steps=[\n",
    "    ('ohe_and_scaling', column_transformer2),\n",
    "    ('regression', Ridge())\n",
    "])\n",
    "\n",
    "\n",
    "\n",
    "model1 = pipeline1.fit(X_train_1, y_train_1)\n",
    "model2 = pipeline2.fit(X_train_2, y_train_2)\n",
    "\n",
    "y_pred_1 = model1.predict(X_test)\n",
    "y_pred_2 = model2.predict(X_test)\n",
    "mask_test = (X_test.OverallQual <= threshold)\n",
    "y_pred = y_pred_1.copy()\n",
    "y_pred[~mask_test] = y_pred_2[~mask_test]\n",
    "\n",
    "print(\"Test RMSE = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Получилось лучше! И это при практически случайном выборе разбиения. Если бы мы поработали над этим получше, то и качество, скорее всего, получилось бы выше."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Перейдём к следующему трюку — бинаризации признаков. Мы выбираем $n$ порогов $t_1, \\dots, t_n$ для признака~$x_j$ и генерируем $n+1$ новый признак: $[x_j \\leq t_1], [t_1 < x_j \\leq t_2], \\dots, [t_{n-1} < x_j \\leq t_n], [x_j > t_n]$. Такое преобразование может неплохо помочь в случае, если целевая переменная нелинейно зависит от одного из признаков. Рассмотрим синтетический пример."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import KBinsDiscretizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-2.0, 2.0)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_plot = np.linspace(0, 1, 10000)\n",
    "\n",
    "X = np.random.uniform(0, 1, size=30)\n",
    "y = np.cos(1.5 * np.pi * X) + np.random.normal(scale=0.1, size=X.shape)\n",
    "\n",
    "fig, axs = plt.subplots(figsize=(16, 4), ncols=2)\n",
    "\n",
    "regr = LinearRegression()\n",
    "regr.fit(X[:, np.newaxis], y)\n",
    "y_pred_regr = regr.predict(x_plot[:, np.newaxis])\n",
    "axs[0].scatter(X[:, np.newaxis], y, label=\"Data\")\n",
    "axs[0].plot(x_plot, y_pred_regr, label=\"Predictions\")\n",
    "axs[0].legend()\n",
    "axs[0].set_title(\"Linear regression on original feature\")\n",
    "axs[0].set_xlabel(\"$X$\")\n",
    "axs[0].set_ylabel(\"$y$\")\n",
    "axs[0].set_ylim(-2, 2)\n",
    "\n",
    "binner = KBinsDiscretizer(n_bins=5, strategy='quantile')\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('binning', binner),\n",
    "    ('regression', LinearRegression())\n",
    "])\n",
    "pipeline.fit(X[:, np.newaxis], y)\n",
    "y_pred_binned = pipeline.predict(x_plot[:, np.newaxis])\n",
    "axs[1].scatter(X[:, np.newaxis], y, label=\"Data\")\n",
    "axs[1].plot(x_plot, y_pred_binned, label=\"Predictions\")\n",
    "axs[1].set_title(\"Linear regression on binned feature\")\n",
    "axs[1].set_xlabel(\"$X$\")\n",
    "axs[1].set_ylabel(\"$y$\")\n",
    "axs[1].set_ylim(-2, 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Видно, что качество модели существенно возрасло. С другой стороны, увеличилось и количество параметров модели (из-за увеличения числа признаков), поэтому при бинаризации важно контролировать переобучение."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Иногда может помочь преобразование целевой переменной. Может оказаться, что по мере роста признаков целевая переменная меняется экспоненциально. Например, может оказаться, что при линейном уменьшении продолжительности видео число его просмотров растёт экспоненциально. Учеть это можно с помощью логарифмирования целевой переменной — ниже синтетический пример с такой ситуацией."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '$y$')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = np.random.exponential(1, size=30)\n",
    "y = np.exp(X) + np.random.normal(scale=0.1, size=X.shape)\n",
    "\n",
    "x_plot = np.linspace(np.min(X), np.max(X), 10000)\n",
    "\n",
    "fig, axs = plt.subplots(figsize=(16, 4), ncols=2)\n",
    "\n",
    "regr = LinearRegression()\n",
    "regr.fit(X[:, np.newaxis], y)\n",
    "y_pred_regr = regr.predict(x_plot[:, np.newaxis])\n",
    "axs[0].scatter(X[:, np.newaxis], y, label=\"Data\")\n",
    "axs[0].plot(x_plot, y_pred_regr, label=\"Predictions\")\n",
    "axs[0].legend()\n",
    "axs[0].set_title(\"Linear regression on original target\")\n",
    "axs[0].set_xlabel(\"$X$\")\n",
    "axs[0].set_ylabel(\"$y$\")\n",
    "\n",
    "y_log = np.log(y)\n",
    "regr.fit(X[:, np.newaxis], y_log)\n",
    "y_pred_log = np.exp(regr.predict(x_plot[:, np.newaxis]))\n",
    "axs[1].scatter(X[:, np.newaxis], y, label=\"Data\")\n",
    "axs[1].plot(x_plot, y_pred_log, label=\"Predictions\")\n",
    "axs[1].set_title(\"Linear regression on log target\")\n",
    "axs[1].set_xlabel(\"$X$\")\n",
    "axs[1].set_ylabel(\"$y$\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Но, конечно, вряд ли в реальных данных будет действительно экспоненциальная связь между целевой переменной и линейной комбинацией признаков. Тем не менее, логарифмирование всё равно может помочь.\n",
    "\n",
    "Ниже пример с данными `boston` по мотивам [примера](https://scikit-learn.org/stable/auto_examples/compose/plot_transformed_target.html) из документации sklearn. Здесь используется несколько более сложное преобразование: вычисляются квантили распределения целевой переменной и квантили нормального распределения, после чего каждое значение целевой переменной переводится в значение, имеющее такую же квантиль у нормального распределения.\n",
    "\n",
    "Обратите внимание на класс `TransformedTargetRegressor`, который как раз удобен для построения моделей на преобразованной целевой переменной."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_boston\n",
    "from sklearn.preprocessing import QuantileTransformer, quantile_transform\n",
    "from sklearn.linear_model import RidgeCV\n",
    "from sklearn.compose import TransformedTargetRegressor\n",
    "from sklearn.metrics import mean_absolute_error, r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = load_boston()\n",
    "target = np.array(dataset.feature_names) == \"DIS\"\n",
    "X = dataset.data[:, np.logical_not(target)]\n",
    "y = dataset.data[:, target].squeeze()\n",
    "y_trans = quantile_transform(dataset.data[:, target],\n",
    "                             n_quantiles=300,\n",
    "                             output_distribution='normal',\n",
    "                             copy=True).squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, (ax0, ax1) = plt.subplots(1, 2, figsize=(15,8))\n",
    "\n",
    "ax0.hist(y, bins=20, density=True)\n",
    "ax0.set_ylabel('Probability')\n",
    "ax0.set_xlabel('Target')\n",
    "ax0.set_title('Target distribution')\n",
    "\n",
    "ax1.hist(y_trans, bins=20, density=True)\n",
    "ax1.set_ylabel('Probability')\n",
    "ax1.set_xlabel('Target')\n",
    "ax1.set_title('Transformed target distribution')\n",
    "\n",
    "f.suptitle(\"Boston housing data: distance to employment centers\", y=0.035)\n",
    "f.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test MAE without target transformation = 0.8837\n",
      "Test RMSE without target transformation = 1.1565\n",
      "Test MAE with target transformation = 0.8093\n",
      "Test RMSE with target transformation = 1.1548\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, (ax0, ax1) = plt.subplots(1, 2, sharey=True, figsize=(15,8))\n",
    "\n",
    "regr = RidgeCV()\n",
    "regr.fit(X_train, y_train)\n",
    "y_pred = regr.predict(X_test)\n",
    "\n",
    "print(\"Test MAE without target transformation = %.4f\" % mean_absolute_error(y_test, y_pred))\n",
    "print(\"Test RMSE without target transformation = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))\n",
    "\n",
    "ax0.scatter(y_test, y_pred)\n",
    "ax0.plot([0, 10], [0, 10], '--k')\n",
    "ax0.set_ylabel('Target predicted')\n",
    "ax0.set_xlabel('True Target')\n",
    "ax0.set_title('Ridge regression \\n without target transformation')\n",
    "ax0.set_xlim([0, 10])\n",
    "ax0.set_ylim([0, 10])\n",
    "\n",
    "regr_trans = TransformedTargetRegressor(\n",
    "    regressor=RidgeCV(),\n",
    "    transformer=QuantileTransformer(n_quantiles=300,\n",
    "                                    output_distribution='normal'))\n",
    "regr_trans.fit(X_train, y_train)\n",
    "y_pred = regr_trans.predict(X_test)\n",
    "\n",
    "print(\"Test MAE with target transformation = %.4f\" % mean_absolute_error(y_test, y_pred))\n",
    "print(\"Test RMSE with target transformation = %.4f\" % mean_squared_error(y_test, y_pred, squared=False))\n",
    "\n",
    "ax1.scatter(y_test, y_pred)\n",
    "ax1.plot([0, 10], [0, 10], '--k')\n",
    "ax1.set_ylabel('Target predicted')\n",
    "ax1.set_xlabel('True Target')\n",
    "ax1.set_title('Ridge regression \\n with target transformation')\n",
    "ax1.set_xlim([0, 10])\n",
    "ax1.set_ylim([0, 10])\n",
    "\n",
    "f.suptitle(\"Boston housing data: distance to employment centers\", y=0.035)\n",
    "f.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Видно, что преобразование целевой переменной дало улучшение, хоть и не очень большое. При этом, как видно из графика, улучшения произошли в основном для небольших значений целевой переменной — именно поэтому изменения больше заметны с точки зрения MAE, а не RMSE."
   ]
  }
 ],
 "metadata": {
  "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}