4e_tree_based_models_regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tree-Based Models for Regression - Predicting a Numeric Indicator of Diabetes Disease Progression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Original Code source: Jaques Grobler: https://scikit-learn.org/stable/auto_examples/linear_model/plot_ols.html\n",
    "# License: BSD 3 clause\n",
    "# Modified by SR\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from sklearn import datasets\n",
    "from sklearn.metrics import mean_absolute_error, mean_absolute_percentage_error, \\\n",
    "                            r2_score, root_mean_squared_error\n",
    "\n",
    "\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from xgboost import XGBRegressor\n",
    "from lightgbm import LGBMRegressor\n",
    "from catboost import CatBoostRegressor\n",
    "from sklearn.ensemble import HistGradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://scikit-learn.org/stable/datasets/toy_dataset.html#diabetes-dataset\n",
    "\n",
    "Ten baseline variables, age, sex, body mass index, average blood pressure, and six blood serum measurements were obtained for each of n = 442 diabetes patients, as well as the response of interest, a quantitative measure of disease progression one year after baseline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>bmi</th>\n",
       "      <th>bp</th>\n",
       "      <th>s1</th>\n",
       "      <th>s2</th>\n",
       "      <th>s3</th>\n",
       "      <th>s4</th>\n",
       "      <th>s5</th>\n",
       "      <th>s6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>59.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>32.1</td>\n",
       "      <td>101.0</td>\n",
       "      <td>157.0</td>\n",
       "      <td>93.2</td>\n",
       "      <td>38.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.8598</td>\n",
       "      <td>87.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>48.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.6</td>\n",
       "      <td>87.0</td>\n",
       "      <td>183.0</td>\n",
       "      <td>103.2</td>\n",
       "      <td>70.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.8918</td>\n",
       "      <td>69.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>72.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.5</td>\n",
       "      <td>93.0</td>\n",
       "      <td>156.0</td>\n",
       "      <td>93.6</td>\n",
       "      <td>41.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.6728</td>\n",
       "      <td>85.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>24.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>25.3</td>\n",
       "      <td>84.0</td>\n",
       "      <td>198.0</td>\n",
       "      <td>131.4</td>\n",
       "      <td>40.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.8903</td>\n",
       "      <td>89.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>101.0</td>\n",
       "      <td>192.0</td>\n",
       "      <td>125.4</td>\n",
       "      <td>52.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.2905</td>\n",
       "      <td>80.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  sex   bmi     bp     s1     s2    s3   s4      s5    s6\n",
       "0  59.0  2.0  32.1  101.0  157.0   93.2  38.0  4.0  4.8598  87.0\n",
       "1  48.0  1.0  21.6   87.0  183.0  103.2  70.0  3.0  3.8918  69.0\n",
       "2  72.0  2.0  30.5   93.0  156.0   93.6  41.0  4.0  4.6728  85.0\n",
       "3  24.0  1.0  25.3   84.0  198.0  131.4  40.0  5.0  4.8903  89.0\n",
       "4  50.0  1.0  23.0  101.0  192.0  125.4  52.0  4.0  4.2905  80.0"
      ]
     },
     "execution_count": 287,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the diabetes dataset\n",
    "diabetes_X, diabetes_y = datasets.load_diabetes(return_X_y=True, as_frame=True, scaled=False)\n",
    "\n",
    "diabetes_X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    151.0\n",
       "1     75.0\n",
       "2    141.0\n",
       "3    206.0\n",
       "4    135.0\n",
       "Name: target, dtype: float64"
      ]
     },
     "execution_count": 288,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes_y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the data into training/testing sets\n",
    "diabetes_X_train, diabetes_X_test, diabetes_y_train, diabetes_y_test = train_test_split(\n",
    "    diabetes_X, diabetes_y,\n",
    "    test_size = 0.25,\n",
    "    random_state=42\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>bmi</th>\n",
       "      <th>bp</th>\n",
       "      <th>s1</th>\n",
       "      <th>s2</th>\n",
       "      <th>s3</th>\n",
       "      <th>s4</th>\n",
       "      <th>s5</th>\n",
       "      <th>s6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>47.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30.3</td>\n",
       "      <td>109.0</td>\n",
       "      <td>207.0</td>\n",
       "      <td>100.2</td>\n",
       "      <td>70.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>5.2149</td>\n",
       "      <td>98.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>66.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.7</td>\n",
       "      <td>126.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>127.8</td>\n",
       "      <td>45.0</td>\n",
       "      <td>4.71</td>\n",
       "      <td>5.2781</td>\n",
       "      <td>101.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>51.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>31.5</td>\n",
       "      <td>93.0</td>\n",
       "      <td>231.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>49.0</td>\n",
       "      <td>4.70</td>\n",
       "      <td>5.2523</td>\n",
       "      <td>117.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>53.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>27.7</td>\n",
       "      <td>95.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>101.8</td>\n",
       "      <td>41.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>5.4638</td>\n",
       "      <td>101.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>24.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>25.3</td>\n",
       "      <td>84.0</td>\n",
       "      <td>198.0</td>\n",
       "      <td>131.4</td>\n",
       "      <td>40.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>4.8903</td>\n",
       "      <td>89.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>22.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>19.3</td>\n",
       "      <td>82.0</td>\n",
       "      <td>156.0</td>\n",
       "      <td>93.2</td>\n",
       "      <td>52.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.9890</td>\n",
       "      <td>71.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>270</th>\n",
       "      <td>50.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>29.2</td>\n",
       "      <td>119.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>85.2</td>\n",
       "      <td>54.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.7362</td>\n",
       "      <td>95.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348</th>\n",
       "      <td>57.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.5</td>\n",
       "      <td>93.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>96.6</td>\n",
       "      <td>71.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.5218</td>\n",
       "      <td>91.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>435</th>\n",
       "      <td>45.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.2</td>\n",
       "      <td>83.0</td>\n",
       "      <td>177.0</td>\n",
       "      <td>118.4</td>\n",
       "      <td>45.0</td>\n",
       "      <td>4.00</td>\n",
       "      <td>4.2195</td>\n",
       "      <td>82.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>23.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>131.4</td>\n",
       "      <td>65.0</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.5850</td>\n",
       "      <td>91.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>331 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      age  sex   bmi     bp     s1     s2    s3    s4      s5     s6\n",
       "16   47.0  1.0  30.3  109.0  207.0  100.2  70.0  3.00  5.2149   98.0\n",
       "408  66.0  1.0  21.7  126.0  212.0  127.8  45.0  4.71  5.2781  101.0\n",
       "432  51.0  1.0  31.5   93.0  231.0  144.0  49.0  4.70  5.2523  117.0\n",
       "316  53.0  2.0  27.7   95.0  190.0  101.8  41.0  5.00  5.4638  101.0\n",
       "3    24.0  1.0  25.3   84.0  198.0  131.4  40.0  5.00  4.8903   89.0\n",
       "..    ...  ...   ...    ...    ...    ...   ...   ...     ...    ...\n",
       "106  22.0  1.0  19.3   82.0  156.0   93.2  52.0  3.00  3.9890   71.0\n",
       "270  50.0  2.0  29.2  119.0  162.0   85.2  54.0  3.00  4.7362   95.0\n",
       "348  57.0  1.0  24.5   93.0  186.0   96.6  71.0  3.00  4.5218   91.0\n",
       "435  45.0  1.0  24.2   83.0  177.0  118.4  45.0  4.00  4.2195   82.0\n",
       "102  23.0  1.0  29.0   90.0  216.0  131.4  65.0  3.00  4.5850   91.0\n",
       "\n",
       "[331 rows x 10 columns]"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes_X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16     166.0\n",
       "408    189.0\n",
       "432    173.0\n",
       "316    220.0\n",
       "3      206.0\n",
       "       ...  \n",
       "106    134.0\n",
       "270    202.0\n",
       "348    148.0\n",
       "435     64.0\n",
       "102    302.0\n",
       "Name: target, Length: 331, dtype: float64"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes_y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 292,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Explore distribution of values being predicted\n",
    "diabetes_y_train.hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 293,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(diabetes_y_train).plot(kind=\"box\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Storing results\n",
    "\n",
    "Create a list to store results in."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_list = []"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [],
   "source": [
    "def standardise_data(X_train, X_test):\n",
    "\n",
    "    # Initialise a new scaling object for normalising input data\n",
    "    sc = StandardScaler()\n",
    "\n",
    "    # Apply the scaler to the training and test sets\n",
    "    train_std=sc.fit_transform(X_train)\n",
    "    test_std=sc.fit_transform(X_test)\n",
    "\n",
    "    return train_std, test_std\n",
    "\n",
    "diabetes_X_train_standardised, diabetes_X_test_standardised = standardise_data(\n",
    "    diabetes_X_train,\n",
    "    diabetes_X_test\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 51.02\n",
      "Mean absolute percentage error: 48.88%\n",
      "Root Mean squared error: 62.00\n",
      "Coefficient of determination: 0.30\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Use only first feature\n",
    "single_feat_train = diabetes_X_train_standardised[:, np.newaxis, 2]\n",
    "single_feat_test = diabetes_X_test_standardised[:, np.newaxis, 2]\n",
    "\n",
    "# Create linear regression object\n",
    "regr = LinearRegression()\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr.fit(single_feat_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr.predict(single_feat_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'Linear Regression',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "# Plot outputs\n",
    "def plot_actual_vs_predicted(actual, predicted):\n",
    "    fig, ax = plt.subplots(figsize=(6, 6))\n",
    "\n",
    "    ax.scatter(actual, predicted, color=\"black\")\n",
    "    ax.axline((1, 1), slope=1)\n",
    "    plt.xlabel('True Values')\n",
    "    plt.ylabel('Predicted Values')\n",
    "    plt.title('True vs Predicted Values')\n",
    "    plt.show()\n",
    "\n",
    "def plot_residuals(actual, predicted):\n",
    "    residuals = actual - predicted\n",
    "\n",
    "    plt.figure(figsize=(10, 5))\n",
    "    plt.hist(residuals, bins=20)\n",
    "    plt.axvline(x = 0, color = 'r')\n",
    "    plt.xlabel('Residual')\n",
    "    plt.ylabel('Frequency')\n",
    "    plt.title('Distribution of Residuals')\n",
    "    plt.show()\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0kAAAHWCAYAAACi1sL/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA8c0lEQVR4nO3deXxMd////+dIZBJkQUKkSGKpWkNpfbRaVCpiKa22uFRD9eqmVY0q6VV7CXXZ2traqxV00erV4qK2WqultXZRtdXWWqKKSBCRvH9/9Ge+Z2TBmGQiHvfbbW6c93mfc14z7yTy9J7zHpsxxggAAAAAIEkq5ukCAAAAAKAwISQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAFAITJ06FDZbLYCuVbz5s3VvHlzx/bq1atls9n02WefFcj1e/TooYiIiAK5lqtSU1P15JNPKjQ0VDabTX379vV0SdnYbDYNHTr0iv0K4mvr0tfQ6tWr8/U6AJDfCEkAkE+SkpJks9kcD19fX4WFhSkmJkZvvvmmzpw545brHD58WEOHDtW2bdvccj53Ksy1XY1Ro0YpKSlJzz77rGbPnq3u3bvn2jciIsJpvEuWLKk777xTs2bNKsCKAQDu4O3pAgCgqBs+fLgiIyOVkZGho0ePavXq1erbt6/Gjx+vBQsWqF69eo6+r732mgYOHHhN5z98+LCGDRumiIgI1a9f/6qPW7Zs2TVdxxV51fbuu+8qKysr32u4HitXrtT//d//aciQIVfVv379+urXr58k6ciRI/rPf/6juLg4paen65///Ge+1Hju3Dl5e/PPOQC4Ez9VASCfxcbGqlGjRo7thIQErVy5Uu3atdMDDzygHTt2yM/PT5Lk7e2d77/wnj17ViVKlJCPj0++XudKihcv7tHrX43k5GTVqlXrqvvfcssteuyxxxzbPXr0UJUqVTRhwoR8C0m+vr75cl4AuJnxdjsA8ID77rtPgwYN0oEDB/TBBx842nO6b2T58uVq2rSpgoKCVKpUKdWoUUOvvvqqpL/vAbnjjjskST179nS81SspKUnS3/cd1alTR5s3b9a9996rEiVKOI69/J6kSzIzM/Xqq68qNDRUJUuW1AMPPKBDhw459YmIiFCPHj2yHWs955Vqy+mepLS0NPXr10+VKlWS3W5XjRo19O9//1vGGKd+NptNzz//vObNm6c6derIbrerdu3aWrJkSc4v+GWSk5PVq1cvlS9fXr6+voqKitLMmTMd+y/dW7Nv3z4tWrTIUfv+/fuv6vyXhISE6LbbbtPevXud2rOysjRx4kTVrl1bvr6+Kl++vJ5++mmdPHnSqd+mTZsUExOj4OBg+fn5KTIyUk888US21+Lye5LWrVunO+64Q76+vqpataqmT5+erbb9+/c7jUde5zxw4ICee+451ahRQ35+fipbtqweeeSRq3o9du/erU6dOik0NFS+vr6qWLGiunTpotOnT1/xWADwFGaSAMBDunfvrldffVXLli3LdZZh+/btateunerVq6fhw4fLbrdrz549+uabbyRJNWvW1PDhwzV48GA99dRTuueeeyRJd911l+McJ06cUGxsrLp06aLHHntM5cuXz7OukSNHymazacCAAUpOTtbEiRMVHR2tbdu2OWa8rsbV1GZljNEDDzygVatWqVevXqpfv76WLl2q/v37648//tCECROc+q9bt06ff/65nnvuOfn7++vNN99Up06ddPDgQZUtWzbXus6dO6fmzZtrz549ev755xUZGam5c+eqR48eOnXqlF588UXVrFlTs2fP1ksvvaSKFSs63kIXEhJy1c9fki5evKjff/9dpUuXdmp/+umnlZSUpJ49e6pPnz7at2+f3n77bW3dulXffPONihcvruTkZLVq1UohISEaOHCggoKCtH//fn3++ed5XvOnn35yHDd06FBdvHhRQ4YMueK452Xjxo369ttv1aVLF1WsWFH79+/X1KlT1bx5c/3yyy8qUaJEjsdduHBBMTExSk9P1wsvvKDQ0FD98ccfWrhwoU6dOqXAwECXawKAfGUAAPlixowZRpLZuHFjrn0CAwNNgwYNHNtDhgwx1h/NEyZMMJLM8ePHcz3Hxo0bjSQzY8aMbPuaNWtmJJlp06bluK9Zs2aO7VWrVhlJ5pZbbjEpKSmO9k8//dRIMpMmTXK0hYeHm7i4uCueM6/a4uLiTHh4uGN73rx5RpJ5/fXXnfo9/PDDxmazmT179jjaJBkfHx+nth9++MFIMm+99Va2a1lNnDjRSDIffPCBo+3ChQumSZMmplSpUk7PPTw83LRt2zbP81n7tmrVyhw/ftwcP37c/PTTT6Z79+5Gkundu7ej39dff20kmQ8//NDp+CVLlji1f/HFF1f8+jHm79diyJAhju2OHTsaX19fc+DAAUfbL7/8Yry8vJy+tvbt25fr2Fx+zrNnz2brs379eiPJzJo1y9F26Wto1apVxhhjtm7daiSZuXPn5vkcAKCw4e12AOBBpUqVynOVu6CgIEnS/PnzXV7kwG63q2fPnlfd//HHH5e/v79j++GHH1aFChX05ZdfunT9q/Xll1/Ky8tLffr0cWrv16+fjDFavHixU3t0dLSqVq3q2K5Xr54CAgL022+/XfE6oaGh6tq1q6OtePHi6tOnj1JTU7VmzRqXn8OyZcsUEhKikJAQ1a1bV7Nnz1bPnj01duxYR5+5c+cqMDBQ999/v/7880/Ho2HDhipVqpRWrVol6f+N/cKFC5WRkXFV18/MzNTSpUvVsWNHVa5c2dFes2ZNxcTEuPy8rDOIGRkZOnHihKpVq6agoCBt2bIl1+MuzRQtXbpUZ8+edfn6AFDQCEkA4EGpqalOgeRynTt31t13360nn3xS5cuXV5cuXfTpp59eU2C65ZZbrmmRhurVqztt22w2VatW7Zrvx7lWBw4cUFhYWLbXo2bNmo79VtYQcEnp0qWz3deT03WqV6+uYsWc/wnM7TrXonHjxlq+fLmWLFmif//73woKCtLJkyedXv/du3fr9OnTKleunCNQXXqkpqYqOTlZktSsWTN16tRJw4YNU3BwsDp06KAZM2YoPT091+sfP35c586dyzaGklSjRg2Xn9e5c+c0ePBgx71iwcHBCgkJ0alTp/K8tygyMlLx8fH6z3/+o+DgYMXExGjy5MncjwSg0OOeJADwkN9//12nT59WtWrVcu3j5+entWvXatWqVVq0aJGWLFmiTz75RPfdd5+WLVsmLy+vK17nWu4julq5fShpZmbmVdXkDrldx1y2yENBCg4OVnR0tCQpJiZGt912m9q1a6dJkyYpPj5e0t+LNpQrV04ffvhhjue4dN/TpQ/23bBhg/73v/9p6dKleuKJJzRu3Dht2LBBpUqVuq5a8xrDy73wwguaMWOG+vbtqyZNmigwMFA2m01dunS5YmAfN26cevToofnz52vZsmXq06ePEhMTtWHDBlWsWPG6ngMA5BdmkgDAQ2bPni1JV3wbVLFixdSyZUuNHz9ev/zyi0aOHKmVK1c63paV2y+7rtq9e7fTtjFGe/bscVqJrnTp0jp16lS2Yy+fhbmW2sLDw3X48OFsbz/89ddfHfvdITw8XLt37872y727ryNJbdu2VbNmzTRq1CilpaVJkqpWraoTJ07o7rvvVnR0dLZHVFSU0zn+7//+TyNHjtSmTZv04Ycfavv27ZozZ06O1wsJCZGfn1+2MZSknTt3Om1fWkzi8nHMaSbts88+U1xcnMaNG6eHH35Y999/v5o2bZrj10BO6tatq9dee01r167V119/rT/++EPTpk27qmMBwBMISQDgAStXrtSIESMUGRmpbt265drvr7/+ytZ26UNZL73tqmTJkpKy/7LrqlmzZjkFlc8++0xHjhxRbGyso61q1arasGGDLly44GhbuHBhtqXCr6W2Nm3aKDMzU2+//bZT+4QJE2Sz2Zyufz3atGmjo0eP6pNPPnG0Xbx4UW+99ZZKlSqlZs2aueU6lwwYMEAnTpzQu+++K0l69NFHlZmZqREjRmTre/HiRcdrdfLkyWyzYpeP/eW8vLwUExOjefPm6eDBg472HTt2aOnSpU59AwICFBwcrLVr1zq1T5kyJcfzXl7LW2+9leOsk1VKSoouXrzo1Fa3bl0VK1Ysz7cNAoCn8XY7AMhnixcv1q+//qqLFy/q2LFjWrlypZYvX67w8HAtWLAgzw8DHT58uNauXau2bdsqPDxcycnJmjJliipWrKimTZtK+juwBAUFadq0afL391fJkiXVuHFjRUZGulRvmTJl1LRpU/Xs2VPHjh3TxIkTVa1aNadlyp988kl99tlnat26tR599FHt3btXH3zwgdNCCtdaW/v27dWiRQv961//0v79+xUVFaVly5Zp/vz56tu3b7Zzu+qpp57S9OnT1aNHD23evFkRERH67LPP9M0332jixIl53iPmitjYWNWpU0fjx49X79691axZMz399NNKTEzUtm3b1KpVKxUvXly7d+/W3LlzNWnSJD388MOaOXOmpkyZogcffFBVq1bVmTNn9O677yogIEBt2rTJ9XrDhg3TkiVLdM899+i5555zBMDatWvrxx9/dOr75JNPavTo0XryySfVqFEjrV27Vrt27cp2znbt2mn27NkKDAxUrVq1tH79en311Vd5LrUu/f2fAc8//7weeeQR3Xrrrbp48aJmz54tLy8vderUybUXFAAKgkfX1gOAIuzSEuCXHj4+PiY0NNTcf//9ZtKkSU5LTV9y+RLgK1asMB06dDBhYWHGx8fHhIWFma5du5pdu3Y5HTd//nxTq1Yt4+3t7bSsc7NmzUzt2rVzrC+3JcA//vhjk5CQYMqVK2f8/PxM27ZtnZaTvmTcuHHmlltuMXa73dx9991m06ZN2c6ZV22XLwFujDFnzpwxL730kgkLCzPFixc31atXN2PHjjVZWVlO/XTZstqX5LY0+eWOHTtmevbsaYKDg42Pj4+pW7dujkthX+sS4Ln1TUpKyrbc9jvvvGMaNmxo/Pz8jL+/v6lbt6555ZVXzOHDh40xxmzZssV07drVVK5c2djtdlOuXDnTrl07s2nTJqdz67Lluo0xZs2aNaZhw4bGx8fHVKlSxUybNi3b15Yxfy/t3atXLxMYGGj8/f3No48+apKTk7Od8+TJk47Xq1SpUiYmJsb8+uuv2V7vy5cA/+2338wTTzxhqlatanx9fU2ZMmVMixYtzFdffXVVrykAeIrNGA/e4QoAAAAAhQz3JAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwKLIf5hsVlaWDh8+LH9/f9lsNk+XAwAAAMBDjDE6c+aMwsLCVKxY7vNFRT4kHT58WJUqVfJ0GQAAAAAKiUOHDqlixYq57i/yIcnf31/S3y9EQECAh6sBAHhUWpoUFvb33w8flkqW9Gw9AIAClZKSokqVKjkyQm6KfEi69Ba7gIAAQhIA3Oy8vP7f3wMCCEkAcJO60m04LNwAAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACAhUdD0tq1a9W+fXuFhYXJZrNp3rx5ufZ95plnZLPZNHHixAKrDwAAAMDNx6MhKS0tTVFRUZo8eXKe/b744gtt2LBBYWFhBVQZAAAAgJuVtycvHhsbq9jY2Dz7/PHHH3rhhRe0dOlStW3btoAqAwAAAHCz8mhIupKsrCx1795d/fv3V+3ata/qmPT0dKWnpzu2U1JS8qs8AAAAAEVQoQ5JY8aMkbe3t/r06XPVxyQmJmrYsGH5WBUAFG4RAxd5ugSH/aN5BwAA4MZTaFe327x5syZNmqSkpCTZbLarPi4hIUGnT592PA4dOpSPVQIAAAAoagptSPr666+VnJysypUry9vbW97e3jpw4ID69euniIiIXI+z2+0KCAhwegAAAADA1Sq0b7fr3r27oqOjndpiYmLUvXt39ezZ00NVAQAAACjqPBqSUlNTtWfPHsf2vn37tG3bNpUpU0aVK1dW2bJlnfoXL15coaGhqlGjRkGXCgAAAOAm4dGQtGnTJrVo0cKxHR8fL0mKi4tTUlKSh6oCAAAAcDPzaEhq3ry5jDFX3X///v35VwwAAAAAqBAv3AAAAAAAnkBIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFh4e7oAACgKIgYu8nQJAADATZhJAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgIVHQ9LatWvVvn17hYWFyWazad68eY59GRkZGjBggOrWrauSJUsqLCxMjz/+uA4fPuy5ggEAAAAUeR4NSWlpaYqKitLkyZOz7Tt79qy2bNmiQYMGacuWLfr888+1c+dOPfDAAx6oFAAAAMDNwtuTF4+NjVVsbGyO+wIDA7V8+XKntrffflt33nmnDh48qMqVK+d4XHp6utLT0x3bKSkp7isYAAAAQJF3Q92TdPr0adlsNgUFBeXaJzExUYGBgY5HpUqVCq5AAAAAADe8GyYknT9/XgMGDFDXrl0VEBCQa7+EhASdPn3a8Th06FABVgkAAADgRufRt9tdrYyMDD366KMyxmjq1Kl59rXb7bLb7QVUGQAAAICiptCHpEsB6cCBA1q5cmWes0gAAAAAcL0KdUi6FJB2796tVatWqWzZsp4uCQAAAEAR59GQlJqaqj179ji29+3bp23btqlMmTKqUKGCHn74YW3ZskULFy5UZmamjh49KkkqU6aMfHx8PFU2AAAAgCLMoyFp06ZNatGihWM7Pj5ekhQXF6ehQ4dqwYIFkqT69es7Hbdq1So1b968oMoEAAAAcBPxaEhq3ry5jDG57s9rHwAAAADkhxtmCXAAAAAAKAiEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAICFt6cLAADgZhQxcJGnS3DYP7qtp0sAgEKFmSQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAuPhqS1a9eqffv2CgsLk81m07x585z2G2M0ePBgVahQQX5+foqOjtbu3bs9UywAAACAm4JHQ1JaWpqioqI0efLkHPe/8cYbevPNNzVt2jR99913KlmypGJiYnT+/PkCrhQAAADAzcLbkxePjY1VbGxsjvuMMZo4caJee+01dejQQZI0a9YslS9fXvPmzVOXLl0KslQAAAAAN4lCe0/Svn37dPToUUVHRzvaAgMD1bhxY61fvz7X49LT05WSkuL0AAAAAICr5dGZpLwcPXpUklS+fHmn9vLlyzv25SQxMVHDhg3L19oAeF7EwEWeLgEAABRRhXYmyVUJCQk6ffq043Ho0CFPlwQAAADgBlJoQ1JoaKgk6dixY07tx44dc+zLid1uV0BAgNMDAAAAAK5WoQ1JkZGRCg0N1YoVKxxtKSkp+u6779SkSRMPVgYAAACgKPPoPUmpqanas2ePY3vfvn3atm2bypQpo8qVK6tv3756/fXXVb16dUVGRmrQoEEKCwtTx44dPVc0AAAAgCLNoyFp06ZNatGihWM7Pj5ekhQXF6ekpCS98sorSktL01NPPaVTp06padOmWrJkiXx9fT1VMgAAAIAizqMhqXnz5jLG5LrfZrNp+PDhGj58eAFWBQAAAOBmVmjvSQIAAAAATyAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFi4FJJ+++03d9cBAAAAAIWCSyGpWrVqatGihT744AOdP3/e3TUBAAAAgMe4FJK2bNmievXqKT4+XqGhoXr66af1/fffu7s2AAAAAChwLoWk+vXra9KkSTp8+LDef/99HTlyRE2bNlWdOnU0fvx4HT9+3N11AgAAAECBuK6FG7y9vfXQQw9p7ty5GjNmjPbs2aOXX35ZlSpV0uOPP64jR464q04AAAAAKBDXFZI2bdqk5557ThUqVND48eP18ssva+/evVq+fLkOHz6sDh06uKtOAAAAACgQ3q4cNH78eM2YMUM7d+5UmzZtNGvWLLVp00bFiv2duSIjI5WUlKSIiAh31goAAAAA+c6lkDR16lQ98cQT6tGjhypUqJBjn3Llyum99967ruIAAAAAoKC5FJJ27959xT4+Pj6Ki4tz5fQAAAAA4DEu3ZM0Y8YMzZ07N1v73LlzNXPmzOsuCgAAAAA8xaWQlJiYqODg4Gzt5cqV06hRo667KAAAAADwFJdC0sGDBxUZGZmtPTw8XAcPHrzuoi7JzMzUoEGDFBkZKT8/P1WtWlUjRoyQMcZt1wAAAAAAK5fuSSpXrpx+/PHHbKvX/fDDDypbtqw76pIkjRkzRlOnTtXMmTNVu3Ztbdq0ST179lRgYKD69OnjtusAAAAAwCUuhaSuXbuqT58+8vf317333itJWrNmjV588UV16dLFbcV9++236tChg9q2bStJioiI0Mcff6zvv//ebdcAAAAAACuX3m43YsQINW7cWC1btpSfn5/8/PzUqlUr3XfffW69J+muu+7SihUrtGvXLkl/z1StW7dOsbGxuR6Tnp6ulJQUpwcAAAAAXC2XZpJ8fHz0ySefaMSIEfrhhx/k5+enunXrKjw83K3FDRw4UCkpKbrtttvk5eWlzMxMjRw5Ut26dcv1mMTERA0bNsytdQAAXBMxcJGnS3DYP7qtp0sAANwgXApJl9x666269dZb3VVLNp9++qk+/PBDffTRR6pdu7a2bdumvn37KiwsLNfPYEpISFB8fLxjOyUlRZUqVcq3GgEAAAAULS6FpMzMTCUlJWnFihVKTk5WVlaW0/6VK1e6pbj+/ftr4MCBjvuc6tatqwMHDigxMTHXkGS322W3291yfQAAAAA3H5dC0osvvqikpCS1bdtWderUkc1mc3ddkqSzZ8+qWDHn26a8vLyyhTIAAAAAcBeXQtKcOXP06aefqk2bNu6ux0n79u01cuRIVa5cWbVr19bWrVs1fvx4PfHEE/l6XQAAAAA3L5cXbqhWrZq7a8nmrbfe0qBBg/Tcc88pOTlZYWFhevrppzV48OB8vzYAAACAm5NLS4D369dPkyZNkjHG3fU48ff318SJE3XgwAGdO3dOe/fu1euvvy4fH598vS4AAACAm5dLM0nr1q3TqlWrtHjxYtWuXVvFixd32v/555+7pTgAAAAAKGguhaSgoCA9+OCD7q4FAAAAADzOpZA0Y8YMd9cBAAAAAIWCS/ckSdLFixf11Vdfafr06Tpz5owk6fDhw0pNTXVbcQAAAABQ0FyaSTpw4IBat26tgwcPKj09Xffff7/8/f01ZswYpaena9q0ae6uEwAAAAAKhEszSS+++KIaNWqkkydPys/Pz9H+4IMPasWKFW4rDgAAAAAKmkszSV9//bW+/fbbbEtxR0RE6I8//nBLYQAAAADgCS7NJGVlZSkzMzNb+++//y5/f//rLgoAAAAAPMWlkNSqVStNnDjRsW2z2ZSamqohQ4aoTZs27qoNAAAAAAqcS2+3GzdunGJiYlSrVi2dP39e//jHP7R7924FBwfr448/dneNAAAAAFBgXApJFStW1A8//KA5c+boxx9/VGpqqnr16qVu3bo5LeQAAAAAADcal0KSJHl7e+uxxx5zZy0AAAAA4HEuhaRZs2bluf/xxx93qRgAAAAA8DSXQtKLL77otJ2RkaGzZ8/Kx8dHJUqUICQBAAAAuGG5tLrdyZMnnR6pqanauXOnmjZtysINAAAAAG5oLoWknFSvXl2jR4/ONssEAAAAADcSt4Uk6e/FHA4fPuzOUwIAAABAgXLpnqQFCxY4bRtjdOTIEb399tu6++673VIYAAAAAHiCSyGpY8eOTts2m00hISG67777NG7cOHfUBQAAAAAe4VJIysrKcncdAHIRMXCRp0tw2D+6radLAFDEFaafeYUNP4OBguPWe5IAAAAA4Ebn0kxSfHz8VfcdP368K5cAAAAAAI9wKSRt3bpVW7duVUZGhmrUqCFJ2rVrl7y8vHT77bc7+tlsNvdUCQAAAAAFxKWQ1L59e/n7+2vmzJkqXbq0pL8/YLZnz56655571K9fP7cWCQAAAAAFxaV7ksaNG6fExERHQJKk0qVL6/XXX2d1OwAAAAA3NJdCUkpKio4fP56t/fjx4zpz5sx1FwUAAAAAnuJSSHrwwQfVs2dPff755/r999/1+++/67///a969eqlhx56yN01AgAAAECBcemepGnTpunll1/WP/7xD2VkZPx9Im9v9erVS2PHjnVrgQAAAABQkFwKSSVKlNCUKVM0duxY7d27V5JUtWpVlSxZ0q3FAQAAAEBBu64Pkz1y5IiOHDmi6tWrq2TJkjLGuKsuAAAAAPAIl0LSiRMn1LJlS916661q06aNjhw5Iknq1asXy38DAAAAuKG5FJJeeuklFS9eXAcPHlSJEiUc7Z07d9aSJUvcVhwAAAAAFDSX7klatmyZli5dqooVKzq1V69eXQcOHHBLYQAAAADgCS7NJKWlpTnNIF3y119/yW63X3dRAAAAAOApLoWke+65R7NmzXJs22w2ZWVl6Y033lCLFi3cVhwAAAAAFDSX3m73xhtvqGXLltq0aZMuXLigV155Rdu3b9dff/2lb775xt01AgAAAECBcWkmqU6dOtq1a5eaNm2qDh06KC0tTQ899JC2bt2qqlWrurtGAAAAACgw1zyTlJGRodatW2vatGn617/+lR81AQAAAIDHXPNMUvHixfXjjz/mRy0AAAAA4HEuvd3uscce03vvvefuWgAAAADA41xauOHixYt6//339dVXX6lhw4YqWbKk0/7x48e7pTgAAAAAKGjXFJJ+++03RURE6Oeff9btt98uSdq1a5dTH5vN5r7qAAAAAKCAXVNIql69uo4cOaJVq1ZJkjp37qw333xT5cuXz5fiJOmPP/7QgAEDtHjxYp09e1bVqlXTjBkz1KhRo3y7JgAAAICb1zWFJGOM0/bixYuVlpbm1oKsTp48qbvvvlstWrTQ4sWLFRISot27d6t06dL5dk0AAAAANzeX7km65PLQ5G5jxoxRpUqVNGPGDEdbZGRkvl4TAAAAwM3tmla3s9ls2e45ys97kBYsWKBGjRrpkUceUbly5dSgQQO9++67eR6Tnp6ulJQUpwcAAAAAXK1rfrtdjx49ZLfbJUnnz5/XM888k211u88//9wtxf3222+aOnWq4uPj9eqrr2rjxo3q06ePfHx8FBcXl+MxiYmJGjZsmFuuD8BZxMBFni4BAAAg311TSLo8mDz22GNuLeZyWVlZatSokUaNGiVJatCggX7++WdNmzYt15CUkJCg+Ph4x3ZKSooqVaqUr3UCAAAAKDquKSRZ7w0qCBUqVFCtWrWc2mrWrKn//ve/uR5jt9sdM10AAAAAcK2u6Z6kgnb33Xdr586dTm27du1SeHi4hyoCAAAAUNQV6pD00ksvacOGDRo1apT27Nmjjz76SO+884569+7t6dIAAAAAFFGFOiTdcccd+uKLL/Txxx+rTp06GjFihCZOnKhu3bp5ujQAAAAARdR1fU5SQWjXrp3atWvn6TIAAAAA3CQK9UwSAAAAABQ0QhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAwtvTBQAAUBAiBi6S34Xz2vH/b9cctETnfHw9WlNhETFwkadLAIBChZkkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAACLGyokjR49WjabTX379vV0KQAAAACKqBsmJG3cuFHTp09XvXr1PF0KAAAAgCLshghJqamp6tatm959912VLl3a0+UAAAAAKMJuiJDUu3dvtW3bVtHR0Vfsm56erpSUFKcHAAAAAFwtb08XcCVz5szRli1btHHjxqvqn5iYqGHDhuVzVQAAAACKqkI9k3To0CG9+OKL+vDDD+Xr63tVxyQkJOj06dOOx6FDh/K5SgAAAABFSaGeSdq8ebOSk5N1++23O9oyMzO1du1avf3220pPT5eXl5fTMXa7XXa7vaBLBQAAAFBEFOqQ1LJlS/30009ObT179tRtt92mAQMGZAtIAAAAAHC9CnVI8vf3V506dZzaSpYsqbJly2ZrBwAAAAB3KNT3JAEAAABAQSvUM0k5Wb16tadLAAAAAFCEMZMEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAICFt6cLAAAAwI0lYuAiT5fgsH90W0+XgCKImSQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAItCHZISExN1xx13yN/fX+XKlVPHjh21c+dOT5cFAAAAoAgr1CFpzZo16t27tzZs2KDly5crIyNDrVq1UlpamqdLAwAAAFBEeXu6gLwsWbLEaTspKUnlypXT5s2bde+993qoKgAAAABFWaEOSZc7ffq0JKlMmTK59klPT1d6erpjOyUlJd/rAgAAAFB03DAhKSsrS3379tXdd9+tOnXq5NovMTFRw4YNK8DKrk3EwEWeLgEAANyA+B0iZ4Xpddk/uq2nS3AoTK+LVLhem6tRqO9Jsurdu7d+/vlnzZkzJ89+CQkJOn36tONx6NChAqoQAAAAQFFwQ8wkPf/881q4cKHWrl2rihUr5tnXbrfLbrcXUGUAAAAAippCHZKMMXrhhRf0xRdfaPXq1YqMjPR0SQAAAACKuEIdknr37q2PPvpI8+fPl7+/v44ePSpJCgwMlJ+fn4erAwAAAFAUFep7kqZOnarTp0+refPmqlChguPxySefeLo0AAAAAEVUoZ5JMsZ4ugQAAAAAN5lCPZMEAAAAAAWNkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgIW3pwsAAAAAioKIgYs8XQLchJkkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCAAAAAAtCEgAAAABYEJIAAAAAwIKQBAAAAAAWhCQAAAAAsCAkAQAAAIAFIQkAAAAALAhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAACLGyIkTZ48WREREfL19VXjxo31/fffe7okAAAAAEVUoQ9Jn3zyieLj4zVkyBBt2bJFUVFRiomJUXJysqdLAwAAAFAEFfqQNH78eP3zn/9Uz549VatWLU2bNk0lSpTQ+++/7+nSAAAAABRB3p4uIC8XLlzQ5s2blZCQ4GgrVqyYoqOjtX79+hyPSU9PV3p6umP79OnTkqSUlJT8LfYqZaWf9XQJAHDTyrxwXpf+NchMP6ssk+XRegDgZlFYfhe/VIcxJs9+hTok/fnnn8rMzFT58uWd2suXL69ff/01x2MSExM1bNiwbO2VKlXKlxoBADeWwEt/mfK4J8sAgJtK4ERPV+DszJkzCgwMzHV/oQ5JrkhISFB8fLxjOysrS3/99ZfKli0rm81WoLWkpKSoUqVKOnTokAICAgr02sh/jG/RxdgWbYxv0cXYFm2Mb9FVkGNrjNGZM2cUFhaWZ79CHZKCg4Pl5eWlY8eOObUfO3ZMoaGhOR5jt9tlt9ud2oKCgvKrxKsSEBDAN3MRxvgWXYxt0cb4Fl2MbdHG+BZdBTW2ec0gXVKoF27w8fFRw4YNtWLFCkdbVlaWVqxYoSZNmniwMgAAAABFVaGeSZKk+Ph4xcXFqVGjRrrzzjs1ceJEpaWlqWfPnp4uDQAAAEARVOhDUufOnXX8+HENHjxYR48eVf369bVkyZJsizkURna7XUOGDMn29j8UDYxv0cXYFm2Mb9HF2BZtjG/RVRjH1mautP4dAAAAANxECvU9SQAAAABQ0AhJAAAAAGBBSAIAAAAAC0ISAAAAAFgQktxk5MiRuuuuu1SiRIlcP7zWZrNle8yZM8epz+rVq3X77bfLbrerWrVqSkpKyv/ikaerGduDBw+qbdu2KlGihMqVK6f+/fvr4sWLTn0Y2xtDREREtu/T0aNHO/X58ccfdc8998jX11eVKlXSG2+84aFqca0mT56siIgI+fr6qnHjxvr+++89XRJcMHTo0Gzfp7fddptj//nz59W7d2+VLVtWpUqVUqdOnbJ9MD0Kh7Vr16p9+/YKCwuTzWbTvHnznPYbYzR48GBVqFBBfn5+io6O1u7du536/PXXX+rWrZsCAgIUFBSkXr16KTU1tQCfBXJzpfHt0aNHtu/l1q1bO/Xx1PgSktzkwoULeuSRR/Tss8/m2W/GjBk6cuSI49GxY0fHvn379qlt27Zq0aKFtm3bpr59++rJJ5/U0qVL87l65OVKY5uZmam2bdvqwoUL+vbbbzVz5kwlJSVp8ODBjj6M7Y1l+PDhTt+nL7zwgmNfSkqKWrVqpfDwcG3evFljx47V0KFD9c4773iwYlyNTz75RPHx8RoyZIi2bNmiqKgoxcTEKDk52dOlwQW1a9d2+j5dt26dY99LL72k//3vf5o7d67WrFmjw4cP66GHHvJgtchNWlqaoqKiNHny5Bz3v/HGG3rzzTc1bdo0fffddypZsqRiYmJ0/vx5R59u3bpp+/btWr58uRYuXKi1a9fqqaeeKqingDxcaXwlqXXr1k7fyx9//LHTfo+Nr4FbzZgxwwQGBua4T5L54osvcj32lVdeMbVr13Zq69y5s4mJiXFjhXBVbmP75ZdfmmLFipmjR4862qZOnWoCAgJMenq6MYaxvZGEh4ebCRMm5Lp/ypQppnTp0o6xNcaYAQMGmBo1ahRAdbged955p+ndu7djOzMz04SFhZnExEQPVgVXDBkyxERFReW479SpU6Z48eJm7ty5jrYdO3YYSWb9+vUFVCFccfnvSVlZWSY0NNSMHTvW0Xbq1Cljt9vNxx9/bIwx5pdffjGSzMaNGx19Fi9ebGw2m/njjz8KrHZcWU6/B8fFxZkOHTrkeownx5eZpALWu3dvBQcH684779T7778vY/mYqvXr1ys6Otqpf0xMjNavX1/QZeIarF+/XnXr1nX6gOOYmBilpKRo+/btjj6M7Y1j9OjRKlu2rBo0aKCxY8c6vXVy/fr1uvfee+Xj4+Noi4mJ0c6dO3Xy5ElPlIurcOHCBW3evNnp+7BYsWKKjo7m+/AGtXv3boWFhalKlSrq1q2bDh48KEnavHmzMjIynMb6tttuU+XKlRnrG8y+fft09OhRp7EMDAxU48aNHWO5fv16BQUFqVGjRo4+0dHRKlasmL777rsCrxnXbvXq1SpXrpxq1KihZ599VidOnHDs8+T4eufr2eFk+PDhuu+++1SiRAktW7ZMzz33nFJTU9WnTx9J0tGjR51+0Zak8uXLKyUlRefOnZOfn58nysYV5DZul/bl1YexLXz69Omj22+/XWXKlNG3336rhIQEHTlyROPHj5f091hGRkY6HWMd79KlSxd4zbiyP//8U5mZmTl+H/76668eqgquaty4sZKSklSjRg0dOXJEw4YN0z333KOff/5ZR48elY+PT7Z7SMuXL+/4mYwbw6Xxyun71vrva7ly5Zz2e3t7q0yZMoz3DaB169Z66KGHFBkZqb179+rVV19VbGys1q9fLy8vL4+OLyEpDwMHDtSYMWPy7LNjxw6nm0XzMmjQIMffGzRooLS0NI0dO9YRklBw3D22KNyuZbzj4+MdbfXq1ZOPj4+efvppJSYmym6353epAK5CbGys4+/16tVT48aNFR4erk8//ZT/dAJuIF26dHH8vW7duqpXr56qVq2q1atXq2XLlh6sjJCUp379+qlHjx559qlSpYrL52/cuLFGjBih9PR02e12hYaGZlt959ixYwoICOCHvpu5c2xDQ0OzrZB1aRxDQ0MdfzK2nnM94924cWNdvHhR+/fvV40aNXIdS+n/jTcKn+DgYHl5eeU4dozbjS8oKEi33nqr9uzZo/vvv18XLlzQqVOnnGaTGOsbz6XxOnbsmCpUqOBoP3bsmOrXr+/oc/niKxcvXtRff/3FeN+AqlSpouDgYO3Zs0ctW7b06PgSkvIQEhKikJCQfDv/tm3bVLp0acf/Tjdp0kRffvmlU5/ly5erSZMm+VbDzcqdY9ukSRONHDlSycnJjinh5cuXKyAgQLVq1XL0YWw953rGe9u2bSpWrJhjbJs0aaJ//etfysjIUPHixSX9PZY1atTgrXaFmI+Pjxo2bKgVK1Y4VhXNysrSihUr9Pzzz3u2OFy31NRU7d27V927d1fDhg1VvHhxrVixQp06dZIk7dy5UwcPHuRn7g0mMjJSoaGhWrFihSMUpaSk6LvvvnOsONukSROdOnVKmzdvVsOGDSVJK1euVFZWlho3buyp0uGi33//XSdOnHCEYo+Ob74uC3ETOXDggNm6dasZNmyYKVWqlNm6davZunWrOXPmjDHGmAULFph3333X/PTTT2b37t1mypQppkSJEmbw4MGOc/z222+mRIkSpn///mbHjh1m8uTJxsvLyyxZssRTTwvmymN78eJFU6dOHdOqVSuzbds2s2TJEhMSEmISEhIc52BsbwzffvutmTBhgtm2bZvZu3ev+eCDD0xISIh5/PHHHX1OnTplypcvb7p3725+/vlnM2fOHFOiRAkzffp0D1aOqzFnzhxjt9tNUlKS+eWXX8xTTz1lgoKCnFamxI2hX79+ZvXq1Wbfvn3mm2++MdHR0SY4ONgkJycbY4x55plnTOXKlc3KlSvNpk2bTJMmTUyTJk08XDVycubMGce/q5LM+PHjzdatW82BAweMMcaMHj3aBAUFmfnz55sff/zRdOjQwURGRppz5845ztG6dWvToEED891335l169aZ6tWrm65du3rqKcEir/E9c+aMefnll8369evNvn37zFdffWVuv/12U716dXP+/HnHOTw1voQkN4mLizOSsj1WrVpljPl7ucL69eubUqVKmZIlS5qoqCgzbdo0k5mZ6XSeVatWmfr16xsfHx9TpUoVM2PGjIJ/MnBypbE1xpj9+/eb2NhY4+fnZ4KDg02/fv1MRkaG03kY28Jv8+bNpnHjxiYwMND4+vqamjVrmlGjRjn9sDbGmB9++ME0bdrU2O12c8stt5jRo0d7qGJcq7feestUrlzZ+Pj4mDvvvNNs2LDB0yXBBZ07dzYVKlQwPj4+5pZbbjGdO3c2e/bscew/d+6cee6550zp0qVNiRIlzIMPPmiOHDniwYqRm1WrVuX4b2xcXJwx5u9lwAcNGmTKly9v7Ha7admypdm5c6fTOU6cOGG6du1qSpUqZQICAkzPnj0d/5EJz8prfM+ePWtatWplQkJCTPHixU14eLj55z//me0/rjw1vjZjLGtQAwAAAMBNjs9JAgAAAAALQhIAAAAAWBCSAAAAAMCCkAQAAAAAFoQkAAAAALAgJAEAAACABSEJAAAAACwISQAAAABgQUgCABQ5+/fvl81m07Zt23Lts3r1atlsNp06dcqt17bZbJo3b55bzwkAKFiEJACAR/To0UM2m002m03FixdXZGSkXnnlFZ0/f/66z12pUiUdOXJEderUcUOlAICbjbenCwAA3Lxat26tGTNmKCMjQ5s3b1ZcXJxsNpvGjBlzXef18vJSaGiom6oEANxsmEkCAHiM3W5XaGioKlWqpI4dOyo6OlrLly+XJGVlZSkxMVGRkZHy8/NTVFSUPvvsM8exJ0+eVLdu3RQSEiI/Pz9Vr15dM2bMkJTz2+2+/PJL3XrrrfLz81OLFi20f/9+p1qGDh2q+vXrO7VNnDhRERERju2NGzfq/vvvV3BwsAIDA9WsWTNt2bLFra8JAMDzCEkAgELh559/1rfffisfHx9JUmJiombNmqVp06Zp+/bteumll/TYY49pzZo1kqRBgwbpl19+0eLFi7Vjxw5NnTpVwcHBOZ770KFDeuihh9S+fXtt27ZNTz75pAYOHHjNNZ45c0ZxcXFat26dNmzYoOrVq6tNmzY6c+aM608cAFDo8HY7AIDHLFy4UKVKldLFixeVnp6uYsWK6e2331Z6erpGjRqlr776Sk2aNJEkValSRevWrdP06dPVrFkzHTx4UA0aNFCjRo0kyWnG53JTp05V1apVNW7cOElSjRo19NNPP13z2/ruu+8+p+133nlHQUFBWrNmjdq1a3dN5wIAFF6EJACAx7Ro0UJTp05VWlqaJkyYIG9vb3Xq1Enbt2/X2bNndf/99zv1v3Dhgho0aCBJevbZZ9WpUydt2bJFrVq1UseOHXXXXXfleJ0dO3aocePGTm2Xwte1OHbsmF577TWtXr1aycnJyszM1NmzZ3Xw4MFrPhcAoPAiJAEAPKZkyZKqVq2aJOn9999XVFSU3nvvPceqdIsWLdItt9zidIzdbpckxcbG6sCBA/ryyy+1fPlytWzZUr1799a///1vl2opVqyYjDFObRkZGU7bcXFxOnHihCZNmqTw8HDZ7XY1adJEFy5ccOmaAIDCiZAEACgUihUrpldffVXx8fHatWuX7Ha7Dh48qGbNmuV6TEhIiOLi4hQXF6d77rlH/fv3zzEk1axZUwsWLHBq27BhQ7ZzHT16VMYY2Ww2Scr2OUvffPONpkyZojZt2kj6+16nP//805WnCwAoxFi4AQBQaDzyyCPy8vLS9OnT9fLLL+ull17SzJkztXfvXm3ZskVvvfWWZs6cKUkaPHiw5s+frz179mj79u1auHChatasmeN5n3nmGe3evVv9+/fXzp079dFHHykpKcmpT/PmzXX8+HG98cYb2rt3ryZPnqzFixc79alevbpmz56tHTt26LvvvlO3bt3k5+eXL68FAMBzCEkAgELD29tbzz//vN544w0lJCRo0KBBSkxMVM2aNdW6dWstWrRIkZGRkiQfHx8lJCSoXr16uvfee+Xl5aU5c+bkeN7KlSvrv//9r+bNm6eoqChNmzZNo0aNcupTs2ZNTZkyRZMnT1ZUVJS+//57vfzyy0593nvvPZ08eVK33367unfvrj59+qhcuXL582IAADzGZi5/AzYAAAAA3MSYSQIAAAAAC0ISAAAAAFgQkgAAAADAgpAEAAAAABaEJAAAAACwICQBAAAAgAUhCQAAAAAsCEkAAAAAYEFIAgAAAAALQhIAAAAAWBCSAAAAAMDi/wOo2ralkgxx0gAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiple Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 42.42\n",
      "Mean absolute percentage error: 40.74%\n",
      "Root Mean squared error: 54.25\n",
      "Coefficient of determination: 0.47\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create linear regression object\n",
    "regr = LinearRegression()\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr.fit(diabetes_X_train_standardised, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr.predict(diabetes_X_test_standardised)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'Multiple Linear Regression',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 57.55\n",
      "Mean absolute percentage error: 50.68%\n",
      "Root Mean squared error: 76.00\n",
      "Coefficient of determination: -0.04\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regr_dt = DecisionTreeRegressor(random_state=42)\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr_dt.fit(diabetes_X_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr_dt.predict(diabetes_X_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'Decision Tree',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1cAAAHWCAYAAACbsXOkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAABCVElEQVR4nO3de1wWZf7/8fcNyA0qBwsUURI8n7GsXExXTRLRTMusWEs0tdZ0y0hL+paH2qI0T6Wp2ypoJ83WtE3D8yFTKw+UmpqaiCbgIRHBBIT5/eHPe7vloODgfaOv5+Mxj4dzzTXXfOZmgvvd3HPdFsMwDAEAAAAAromLowsAAAAAgBsB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAquLFjx8pisVyXY3Xs2FEdO3a0ra9bt04Wi0Wff/75dTl+//79FRwcfF2OVVZZWVkaNGiQAgICZLFYNHz4cEeXVIjFYtHYsWOv2O96XFuXrqF169aV63EA4HogXAGAE0lISJDFYrEtHh4eCgwMVEREhN59912dPXvWlOMcO3ZMY8eOVVJSkinjmcmZa7sab775phISEjRkyBB9+OGHeuKJJ4rtGxwcbPfzrlKliu6++27NmzfvOlYMADCLm6MLAAAU9tprrykkJER5eXlKS0vTunXrNHz4cE2aNElffvmlWrZsaev7yiuvaNSoUaUa/9ixYxo3bpyCg4PVqlWrq95vxYoVpTpOWZRU2wcffKCCgoJyr+FarFmzRn/5y180ZsyYq+rfqlUrvfDCC5Kk1NRU/fvf/1Z0dLRycnI0ePDgcqnxjz/+kJsbbwEAwGz8ZgUAJxQZGak777zTth4bG6s1a9bo/vvv1wMPPKA9e/bI09NTkuTm5lbub5TPnTunypUry93dvVyPcyWVKlVy6PGvxvHjx9W0adOr7l+rVi09/vjjtvX+/furbt26mjx5crmFKw8Pj3IZFwBudnwsEAAqiHvvvVevvvqqDh8+rI8++sjWXtRzMStXrlS7du3k6+urqlWrqlGjRnr55ZclXXzG5a677pIkDRgwwPaRtISEBEkXn6tq3ry5tm3bpr/+9a+qXLmybd/Ln7m6JD8/Xy+//LICAgJUpUoVPfDAAzpy5Ihdn+DgYPXv37/Qvn8e80q1FfXMVXZ2tl544QUFBQXJarWqUaNGeuedd2QYhl0/i8WiYcOGafHixWrevLmsVquaNWumxMTEol/wyxw/flwDBw5UjRo15OHhodDQUM2dO9e2/dKzQ4cOHdLSpUtttScnJ1/V+Jf4+/urcePGOnjwoF17QUGBpkyZombNmsnDw0M1atTQ008/rdOnT9v127p1qyIiIuTn5ydPT0+FhIToySefLPRaXP7M1caNG3XXXXfJw8ND9erV06xZswrVlpycbPfzKGnMw4cP65lnnlGjRo3k6empW2+9VX369Lmq12P//v3q3bu3AgIC5OHhodq1a+uxxx7TmTNnrrgvADgSd64AoAJ54okn9PLLL2vFihXF3tXYvXu37r//frVs2VKvvfaarFarDhw4oG+//VaS1KRJE7322msaPXq0nnrqKbVv316S1LZtW9sYp06dUmRkpB577DE9/vjjqlGjRol1vfHGG7JYLHrppZd0/PhxTZkyReHh4UpKSrLdYbsaV1PbnxmGoQceeEBr167VwIED1apVKy1fvlwjR47Ub7/9psmTJ9v137hxoxYtWqRnnnlGXl5eevfdd9W7d2+lpKTo1ltvLbauP/74Qx07dtSBAwc0bNgwhYSEaOHCherfv78yMjL03HPPqUmTJvrwww/1/PPPq3bt2raP+vn7+1/1+UvShQsXdPToUVWrVs2u/emnn1ZCQoIGDBigZ599VocOHdK0adO0Y8cOffvtt6pUqZKOHz+uLl26yN/fX6NGjZKvr6+Sk5O1aNGiEo+5c+dO235jx47VhQsXNGbMmCv+3Evyww8/aNOmTXrsscdUu3ZtJScna8aMGerYsaN+/vlnVa5cucj9cnNzFRERoZycHP3jH/9QQECAfvvtN3311VfKyMiQj49PmWsCgHJnAACcRnx8vCHJ+OGHH4rt4+PjY9x+++229TFjxhh//nU+efJkQ5Jx4sSJYsf44YcfDElGfHx8oW0dOnQwJBkzZ84scluHDh1s62vXrjUkGbVq1TIyMzNt7Z999pkhyZg6daqtrU6dOkZ0dPQVxyyptujoaKNOnTq29cWLFxuSjH/+8592/R5++GHDYrEYBw4csLVJMtzd3e3afvzxR0OS8d577xU61p9NmTLFkGR89NFHtrbc3FwjLCzMqFq1qt2516lTx+jevXuJ4/25b5cuXYwTJ04YJ06cMHbu3Gk88cQThiRj6NChtn7ffPONIcn4+OOP7fZPTEy0a//iiy+ueP0YxsXXYsyYMbb1Xr16GR4eHsbhw4dtbT///LPh6upqd20dOnSo2J/N5WOeO3euUJ/Nmzcbkox58+bZ2i5dQ2vXrjUMwzB27NhhSDIWLlxY4jkAgDPiY4EAUMFUrVq1xFkDfX19JUlLliwp8+QPVqtVAwYMuOr+/fr1k5eXl2394YcfVs2aNbVs2bIyHf9qLVu2TK6urnr22Wft2l944QUZhqGvv/7arj08PFz16tWzrbds2VLe3t769ddfr3icgIAARUVF2doqVaqkZ599VllZWVq/fn2Zz2HFihXy9/eXv7+/WrRooQ8//FADBgzQhAkTbH0WLlwoHx8f3XfffTp58qRtad26tapWraq1a9dK+t/P/quvvlJeXt5VHT8/P1/Lly9Xr169dNttt9namzRpooiIiDKf15/vWObl5enUqVOqX7++fH19tX379mL3u3Rnavny5Tp37lyZjw8AjkC4AoAKJisryy7IXO7RRx/VPffco0GDBqlGjRp67LHH9Nlnn5UqaNWqVatUk1c0aNDAbt1isah+/fqlft6otA4fPqzAwMBCr0eTJk1s2//sz+HhkmrVqhV6bqmo4zRo0EAuLvZ/Nos7Tmm0adNGK1euVGJiot555x35+vrq9OnTdq///v37debMGVWvXt0WxC4tWVlZOn78uCSpQ4cO6t27t8aNGyc/Pz/17NlT8fHxysnJKfb4J06c0B9//FHoZyhJjRo1KvN5/fHHHxo9erTtWTg/Pz/5+/srIyOjxGenQkJCFBMTo3//+9/y8/NTRESEpk+fzvNWACoEnrkCgArk6NGjOnPmjOrXr19sH09PT23YsEFr167V0qVLlZiYqAULFujee+/VihUr5OrqesXjlOY5qatV3JfR5ufnX1VNZijuOMZlk19cT35+fgoPD5ckRUREqHHjxrr//vs1depUxcTESLo4mUX16tX18ccfFznGpee6Ln2h85YtW/Tf//5Xy5cv15NPPqmJEydqy5Ytqlq16jXVWtLP8HL/+Mc/FB8fr+HDhyssLEw+Pj6yWCx67LHHrhj0J06cqP79+2vJkiVasWKFnn32WcXFxWnLli2qXbv2NZ0DAJQn7lwBQAXy4YcfStIVP67l4uKizp07a9KkSfr555/1xhtvaM2aNbaPjxX3Jrms9u/fb7duGIYOHDhgN7NftWrVlJGRUWjfy+/6lKa2OnXq6NixY4U+Jrl3717bdjPUqVNH+/fvLxQKzD6OJHXv3l0dOnTQm2++qezsbElSvXr1dOrUKd1zzz0KDw8vtISGhtqN8Ze//EVvvPGGtm7dqo8//li7d+/W/Pnzizyev7+/PD09C/0MJWnfvn1265cm2bj851jUnbvPP/9c0dHRmjhxoh5++GHdd999ateuXZHXQFFatGihV155RRs2bNA333yj3377TTNnzryqfQHAUQhXAFBBrFmzRq+//rpCQkLUt2/fYvv9/vvvhdoufRnvpY+HValSRVLhN8llNW/ePLuA8/nnnys1NVWRkZG2tnr16mnLli3Kzc21tX311VeFpmwvTW3dunVTfn6+pk2bZtc+efJkWSwWu+Nfi27duiktLU0LFiywtV24cEHvvfeeqlatqg4dOphynEteeuklnTp1Sh988IEk6ZFHHlF+fr5ef/31Qn0vXLhge61Onz5d6C7c5T/7y7m6uioiIkKLFy9WSkqKrX3Pnj1avny5XV9vb2/5+flpw4YNdu3vv/9+keNeXst7771X5F2uP8vMzNSFCxfs2lq0aCEXF5cSP94IAM6AjwUCgBP6+uuvtXfvXl24cEHp6elas2aNVq5cqTp16ujLL78s8UtgX3vtNW3YsEHdu3dXnTp1dPz4cb3//vuqXbu22rVrJ+li0PH19dXMmTPl5eWlKlWqqE2bNgoJCSlTvbfccovatWunAQMGKD09XVOmTFH9+vXtposfNGiQPv/8c3Xt2lWPPPKIDh48qI8++shugonS1tajRw916tRJ//d//6fk5GSFhoZqxYoVWrJkiYYPH15o7LJ66qmnNGvWLPXv31/btm1TcHCwPv/8c3377beaMmVKic/AlUVkZKSaN2+uSZMmaejQoerQoYOefvppxcXFKSkpSV26dFGlSpW0f/9+LVy4UFOnTtXDDz+suXPn6v3339eDDz6oevXq6ezZs/rggw/k7e2tbt26FXu8cePGKTExUe3bt9czzzxjC47NmjXTTz/9ZNd30KBBeuuttzRo0CDdeeed2rBhg3755ZdCY95///368MMP5ePjo6ZNm2rz5s1atWpViVPeSxf/J8KwYcPUp08fNWzYUBcuXNCHH34oV1dX9e7du2wvKABcLw6dqxAAYOfSVOyXFnd3dyMgIMC47777jKlTp9pN+X3J5VOxr1692ujZs6cRGBhouLu7G4GBgUZUVJTxyy+/2O23ZMkSo2nTpoabm5vd9NodOnQwmjVrVmR9xU3F/umnnxqxsbFG9erVDU9PT6N79+5203pfMnHiRKNWrVqG1Wo17rnnHmPr1q2FxiyptsunYjcMwzh79qzx/PPPG4GBgUalSpWMBg0aGBMmTDAKCgrs+umy6c0vKW6K+Mulp6cbAwYMMPz8/Ax3d3ejRYsWRU5JXtqp2Ivrm5CQUGja83/9619G69atDU9PT8PLy8to0aKF8eKLLxrHjh0zDMMwtm/fbkRFRRm33XabYbVajerVqxv333+/sXXrVruxddm06YZhGOvXrzdat25tuLu7G3Xr1jVmzpxZ6NoyjItTrA8cONDw8fExvLy8jEceecQ4fvx4oTFPnz5te72qVq1qREREGHv37i30el8+Ffuvv/5qPPnkk0a9evUMDw8P45ZbbjE6depkrFq16qpeUwBwJIthOPApXgAAAAC4QfDMFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAm4EuEi1BQUKBjx47Jy8tLFovF0eUAAAAAcBDDMHT27FkFBgbKxaXke1OEqyIcO3ZMQUFBji4DAAAAgJM4cuSIateuXWIfwlURvLy8JF18Ab29vR1cDQCgQsjOlgIDL/772DGpShXH1gMAMEVmZqaCgoJsGaEkhKsiXPoooLe3N+EKAHB1XF3/929vb8IVANxgruZxISa0AAAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwgZujCwAAwJkFj1p6Vf08c89rz///d5NXE/WHu4fptSS/1d30MQEA5uHOFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmMCh4WrDhg3q0aOHAgMDZbFYtHjxYrvtFoulyGXChAnFjjl27NhC/Rs3blzOZwIAAADgZufQcJWdna3Q0FBNnz69yO2pqal2y5w5c2SxWNS7d+8Sx23WrJndfhs3biyP8gEAAADAxqHfcxUZGanIyMhitwcEBNitL1myRJ06dVLdunVLHNfNza3QvgAAAABQnirMM1fp6elaunSpBg4ceMW++/fvV2BgoOrWrau+ffsqJSWlxP45OTnKzMy0WwAAAACgNCpMuJo7d668vLz00EMPldivTZs2SkhIUGJiombMmKFDhw6pffv2Onv2bLH7xMXFycfHx7YEBQWZXT4AAACAG1yFCVdz5sxR37595eHhUWK/yMhI9enTRy1btlRERISWLVumjIwMffbZZ8XuExsbqzNnztiWI0eOmF0+AAAAgBucQ5+5ulrffPON9u3bpwULFpR6X19fXzVs2FAHDhwoto/VapXVar2WEgEAAADc5CrEnavZs2erdevWCg0NLfW+WVlZOnjwoGrWrFkOlQEAAADARQ4NV1lZWUpKSlJSUpIk6dChQ0pKSrKbgCIzM1MLFy7UoEGDihyjc+fOmjZtmm19xIgRWr9+vZKTk7Vp0yY9+OCDcnV1VVRUVLmeCwAAAICbm0M/Frh161Z16tTJth4TEyNJio6OVkJCgiRp/vz5Mgyj2HB08OBBnTx50rZ+9OhRRUVF6dSpU/L391e7du20ZcsW+fv7l9+JAAAAALjpOTRcdezYUYZhlNjnqaee0lNPPVXs9uTkZLv1+fPnm1EaAAAAAJRKhXjmCgAAAACcHeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEDg1XGzZsUI8ePRQYGCiLxaLFixfbbe/fv78sFovd0rVr1yuOO336dAUHB8vDw0Nt2rTR999/X05nAAAAAAAXOTRcZWdnKzQ0VNOnTy+2T9euXZWammpbPv300xLHXLBggWJiYjRmzBht375doaGhioiI0PHjx80uHwAAAABs3Bx58MjISEVGRpbYx2q1KiAg4KrHnDRpkgYPHqwBAwZIkmbOnKmlS5dqzpw5GjVq1DXVCwAAAADFcfpnrtatW6fq1aurUaNGGjJkiE6dOlVs39zcXG3btk3h4eG2NhcXF4WHh2vz5s3F7peTk6PMzEy7BQAAAABKw6nDVdeuXTVv3jytXr1ab7/9ttavX6/IyEjl5+cX2f/kyZPKz89XjRo17Npr1KihtLS0Yo8TFxcnHx8f2xIUFGTqeQAAAAC48Tn0Y4FX8thjj9n+3aJFC7Vs2VL16tXTunXr1LlzZ9OOExsbq5iYGNt6ZmYmAQsAAABAqTj1navL1a1bV35+fjpw4ECR2/38/OTq6qr09HS79vT09BKf27JarfL29rZbAAAAAKA0KlS4Onr0qE6dOqWaNWsWud3d3V2tW7fW6tWrbW0FBQVavXq1wsLCrleZAAAAAG5CDg1XWVlZSkpKUlJSkiTp0KFDSkpKUkpKirKysjRy5Eht2bJFycnJWr16tXr27Kn69esrIiLCNkbnzp01bdo023pMTIw++OADzZ07V3v27NGQIUOUnZ1tmz0QAAAAAMqDQ5+52rp1qzp16mRbv/TcU3R0tGbMmKGffvpJc+fOVUZGhgIDA9WlSxe9/vrrslqttn0OHjyokydP2tYfffRRnThxQqNHj1ZaWppatWqlxMTEQpNcAAAAAICZHBquOnbsKMMwit2+fPnyK46RnJxcqG3YsGEaNmzYtZQGAAAAAKVSoZ65AgAAAABnRbgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABO4OboAAABwdYJHLXV0CU4p+a3uji4BACRx5woAAAAATEG4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEbo4uAACAPwsetdTRJQAAUCbcuQIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABM4NFxt2LBBPXr0UGBgoCwWixYvXmzblpeXp5deekktWrRQlSpVFBgYqH79+unYsWMljjl27FhZLBa7pXHjxuV8JgAAAABudg4NV9nZ2QoNDdX06dMLbTt37py2b9+uV199Vdu3b9eiRYu0b98+PfDAA1cct1mzZkpNTbUtGzduLI/yAQAAAMDGod9zFRkZqcjIyCK3+fj4aOXKlXZt06ZN0913362UlBTddtttxY7r5uamgIAAU2sFAAAAgJJUqGeuzpw5I4vFIl9f3xL77d+/X4GBgapbt6769u2rlJSUEvvn5OQoMzPTbgEAAACA0qgw4er8+fN66aWXFBUVJW9v72L7tWnTRgkJCUpMTNSMGTN06NAhtW/fXmfPni12n7i4OPn4+NiWoKCg8jgFAAAAADewChGu8vLy9Mgjj8gwDM2YMaPEvpGRkerTp49atmypiIgILVu2TBkZGfrss8+K3Sc2NlZnzpyxLUeOHDH7FAAAAADc4Bz6zNXVuBSsDh8+rDVr1pR416oovr6+atiwoQ4cOFBsH6vVKqvVeq2lAgAAALiJOfWdq0vBav/+/Vq1apVuvfXWUo+RlZWlgwcPqmbNmuVQIQAAAABc5NBwlZWVpaSkJCUlJUmSDh06pKSkJKWkpCgvL08PP/ywtm7dqo8//lj5+flKS0tTWlqacnNzbWN07txZ06ZNs62PGDFC69evV3JysjZt2qQHH3xQrq6uioqKut6nBwAAAOAm4tCPBW7dulWdOnWyrcfExEiSoqOjNXbsWH355ZeSpFatWtntt3btWnXs2FGSdPDgQZ08edK27ejRo4qKitKpU6fk7++vdu3aacuWLfL39y/fkwEAAABwU3NouOrYsaMMwyh2e0nbLklOTrZbnz9//rWWBQAAAACl5tTPXAEAAABARUG4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABGUKV7/++qvZdQAAAABAhVamcFW/fn116tRJH330kc6fP292TQAAAABQ4ZQpXG3fvl0tW7ZUTEyMAgIC9PTTT+v77783uzYAAAAAqDDKFK5atWqlqVOn6tixY5ozZ45SU1PVrl07NW/eXJMmTdKJEyfMrhMAAAAAnNo1TWjh5uamhx56SAsXLtTbb7+tAwcOaMSIEQoKClK/fv2UmppqVp0AAAAA4NSuKVxt3bpVzzzzjGrWrKlJkyZpxIgROnjwoFauXKljx46pZ8+eZtUJAAAAAE7NrSw7TZo0SfHx8dq3b5+6deumefPmqVu3bnJxuZjVQkJClJCQoODgYDNrBQAAAACnVaZwNWPGDD355JPq37+/atasWWSf6tWra/bs2ddUHAAAAABUFGUKV/v3779iH3d3d0VHR5dleAAAAACocMr0zFV8fLwWLlxYqH3hwoWaO3fuNRcFAAAAABVNmcJVXFyc/Pz8CrVXr15db7755jUXBQAAAAAVTZnCVUpKikJCQgq116lTRykpKddcFAAAAABUNGUKV9WrV9dPP/1UqP3HH3/Urbfees1FAQAAAEBFU6ZwFRUVpWeffVZr165Vfn6+8vPztWbNGj333HN67LHHzK4RAAAAAJxemWYLfP3115WcnKzOnTvLze3iEAUFBerXrx/PXAEAAAC4KZUpXLm7u2vBggV6/fXX9eOPP8rT01MtWrRQnTp1zK4PAAAAACqEMoWrSxo2bKiGDRuaVQsAAAAAVFhleuYqPz9fs2fP1t/+9jeFh4fr3nvvtVuu1oYNG9SjRw8FBgbKYrFo8eLFdtsNw9Do0aNVs2ZNeXp6Kjw8/Kq+wHj69OkKDg6Wh4eH2rRpo++//760pwgAAAAApVKmcPXcc8/pueeeU35+vpo3b67Q0FC75WplZ2crNDRU06dPL3L7+PHj9e6772rmzJn67rvvVKVKFUVEROj8+fPFjrlgwQLFxMRozJgx2r59u0JDQxUREaHjx4+X+jwBAAAA4GqV6WOB8+fP12effaZu3bpd08EjIyMVGRlZ5DbDMDRlyhS98sor6tmzpyRp3rx5qlGjhhYvXlzsrISTJk3S4MGDNWDAAEnSzJkztXTpUs2ZM0ejRo26pnoBAAAAoDhlunPl7u6u+vXrm12LnUOHDiktLU3h4eG2Nh8fH7Vp00abN28ucp/c3Fxt27bNbh8XFxeFh4cXu48k5eTkKDMz024BAAAAgNIo052rF154QVOnTtW0adNksVjMrkmSlJaWJkmqUaOGXXuNGjVs2y538uRJ5efnF7nP3r17iz1WXFycxo0bd40VAwAARwgetdTRJdgkv9Xd0SUAcKAyhauNGzdq7dq1+vrrr9WsWTNVqlTJbvuiRYtMKe56iY2NVUxMjG09MzNTQUFBDqwIAAAAQEVTpnDl6+urBx980Oxa7AQEBEiS0tPTVbNmTVt7enq6WrVqVeQ+fn5+cnV1VXp6ul17enq6bbyiWK1WWa3Way8aAAAAwE2rTOEqPj7e7DoKCQkJUUBAgFavXm0LU5mZmfruu+80ZMiQIvdxd3dX69attXr1avXq1UuSVFBQoNWrV2vYsGHlXjMAAACAm1eZJrSQpAsXLmjVqlWaNWuWzp49K0k6duyYsrKyrnqMrKwsJSUlKSkpSdLFSSySkpKUkpIii8Wi4cOH65///Ke+/PJL7dy5U/369VNgYKAtOElS586dNW3aNNt6TEyMPvjgA82dO1d79uzRkCFDlJ2dbZs9EAAAAADKQ5nuXB0+fFhdu3ZVSkqKcnJydN9998nLy0tvv/22cnJyNHPmzKsaZ+vWrerUqZNt/dJzT9HR0UpISNCLL76o7OxsPfXUU8rIyFC7du2UmJgoDw8P2z4HDx7UyZMnbeuPPvqoTpw4odGjRystLU2tWrVSYmJioUkuAAAAAMBMFsMwjNLu1KtXL3l5eWn27Nm69dZb9eOPP6pu3bpat26dBg8erP3795dHrddNZmamfHx8dObMGXl7ezu6HAC4qTjTzG+l4Zl7XnsmPyxJavL85/rD3eMKe+BGxGyBwI2nNNmgTHeuvvnmG23atEnu7u527cHBwfrtt9/KMiQAAAAAVGhleuaqoKBA+fn5hdqPHj0qLy+vay4KAAAAACqaMoWrLl26aMqUKbZ1i8WirKwsjRkzRt26dTOrNgAAAACoMMr0scCJEycqIiJCTZs21fnz5/W3v/1N+/fvl5+fnz799FOzawQAAAAAp1emcFW7dm39+OOPmj9/vn766SdlZWVp4MCB6tu3rzw9Pc2uEQAAAACcXpnClSS5ubnp8ccfN7MWAAAAAKiwyhSu5s2bV+L2fv36lakYAAAAAKioyhSunnvuObv1vLw8nTt3Tu7u7qpcuTLhCgAAAMBNp0yzBZ4+fdpuycrK0r59+9SuXTsmtAAAAABwUypTuCpKgwYN9NZbbxW6qwUAAAAANwPTwpV0cZKLY8eOmTkkAAAAAFQIZXrm6ssvv7RbNwxDqampmjZtmu655x5TCgMAAACAiqRM4apXr1526xaLRf7+/rr33ns1ceJEM+oCAAAAgAqlTOGqoKDA7DoAAA4UPGqpo0sAAKDCM/WZKwAAAAC4WZXpzlVMTMxV9500aVJZDgEAAAAAFUqZwtWOHTu0Y8cO5eXlqVGjRpKkX375Ra6urrrjjjts/SwWizlVAgAAAICTK1O46tGjh7y8vDR37lxVq1ZN0sUvFh4wYIDat2+vF154wdQiAQAAAMDZlemZq4kTJyouLs4WrCSpWrVq+uc//8lsgQAAAABuSmUKV5mZmTpx4kSh9hMnTujs2bPXXBQAAAAAVDRlClcPPvigBgwYoEWLFuno0aM6evSo/vOf/2jgwIF66KGHzK4RAAAAAJxemZ65mjlzpkaMGKG//e1vysvLuziQm5sGDhyoCRMmmFogAAAAAFQEZQpXlStX1vvvv68JEybo4MGDkqR69eqpSpUqphYHAAAAABXFNX2JcGpqqlJTU9WgQQNVqVJFhmGYVRcAAAAAVChlClenTp1S586d1bBhQ3Xr1k2pqamSpIEDBzINOwAAAICbUpnC1fPPP69KlSopJSVFlStXtrU/+uijSkxMNK04AAAAAKgoyvTM1YoVK7R8+XLVrl3brr1BgwY6fPiwKYUBAAAAQEVSpjtX2dnZdnesLvn9999ltVqvuSgAAAAAqGjKFK7at2+vefPm2dYtFosKCgo0fvx4derUybTiAAAAAKCiKNPHAsePH6/OnTtr69atys3N1Ysvvqjdu3fr999/17fffmt2jQAAAADg9Mp056p58+b65Zdf1K5dO/Xs2VPZ2dl66KGHtGPHDtWrV8/sGgEAAADA6ZX6zlVeXp66du2qmTNn6v/+7//KoyYAAAAAqHBKfeeqUqVK+umnn8qjFgAAAACosMr0scDHH39cs2fPNrsWAAAAAKiwyjShxYULFzRnzhytWrVKrVu3VpUqVey2T5o0yZTiAAAAAKCiKFW4+vXXXxUcHKxdu3bpjjvukCT98ssvdn0sFot51QEAAABABVGqcNWgQQOlpqZq7dq1kqRHH31U7777rmrUqFEuxQEAAABARVGqZ64Mw7Bb//rrr5WdnW1qQQAAAABQEZVpQotLLg9bAAAAAHCzKlW4slgshZ6p4hkrAAAAACjlM1eGYah///6yWq2SpPPnz+vvf/97odkCFy1aZFqBwcHBOnz4cKH2Z555RtOnTy/UnpCQoAEDBti1Wa1WnT9/3rSaAAAAAOBypQpX0dHRduuPP/64qcUU5YcfflB+fr5tfdeuXbrvvvvUp0+fYvfx9vbWvn37bOvcXQMAAABQ3koVruLj48urjmL5+/vbrb/11luqV6+eOnToUOw+FotFAQEB5V0aAAAAANhc04QW11tubq4++ugjPfnkkyXejcrKylKdOnUUFBSknj17avfu3SWOm5OTo8zMTLsFAAAAAEqjQoWrxYsXKyMjQ/379y+2T6NGjTRnzhwtWbJEH330kQoKCtS2bVsdPXq02H3i4uLk4+NjW4KCgsqhegAAAAA3sgoVrmbPnq3IyEgFBgYW2ycsLEz9+vVTq1at1KFDBy1atEj+/v6aNWtWsfvExsbqzJkztuXIkSPlUT4AAACAG1ipnrlypMOHD2vVqlWlnomwUqVKuv3223XgwIFi+1itVtsMiAAAAABQFhXmzlV8fLyqV6+u7t27l2q//Px87dy5UzVr1iynygAAAACggoSrgoICxcfHKzo6Wm5u9jfb+vXrp9jYWNv6a6+9phUrVujXX3/V9u3b9fjjj+vw4cMaNGjQ9S4bAAAAwE2kQnwscNWqVUpJSdGTTz5ZaFtKSopcXP6XEU+fPq3BgwcrLS1N1apVU+vWrbVp0yY1bdr0epYMAAAA4CZTIcJVly5dZBhGkdvWrVtntz558mRNnjz5OlQFAAAAAP9TIT4WCAAAAADOjnAFAAAAACYgXAEAAACACSrEM1cAcCMKHrXU0SUAAAATcecKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADCBm6MLAHBjCx611NEl2El+q7ujSwAAADco7lwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAAAAAJiBcAQAAAIAJCFcAAAAAYALCFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJjAqcPV2LFjZbFY7JbGjRuXuM/ChQvVuHFjeXh4qEWLFlq2bNl1qhYAAADAzcypw5UkNWvWTKmpqbZl48aNxfbdtGmToqKiNHDgQO3YsUO9evVSr169tGvXrutYMQAAAICbkdOHKzc3NwUEBNgWPz+/YvtOnTpVXbt21ciRI9WkSRO9/vrruuOOOzRt2rTrWDEAAACAm5HTh6v9+/crMDBQdevWVd++fZWSklJs382bNys8PNyuLSIiQps3by7xGDk5OcrMzLRbAAAAAKA03BxdQEnatGmjhIQENWrUSKmpqRo3bpzat2+vXbt2ycvLq1D/tLQ01ahRw66tRo0aSktLK/E4cXFxGjdunKm1A3BOwaOWOroEAABwg3LqO1eRkZHq06ePWrZsqYiICC1btkwZGRn67LPPTD1ObGyszpw5Y1uOHDli6vgAAAAAbnxOfefqcr6+vmrYsKEOHDhQ5PaAgAClp6fbtaWnpysgIKDEca1Wq6xWq2l1AgAAALj5OPWdq8tlZWXp4MGDqlmzZpHbw8LCtHr1aru2lStXKiws7HqUBwAAAOAm5tThasSIEVq/fr2Sk5O1adMmPfjgg3J1dVVUVJQkqV+/foqNjbX1f+6555SYmKiJEydq7969Gjt2rLZu3aphw4Y56hQAAAAA3CSc+mOBR48eVVRUlE6dOiV/f3+1a9dOW7Zskb+/vyQpJSVFLi7/y4dt27bVJ598oldeeUUvv/yyGjRooMWLF6t58+aOOgUAAAAANwmnDlfz588vcfu6desKtfXp00d9+vQpp4oAAAAAoGhO/bFAAAAAAKgoCFcAAAAAYALCFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAAAAAJiBcAQAAAIAJCFcAAAAAYALCFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAAAAAJiBcAQAAAIAJCFcAAAAAYALCFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAjdHFwDAfMGjljq6BAC4KfH7t3jJb3V3dAlAuePOFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmMCpw1VcXJzuuusueXl5qXr16urVq5f27dtX4j4JCQmyWCx2i4eHx3WqGAAAAMDNyqnD1fr16zV06FBt2bJFK1euVF5enrp06aLs7OwS9/P29lZqaqptOXz48HWqGAAAAMDNyqm/5yoxMdFuPSEhQdWrV9e2bdv017/+tdj9LBaLAgICyrs8AAAAALBx6jtXlztz5owk6ZZbbimxX1ZWlurUqaOgoCD17NlTu3fvLrF/Tk6OMjMz7RYAAAAAKI0KE64KCgo0fPhw3XPPPWrevHmx/Ro1aqQ5c+ZoyZIl+uijj1RQUKC2bdvq6NGjxe4TFxcnHx8f2xIUFFQepwAAAADgBlZhwtXQoUO1a9cuzZ8/v8R+YWFh6tevn1q1aqUOHTpo0aJF8vf316xZs4rdJzY2VmfOnLEtR44cMbt8AAAAADc4p37m6pJhw4bpq6++0oYNG1S7du1S7VupUiXdfvvtOnDgQLF9rFarrFbrtZYJAAAA4Cbm1HeuDMPQsGHD9MUXX2jNmjUKCQkp9Rj5+fnauXOnatasWQ4VAgAAAMBFTn3naujQofrkk0+0ZMkSeXl5KS0tTZLk4+MjT09PSVK/fv1Uq1YtxcXFSZJee+01/eUvf1H9+vWVkZGhCRMm6PDhwxo0aJDDzgMAAADAjc+pw9WMGTMkSR07drRrj4+PV//+/SVJKSkpcnH53w2406dPa/DgwUpLS1O1atXUunVrbdq0SU2bNr1eZQMAAAC4CTl1uDIM44p91q1bZ7c+efJkTZ48uZwqAgAAAICiOfUzVwAAAABQURCuAAAAAMAEhCsAAAAAMIFTP3OFi4JHLXV0CTbJb3V3dAkAAACAU+LOFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAAAAAJiBcAQAAAIAJCFcAAAAAYALCFQAAAACYgHAFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAncHF0AKpbgUUsdXYKd5Le6O7oEAABwFZztPYSzcLb3Ms70c3K21+ZqcOcKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMAHhCgAAAABMQLgCAAAAABMQrgAAAADABIQrAAAAADAB4QoAAAAATEC4AgAAAAATEK4AAAAAwASEKwAAAAAwAeEKAAAAAExAuAIAAAAAExCuAAAAAMAEhCsAAAAAMEGFCFfTp09XcHCwPDw81KZNG33//fcl9l+4cKEaN24sDw8PtWjRQsuWLbtOlQIAAAC4WTl9uFqwYIFiYmI0ZswYbd++XaGhoYqIiNDx48eL7L9p0yZFRUVp4MCB2rFjh3r16qVevXpp165d17lyAAAAADcTpw9XkyZN0uDBgzVgwAA1bdpUM2fOVOXKlTVnzpwi+0+dOlVdu3bVyJEj1aRJE73++uu64447NG3atOtcOQAAAICbiZujCyhJbm6utm3bptjYWFubi4uLwsPDtXnz5iL32bx5s2JiYuzaIiIitHjx4mKPk5OTo5ycHNv6mTNnJEmZmZnXUL15CnLOOboEp+UsPyNnwzUDXH/5ued16TdSfs45FRgFDq0HQMXgbO9lnOk9hLO8NpfqMAzjin2dOlydPHlS+fn5qlGjhl17jRo1tHfv3iL3SUtLK7J/WlpasceJi4vTuHHjCrUHBQWVoWpcTz5THF0BAPyPz6V/vN/PkWUAqEB4L1M8Z3ttzp49Kx8fnxL7OHW4ul5iY2Pt7nYVFBTo999/16233iqLxeLAym5emZmZCgoK0pEjR+Tt7e3ocnAD4dpCeeL6Qnnh2kJ54dq6MsMwdPbsWQUGBl6xr1OHKz8/P7m6uio9Pd2uPT09XQEBAUXuExAQUKr+kmS1WmW1Wu3afH19y1Y0TOXt7c1/6CgXXFsoT1xfKC9cWygvXFslu9Idq0ucekILd3d3tW7dWqtXr7a1FRQUaPXq1QoLCytyn7CwMLv+krRy5cpi+wMAAACAGZz6zpUkxcTEKDo6WnfeeafuvvtuTZkyRdnZ2RowYIAkqV+/fqpVq5bi4uIkSc8995w6dOigiRMnqnv37po/f762bt2qf/3rX448DQAAAAA3OKcPV48++qhOnDih0aNHKy0tTa1atVJiYqJt0oqUlBS5uPzvBlzbtm31ySef6JVXXtHLL7+sBg0aaPHixWrevLmjTgFlYLVaNWbMmEIf1wSuFdcWyhPXF8oL1xbKC9eWuSzG1cwpCAAAAAAokVM/cwUAAAAAFQXhCgAAAABMQLgCAAAAABMQrgAAAADABIQrOFRycrIGDhyokJAQeXp6ql69ehozZoxyc3Pt+v30009q3769PDw8FBQUpPHjxxcaa+HChWrcuLE8PDzUokULLVu27HqdBpzUG2+8obZt26py5crFfjF4SkqKunfvrsqVK6t69eoaOXKkLly4YNdn3bp1uuOOO2S1WlW/fn0lJCSUf/GocKZPn67g4GB5eHioTZs2+v777x1dEpzchg0b1KNHDwUGBspisWjx4sV22w3D0OjRo1WzZk15enoqPDxc+/fvt+vz+++/q2/fvvL29pavr68GDhyorKys63gWcEZxcXG666675OXlperVq6tXr17at2+fXZ/z589r6NChuvXWW1W1alX17t1b6enpdn2u5m8k7BGu4FB79+5VQUGBZs2apd27d2vy5MmaOXOmXn75ZVufzMxMdenSRXXq1NG2bds0YcIEjR071u67yzZt2qSoqCgNHDhQO3bsUK9evdSrVy/t2rXLEacFJ5Gbm6s+ffpoyJAhRW7Pz89X9+7dlZubq02bNmnu3LlKSEjQ6NGjbX0OHTqk7t27q1OnTkpKStLw4cM1aNAgLV++/HqdBiqABQsWKCYmRmPGjNH27dsVGhqqiIgIHT9+3NGlwYllZ2crNDRU06dPL3L7+PHj9e6772rmzJn67rvvVKVKFUVEROj8+fO2Pn379tXu3bu1cuVKffXVV9qwYYOeeuqp63UKcFLr16/X0KFDtWXLFq1cuVJ5eXnq0qWLsrOzbX2ef/55/fe//9XChQu1fv16HTt2TA899JBt+9X8jUQRDMDJjB8/3ggJCbGtv//++0a1atWMnJwcW9tLL71kNGrUyLb+yCOPGN27d7cbp02bNsbTTz9d/gXD6cXHxxs+Pj6F2pctW2a4uLgYaWlptrYZM2YY3t7etuvtxRdfNJo1a2a336OPPmpERESUa82oWO6++25j6NChtvX8/HwjMDDQiIuLc2BVqEgkGV988YVtvaCgwAgICDAmTJhga8vIyDCsVqvx6aefGoZhGD///LMhyfjhhx9sfb7++mvDYrEYv/3223WrHc7v+PHjhiRj/fr1hmFcvJYqVapkLFy40NZnz549hiRj8+bNhmFc3d9IFMadKzidM2fO6JZbbrGtb968WX/961/l7u5ua4uIiNC+fft0+vRpW5/w8HC7cSIiIrR58+brUzQqpM2bN6tFixa2LyWXLl43mZmZ2r17t60P1xZKkpubq23bttldJy4uLgoPD+c6QZkdOnRIaWlpdteVj4+P2rRpY7uuNm/eLF9fX9155522PuHh4XJxcdF333133WuG8zpz5owk2d5fbdu2TXl5eXbXV+PGjXXbbbfZXV9X+huJwghXcCoHDhzQe++9p6efftrWlpaWZvcftiTbelpaWol9Lm0HinIt11ZmZqb++OOP61MonNrJkyeVn5/P7yCY6tK1U9J1lZaWpurVq9ttd3Nz0y233MK1B5uCggINHz5c99xzj5o3by7p4rXj7u5e6Hnky6+vK/2NRGGEK5SLUaNGyWKxlLjs3bvXbp/ffvtNXbt2VZ8+fTR48GAHVQ5nV5ZrCwCAm9XQoUO1a9cuzZ8/39Gl3BTcHF0AbkwvvPCC+vfvX2KfunXr2v597NgxderUSW3btrWbqEKSAgICCs1ec2k9ICCgxD6XtuPGUdprqyQBAQGFZnS72mvL29tbnp6eV1k1bmR+fn5ydXXldxBMdenaSU9PV82aNW3t6enpatWqla3P5ZOmXLhwQb///jvXHiRJw4YNs010Urt2bVt7QECAcnNzlZGRYXf36s+/t67mbyQK484VyoW/v78aN25c4nLpGarffvtNHTt2VOvWrRUfHy8XF/vLMiwsTBs2bFBeXp6tbeXKlWrUqJGqVatm67N69Wq7/VauXKmwsLByPlNcb6W5tq4kLCxMO3futHtzsnLlSnl7e6tp06a2PlxbKIm7u7tat25td50UFBRo9erVXCcos5CQEAUEBNhdV5mZmfruu+9s11VYWJgyMjK0bds2W581a9aooKBAbdq0ue41w3kYhqFhw4bpiy++0Jo1axQSEmK3vXXr1qpUqZLd9bVv3z6lpKTYXV9X+huJIjh6Rg3c3I4ePWrUr1/f6Ny5s3H06FEjNTXVtlySkZFh1KhRw3jiiSeMXbt2GfPnzzcqV65szJo1y9bn22+/Ndzc3Ix33nnH2LNnjzFmzBijUqVKxs6dOx1xWnAShw8fNnbs2GGMGzfOqFq1qrFjxw5jx44dxtmzZw3DMIwLFy4YzZs3N7p06WIkJSUZiYmJhr+/vxEbG2sb49dffzUqV65sjBw50tizZ48xffp0w9XV1UhMTHTUacEJzZ8/37BarUZCQoLx888/G0899ZTh6+trN8sWcLmzZ8/afi9JMiZNmmTs2LHDOHz4sGEYhvHWW28Zvr6+xpIlS4yffvrJ6NmzpxESEmL88ccftjG6du1q3H777cZ3331nbNy40WjQoIERFRXlqFOCkxgyZIjh4+NjrFu3zu691blz52x9/v73vxu33XabsWbNGmPr1q1GWFiYERYWZtt+NX8jURjhCg4VHx9vSCpy+bMff/zRaNeunWG1Wo1atWoZb731VqGxPvvsM6Nhw4aGu7u70axZM2Pp0qXX6zTgpKKjo4u8ttauXWvrk5ycbERGRhqenp6Gn5+f8cILLxh5eXl246xdu9Zo1aqV4e7ubtStW9eIj4+/vieCCuG9994zbrvtNsPd3d24++67jS1btji6JDi5tWvXFvk7Kjo62jCMi9Oxv/rqq0aNGjUMq9VqdO7c2di3b5/dGKdOnTKioqKMqlWrGt7e3saAAQNs/wMJN6/i3lv9+e/XH3/8YTzzzDNGtWrVjMqVKxsPPvig3f/cNoyr+xsJexbDMIzreKMMAAAAAG5IPHMFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAwP+XnJwsi8WipKSkYvusW7dOFotFGRkZph7bYrFo8eLFpo4JALi+CFcAgAqlf//+slgsslgsqlSpkkJCQvTiiy/q/Pnz1zx2UFCQUlNT1bx5cxMqBQDcbNwcXQAAAKXVtWtXxcfHKy8vT9u2bVN0dLQsFovefvvtaxrX1dVVAQEBJlUJALjZcOcKAFDhWK1WBQQEKCgoSL169VJ4eLhWrlwpSSooKFBcXJxCQkLk6emp0NBQff7557Z9T58+rb59+8rf31+enp5q0KCB4uPjJRX9scBly5apYcOG8vT0VKdOnZScnGxXy9ixY9WqVSu7tilTpig4ONi2/sMPP+i+++6Tn5+ffHx81KFDB23fvt3U1wQA4HiEKwBAhbZr1y5t2rRJ7u7ukqS4uDjNmzdPM2fO1O7du/X888/r8ccf1/r16yVJr776qn7++Wd9/fXX2rNnj2bMmCE/P78ixz5y5Igeeugh9ejRQ0lJSRo0aJBGjRpV6hrPnj2r6Ohobdy4UVu2bFGDBg3UrVs3nT17tuwnDgBwOnwsEABQ4Xz11VeqWrWqLly4oJycHLm4uGjatGnKycnRm2++qVWrViksLEySVLduXW3cuFGzZs1Shw4dlJKSottvv1133nmnJNndYbrcjBkzVK9ePU2cOFGS1KhRI+3cubPUHz+899577db/9a9/ydfXV+vXr9f9999fqrEAAM6LcAUAqHA6deqkGTNmKDs7W5MnT5abm5t69+6t3bt369y5c7rvvvvs+ufm5ur222+XJA0ZMkS9e/fW9u3b1aVLF/Xq1Utt27Yt8jh79uxRmzZt7NouhbbSSE9P1yuvvKJ169bp+PHjys/P17lz55SSklLqsQAAzotwBQCocKpUqaL69etLkubMmaPQ0FDNnj3bNsvf0qVLVatWLbt9rFarJCkyMlKHDx/WsmXLtHLlSnXu3FlDhw7VO++8U6ZaXFxcZBiGXVteXp7denR0tE6dOqWpU6eqTp06slqtCgsLU25ubpmOCQBwToQrAECF5uLiopdfflkxMTH65ZdfZLValZKSog4dOhS7j7+/v6KjoxUdHa327dtr5MiRRYarJk2a6Msvv7Rr27JlS6Gx0tLSZBiGLBaLJBX6nqxvv/1W77//vrp16ybp4rNcJ0+eLMvpAgCcGBNaAAAqvD59+sjV1VWzZs3SiBEj9Pzzz2vu3Lk6ePCgtm/frvfee09z586VJI0ePVpLlizRgQMHtHv3bn311Vdq0qRJkeP+/e9/1/79+zVy5Ejt27dPn3zyiRISEuz6dOzYUSdOnND48eN18OBBTZ8+XV9//bVdnwYNGujDDz/Unj179N1336lv377y9PQsl9cCAOA4hCsAQIXn5uamYcOGafz48YqNjdWrr76quLg4NWnSRF27dtXSpUsVEhIiSXJ3d1dsbKxatmypv/71r3J1ddX8+fOLHPe2227Tf/7zHy1evFihoaGaOXOm3nzzTbs+TZo00fvvv6/p06crNDRU33//vUaMGGHXZ/bs2Tp9+rTuuOMOPfHEE3r22WdVvXr18nkxAAAOYzEu/6A4AAAAAKDUuHMFAAAAACYgXAEAAACACQhXAAAAAGACwhUAAAAAmIBwBQAAAAAmIFwBAAAAgAkIVwAAAABgAsIVAAAAAJiAcAUAAAAAJiBcAQAAAIAJCFcAAAAAYIL/BxaqpC3EKwqQAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 43.86\n",
      "Mean absolute percentage error: 40.34%\n",
      "Root Mean squared error: 54.96\n",
      "Coefficient of determination: 0.45\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regr_rf = RandomForestRegressor(random_state=42)\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr_rf.fit(diabetes_X_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr_rf.predict(diabetes_X_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'Random Forest',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 46.59\n",
      "Mean absolute percentage error: 39.84%\n",
      "Root Mean squared error: 58.01\n",
      "Coefficient of determination: 0.39\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regr_xg = XGBRegressor(random_state=42)\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr_xg.fit(diabetes_X_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr_xg.predict(diabetes_X_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'XGBoost',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Light GBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 45.91\n",
      "Mean absolute percentage error: 40.71%\n",
      "Root Mean squared error: 57.97\n",
      "Coefficient of determination: 0.39\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regr_lgbm = LGBMRegressor(random_state=42)\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr_lgbm.fit(diabetes_X_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr_lgbm.predict(diabetes_X_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'Light GBM',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Catboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning rate set to 0.034381\n",
      "0:\tlearn: 76.7698931\ttotal: 1.5ms\tremaining: 1.5s\n",
      "1:\tlearn: 75.7552747\ttotal: 2.5ms\tremaining: 1.25s\n",
      "2:\tlearn: 74.8136416\ttotal: 3.54ms\tremaining: 1.18s\n",
      "3:\tlearn: 73.9549588\ttotal: 4.85ms\tremaining: 1.21s\n",
      "4:\tlearn: 73.0782030\ttotal: 6.35ms\tremaining: 1.26s\n",
      "5:\tlearn: 72.1682178\ttotal: 7.69ms\tremaining: 1.27s\n",
      "6:\tlearn: 71.3682200\ttotal: 8.78ms\tremaining: 1.25s\n",
      "7:\tlearn: 70.5608211\ttotal: 9.75ms\tremaining: 1.21s\n",
      "8:\tlearn: 69.7224552\ttotal: 10.9ms\tremaining: 1.2s\n",
      "9:\tlearn: 68.9543606\ttotal: 12.2ms\tremaining: 1.21s\n",
      "10:\tlearn: 68.2429798\ttotal: 13.5ms\tremaining: 1.21s\n",
      "11:\tlearn: 67.5175992\ttotal: 14.8ms\tremaining: 1.22s\n",
      "12:\tlearn: 66.8114535\ttotal: 16.1ms\tremaining: 1.22s\n",
      "13:\tlearn: 66.1442549\ttotal: 17.3ms\tremaining: 1.22s\n",
      "14:\tlearn: 65.5281514\ttotal: 18.5ms\tremaining: 1.22s\n",
      "15:\tlearn: 64.8807317\ttotal: 19.7ms\tremaining: 1.21s\n",
      "16:\tlearn: 64.3379117\ttotal: 21.1ms\tremaining: 1.22s\n",
      "17:\tlearn: 63.7508639\ttotal: 22.4ms\tremaining: 1.22s\n",
      "18:\tlearn: 63.2097439\ttotal: 23.6ms\tremaining: 1.22s\n",
      "19:\tlearn: 62.7139608\ttotal: 24.7ms\tremaining: 1.21s\n",
      "20:\tlearn: 62.2236234\ttotal: 26ms\tremaining: 1.21s\n",
      "21:\tlearn: 61.7820840\ttotal: 27.1ms\tremaining: 1.2s\n",
      "22:\tlearn: 61.2214417\ttotal: 28ms\tremaining: 1.19s\n",
      "23:\tlearn: 60.7727226\ttotal: 28.9ms\tremaining: 1.18s\n",
      "24:\tlearn: 60.3932562\ttotal: 29.9ms\tremaining: 1.16s\n",
      "25:\tlearn: 59.9563855\ttotal: 30.8ms\tremaining: 1.15s\n",
      "26:\tlearn: 59.5363233\ttotal: 31.9ms\tremaining: 1.15s\n",
      "27:\tlearn: 59.1791749\ttotal: 33.1ms\tremaining: 1.15s\n",
      "28:\tlearn: 58.6907023\ttotal: 34.1ms\tremaining: 1.14s\n",
      "29:\tlearn: 58.2293662\ttotal: 35ms\tremaining: 1.13s\n",
      "30:\tlearn: 57.8130967\ttotal: 36ms\tremaining: 1.13s\n",
      "31:\tlearn: 57.4281744\ttotal: 37.3ms\tremaining: 1.13s\n",
      "32:\tlearn: 57.1725832\ttotal: 38.5ms\tremaining: 1.13s\n",
      "33:\tlearn: 56.7632606\ttotal: 39.6ms\tremaining: 1.13s\n",
      "34:\tlearn: 56.4138620\ttotal: 40.7ms\tremaining: 1.12s\n",
      "35:\tlearn: 56.0466281\ttotal: 42ms\tremaining: 1.12s\n",
      "36:\tlearn: 55.7211225\ttotal: 43.1ms\tremaining: 1.12s\n",
      "37:\tlearn: 55.4148873\ttotal: 44.2ms\tremaining: 1.12s\n",
      "38:\tlearn: 55.1553516\ttotal: 45.2ms\tremaining: 1.11s\n",
      "39:\tlearn: 54.8464593\ttotal: 46.2ms\tremaining: 1.11s\n",
      "40:\tlearn: 54.5206398\ttotal: 47.3ms\tremaining: 1.11s\n",
      "41:\tlearn: 54.2016506\ttotal: 48.3ms\tremaining: 1.1s\n",
      "42:\tlearn: 53.9583345\ttotal: 49.2ms\tremaining: 1.09s\n",
      "43:\tlearn: 53.6716694\ttotal: 50.7ms\tremaining: 1.1s\n",
      "44:\tlearn: 53.4208321\ttotal: 51.9ms\tremaining: 1.1s\n",
      "45:\tlearn: 53.1852474\ttotal: 53ms\tremaining: 1.1s\n",
      "46:\tlearn: 52.8905478\ttotal: 54ms\tremaining: 1.09s\n",
      "47:\tlearn: 52.7149197\ttotal: 55ms\tremaining: 1.09s\n",
      "48:\tlearn: 52.5124637\ttotal: 56.1ms\tremaining: 1.09s\n",
      "49:\tlearn: 52.3121632\ttotal: 57.1ms\tremaining: 1.08s\n",
      "50:\tlearn: 52.1629426\ttotal: 58ms\tremaining: 1.08s\n",
      "51:\tlearn: 51.9156136\ttotal: 59ms\tremaining: 1.07s\n",
      "52:\tlearn: 51.6313901\ttotal: 59.8ms\tremaining: 1.07s\n",
      "53:\tlearn: 51.3785559\ttotal: 60.8ms\tremaining: 1.06s\n",
      "54:\tlearn: 51.2121959\ttotal: 61.6ms\tremaining: 1.06s\n",
      "55:\tlearn: 51.0247514\ttotal: 62.4ms\tremaining: 1.05s\n",
      "56:\tlearn: 50.7958459\ttotal: 63.4ms\tremaining: 1.05s\n",
      "57:\tlearn: 50.6322565\ttotal: 64.2ms\tremaining: 1.04s\n",
      "58:\tlearn: 50.4693849\ttotal: 65.1ms\tremaining: 1.04s\n",
      "59:\tlearn: 50.2729021\ttotal: 66.1ms\tremaining: 1.03s\n",
      "60:\tlearn: 50.0652141\ttotal: 67ms\tremaining: 1.03s\n",
      "61:\tlearn: 49.9072643\ttotal: 67.9ms\tremaining: 1.03s\n",
      "62:\tlearn: 49.7088928\ttotal: 68.8ms\tremaining: 1.02s\n",
      "63:\tlearn: 49.4991804\ttotal: 69.7ms\tremaining: 1.02s\n",
      "64:\tlearn: 49.3220877\ttotal: 70.5ms\tremaining: 1.01s\n",
      "65:\tlearn: 49.1778938\ttotal: 71.3ms\tremaining: 1.01s\n",
      "66:\tlearn: 49.0257537\ttotal: 72.2ms\tremaining: 1s\n",
      "67:\tlearn: 48.8766285\ttotal: 73ms\tremaining: 1s\n",
      "68:\tlearn: 48.7340804\ttotal: 73.8ms\tremaining: 996ms\n",
      "69:\tlearn: 48.6003848\ttotal: 74.6ms\tremaining: 991ms\n",
      "70:\tlearn: 48.4638499\ttotal: 75.4ms\tremaining: 986ms\n",
      "71:\tlearn: 48.3048779\ttotal: 76.2ms\tremaining: 982ms\n",
      "72:\tlearn: 48.1774846\ttotal: 77ms\tremaining: 978ms\n",
      "73:\tlearn: 48.0137795\ttotal: 77.8ms\tremaining: 973ms\n",
      "74:\tlearn: 47.8542608\ttotal: 78.7ms\tremaining: 971ms\n",
      "75:\tlearn: 47.6948567\ttotal: 79.7ms\tremaining: 969ms\n",
      "76:\tlearn: 47.5868712\ttotal: 80.7ms\tremaining: 967ms\n",
      "77:\tlearn: 47.4238916\ttotal: 81.8ms\tremaining: 966ms\n",
      "78:\tlearn: 47.2951391\ttotal: 82.9ms\tremaining: 966ms\n",
      "79:\tlearn: 47.1670664\ttotal: 83.8ms\tremaining: 963ms\n",
      "80:\tlearn: 46.9672861\ttotal: 84.7ms\tremaining: 961ms\n",
      "81:\tlearn: 46.8544297\ttotal: 85.6ms\tremaining: 958ms\n",
      "82:\tlearn: 46.7604885\ttotal: 86.6ms\tremaining: 956ms\n",
      "83:\tlearn: 46.5980069\ttotal: 87.4ms\tremaining: 953ms\n",
      "84:\tlearn: 46.4970350\ttotal: 88.2ms\tremaining: 950ms\n",
      "85:\tlearn: 46.3157959\ttotal: 89.1ms\tremaining: 947ms\n",
      "86:\tlearn: 46.1764093\ttotal: 89.9ms\tremaining: 943ms\n",
      "87:\tlearn: 46.0119289\ttotal: 90.8ms\tremaining: 941ms\n",
      "88:\tlearn: 45.9075921\ttotal: 91.7ms\tremaining: 938ms\n",
      "89:\tlearn: 45.7041591\ttotal: 92.7ms\tremaining: 937ms\n",
      "90:\tlearn: 45.6230779\ttotal: 93.6ms\tremaining: 935ms\n",
      "91:\tlearn: 45.4441067\ttotal: 94.5ms\tremaining: 933ms\n",
      "92:\tlearn: 45.3302847\ttotal: 95.5ms\tremaining: 931ms\n",
      "93:\tlearn: 45.2266465\ttotal: 96.5ms\tremaining: 930ms\n",
      "94:\tlearn: 45.1188503\ttotal: 97.5ms\tremaining: 929ms\n",
      "95:\tlearn: 45.0070179\ttotal: 98.5ms\tremaining: 928ms\n",
      "96:\tlearn: 44.9073280\ttotal: 99.5ms\tremaining: 926ms\n",
      "97:\tlearn: 44.8299073\ttotal: 100ms\tremaining: 925ms\n",
      "98:\tlearn: 44.7113215\ttotal: 101ms\tremaining: 923ms\n",
      "99:\tlearn: 44.6274691\ttotal: 102ms\tremaining: 921ms\n",
      "100:\tlearn: 44.5478216\ttotal: 103ms\tremaining: 920ms\n",
      "101:\tlearn: 44.4544657\ttotal: 104ms\tremaining: 918ms\n",
      "102:\tlearn: 44.2905178\ttotal: 105ms\tremaining: 915ms\n",
      "103:\tlearn: 44.2331215\ttotal: 106ms\tremaining: 914ms\n",
      "104:\tlearn: 44.1630763\ttotal: 107ms\tremaining: 913ms\n",
      "105:\tlearn: 44.0799029\ttotal: 108ms\tremaining: 910ms\n",
      "106:\tlearn: 43.9771810\ttotal: 109ms\tremaining: 910ms\n",
      "107:\tlearn: 43.8666901\ttotal: 110ms\tremaining: 909ms\n",
      "108:\tlearn: 43.7729359\ttotal: 111ms\tremaining: 907ms\n",
      "109:\tlearn: 43.6575975\ttotal: 112ms\tremaining: 905ms\n",
      "110:\tlearn: 43.5374373\ttotal: 113ms\tremaining: 903ms\n",
      "111:\tlearn: 43.4217869\ttotal: 114ms\tremaining: 901ms\n",
      "112:\tlearn: 43.2726354\ttotal: 115ms\tremaining: 900ms\n",
      "113:\tlearn: 43.1591655\ttotal: 116ms\tremaining: 898ms\n",
      "114:\tlearn: 43.1050945\ttotal: 117ms\tremaining: 897ms\n",
      "115:\tlearn: 42.9959354\ttotal: 117ms\tremaining: 895ms\n",
      "116:\tlearn: 42.9189572\ttotal: 118ms\tremaining: 893ms\n",
      "117:\tlearn: 42.7879437\ttotal: 119ms\tremaining: 891ms\n",
      "118:\tlearn: 42.7335824\ttotal: 120ms\tremaining: 889ms\n",
      "119:\tlearn: 42.6647507\ttotal: 121ms\tremaining: 886ms\n",
      "120:\tlearn: 42.5460610\ttotal: 122ms\tremaining: 884ms\n",
      "121:\tlearn: 42.4575359\ttotal: 123ms\tremaining: 883ms\n",
      "122:\tlearn: 42.3488106\ttotal: 124ms\tremaining: 882ms\n",
      "123:\tlearn: 42.2746008\ttotal: 125ms\tremaining: 881ms\n",
      "124:\tlearn: 42.1997445\ttotal: 126ms\tremaining: 879ms\n",
      "125:\tlearn: 42.0779892\ttotal: 126ms\tremaining: 877ms\n",
      "126:\tlearn: 41.9835915\ttotal: 128ms\tremaining: 876ms\n",
      "127:\tlearn: 41.8873330\ttotal: 129ms\tremaining: 876ms\n",
      "128:\tlearn: 41.8151306\ttotal: 130ms\tremaining: 875ms\n",
      "129:\tlearn: 41.7505622\ttotal: 131ms\tremaining: 874ms\n",
      "130:\tlearn: 41.6331956\ttotal: 132ms\tremaining: 874ms\n",
      "131:\tlearn: 41.5887267\ttotal: 133ms\tremaining: 872ms\n",
      "132:\tlearn: 41.4726460\ttotal: 134ms\tremaining: 871ms\n",
      "133:\tlearn: 41.4357049\ttotal: 135ms\tremaining: 869ms\n",
      "134:\tlearn: 41.3683909\ttotal: 135ms\tremaining: 867ms\n",
      "135:\tlearn: 41.3529582\ttotal: 136ms\tremaining: 862ms\n",
      "136:\tlearn: 41.2664667\ttotal: 137ms\tremaining: 860ms\n",
      "137:\tlearn: 41.1269202\ttotal: 137ms\tremaining: 858ms\n",
      "138:\tlearn: 41.0353019\ttotal: 138ms\tremaining: 857ms\n",
      "139:\tlearn: 40.9380830\ttotal: 139ms\tremaining: 856ms\n",
      "140:\tlearn: 40.8300088\ttotal: 140ms\tremaining: 855ms\n",
      "141:\tlearn: 40.7284181\ttotal: 141ms\tremaining: 853ms\n",
      "142:\tlearn: 40.6691379\ttotal: 142ms\tremaining: 853ms\n",
      "143:\tlearn: 40.5847037\ttotal: 144ms\tremaining: 853ms\n",
      "144:\tlearn: 40.5070803\ttotal: 145ms\tremaining: 853ms\n",
      "145:\tlearn: 40.4541409\ttotal: 146ms\tremaining: 851ms\n",
      "146:\tlearn: 40.3628197\ttotal: 146ms\tremaining: 849ms\n",
      "147:\tlearn: 40.2947908\ttotal: 147ms\tremaining: 847ms\n",
      "148:\tlearn: 40.2383656\ttotal: 148ms\tremaining: 845ms\n",
      "149:\tlearn: 40.1418319\ttotal: 149ms\tremaining: 844ms\n",
      "150:\tlearn: 40.0635154\ttotal: 150ms\tremaining: 843ms\n",
      "151:\tlearn: 40.0195394\ttotal: 151ms\tremaining: 841ms\n",
      "152:\tlearn: 39.9984043\ttotal: 152ms\tremaining: 841ms\n",
      "153:\tlearn: 39.9361859\ttotal: 153ms\tremaining: 840ms\n",
      "154:\tlearn: 39.8484358\ttotal: 154ms\tremaining: 838ms\n",
      "155:\tlearn: 39.7381342\ttotal: 155ms\tremaining: 836ms\n",
      "156:\tlearn: 39.6576067\ttotal: 155ms\tremaining: 835ms\n",
      "157:\tlearn: 39.6234486\ttotal: 156ms\tremaining: 833ms\n",
      "158:\tlearn: 39.5537998\ttotal: 157ms\tremaining: 832ms\n",
      "159:\tlearn: 39.4925269\ttotal: 158ms\tremaining: 831ms\n",
      "160:\tlearn: 39.4231701\ttotal: 159ms\tremaining: 830ms\n",
      "161:\tlearn: 39.3822348\ttotal: 160ms\tremaining: 829ms\n",
      "162:\tlearn: 39.3088103\ttotal: 161ms\tremaining: 829ms\n",
      "163:\tlearn: 39.2205778\ttotal: 162ms\tremaining: 828ms\n",
      "164:\tlearn: 39.1371691\ttotal: 163ms\tremaining: 826ms\n",
      "165:\tlearn: 39.1192417\ttotal: 164ms\tremaining: 825ms\n",
      "166:\tlearn: 39.0395533\ttotal: 165ms\tremaining: 823ms\n",
      "167:\tlearn: 38.9513121\ttotal: 166ms\tremaining: 822ms\n",
      "168:\tlearn: 38.9051022\ttotal: 167ms\tremaining: 821ms\n",
      "169:\tlearn: 38.8698619\ttotal: 168ms\tremaining: 821ms\n",
      "170:\tlearn: 38.7690720\ttotal: 169ms\tremaining: 820ms\n",
      "171:\tlearn: 38.7107433\ttotal: 170ms\tremaining: 818ms\n",
      "172:\tlearn: 38.5816804\ttotal: 171ms\tremaining: 817ms\n",
      "173:\tlearn: 38.4737286\ttotal: 172ms\tremaining: 816ms\n",
      "174:\tlearn: 38.4367775\ttotal: 173ms\tremaining: 815ms\n",
      "175:\tlearn: 38.3782871\ttotal: 174ms\tremaining: 814ms\n",
      "176:\tlearn: 38.3075423\ttotal: 175ms\tremaining: 813ms\n",
      "177:\tlearn: 38.2641161\ttotal: 176ms\tremaining: 812ms\n",
      "178:\tlearn: 38.1671217\ttotal: 177ms\tremaining: 811ms\n",
      "179:\tlearn: 38.0724997\ttotal: 178ms\tremaining: 809ms\n",
      "180:\tlearn: 38.0517931\ttotal: 179ms\tremaining: 808ms\n",
      "181:\tlearn: 37.9349784\ttotal: 179ms\tremaining: 807ms\n",
      "182:\tlearn: 37.8350153\ttotal: 180ms\tremaining: 805ms\n",
      "183:\tlearn: 37.7654821\ttotal: 181ms\tremaining: 804ms\n",
      "184:\tlearn: 37.6858220\ttotal: 182ms\tremaining: 803ms\n",
      "185:\tlearn: 37.5583218\ttotal: 183ms\tremaining: 801ms\n",
      "186:\tlearn: 37.4254625\ttotal: 184ms\tremaining: 800ms\n",
      "187:\tlearn: 37.4015333\ttotal: 185ms\tremaining: 798ms\n",
      "188:\tlearn: 37.3565400\ttotal: 186ms\tremaining: 797ms\n",
      "189:\tlearn: 37.2798855\ttotal: 187ms\tremaining: 796ms\n",
      "190:\tlearn: 37.1737303\ttotal: 188ms\tremaining: 794ms\n",
      "191:\tlearn: 37.0917386\ttotal: 189ms\tremaining: 794ms\n",
      "192:\tlearn: 36.9876853\ttotal: 190ms\tremaining: 793ms\n",
      "193:\tlearn: 36.9347396\ttotal: 191ms\tremaining: 792ms\n",
      "194:\tlearn: 36.8560969\ttotal: 192ms\tremaining: 791ms\n",
      "195:\tlearn: 36.7820441\ttotal: 193ms\tremaining: 790ms\n",
      "196:\tlearn: 36.7382448\ttotal: 193ms\tremaining: 788ms\n",
      "197:\tlearn: 36.6888071\ttotal: 194ms\tremaining: 787ms\n",
      "198:\tlearn: 36.6077915\ttotal: 196ms\tremaining: 788ms\n",
      "199:\tlearn: 36.5194556\ttotal: 197ms\tremaining: 788ms\n",
      "200:\tlearn: 36.4391372\ttotal: 198ms\tremaining: 787ms\n",
      "201:\tlearn: 36.3613057\ttotal: 199ms\tremaining: 786ms\n",
      "202:\tlearn: 36.3233642\ttotal: 200ms\tremaining: 785ms\n",
      "203:\tlearn: 36.2923866\ttotal: 201ms\tremaining: 783ms\n",
      "204:\tlearn: 36.1969353\ttotal: 202ms\tremaining: 782ms\n",
      "205:\tlearn: 36.1045298\ttotal: 202ms\tremaining: 780ms\n",
      "206:\tlearn: 36.0651219\ttotal: 203ms\tremaining: 779ms\n",
      "207:\tlearn: 36.0293077\ttotal: 204ms\tremaining: 778ms\n",
      "208:\tlearn: 35.9293003\ttotal: 205ms\tremaining: 777ms\n",
      "209:\tlearn: 35.8540193\ttotal: 206ms\tremaining: 776ms\n",
      "210:\tlearn: 35.7344981\ttotal: 207ms\tremaining: 775ms\n",
      "211:\tlearn: 35.6745242\ttotal: 208ms\tremaining: 774ms\n",
      "212:\tlearn: 35.5702305\ttotal: 209ms\tremaining: 772ms\n",
      "213:\tlearn: 35.4581545\ttotal: 210ms\tremaining: 772ms\n",
      "214:\tlearn: 35.3961630\ttotal: 211ms\tremaining: 771ms\n",
      "215:\tlearn: 35.3470567\ttotal: 212ms\tremaining: 770ms\n",
      "216:\tlearn: 35.3164110\ttotal: 213ms\tremaining: 769ms\n",
      "217:\tlearn: 35.3022169\ttotal: 214ms\tremaining: 768ms\n",
      "218:\tlearn: 35.2201926\ttotal: 215ms\tremaining: 768ms\n",
      "219:\tlearn: 35.1870608\ttotal: 216ms\tremaining: 767ms\n",
      "220:\tlearn: 35.1522341\ttotal: 217ms\tremaining: 766ms\n",
      "221:\tlearn: 35.1144408\ttotal: 218ms\tremaining: 765ms\n",
      "222:\tlearn: 35.0458040\ttotal: 219ms\tremaining: 764ms\n",
      "223:\tlearn: 35.0073791\ttotal: 220ms\tremaining: 763ms\n",
      "224:\tlearn: 34.9968960\ttotal: 221ms\tremaining: 762ms\n",
      "225:\tlearn: 34.9134931\ttotal: 222ms\tremaining: 761ms\n",
      "226:\tlearn: 34.8522448\ttotal: 223ms\tremaining: 761ms\n",
      "227:\tlearn: 34.7182768\ttotal: 224ms\tremaining: 759ms\n",
      "228:\tlearn: 34.7080690\ttotal: 225ms\tremaining: 758ms\n",
      "229:\tlearn: 34.6158693\ttotal: 226ms\tremaining: 757ms\n",
      "230:\tlearn: 34.5994554\ttotal: 227ms\tremaining: 756ms\n",
      "231:\tlearn: 34.5231509\ttotal: 228ms\tremaining: 755ms\n",
      "232:\tlearn: 34.4521360\ttotal: 229ms\tremaining: 753ms\n",
      "233:\tlearn: 34.4377034\ttotal: 230ms\tremaining: 752ms\n",
      "234:\tlearn: 34.4019649\ttotal: 231ms\tremaining: 751ms\n",
      "235:\tlearn: 34.3384892\ttotal: 231ms\tremaining: 749ms\n",
      "236:\tlearn: 34.2648958\ttotal: 232ms\tremaining: 748ms\n",
      "237:\tlearn: 34.1924062\ttotal: 233ms\tremaining: 746ms\n",
      "238:\tlearn: 34.0597921\ttotal: 234ms\tremaining: 745ms\n",
      "239:\tlearn: 33.9936793\ttotal: 235ms\tremaining: 744ms\n",
      "240:\tlearn: 33.9363340\ttotal: 236ms\tremaining: 743ms\n",
      "241:\tlearn: 33.8403938\ttotal: 237ms\tremaining: 743ms\n",
      "242:\tlearn: 33.8253175\ttotal: 238ms\tremaining: 742ms\n",
      "243:\tlearn: 33.7073406\ttotal: 239ms\tremaining: 741ms\n",
      "244:\tlearn: 33.6338510\ttotal: 240ms\tremaining: 739ms\n",
      "245:\tlearn: 33.6226077\ttotal: 241ms\tremaining: 738ms\n",
      "246:\tlearn: 33.5647316\ttotal: 242ms\tremaining: 736ms\n",
      "247:\tlearn: 33.4532943\ttotal: 242ms\tremaining: 735ms\n",
      "248:\tlearn: 33.3957704\ttotal: 243ms\tremaining: 734ms\n",
      "249:\tlearn: 33.3218832\ttotal: 244ms\tremaining: 733ms\n",
      "250:\tlearn: 33.2823774\ttotal: 245ms\tremaining: 732ms\n",
      "251:\tlearn: 33.1949668\ttotal: 246ms\tremaining: 731ms\n",
      "252:\tlearn: 33.1268050\ttotal: 247ms\tremaining: 729ms\n",
      "253:\tlearn: 33.0295259\ttotal: 248ms\tremaining: 728ms\n",
      "254:\tlearn: 32.8956135\ttotal: 249ms\tremaining: 727ms\n",
      "255:\tlearn: 32.8307975\ttotal: 250ms\tremaining: 726ms\n",
      "256:\tlearn: 32.7931065\ttotal: 251ms\tremaining: 725ms\n",
      "257:\tlearn: 32.7541666\ttotal: 252ms\tremaining: 725ms\n",
      "258:\tlearn: 32.6455644\ttotal: 253ms\tremaining: 724ms\n",
      "259:\tlearn: 32.5994355\ttotal: 254ms\tremaining: 723ms\n",
      "260:\tlearn: 32.5175453\ttotal: 255ms\tremaining: 721ms\n",
      "261:\tlearn: 32.4368870\ttotal: 256ms\tremaining: 720ms\n",
      "262:\tlearn: 32.4094157\ttotal: 256ms\tremaining: 718ms\n",
      "263:\tlearn: 32.3119991\ttotal: 257ms\tremaining: 717ms\n",
      "264:\tlearn: 32.2503354\ttotal: 258ms\tremaining: 715ms\n",
      "265:\tlearn: 32.1916534\ttotal: 259ms\tremaining: 714ms\n",
      "266:\tlearn: 32.1189335\ttotal: 260ms\tremaining: 713ms\n",
      "267:\tlearn: 32.0524602\ttotal: 261ms\tremaining: 712ms\n",
      "268:\tlearn: 31.9945096\ttotal: 261ms\tremaining: 710ms\n",
      "269:\tlearn: 31.9432231\ttotal: 262ms\tremaining: 709ms\n",
      "270:\tlearn: 31.8558105\ttotal: 263ms\tremaining: 708ms\n",
      "271:\tlearn: 31.7245360\ttotal: 264ms\tremaining: 707ms\n",
      "272:\tlearn: 31.6810521\ttotal: 265ms\tremaining: 706ms\n",
      "273:\tlearn: 31.5905310\ttotal: 267ms\tremaining: 706ms\n",
      "274:\tlearn: 31.5814820\ttotal: 268ms\tremaining: 706ms\n",
      "275:\tlearn: 31.5287706\ttotal: 269ms\tremaining: 705ms\n",
      "276:\tlearn: 31.3911913\ttotal: 269ms\tremaining: 703ms\n",
      "277:\tlearn: 31.2843119\ttotal: 270ms\tremaining: 702ms\n",
      "278:\tlearn: 31.2729218\ttotal: 271ms\tremaining: 701ms\n",
      "279:\tlearn: 31.2064306\ttotal: 272ms\tremaining: 701ms\n",
      "280:\tlearn: 31.1236218\ttotal: 273ms\tremaining: 699ms\n",
      "281:\tlearn: 31.0698615\ttotal: 274ms\tremaining: 698ms\n",
      "282:\tlearn: 30.9990092\ttotal: 275ms\tremaining: 697ms\n",
      "283:\tlearn: 30.9401974\ttotal: 276ms\tremaining: 695ms\n",
      "284:\tlearn: 30.9325154\ttotal: 277ms\tremaining: 694ms\n",
      "285:\tlearn: 30.8544742\ttotal: 278ms\tremaining: 693ms\n",
      "286:\tlearn: 30.7972384\ttotal: 279ms\tremaining: 692ms\n",
      "287:\tlearn: 30.7880988\ttotal: 279ms\tremaining: 691ms\n",
      "288:\tlearn: 30.7320574\ttotal: 280ms\tremaining: 690ms\n",
      "289:\tlearn: 30.7239007\ttotal: 281ms\tremaining: 689ms\n",
      "290:\tlearn: 30.6605866\ttotal: 282ms\tremaining: 688ms\n",
      "291:\tlearn: 30.6308656\ttotal: 283ms\tremaining: 687ms\n",
      "292:\tlearn: 30.5739096\ttotal: 284ms\tremaining: 686ms\n",
      "293:\tlearn: 30.5626222\ttotal: 285ms\tremaining: 685ms\n",
      "294:\tlearn: 30.5568473\ttotal: 286ms\tremaining: 684ms\n",
      "295:\tlearn: 30.5015496\ttotal: 287ms\tremaining: 683ms\n",
      "296:\tlearn: 30.4270317\ttotal: 288ms\tremaining: 682ms\n",
      "297:\tlearn: 30.3754589\ttotal: 289ms\tremaining: 681ms\n",
      "298:\tlearn: 30.3378030\ttotal: 290ms\tremaining: 680ms\n",
      "299:\tlearn: 30.2313208\ttotal: 291ms\tremaining: 679ms\n",
      "300:\tlearn: 30.1393072\ttotal: 292ms\tremaining: 678ms\n",
      "301:\tlearn: 30.0527433\ttotal: 293ms\tremaining: 676ms\n",
      "302:\tlearn: 29.9579479\ttotal: 294ms\tremaining: 676ms\n",
      "303:\tlearn: 29.9058132\ttotal: 295ms\tremaining: 675ms\n",
      "304:\tlearn: 29.8985179\ttotal: 296ms\tremaining: 674ms\n",
      "305:\tlearn: 29.8111192\ttotal: 297ms\tremaining: 673ms\n",
      "306:\tlearn: 29.7506065\ttotal: 298ms\tremaining: 673ms\n",
      "307:\tlearn: 29.6532751\ttotal: 299ms\tremaining: 672ms\n",
      "308:\tlearn: 29.6013337\ttotal: 300ms\tremaining: 671ms\n",
      "309:\tlearn: 29.5949394\ttotal: 301ms\tremaining: 670ms\n",
      "310:\tlearn: 29.5370916\ttotal: 302ms\tremaining: 668ms\n",
      "311:\tlearn: 29.5077830\ttotal: 303ms\tremaining: 667ms\n",
      "312:\tlearn: 29.4248526\ttotal: 304ms\tremaining: 666ms\n",
      "313:\tlearn: 29.3805316\ttotal: 305ms\tremaining: 666ms\n",
      "314:\tlearn: 29.2893527\ttotal: 306ms\tremaining: 665ms\n",
      "315:\tlearn: 29.2432232\ttotal: 307ms\tremaining: 664ms\n",
      "316:\tlearn: 29.2373361\ttotal: 308ms\tremaining: 663ms\n",
      "317:\tlearn: 29.1224803\ttotal: 309ms\tremaining: 662ms\n",
      "318:\tlearn: 29.0765583\ttotal: 309ms\tremaining: 661ms\n",
      "319:\tlearn: 28.9848827\ttotal: 310ms\tremaining: 660ms\n",
      "320:\tlearn: 28.9083749\ttotal: 311ms\tremaining: 658ms\n",
      "321:\tlearn: 28.8076597\ttotal: 312ms\tremaining: 658ms\n",
      "322:\tlearn: 28.7360412\ttotal: 313ms\tremaining: 657ms\n",
      "323:\tlearn: 28.7158107\ttotal: 314ms\tremaining: 656ms\n",
      "324:\tlearn: 28.6685417\ttotal: 315ms\tremaining: 655ms\n",
      "325:\tlearn: 28.6477891\ttotal: 316ms\tremaining: 654ms\n",
      "326:\tlearn: 28.6027510\ttotal: 317ms\tremaining: 653ms\n",
      "327:\tlearn: 28.5511355\ttotal: 318ms\tremaining: 652ms\n",
      "328:\tlearn: 28.5440701\ttotal: 319ms\tremaining: 651ms\n",
      "329:\tlearn: 28.4708110\ttotal: 320ms\tremaining: 650ms\n",
      "330:\tlearn: 28.4645627\ttotal: 321ms\tremaining: 649ms\n",
      "331:\tlearn: 28.3704438\ttotal: 322ms\tremaining: 648ms\n",
      "332:\tlearn: 28.3634241\ttotal: 323ms\tremaining: 648ms\n",
      "333:\tlearn: 28.3186244\ttotal: 324ms\tremaining: 647ms\n",
      "334:\tlearn: 28.2613764\ttotal: 325ms\tremaining: 646ms\n",
      "335:\tlearn: 28.1538932\ttotal: 326ms\tremaining: 645ms\n",
      "336:\tlearn: 28.1223246\ttotal: 327ms\tremaining: 644ms\n",
      "337:\tlearn: 28.0715556\ttotal: 328ms\tremaining: 643ms\n",
      "338:\tlearn: 28.0045095\ttotal: 329ms\tremaining: 642ms\n",
      "339:\tlearn: 27.9469559\ttotal: 330ms\tremaining: 641ms\n",
      "340:\tlearn: 27.9075170\ttotal: 331ms\tremaining: 640ms\n",
      "341:\tlearn: 27.8393856\ttotal: 332ms\tremaining: 639ms\n",
      "342:\tlearn: 27.8084813\ttotal: 333ms\tremaining: 638ms\n",
      "343:\tlearn: 27.8006204\ttotal: 334ms\tremaining: 636ms\n",
      "344:\tlearn: 27.7940121\ttotal: 335ms\tremaining: 635ms\n",
      "345:\tlearn: 27.7897137\ttotal: 335ms\tremaining: 634ms\n",
      "346:\tlearn: 27.7276765\ttotal: 336ms\tremaining: 633ms\n",
      "347:\tlearn: 27.6434422\ttotal: 337ms\tremaining: 632ms\n",
      "348:\tlearn: 27.6096334\ttotal: 338ms\tremaining: 631ms\n",
      "349:\tlearn: 27.5187315\ttotal: 339ms\tremaining: 630ms\n",
      "350:\tlearn: 27.4845217\ttotal: 340ms\tremaining: 628ms\n",
      "351:\tlearn: 27.4200688\ttotal: 341ms\tremaining: 627ms\n",
      "352:\tlearn: 27.3478667\ttotal: 342ms\tremaining: 626ms\n",
      "353:\tlearn: 27.2974769\ttotal: 343ms\tremaining: 625ms\n",
      "354:\tlearn: 27.2468761\ttotal: 344ms\tremaining: 624ms\n",
      "355:\tlearn: 27.1848065\ttotal: 344ms\tremaining: 623ms\n",
      "356:\tlearn: 27.1263009\ttotal: 345ms\tremaining: 622ms\n",
      "357:\tlearn: 27.0822692\ttotal: 346ms\tremaining: 621ms\n",
      "358:\tlearn: 27.0691253\ttotal: 347ms\tremaining: 620ms\n",
      "359:\tlearn: 27.0269038\ttotal: 348ms\tremaining: 619ms\n",
      "360:\tlearn: 26.9952141\ttotal: 349ms\tremaining: 618ms\n",
      "361:\tlearn: 26.9890372\ttotal: 350ms\tremaining: 617ms\n",
      "362:\tlearn: 26.9465768\ttotal: 351ms\tremaining: 616ms\n",
      "363:\tlearn: 26.9044433\ttotal: 352ms\tremaining: 615ms\n",
      "364:\tlearn: 26.8682630\ttotal: 353ms\tremaining: 614ms\n",
      "365:\tlearn: 26.8618548\ttotal: 354ms\tremaining: 612ms\n",
      "366:\tlearn: 26.8144006\ttotal: 354ms\tremaining: 611ms\n",
      "367:\tlearn: 26.8093179\ttotal: 355ms\tremaining: 610ms\n",
      "368:\tlearn: 26.7618810\ttotal: 356ms\tremaining: 609ms\n",
      "369:\tlearn: 26.7571170\ttotal: 357ms\tremaining: 608ms\n",
      "370:\tlearn: 26.7043205\ttotal: 358ms\tremaining: 607ms\n",
      "371:\tlearn: 26.6378026\ttotal: 359ms\tremaining: 606ms\n",
      "372:\tlearn: 26.5514300\ttotal: 360ms\tremaining: 605ms\n",
      "373:\tlearn: 26.4460130\ttotal: 361ms\tremaining: 604ms\n",
      "374:\tlearn: 26.4398424\ttotal: 362ms\tremaining: 603ms\n",
      "375:\tlearn: 26.3852939\ttotal: 363ms\tremaining: 602ms\n",
      "376:\tlearn: 26.3344798\ttotal: 364ms\tremaining: 601ms\n",
      "377:\tlearn: 26.3272707\ttotal: 365ms\tremaining: 600ms\n",
      "378:\tlearn: 26.2989990\ttotal: 365ms\tremaining: 599ms\n",
      "379:\tlearn: 26.2476501\ttotal: 366ms\tremaining: 598ms\n",
      "380:\tlearn: 26.1918108\ttotal: 367ms\tremaining: 596ms\n",
      "381:\tlearn: 26.1154654\ttotal: 368ms\tremaining: 595ms\n",
      "382:\tlearn: 26.1087670\ttotal: 369ms\tremaining: 594ms\n",
      "383:\tlearn: 26.1040716\ttotal: 370ms\tremaining: 593ms\n",
      "384:\tlearn: 26.0318993\ttotal: 371ms\tremaining: 592ms\n",
      "385:\tlearn: 25.9751131\ttotal: 372ms\tremaining: 591ms\n",
      "386:\tlearn: 25.9379526\ttotal: 373ms\tremaining: 590ms\n",
      "387:\tlearn: 25.9287243\ttotal: 374ms\tremaining: 589ms\n",
      "388:\tlearn: 25.8751276\ttotal: 375ms\tremaining: 588ms\n",
      "389:\tlearn: 25.8300439\ttotal: 376ms\tremaining: 587ms\n",
      "390:\tlearn: 25.7598570\ttotal: 376ms\tremaining: 586ms\n",
      "391:\tlearn: 25.7332718\ttotal: 377ms\tremaining: 585ms\n",
      "392:\tlearn: 25.6894335\ttotal: 378ms\tremaining: 584ms\n",
      "393:\tlearn: 25.6840453\ttotal: 379ms\tremaining: 583ms\n",
      "394:\tlearn: 25.6672354\ttotal: 380ms\tremaining: 582ms\n",
      "395:\tlearn: 25.6613645\ttotal: 381ms\tremaining: 581ms\n",
      "396:\tlearn: 25.6040062\ttotal: 382ms\tremaining: 580ms\n",
      "397:\tlearn: 25.5713619\ttotal: 383ms\tremaining: 579ms\n",
      "398:\tlearn: 25.5146095\ttotal: 384ms\tremaining: 578ms\n",
      "399:\tlearn: 25.4836363\ttotal: 384ms\tremaining: 577ms\n",
      "400:\tlearn: 25.4311068\ttotal: 385ms\tremaining: 576ms\n",
      "401:\tlearn: 25.3692198\ttotal: 386ms\tremaining: 575ms\n",
      "402:\tlearn: 25.3480067\ttotal: 387ms\tremaining: 573ms\n",
      "403:\tlearn: 25.3052966\ttotal: 388ms\tremaining: 572ms\n",
      "404:\tlearn: 25.2519111\ttotal: 389ms\tremaining: 571ms\n",
      "405:\tlearn: 25.2460603\ttotal: 390ms\tremaining: 571ms\n",
      "406:\tlearn: 25.2144729\ttotal: 391ms\tremaining: 570ms\n",
      "407:\tlearn: 25.1347374\ttotal: 392ms\tremaining: 569ms\n",
      "408:\tlearn: 25.0284719\ttotal: 394ms\tremaining: 569ms\n",
      "409:\tlearn: 24.9752360\ttotal: 395ms\tremaining: 568ms\n",
      "410:\tlearn: 24.9260888\ttotal: 396ms\tremaining: 567ms\n",
      "411:\tlearn: 24.8766066\ttotal: 396ms\tremaining: 566ms\n",
      "412:\tlearn: 24.8719284\ttotal: 397ms\tremaining: 565ms\n",
      "413:\tlearn: 24.7976067\ttotal: 398ms\tremaining: 564ms\n",
      "414:\tlearn: 24.7491830\ttotal: 399ms\tremaining: 563ms\n",
      "415:\tlearn: 24.7333382\ttotal: 400ms\tremaining: 561ms\n",
      "416:\tlearn: 24.6709770\ttotal: 401ms\tremaining: 560ms\n",
      "417:\tlearn: 24.6271914\ttotal: 402ms\tremaining: 559ms\n",
      "418:\tlearn: 24.5975426\ttotal: 403ms\tremaining: 558ms\n",
      "419:\tlearn: 24.5646781\ttotal: 404ms\tremaining: 557ms\n",
      "420:\tlearn: 24.5254733\ttotal: 405ms\tremaining: 557ms\n",
      "421:\tlearn: 24.4870436\ttotal: 406ms\tremaining: 556ms\n",
      "422:\tlearn: 24.4832883\ttotal: 407ms\tremaining: 555ms\n",
      "423:\tlearn: 24.4214499\ttotal: 408ms\tremaining: 554ms\n",
      "424:\tlearn: 24.4178535\ttotal: 409ms\tremaining: 553ms\n",
      "425:\tlearn: 24.4120127\ttotal: 409ms\tremaining: 552ms\n",
      "426:\tlearn: 24.3547134\ttotal: 410ms\tremaining: 551ms\n",
      "427:\tlearn: 24.3406164\ttotal: 411ms\tremaining: 550ms\n",
      "428:\tlearn: 24.2735838\ttotal: 412ms\tremaining: 549ms\n",
      "429:\tlearn: 24.1751947\ttotal: 413ms\tremaining: 548ms\n",
      "430:\tlearn: 24.1377276\ttotal: 414ms\tremaining: 547ms\n",
      "431:\tlearn: 24.1304914\ttotal: 415ms\tremaining: 546ms\n",
      "432:\tlearn: 24.1223580\ttotal: 416ms\tremaining: 545ms\n",
      "433:\tlearn: 24.0860040\ttotal: 417ms\tremaining: 544ms\n",
      "434:\tlearn: 24.0466889\ttotal: 418ms\tremaining: 543ms\n",
      "435:\tlearn: 23.9919742\ttotal: 419ms\tremaining: 542ms\n",
      "436:\tlearn: 23.9457932\ttotal: 420ms\tremaining: 541ms\n",
      "437:\tlearn: 23.8815083\ttotal: 421ms\tremaining: 540ms\n",
      "438:\tlearn: 23.8781224\ttotal: 422ms\tremaining: 540ms\n",
      "439:\tlearn: 23.8259252\ttotal: 423ms\tremaining: 539ms\n",
      "440:\tlearn: 23.7926954\ttotal: 424ms\tremaining: 538ms\n",
      "441:\tlearn: 23.7407856\ttotal: 425ms\tremaining: 537ms\n",
      "442:\tlearn: 23.7320938\ttotal: 426ms\tremaining: 536ms\n",
      "443:\tlearn: 23.6915647\ttotal: 427ms\tremaining: 535ms\n",
      "444:\tlearn: 23.6792172\ttotal: 428ms\tremaining: 534ms\n",
      "445:\tlearn: 23.6210072\ttotal: 429ms\tremaining: 533ms\n",
      "446:\tlearn: 23.6161297\ttotal: 430ms\tremaining: 532ms\n",
      "447:\tlearn: 23.5496392\ttotal: 431ms\tremaining: 531ms\n",
      "448:\tlearn: 23.5328028\ttotal: 432ms\tremaining: 530ms\n",
      "449:\tlearn: 23.4639952\ttotal: 433ms\tremaining: 529ms\n",
      "450:\tlearn: 23.4092092\ttotal: 433ms\tremaining: 528ms\n",
      "451:\tlearn: 23.3452766\ttotal: 434ms\tremaining: 527ms\n",
      "452:\tlearn: 23.3140723\ttotal: 436ms\tremaining: 526ms\n",
      "453:\tlearn: 23.2249015\ttotal: 437ms\tremaining: 525ms\n",
      "454:\tlearn: 23.1744393\ttotal: 438ms\tremaining: 524ms\n",
      "455:\tlearn: 23.1699584\ttotal: 439ms\tremaining: 523ms\n",
      "456:\tlearn: 23.1302089\ttotal: 440ms\tremaining: 522ms\n",
      "457:\tlearn: 23.0294126\ttotal: 440ms\tremaining: 521ms\n",
      "458:\tlearn: 22.9840724\ttotal: 441ms\tremaining: 520ms\n",
      "459:\tlearn: 22.9358059\ttotal: 442ms\tremaining: 519ms\n",
      "460:\tlearn: 22.8968573\ttotal: 443ms\tremaining: 518ms\n",
      "461:\tlearn: 22.8930427\ttotal: 444ms\tremaining: 517ms\n",
      "462:\tlearn: 22.8331553\ttotal: 445ms\tremaining: 516ms\n",
      "463:\tlearn: 22.7810277\ttotal: 446ms\tremaining: 515ms\n",
      "464:\tlearn: 22.7759205\ttotal: 447ms\tremaining: 514ms\n",
      "465:\tlearn: 22.7720110\ttotal: 448ms\tremaining: 513ms\n",
      "466:\tlearn: 22.7024479\ttotal: 449ms\tremaining: 512ms\n",
      "467:\tlearn: 22.6995010\ttotal: 450ms\tremaining: 511ms\n",
      "468:\tlearn: 22.6619011\ttotal: 451ms\tremaining: 510ms\n",
      "469:\tlearn: 22.6081489\ttotal: 452ms\tremaining: 510ms\n",
      "470:\tlearn: 22.5927501\ttotal: 453ms\tremaining: 509ms\n",
      "471:\tlearn: 22.5346437\ttotal: 454ms\tremaining: 508ms\n",
      "472:\tlearn: 22.4779136\ttotal: 455ms\tremaining: 507ms\n",
      "473:\tlearn: 22.4276706\ttotal: 456ms\tremaining: 506ms\n",
      "474:\tlearn: 22.4199813\ttotal: 457ms\tremaining: 505ms\n",
      "475:\tlearn: 22.3876330\ttotal: 457ms\tremaining: 504ms\n",
      "476:\tlearn: 22.3528581\ttotal: 458ms\tremaining: 503ms\n",
      "477:\tlearn: 22.3490392\ttotal: 459ms\tremaining: 502ms\n",
      "478:\tlearn: 22.3031423\ttotal: 460ms\tremaining: 501ms\n",
      "479:\tlearn: 22.3007944\ttotal: 461ms\tremaining: 500ms\n",
      "480:\tlearn: 22.2542492\ttotal: 462ms\tremaining: 499ms\n",
      "481:\tlearn: 22.2209373\ttotal: 463ms\tremaining: 498ms\n",
      "482:\tlearn: 22.2123723\ttotal: 464ms\tremaining: 496ms\n",
      "483:\tlearn: 22.1482190\ttotal: 465ms\tremaining: 495ms\n",
      "484:\tlearn: 22.1339636\ttotal: 466ms\tremaining: 495ms\n",
      "485:\tlearn: 22.0631395\ttotal: 467ms\tremaining: 494ms\n",
      "486:\tlearn: 22.0588600\ttotal: 468ms\tremaining: 493ms\n",
      "487:\tlearn: 22.0347781\ttotal: 469ms\tremaining: 492ms\n",
      "488:\tlearn: 22.0277106\ttotal: 470ms\tremaining: 491ms\n",
      "489:\tlearn: 22.0139632\ttotal: 470ms\tremaining: 490ms\n",
      "490:\tlearn: 21.9731526\ttotal: 471ms\tremaining: 489ms\n",
      "491:\tlearn: 21.9436288\ttotal: 472ms\tremaining: 488ms\n",
      "492:\tlearn: 21.8956250\ttotal: 473ms\tremaining: 487ms\n",
      "493:\tlearn: 21.8193608\ttotal: 474ms\tremaining: 485ms\n",
      "494:\tlearn: 21.8167284\ttotal: 475ms\tremaining: 485ms\n",
      "495:\tlearn: 21.7766221\ttotal: 476ms\tremaining: 483ms\n",
      "496:\tlearn: 21.7498528\ttotal: 477ms\tremaining: 482ms\n",
      "497:\tlearn: 21.7303242\ttotal: 477ms\tremaining: 481ms\n",
      "498:\tlearn: 21.6849682\ttotal: 478ms\tremaining: 480ms\n",
      "499:\tlearn: 21.6427310\ttotal: 479ms\tremaining: 479ms\n",
      "500:\tlearn: 21.5528544\ttotal: 480ms\tremaining: 478ms\n",
      "501:\tlearn: 21.5054316\ttotal: 481ms\tremaining: 478ms\n",
      "502:\tlearn: 21.4760129\ttotal: 482ms\tremaining: 477ms\n",
      "503:\tlearn: 21.4490218\ttotal: 483ms\tremaining: 476ms\n",
      "504:\tlearn: 21.4265026\ttotal: 484ms\tremaining: 475ms\n",
      "505:\tlearn: 21.3902548\ttotal: 485ms\tremaining: 473ms\n",
      "506:\tlearn: 21.3540889\ttotal: 486ms\tremaining: 472ms\n",
      "507:\tlearn: 21.3198302\ttotal: 487ms\tremaining: 471ms\n",
      "508:\tlearn: 21.2740171\ttotal: 488ms\tremaining: 470ms\n",
      "509:\tlearn: 21.2345957\ttotal: 488ms\tremaining: 469ms\n",
      "510:\tlearn: 21.1872209\ttotal: 489ms\tremaining: 468ms\n",
      "511:\tlearn: 21.1695153\ttotal: 490ms\tremaining: 467ms\n",
      "512:\tlearn: 21.1522472\ttotal: 491ms\tremaining: 466ms\n",
      "513:\tlearn: 21.1497867\ttotal: 492ms\tremaining: 465ms\n",
      "514:\tlearn: 21.1015735\ttotal: 493ms\tremaining: 464ms\n",
      "515:\tlearn: 21.0385321\ttotal: 494ms\tremaining: 464ms\n",
      "516:\tlearn: 20.9991473\ttotal: 495ms\tremaining: 463ms\n",
      "517:\tlearn: 20.9703746\ttotal: 496ms\tremaining: 462ms\n",
      "518:\tlearn: 20.9222185\ttotal: 497ms\tremaining: 461ms\n",
      "519:\tlearn: 20.8999891\ttotal: 498ms\tremaining: 460ms\n",
      "520:\tlearn: 20.8554804\ttotal: 499ms\tremaining: 459ms\n",
      "521:\tlearn: 20.8128166\ttotal: 500ms\tremaining: 458ms\n",
      "522:\tlearn: 20.7911705\ttotal: 501ms\tremaining: 457ms\n",
      "523:\tlearn: 20.7460615\ttotal: 501ms\tremaining: 456ms\n",
      "524:\tlearn: 20.7055146\ttotal: 502ms\tremaining: 455ms\n",
      "525:\tlearn: 20.6579428\ttotal: 504ms\tremaining: 454ms\n",
      "526:\tlearn: 20.6358801\ttotal: 504ms\tremaining: 453ms\n",
      "527:\tlearn: 20.5876242\ttotal: 505ms\tremaining: 452ms\n",
      "528:\tlearn: 20.5663373\ttotal: 506ms\tremaining: 451ms\n",
      "529:\tlearn: 20.5187919\ttotal: 507ms\tremaining: 450ms\n",
      "530:\tlearn: 20.4578327\ttotal: 508ms\tremaining: 449ms\n",
      "531:\tlearn: 20.3828339\ttotal: 509ms\tremaining: 448ms\n",
      "532:\tlearn: 20.3633039\ttotal: 510ms\tremaining: 447ms\n",
      "533:\tlearn: 20.3358557\ttotal: 511ms\tremaining: 446ms\n",
      "534:\tlearn: 20.3215443\ttotal: 512ms\tremaining: 445ms\n",
      "535:\tlearn: 20.3123640\ttotal: 513ms\tremaining: 444ms\n",
      "536:\tlearn: 20.2915068\ttotal: 514ms\tremaining: 443ms\n",
      "537:\tlearn: 20.2365966\ttotal: 515ms\tremaining: 442ms\n",
      "538:\tlearn: 20.2115139\ttotal: 516ms\tremaining: 441ms\n",
      "539:\tlearn: 20.2054815\ttotal: 517ms\tremaining: 440ms\n",
      "540:\tlearn: 20.1590883\ttotal: 518ms\tremaining: 439ms\n",
      "541:\tlearn: 20.1547842\ttotal: 519ms\tremaining: 438ms\n",
      "542:\tlearn: 20.1279782\ttotal: 520ms\tremaining: 437ms\n",
      "543:\tlearn: 20.1237081\ttotal: 521ms\tremaining: 436ms\n",
      "544:\tlearn: 20.0569792\ttotal: 522ms\tremaining: 435ms\n",
      "545:\tlearn: 20.0329940\ttotal: 522ms\tremaining: 434ms\n",
      "546:\tlearn: 19.9896129\ttotal: 523ms\tremaining: 433ms\n",
      "547:\tlearn: 19.9398546\ttotal: 524ms\tremaining: 432ms\n",
      "548:\tlearn: 19.8595835\ttotal: 525ms\tremaining: 431ms\n",
      "549:\tlearn: 19.8409663\ttotal: 526ms\tremaining: 430ms\n",
      "550:\tlearn: 19.8053021\ttotal: 527ms\tremaining: 430ms\n",
      "551:\tlearn: 19.7454656\ttotal: 529ms\tremaining: 429ms\n",
      "552:\tlearn: 19.6980410\ttotal: 530ms\tremaining: 428ms\n",
      "553:\tlearn: 19.6626935\ttotal: 531ms\tremaining: 427ms\n",
      "554:\tlearn: 19.6266865\ttotal: 531ms\tremaining: 426ms\n",
      "555:\tlearn: 19.5898795\ttotal: 533ms\tremaining: 425ms\n",
      "556:\tlearn: 19.5751403\ttotal: 533ms\tremaining: 424ms\n",
      "557:\tlearn: 19.5731983\ttotal: 534ms\tremaining: 423ms\n",
      "558:\tlearn: 19.5371056\ttotal: 535ms\tremaining: 422ms\n",
      "559:\tlearn: 19.5048042\ttotal: 536ms\tremaining: 421ms\n",
      "560:\tlearn: 19.5012571\ttotal: 537ms\tremaining: 420ms\n",
      "561:\tlearn: 19.4832223\ttotal: 538ms\tremaining: 419ms\n",
      "562:\tlearn: 19.4600865\ttotal: 539ms\tremaining: 418ms\n",
      "563:\tlearn: 19.4384968\ttotal: 540ms\tremaining: 417ms\n",
      "564:\tlearn: 19.4361093\ttotal: 541ms\tremaining: 416ms\n",
      "565:\tlearn: 19.4110086\ttotal: 542ms\tremaining: 415ms\n",
      "566:\tlearn: 19.3699414\ttotal: 543ms\tremaining: 414ms\n",
      "567:\tlearn: 19.3500309\ttotal: 544ms\tremaining: 414ms\n",
      "568:\tlearn: 19.2982027\ttotal: 545ms\tremaining: 413ms\n",
      "569:\tlearn: 19.2285478\ttotal: 546ms\tremaining: 412ms\n",
      "570:\tlearn: 19.2107074\ttotal: 547ms\tremaining: 411ms\n",
      "571:\tlearn: 19.1894003\ttotal: 548ms\tremaining: 410ms\n",
      "572:\tlearn: 19.1858983\ttotal: 549ms\tremaining: 409ms\n",
      "573:\tlearn: 19.1385056\ttotal: 549ms\tremaining: 408ms\n",
      "574:\tlearn: 19.1160386\ttotal: 550ms\tremaining: 407ms\n",
      "575:\tlearn: 19.0988308\ttotal: 551ms\tremaining: 406ms\n",
      "576:\tlearn: 19.0783785\ttotal: 552ms\tremaining: 405ms\n",
      "577:\tlearn: 19.0231023\ttotal: 553ms\tremaining: 404ms\n",
      "578:\tlearn: 19.0036597\ttotal: 554ms\tremaining: 403ms\n",
      "579:\tlearn: 18.9799549\ttotal: 555ms\tremaining: 402ms\n",
      "580:\tlearn: 18.9711749\ttotal: 556ms\tremaining: 401ms\n",
      "581:\tlearn: 18.9443614\ttotal: 557ms\tremaining: 400ms\n",
      "582:\tlearn: 18.9191490\ttotal: 558ms\tremaining: 399ms\n",
      "583:\tlearn: 18.8751081\ttotal: 559ms\tremaining: 398ms\n",
      "584:\tlearn: 18.8293391\ttotal: 560ms\tremaining: 397ms\n",
      "585:\tlearn: 18.7965695\ttotal: 561ms\tremaining: 396ms\n",
      "586:\tlearn: 18.7819962\ttotal: 561ms\tremaining: 395ms\n",
      "587:\tlearn: 18.7304407\ttotal: 562ms\tremaining: 394ms\n",
      "588:\tlearn: 18.6926858\ttotal: 563ms\tremaining: 393ms\n",
      "589:\tlearn: 18.6510988\ttotal: 564ms\tremaining: 392ms\n",
      "590:\tlearn: 18.6365421\ttotal: 565ms\tremaining: 391ms\n",
      "591:\tlearn: 18.6016154\ttotal: 566ms\tremaining: 390ms\n",
      "592:\tlearn: 18.5842759\ttotal: 567ms\tremaining: 389ms\n",
      "593:\tlearn: 18.5401411\ttotal: 568ms\tremaining: 388ms\n",
      "594:\tlearn: 18.5373289\ttotal: 569ms\tremaining: 387ms\n",
      "595:\tlearn: 18.5086795\ttotal: 570ms\tremaining: 386ms\n",
      "596:\tlearn: 18.4493193\ttotal: 571ms\tremaining: 386ms\n",
      "597:\tlearn: 18.4063125\ttotal: 572ms\tremaining: 385ms\n",
      "598:\tlearn: 18.3520546\ttotal: 573ms\tremaining: 384ms\n",
      "599:\tlearn: 18.3180003\ttotal: 574ms\tremaining: 383ms\n",
      "600:\tlearn: 18.3041324\ttotal: 575ms\tremaining: 382ms\n",
      "601:\tlearn: 18.2722337\ttotal: 576ms\tremaining: 381ms\n",
      "602:\tlearn: 18.2287353\ttotal: 577ms\tremaining: 380ms\n",
      "603:\tlearn: 18.1832568\ttotal: 578ms\tremaining: 379ms\n",
      "604:\tlearn: 18.1805032\ttotal: 579ms\tremaining: 378ms\n",
      "605:\tlearn: 18.1520544\ttotal: 580ms\tremaining: 377ms\n",
      "606:\tlearn: 18.1354692\ttotal: 581ms\tremaining: 376ms\n",
      "607:\tlearn: 18.1264107\ttotal: 581ms\tremaining: 375ms\n",
      "608:\tlearn: 18.1026188\ttotal: 583ms\tremaining: 374ms\n",
      "609:\tlearn: 18.0549990\ttotal: 584ms\tremaining: 373ms\n",
      "610:\tlearn: 18.0037741\ttotal: 585ms\tremaining: 372ms\n",
      "611:\tlearn: 17.9604370\ttotal: 586ms\tremaining: 371ms\n",
      "612:\tlearn: 17.9285276\ttotal: 587ms\tremaining: 370ms\n",
      "613:\tlearn: 17.9213726\ttotal: 588ms\tremaining: 369ms\n",
      "614:\tlearn: 17.9066908\ttotal: 589ms\tremaining: 368ms\n",
      "615:\tlearn: 17.9030634\ttotal: 590ms\tremaining: 368ms\n",
      "616:\tlearn: 17.8900287\ttotal: 591ms\tremaining: 367ms\n",
      "617:\tlearn: 17.8829280\ttotal: 592ms\tremaining: 366ms\n",
      "618:\tlearn: 17.8388985\ttotal: 593ms\tremaining: 365ms\n",
      "619:\tlearn: 17.7928862\ttotal: 593ms\tremaining: 364ms\n",
      "620:\tlearn: 17.7466362\ttotal: 594ms\tremaining: 363ms\n",
      "621:\tlearn: 17.7104434\ttotal: 595ms\tremaining: 362ms\n",
      "622:\tlearn: 17.6661359\ttotal: 596ms\tremaining: 361ms\n",
      "623:\tlearn: 17.6501147\ttotal: 597ms\tremaining: 360ms\n",
      "624:\tlearn: 17.6396963\ttotal: 598ms\tremaining: 359ms\n",
      "625:\tlearn: 17.6102058\ttotal: 599ms\tremaining: 358ms\n",
      "626:\tlearn: 17.6078614\ttotal: 600ms\tremaining: 357ms\n",
      "627:\tlearn: 17.5630963\ttotal: 601ms\tremaining: 356ms\n",
      "628:\tlearn: 17.5522732\ttotal: 602ms\tremaining: 355ms\n",
      "629:\tlearn: 17.5429766\ttotal: 603ms\tremaining: 354ms\n",
      "630:\tlearn: 17.5297201\ttotal: 604ms\tremaining: 353ms\n",
      "631:\tlearn: 17.5253753\ttotal: 605ms\tremaining: 352ms\n",
      "632:\tlearn: 17.4715324\ttotal: 606ms\tremaining: 351ms\n",
      "633:\tlearn: 17.4328538\ttotal: 607ms\tremaining: 351ms\n",
      "634:\tlearn: 17.4197552\ttotal: 608ms\tremaining: 350ms\n",
      "635:\tlearn: 17.3686977\ttotal: 609ms\tremaining: 349ms\n",
      "636:\tlearn: 17.3455865\ttotal: 610ms\tremaining: 348ms\n",
      "637:\tlearn: 17.2964399\ttotal: 611ms\tremaining: 347ms\n",
      "638:\tlearn: 17.2723522\ttotal: 612ms\tremaining: 346ms\n",
      "639:\tlearn: 17.2310604\ttotal: 613ms\tremaining: 345ms\n",
      "640:\tlearn: 17.1793967\ttotal: 614ms\tremaining: 344ms\n",
      "641:\tlearn: 17.1565288\ttotal: 615ms\tremaining: 343ms\n",
      "642:\tlearn: 17.1317463\ttotal: 616ms\tremaining: 342ms\n",
      "643:\tlearn: 17.1118715\ttotal: 617ms\tremaining: 341ms\n",
      "644:\tlearn: 17.0735889\ttotal: 618ms\tremaining: 340ms\n",
      "645:\tlearn: 17.0529158\ttotal: 618ms\tremaining: 339ms\n",
      "646:\tlearn: 17.0329358\ttotal: 619ms\tremaining: 338ms\n",
      "647:\tlearn: 16.9940748\ttotal: 620ms\tremaining: 337ms\n",
      "648:\tlearn: 16.9521366\ttotal: 621ms\tremaining: 336ms\n",
      "649:\tlearn: 16.9500090\ttotal: 622ms\tremaining: 335ms\n",
      "650:\tlearn: 16.9398288\ttotal: 623ms\tremaining: 334ms\n",
      "651:\tlearn: 16.9269785\ttotal: 624ms\tremaining: 333ms\n",
      "652:\tlearn: 16.9251987\ttotal: 625ms\tremaining: 332ms\n",
      "653:\tlearn: 16.8865075\ttotal: 626ms\tremaining: 331ms\n",
      "654:\tlearn: 16.8336257\ttotal: 627ms\tremaining: 330ms\n",
      "655:\tlearn: 16.7745723\ttotal: 628ms\tremaining: 329ms\n",
      "656:\tlearn: 16.7453018\ttotal: 629ms\tremaining: 328ms\n",
      "657:\tlearn: 16.7132547\ttotal: 630ms\tremaining: 327ms\n",
      "658:\tlearn: 16.6688611\ttotal: 631ms\tremaining: 326ms\n",
      "659:\tlearn: 16.6336092\ttotal: 631ms\tremaining: 325ms\n",
      "660:\tlearn: 16.6224205\ttotal: 632ms\tremaining: 324ms\n",
      "661:\tlearn: 16.6047201\ttotal: 633ms\tremaining: 323ms\n",
      "662:\tlearn: 16.5744777\ttotal: 634ms\tremaining: 322ms\n",
      "663:\tlearn: 16.5205394\ttotal: 635ms\tremaining: 321ms\n",
      "664:\tlearn: 16.4841640\ttotal: 636ms\tremaining: 320ms\n",
      "665:\tlearn: 16.4718565\ttotal: 637ms\tremaining: 320ms\n",
      "666:\tlearn: 16.4506629\ttotal: 638ms\tremaining: 319ms\n",
      "667:\tlearn: 16.4434901\ttotal: 639ms\tremaining: 318ms\n",
      "668:\tlearn: 16.4390434\ttotal: 640ms\tremaining: 317ms\n",
      "669:\tlearn: 16.4033147\ttotal: 641ms\tremaining: 316ms\n",
      "670:\tlearn: 16.3914168\ttotal: 642ms\tremaining: 315ms\n",
      "671:\tlearn: 16.3883133\ttotal: 643ms\tremaining: 314ms\n",
      "672:\tlearn: 16.3619882\ttotal: 644ms\tremaining: 313ms\n",
      "673:\tlearn: 16.3187350\ttotal: 645ms\tremaining: 312ms\n",
      "674:\tlearn: 16.3071318\ttotal: 646ms\tremaining: 311ms\n",
      "675:\tlearn: 16.2799616\ttotal: 646ms\tremaining: 310ms\n",
      "676:\tlearn: 16.2543052\ttotal: 647ms\tremaining: 309ms\n",
      "677:\tlearn: 16.2247365\ttotal: 648ms\tremaining: 308ms\n",
      "678:\tlearn: 16.2224638\ttotal: 649ms\tremaining: 307ms\n",
      "679:\tlearn: 16.1731674\ttotal: 651ms\tremaining: 306ms\n",
      "680:\tlearn: 16.1714538\ttotal: 652ms\tremaining: 305ms\n",
      "681:\tlearn: 16.1613971\ttotal: 652ms\tremaining: 304ms\n",
      "682:\tlearn: 16.1147448\ttotal: 653ms\tremaining: 303ms\n",
      "683:\tlearn: 16.0696533\ttotal: 654ms\tremaining: 302ms\n",
      "684:\tlearn: 16.0237966\ttotal: 655ms\tremaining: 301ms\n",
      "685:\tlearn: 16.0150106\ttotal: 656ms\tremaining: 300ms\n",
      "686:\tlearn: 15.9803812\ttotal: 657ms\tremaining: 299ms\n",
      "687:\tlearn: 15.9782292\ttotal: 658ms\tremaining: 298ms\n",
      "688:\tlearn: 15.9350017\ttotal: 659ms\tremaining: 297ms\n",
      "689:\tlearn: 15.9165426\ttotal: 660ms\tremaining: 296ms\n",
      "690:\tlearn: 15.9147553\ttotal: 660ms\tremaining: 295ms\n",
      "691:\tlearn: 15.9021542\ttotal: 661ms\tremaining: 294ms\n",
      "692:\tlearn: 15.8686086\ttotal: 662ms\tremaining: 293ms\n",
      "693:\tlearn: 15.8668570\ttotal: 663ms\tremaining: 292ms\n",
      "694:\tlearn: 15.8137514\ttotal: 664ms\tremaining: 291ms\n",
      "695:\tlearn: 15.8034165\ttotal: 665ms\tremaining: 290ms\n",
      "696:\tlearn: 15.7543075\ttotal: 666ms\tremaining: 289ms\n",
      "697:\tlearn: 15.7227008\ttotal: 667ms\tremaining: 289ms\n",
      "698:\tlearn: 15.6928794\ttotal: 668ms\tremaining: 288ms\n",
      "699:\tlearn: 15.6511903\ttotal: 669ms\tremaining: 287ms\n",
      "700:\tlearn: 15.6321406\ttotal: 670ms\tremaining: 286ms\n",
      "701:\tlearn: 15.6043360\ttotal: 671ms\tremaining: 285ms\n",
      "702:\tlearn: 15.5952870\ttotal: 672ms\tremaining: 284ms\n",
      "703:\tlearn: 15.5905096\ttotal: 673ms\tremaining: 283ms\n",
      "704:\tlearn: 15.5836839\ttotal: 673ms\tremaining: 282ms\n",
      "705:\tlearn: 15.5652754\ttotal: 674ms\tremaining: 281ms\n",
      "706:\tlearn: 15.5282494\ttotal: 675ms\tremaining: 280ms\n",
      "707:\tlearn: 15.4869307\ttotal: 676ms\tremaining: 279ms\n",
      "708:\tlearn: 15.4644509\ttotal: 677ms\tremaining: 278ms\n",
      "709:\tlearn: 15.4380354\ttotal: 678ms\tremaining: 277ms\n",
      "710:\tlearn: 15.4270000\ttotal: 679ms\tremaining: 276ms\n",
      "711:\tlearn: 15.4189790\ttotal: 680ms\tremaining: 275ms\n",
      "712:\tlearn: 15.4178505\ttotal: 680ms\tremaining: 274ms\n",
      "713:\tlearn: 15.3923931\ttotal: 681ms\tremaining: 273ms\n",
      "714:\tlearn: 15.3478757\ttotal: 682ms\tremaining: 272ms\n",
      "715:\tlearn: 15.3082102\ttotal: 683ms\tremaining: 271ms\n",
      "716:\tlearn: 15.2968684\ttotal: 685ms\tremaining: 270ms\n",
      "717:\tlearn: 15.2819067\ttotal: 686ms\tremaining: 269ms\n",
      "718:\tlearn: 15.2708815\ttotal: 687ms\tremaining: 268ms\n",
      "719:\tlearn: 15.2426082\ttotal: 688ms\tremaining: 267ms\n",
      "720:\tlearn: 15.2010081\ttotal: 689ms\tremaining: 266ms\n",
      "721:\tlearn: 15.1757935\ttotal: 690ms\tremaining: 266ms\n",
      "722:\tlearn: 15.1496573\ttotal: 691ms\tremaining: 265ms\n",
      "723:\tlearn: 15.1069891\ttotal: 691ms\tremaining: 264ms\n",
      "724:\tlearn: 15.0957545\ttotal: 692ms\tremaining: 263ms\n",
      "725:\tlearn: 15.0678419\ttotal: 693ms\tremaining: 262ms\n",
      "726:\tlearn: 15.0323543\ttotal: 694ms\tremaining: 261ms\n",
      "727:\tlearn: 15.0099373\ttotal: 695ms\tremaining: 260ms\n",
      "728:\tlearn: 14.9804890\ttotal: 696ms\tremaining: 259ms\n",
      "729:\tlearn: 14.9526908\ttotal: 698ms\tremaining: 258ms\n",
      "730:\tlearn: 14.9302848\ttotal: 699ms\tremaining: 257ms\n",
      "731:\tlearn: 14.8966998\ttotal: 700ms\tremaining: 256ms\n",
      "732:\tlearn: 14.8699435\ttotal: 701ms\tremaining: 255ms\n",
      "733:\tlearn: 14.8304247\ttotal: 702ms\tremaining: 254ms\n",
      "734:\tlearn: 14.7929834\ttotal: 703ms\tremaining: 253ms\n",
      "735:\tlearn: 14.7696209\ttotal: 704ms\tremaining: 253ms\n",
      "736:\tlearn: 14.7562766\ttotal: 705ms\tremaining: 252ms\n",
      "737:\tlearn: 14.7213105\ttotal: 706ms\tremaining: 251ms\n",
      "738:\tlearn: 14.7000863\ttotal: 707ms\tremaining: 250ms\n",
      "739:\tlearn: 14.6493550\ttotal: 708ms\tremaining: 249ms\n",
      "740:\tlearn: 14.6335162\ttotal: 709ms\tremaining: 248ms\n",
      "741:\tlearn: 14.5998758\ttotal: 710ms\tremaining: 247ms\n",
      "742:\tlearn: 14.5869238\ttotal: 711ms\tremaining: 246ms\n",
      "743:\tlearn: 14.5759591\ttotal: 712ms\tremaining: 245ms\n",
      "744:\tlearn: 14.5547290\ttotal: 714ms\tremaining: 244ms\n",
      "745:\tlearn: 14.5283070\ttotal: 715ms\tremaining: 243ms\n",
      "746:\tlearn: 14.4853954\ttotal: 716ms\tremaining: 242ms\n",
      "747:\tlearn: 14.4757814\ttotal: 717ms\tremaining: 241ms\n",
      "748:\tlearn: 14.4354938\ttotal: 718ms\tremaining: 241ms\n",
      "749:\tlearn: 14.4036861\ttotal: 719ms\tremaining: 240ms\n",
      "750:\tlearn: 14.3687833\ttotal: 720ms\tremaining: 239ms\n",
      "751:\tlearn: 14.3437669\ttotal: 721ms\tremaining: 238ms\n",
      "752:\tlearn: 14.3209004\ttotal: 722ms\tremaining: 237ms\n",
      "753:\tlearn: 14.2981607\ttotal: 723ms\tremaining: 236ms\n",
      "754:\tlearn: 14.2959300\ttotal: 724ms\tremaining: 235ms\n",
      "755:\tlearn: 14.2680160\ttotal: 724ms\tremaining: 234ms\n",
      "756:\tlearn: 14.2535389\ttotal: 725ms\tremaining: 233ms\n",
      "757:\tlearn: 14.2501520\ttotal: 727ms\tremaining: 232ms\n",
      "758:\tlearn: 14.2260538\ttotal: 728ms\tremaining: 231ms\n",
      "759:\tlearn: 14.2234442\ttotal: 729ms\tremaining: 230ms\n",
      "760:\tlearn: 14.1873135\ttotal: 730ms\tremaining: 229ms\n",
      "761:\tlearn: 14.1765115\ttotal: 731ms\tremaining: 228ms\n",
      "762:\tlearn: 14.1697434\ttotal: 731ms\tremaining: 227ms\n",
      "763:\tlearn: 14.1365525\ttotal: 732ms\tremaining: 226ms\n",
      "764:\tlearn: 14.1015278\ttotal: 733ms\tremaining: 225ms\n",
      "765:\tlearn: 14.0622675\ttotal: 734ms\tremaining: 224ms\n",
      "766:\tlearn: 14.0409553\ttotal: 735ms\tremaining: 223ms\n",
      "767:\tlearn: 14.0337570\ttotal: 736ms\tremaining: 222ms\n",
      "768:\tlearn: 13.9910329\ttotal: 737ms\tremaining: 221ms\n",
      "769:\tlearn: 13.9510113\ttotal: 738ms\tremaining: 220ms\n",
      "770:\tlearn: 13.9174679\ttotal: 739ms\tremaining: 220ms\n",
      "771:\tlearn: 13.8864002\ttotal: 740ms\tremaining: 219ms\n",
      "772:\tlearn: 13.8446475\ttotal: 741ms\tremaining: 218ms\n",
      "773:\tlearn: 13.8397236\ttotal: 742ms\tremaining: 217ms\n",
      "774:\tlearn: 13.8029820\ttotal: 743ms\tremaining: 216ms\n",
      "775:\tlearn: 13.7857820\ttotal: 744ms\tremaining: 215ms\n",
      "776:\tlearn: 13.7646024\ttotal: 745ms\tremaining: 214ms\n",
      "777:\tlearn: 13.7486851\ttotal: 746ms\tremaining: 213ms\n",
      "778:\tlearn: 13.7014327\ttotal: 747ms\tremaining: 212ms\n",
      "779:\tlearn: 13.6934358\ttotal: 747ms\tremaining: 211ms\n",
      "780:\tlearn: 13.6683386\ttotal: 748ms\tremaining: 210ms\n",
      "781:\tlearn: 13.6566338\ttotal: 749ms\tremaining: 209ms\n",
      "782:\tlearn: 13.6444486\ttotal: 750ms\tremaining: 208ms\n",
      "783:\tlearn: 13.6250279\ttotal: 751ms\tremaining: 207ms\n",
      "784:\tlearn: 13.5875176\ttotal: 752ms\tremaining: 206ms\n",
      "785:\tlearn: 13.5783446\ttotal: 752ms\tremaining: 205ms\n",
      "786:\tlearn: 13.5576008\ttotal: 753ms\tremaining: 204ms\n",
      "787:\tlearn: 13.5477919\ttotal: 754ms\tremaining: 203ms\n",
      "788:\tlearn: 13.5055189\ttotal: 755ms\tremaining: 202ms\n",
      "789:\tlearn: 13.4644475\ttotal: 756ms\tremaining: 201ms\n",
      "790:\tlearn: 13.4570524\ttotal: 757ms\tremaining: 200ms\n",
      "791:\tlearn: 13.4244055\ttotal: 758ms\tremaining: 199ms\n",
      "792:\tlearn: 13.4171601\ttotal: 759ms\tremaining: 198ms\n",
      "793:\tlearn: 13.3977292\ttotal: 760ms\tremaining: 197ms\n",
      "794:\tlearn: 13.3911146\ttotal: 761ms\tremaining: 196ms\n",
      "795:\tlearn: 13.3816584\ttotal: 762ms\tremaining: 195ms\n",
      "796:\tlearn: 13.3539009\ttotal: 763ms\tremaining: 194ms\n",
      "797:\tlearn: 13.3166557\ttotal: 764ms\tremaining: 193ms\n",
      "798:\tlearn: 13.2942330\ttotal: 765ms\tremaining: 192ms\n",
      "799:\tlearn: 13.2692862\ttotal: 765ms\tremaining: 191ms\n",
      "800:\tlearn: 13.2547553\ttotal: 766ms\tremaining: 190ms\n",
      "801:\tlearn: 13.2259343\ttotal: 767ms\tremaining: 189ms\n",
      "802:\tlearn: 13.2246521\ttotal: 768ms\tremaining: 188ms\n",
      "803:\tlearn: 13.2119291\ttotal: 769ms\tremaining: 188ms\n",
      "804:\tlearn: 13.1793474\ttotal: 771ms\tremaining: 187ms\n",
      "805:\tlearn: 13.1652542\ttotal: 772ms\tremaining: 186ms\n",
      "806:\tlearn: 13.1293534\ttotal: 773ms\tremaining: 185ms\n",
      "807:\tlearn: 13.1211653\ttotal: 774ms\tremaining: 184ms\n",
      "808:\tlearn: 13.1110732\ttotal: 775ms\tremaining: 183ms\n",
      "809:\tlearn: 13.1056131\ttotal: 776ms\tremaining: 182ms\n",
      "810:\tlearn: 13.0704954\ttotal: 776ms\tremaining: 181ms\n",
      "811:\tlearn: 13.0482918\ttotal: 777ms\tremaining: 180ms\n",
      "812:\tlearn: 13.0313717\ttotal: 778ms\tremaining: 179ms\n",
      "813:\tlearn: 13.0295572\ttotal: 779ms\tremaining: 178ms\n",
      "814:\tlearn: 13.0076548\ttotal: 780ms\tremaining: 177ms\n",
      "815:\tlearn: 13.0056559\ttotal: 781ms\tremaining: 176ms\n",
      "816:\tlearn: 12.9992709\ttotal: 782ms\tremaining: 175ms\n",
      "817:\tlearn: 12.9824007\ttotal: 783ms\tremaining: 174ms\n",
      "818:\tlearn: 12.9670734\ttotal: 784ms\tremaining: 173ms\n",
      "819:\tlearn: 12.9372833\ttotal: 785ms\tremaining: 172ms\n",
      "820:\tlearn: 12.9087596\ttotal: 786ms\tremaining: 171ms\n",
      "821:\tlearn: 12.9075563\ttotal: 787ms\tremaining: 170ms\n",
      "822:\tlearn: 12.8694642\ttotal: 788ms\tremaining: 169ms\n",
      "823:\tlearn: 12.8683302\ttotal: 789ms\tremaining: 168ms\n",
      "824:\tlearn: 12.8490869\ttotal: 790ms\tremaining: 168ms\n",
      "825:\tlearn: 12.8290022\ttotal: 791ms\tremaining: 167ms\n",
      "826:\tlearn: 12.8107471\ttotal: 792ms\tremaining: 166ms\n",
      "827:\tlearn: 12.7834375\ttotal: 793ms\tremaining: 165ms\n",
      "828:\tlearn: 12.7492612\ttotal: 794ms\tremaining: 164ms\n",
      "829:\tlearn: 12.7143908\ttotal: 795ms\tremaining: 163ms\n",
      "830:\tlearn: 12.6794802\ttotal: 796ms\tremaining: 162ms\n",
      "831:\tlearn: 12.6575840\ttotal: 796ms\tremaining: 161ms\n",
      "832:\tlearn: 12.6414062\ttotal: 797ms\tremaining: 160ms\n",
      "833:\tlearn: 12.6323550\ttotal: 798ms\tremaining: 159ms\n",
      "834:\tlearn: 12.6004844\ttotal: 799ms\tremaining: 158ms\n",
      "835:\tlearn: 12.5992796\ttotal: 800ms\tremaining: 157ms\n",
      "836:\tlearn: 12.5664653\ttotal: 801ms\tremaining: 156ms\n",
      "837:\tlearn: 12.5389714\ttotal: 802ms\tremaining: 155ms\n",
      "838:\tlearn: 12.5116178\ttotal: 803ms\tremaining: 154ms\n",
      "839:\tlearn: 12.4901320\ttotal: 803ms\tremaining: 153ms\n",
      "840:\tlearn: 12.4812527\ttotal: 805ms\tremaining: 152ms\n",
      "841:\tlearn: 12.4562707\ttotal: 805ms\tremaining: 151ms\n",
      "842:\tlearn: 12.4209878\ttotal: 806ms\tremaining: 150ms\n",
      "843:\tlearn: 12.4015992\ttotal: 807ms\tremaining: 149ms\n",
      "844:\tlearn: 12.3675998\ttotal: 808ms\tremaining: 148ms\n",
      "845:\tlearn: 12.3521691\ttotal: 809ms\tremaining: 147ms\n",
      "846:\tlearn: 12.3426886\ttotal: 810ms\tremaining: 146ms\n",
      "847:\tlearn: 12.3188571\ttotal: 811ms\tremaining: 145ms\n",
      "848:\tlearn: 12.2937388\ttotal: 812ms\tremaining: 144ms\n",
      "849:\tlearn: 12.2724402\ttotal: 813ms\tremaining: 143ms\n",
      "850:\tlearn: 12.2342559\ttotal: 814ms\tremaining: 142ms\n",
      "851:\tlearn: 12.2159725\ttotal: 815ms\tremaining: 142ms\n",
      "852:\tlearn: 12.2147289\ttotal: 816ms\tremaining: 141ms\n",
      "853:\tlearn: 12.1935801\ttotal: 817ms\tremaining: 140ms\n",
      "854:\tlearn: 12.1795597\ttotal: 818ms\tremaining: 139ms\n",
      "855:\tlearn: 12.1489328\ttotal: 819ms\tremaining: 138ms\n",
      "856:\tlearn: 12.1476791\ttotal: 820ms\tremaining: 137ms\n",
      "857:\tlearn: 12.1143878\ttotal: 821ms\tremaining: 136ms\n",
      "858:\tlearn: 12.1133877\ttotal: 821ms\tremaining: 135ms\n",
      "859:\tlearn: 12.0927666\ttotal: 822ms\tremaining: 134ms\n",
      "860:\tlearn: 12.0908934\ttotal: 823ms\tremaining: 133ms\n",
      "861:\tlearn: 12.0667739\ttotal: 824ms\tremaining: 132ms\n",
      "862:\tlearn: 12.0312513\ttotal: 825ms\tremaining: 131ms\n",
      "863:\tlearn: 12.0298478\ttotal: 826ms\tremaining: 130ms\n",
      "864:\tlearn: 12.0174484\ttotal: 827ms\tremaining: 129ms\n",
      "865:\tlearn: 12.0093901\ttotal: 828ms\tremaining: 128ms\n",
      "866:\tlearn: 11.9731302\ttotal: 829ms\tremaining: 127ms\n",
      "867:\tlearn: 11.9617694\ttotal: 830ms\tremaining: 126ms\n",
      "868:\tlearn: 11.9536413\ttotal: 831ms\tremaining: 125ms\n",
      "869:\tlearn: 11.9234807\ttotal: 832ms\tremaining: 124ms\n",
      "870:\tlearn: 11.8995296\ttotal: 833ms\tremaining: 123ms\n",
      "871:\tlearn: 11.8985399\ttotal: 833ms\tremaining: 122ms\n",
      "872:\tlearn: 11.8748993\ttotal: 834ms\tremaining: 121ms\n",
      "873:\tlearn: 11.8473937\ttotal: 835ms\tremaining: 120ms\n",
      "874:\tlearn: 11.7980081\ttotal: 836ms\tremaining: 119ms\n",
      "875:\tlearn: 11.7968597\ttotal: 837ms\tremaining: 119ms\n",
      "876:\tlearn: 11.7952536\ttotal: 838ms\tremaining: 118ms\n",
      "877:\tlearn: 11.7765712\ttotal: 839ms\tremaining: 117ms\n",
      "878:\tlearn: 11.7433623\ttotal: 840ms\tremaining: 116ms\n",
      "879:\tlearn: 11.7220227\ttotal: 841ms\tremaining: 115ms\n",
      "880:\tlearn: 11.7079171\ttotal: 842ms\tremaining: 114ms\n",
      "881:\tlearn: 11.6892547\ttotal: 843ms\tremaining: 113ms\n",
      "882:\tlearn: 11.6752283\ttotal: 844ms\tremaining: 112ms\n",
      "883:\tlearn: 11.6465248\ttotal: 845ms\tremaining: 111ms\n",
      "884:\tlearn: 11.6199604\ttotal: 846ms\tremaining: 110ms\n",
      "885:\tlearn: 11.5840290\ttotal: 847ms\tremaining: 109ms\n",
      "886:\tlearn: 11.5665343\ttotal: 848ms\tremaining: 108ms\n",
      "887:\tlearn: 11.5653334\ttotal: 849ms\tremaining: 107ms\n",
      "888:\tlearn: 11.5409858\ttotal: 850ms\tremaining: 106ms\n",
      "889:\tlearn: 11.5126963\ttotal: 850ms\tremaining: 105ms\n",
      "890:\tlearn: 11.4937102\ttotal: 851ms\tremaining: 104ms\n",
      "891:\tlearn: 11.4652907\ttotal: 852ms\tremaining: 103ms\n",
      "892:\tlearn: 11.4406454\ttotal: 853ms\tremaining: 102ms\n",
      "893:\tlearn: 11.4157600\ttotal: 854ms\tremaining: 101ms\n",
      "894:\tlearn: 11.4145225\ttotal: 855ms\tremaining: 100ms\n",
      "895:\tlearn: 11.3944764\ttotal: 856ms\tremaining: 99.4ms\n",
      "896:\tlearn: 11.3718753\ttotal: 857ms\tremaining: 98.4ms\n",
      "897:\tlearn: 11.3698629\ttotal: 858ms\tremaining: 97.5ms\n",
      "898:\tlearn: 11.3652583\ttotal: 859ms\tremaining: 96.5ms\n",
      "899:\tlearn: 11.3403921\ttotal: 860ms\tremaining: 95.6ms\n",
      "900:\tlearn: 11.3188723\ttotal: 861ms\tremaining: 94.6ms\n",
      "901:\tlearn: 11.2825028\ttotal: 862ms\tremaining: 93.6ms\n",
      "902:\tlearn: 11.2756316\ttotal: 863ms\tremaining: 92.7ms\n",
      "903:\tlearn: 11.2746247\ttotal: 864ms\tremaining: 91.7ms\n",
      "904:\tlearn: 11.2506409\ttotal: 864ms\tremaining: 90.7ms\n",
      "905:\tlearn: 11.2459222\ttotal: 865ms\tremaining: 89.8ms\n",
      "906:\tlearn: 11.2231937\ttotal: 866ms\tremaining: 88.8ms\n",
      "907:\tlearn: 11.2191230\ttotal: 867ms\tremaining: 87.8ms\n",
      "908:\tlearn: 11.1878231\ttotal: 868ms\tremaining: 86.9ms\n",
      "909:\tlearn: 11.1635473\ttotal: 869ms\tremaining: 86ms\n",
      "910:\tlearn: 11.1593446\ttotal: 870ms\tremaining: 85ms\n",
      "911:\tlearn: 11.1312444\ttotal: 871ms\tremaining: 84.1ms\n",
      "912:\tlearn: 11.1178570\ttotal: 872ms\tremaining: 83.1ms\n",
      "913:\tlearn: 11.1042827\ttotal: 873ms\tremaining: 82.1ms\n",
      "914:\tlearn: 11.0835715\ttotal: 874ms\tremaining: 81.2ms\n",
      "915:\tlearn: 11.0643196\ttotal: 875ms\tremaining: 80.2ms\n",
      "916:\tlearn: 11.0602889\ttotal: 876ms\tremaining: 79.3ms\n",
      "917:\tlearn: 11.0272219\ttotal: 877ms\tremaining: 78.3ms\n",
      "918:\tlearn: 10.9986897\ttotal: 878ms\tremaining: 77.3ms\n",
      "919:\tlearn: 10.9792479\ttotal: 879ms\tremaining: 76.4ms\n",
      "920:\tlearn: 10.9535501\ttotal: 879ms\tremaining: 75.4ms\n",
      "921:\tlearn: 10.9275044\ttotal: 880ms\tremaining: 74.5ms\n",
      "922:\tlearn: 10.9219224\ttotal: 881ms\tremaining: 73.5ms\n",
      "923:\tlearn: 10.9035961\ttotal: 882ms\tremaining: 72.6ms\n",
      "924:\tlearn: 10.8891537\ttotal: 884ms\tremaining: 71.6ms\n",
      "925:\tlearn: 10.8702443\ttotal: 885ms\tremaining: 70.7ms\n",
      "926:\tlearn: 10.8421716\ttotal: 886ms\tremaining: 69.7ms\n",
      "927:\tlearn: 10.8203110\ttotal: 886ms\tremaining: 68.8ms\n",
      "928:\tlearn: 10.7915801\ttotal: 888ms\tremaining: 67.8ms\n",
      "929:\tlearn: 10.7652304\ttotal: 888ms\tremaining: 66.9ms\n",
      "930:\tlearn: 10.7342804\ttotal: 889ms\tremaining: 65.9ms\n",
      "931:\tlearn: 10.7211866\ttotal: 890ms\tremaining: 65ms\n",
      "932:\tlearn: 10.7047031\ttotal: 891ms\tremaining: 64ms\n",
      "933:\tlearn: 10.6961734\ttotal: 892ms\tremaining: 63ms\n",
      "934:\tlearn: 10.6841913\ttotal: 893ms\tremaining: 62.1ms\n",
      "935:\tlearn: 10.6763968\ttotal: 894ms\tremaining: 61.1ms\n",
      "936:\tlearn: 10.6645409\ttotal: 895ms\tremaining: 60.2ms\n",
      "937:\tlearn: 10.6631696\ttotal: 896ms\tremaining: 59.2ms\n",
      "938:\tlearn: 10.6586104\ttotal: 896ms\tremaining: 58.2ms\n",
      "939:\tlearn: 10.6395476\ttotal: 897ms\tremaining: 57.3ms\n",
      "940:\tlearn: 10.6354434\ttotal: 898ms\tremaining: 56.3ms\n",
      "941:\tlearn: 10.6184324\ttotal: 899ms\tremaining: 55.4ms\n",
      "942:\tlearn: 10.6055869\ttotal: 900ms\tremaining: 54.4ms\n",
      "943:\tlearn: 10.5965561\ttotal: 901ms\tremaining: 53.5ms\n",
      "944:\tlearn: 10.5955793\ttotal: 902ms\tremaining: 52.5ms\n",
      "945:\tlearn: 10.5947611\ttotal: 903ms\tremaining: 51.6ms\n",
      "946:\tlearn: 10.5693144\ttotal: 904ms\tremaining: 50.6ms\n",
      "947:\tlearn: 10.5637694\ttotal: 905ms\tremaining: 49.6ms\n",
      "948:\tlearn: 10.5374266\ttotal: 906ms\tremaining: 48.7ms\n",
      "949:\tlearn: 10.5099668\ttotal: 907ms\tremaining: 47.7ms\n",
      "950:\tlearn: 10.5086617\ttotal: 908ms\tremaining: 46.8ms\n",
      "951:\tlearn: 10.5067682\ttotal: 909ms\tremaining: 45.8ms\n",
      "952:\tlearn: 10.4925699\ttotal: 909ms\tremaining: 44.9ms\n",
      "953:\tlearn: 10.4716689\ttotal: 911ms\tremaining: 43.9ms\n",
      "954:\tlearn: 10.4379081\ttotal: 912ms\tremaining: 43ms\n",
      "955:\tlearn: 10.4006751\ttotal: 913ms\tremaining: 42ms\n",
      "956:\tlearn: 10.3836660\ttotal: 913ms\tremaining: 41ms\n",
      "957:\tlearn: 10.3575722\ttotal: 915ms\tremaining: 40.1ms\n",
      "958:\tlearn: 10.3478138\ttotal: 916ms\tremaining: 39.1ms\n",
      "959:\tlearn: 10.3187226\ttotal: 916ms\tremaining: 38.2ms\n",
      "960:\tlearn: 10.3022324\ttotal: 918ms\tremaining: 37.2ms\n",
      "961:\tlearn: 10.2997720\ttotal: 919ms\tremaining: 36.3ms\n",
      "962:\tlearn: 10.2989286\ttotal: 920ms\tremaining: 35.3ms\n",
      "963:\tlearn: 10.2895343\ttotal: 921ms\tremaining: 34.4ms\n",
      "964:\tlearn: 10.2815641\ttotal: 922ms\tremaining: 33.4ms\n",
      "965:\tlearn: 10.2625952\ttotal: 923ms\tremaining: 32.5ms\n",
      "966:\tlearn: 10.2506695\ttotal: 924ms\tremaining: 31.5ms\n",
      "967:\tlearn: 10.2397609\ttotal: 925ms\tremaining: 30.6ms\n",
      "968:\tlearn: 10.2154145\ttotal: 927ms\tremaining: 29.6ms\n",
      "969:\tlearn: 10.2143644\ttotal: 928ms\tremaining: 28.7ms\n",
      "970:\tlearn: 10.1981073\ttotal: 929ms\tremaining: 27.7ms\n",
      "971:\tlearn: 10.1920221\ttotal: 930ms\tremaining: 26.8ms\n",
      "972:\tlearn: 10.1806991\ttotal: 931ms\tremaining: 25.8ms\n",
      "973:\tlearn: 10.1597212\ttotal: 933ms\tremaining: 24.9ms\n",
      "974:\tlearn: 10.1358828\ttotal: 934ms\tremaining: 23.9ms\n",
      "975:\tlearn: 10.1222784\ttotal: 935ms\tremaining: 23ms\n",
      "976:\tlearn: 10.1128777\ttotal: 936ms\tremaining: 22ms\n",
      "977:\tlearn: 10.1065225\ttotal: 937ms\tremaining: 21.1ms\n",
      "978:\tlearn: 10.0954545\ttotal: 938ms\tremaining: 20.1ms\n",
      "979:\tlearn: 10.0725280\ttotal: 939ms\tremaining: 19.2ms\n",
      "980:\tlearn: 10.0443783\ttotal: 940ms\tremaining: 18.2ms\n",
      "981:\tlearn: 10.0421942\ttotal: 941ms\tremaining: 17.2ms\n",
      "982:\tlearn: 10.0135134\ttotal: 941ms\tremaining: 16.3ms\n",
      "983:\tlearn: 10.0009906\ttotal: 942ms\tremaining: 15.3ms\n",
      "984:\tlearn: 9.9894585\ttotal: 943ms\tremaining: 14.4ms\n",
      "985:\tlearn: 9.9872832\ttotal: 944ms\tremaining: 13.4ms\n",
      "986:\tlearn: 9.9654572\ttotal: 945ms\tremaining: 12.4ms\n",
      "987:\tlearn: 9.9461931\ttotal: 946ms\tremaining: 11.5ms\n",
      "988:\tlearn: 9.9451290\ttotal: 947ms\tremaining: 10.5ms\n",
      "989:\tlearn: 9.9431210\ttotal: 948ms\tremaining: 9.57ms\n",
      "990:\tlearn: 9.9299945\ttotal: 949ms\tremaining: 8.62ms\n",
      "991:\tlearn: 9.9131150\ttotal: 950ms\tremaining: 7.66ms\n",
      "992:\tlearn: 9.9123473\ttotal: 950ms\tremaining: 6.7ms\n",
      "993:\tlearn: 9.8978089\ttotal: 951ms\tremaining: 5.74ms\n",
      "994:\tlearn: 9.8841806\ttotal: 952ms\tremaining: 4.79ms\n",
      "995:\tlearn: 9.8638563\ttotal: 953ms\tremaining: 3.83ms\n",
      "996:\tlearn: 9.8491248\ttotal: 954ms\tremaining: 2.87ms\n",
      "997:\tlearn: 9.8427010\ttotal: 955ms\tremaining: 1.91ms\n",
      "998:\tlearn: 9.8417925\ttotal: 956ms\tremaining: 957us\n",
      "999:\tlearn: 9.8380086\ttotal: 957ms\tremaining: 0us\n",
      "Mean absolute error: 43.06\n",
      "Mean absolute percentage error: 39.39%\n",
      "Root Mean squared error: 55.14\n",
      "Coefficient of determination: 0.45\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhwAAAIjCAYAAABI0sIEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAACF90lEQVR4nO3dd1xTV/8H8E8IWwRElpII7olaB5T6WHHU1UGL1r3psLV9tNbW2mXtstNqnw473C3WAdY+dloNjqq4d10UFFEUB3smOb8//JHHSIAEktyMz/v1yuslyc3NN9cL53vPPed7ZEIIASIiIiILcpE6ACIiInJ8TDiIiIjI4phwEBERkcUx4SAiIiKLY8JBREREFseEg4iIiCyOCQcRERFZHBMOIiIisjgmHERERGRxTDiIyCYtX74cMpkMGRkZuudiY2MRGxsrWUx3MhSjNaSkpEAmkyElJcWqn0tUH0w4iO4gk8mMejj6H/uIiAi97xscHIzevXtjw4YNUodmkuLiYrzxxhuS/X917twZzZo1Q02rSPTq1QshISFQq9VWjIzIulylDoDI1qxatUrv55UrV2Lz5s1Vnm/fvr01w5JE165d8fzzzwMALl26hK+++grx8fH48ssvMXXqVKvH88cff5j8nuLiYsybNw8AJOkdGTt2LF566SXs2LED9957b5XXMzIysHv3bjzzzDNwdeWfZHJcPLuJ7jBu3Di9n/fs2YPNmzdXef5OxcXF8Pb2tmRoVhcWFqb3vSdMmIBWrVrhk08+qTbhUKvV0Gq1cHd3N3s8ltinpY0ZMwZz5sxBYmKiwYRj9erVEEJg7NixEkRHZD28pUJUB7GxsejUqRMOHDiAe++9F97e3nj55ZcB3Lol88Ybb1R5T0REBCZNmqT3XG5uLmbMmAGlUgkPDw+0atUK77//PrRabY2f/8ADD6BFixYGX4uJiUGPHj10P2/evBn/+te/4O/vDx8fH7Rt21YXq6lCQ0PRvn17pKenA7h1dS6TyfDRRx9h4cKFaNmyJTw8PHDy5EkAwKlTpzB8+HAEBATA09MTPXr0wE8//VRlvydOnEC/fv3g5eUFhUKBt99+2+AxMDSGo7S0FG+88QbatGkDT09PNGnSBPHx8UhLS0NGRgaCgoIAAPPmzdPdHrr9/8fcMd5JqVTi3nvvxfr161FRUVHl9cTERLRs2RLR0dE4f/48nn76abRt2xZeXl5o3LgxHn30UaPGiBg6vwDDx6ysrAxz585Fq1at4OHhAaVSiRdffBFlZWV625nz3CFiDwdRHV2/fh1DhgzBqFGjMG7cOISEhJj0/uLiYvTp0wdZWVl48skn0axZM+zatQtz5szB5cuXsXDhwmrfO3LkSEyYMAH79u1Dz549dc+fP38ee/bswYcffgjgViP5wAMPoHPnznjzzTfh4eGBc+fO4a+//qrTd66oqEBmZiYaN26s9/yyZctQWlqKJ554Ah4eHggICMCJEyfQq1cvhIWF4aWXXkKDBg2wdu1aPPzww0hKSsIjjzwCAMjOzkbfvn2hVqt123399dfw8vKqNR6NRoMHHngAW7ZswahRozB9+nQUFBRg8+bNOH78OAYMGIAvv/wSTz31FB555BHEx8cDuDWuovL4WDpG4NZtlSeeeAK///47HnjgAd3zx44dw/Hjx/H6668DAPbt24ddu3Zh1KhRUCgUyMjIwJdffonY2FicPHnSLD1oWq0WDz30EHbu3IknnngC7du3x7Fjx/DJJ5/gzJkz+PHHH3XHxpznDhEEEdVo2rRp4s5flT59+ggAYvHixVW2ByDmzp1b5fnw8HAxceJE3c9vvfWWaNCggThz5ozedi+99JKQy+XiwoUL1caUl5cnPDw8xPPPP6/3/AcffCBkMpk4f/68EEKITz75RAAQOTk5tX1Ng/EOHDhQ5OTkiJycHHHkyBExatQoAUA8++yzQggh0tPTBQDh6+srrl69qvf+/v37i8jISFFaWqp7TqvVinvuuUe0bt1a99yMGTMEAJGamqp77urVq8LPz08AEOnp6brn+/TpI/r06aP7eenSpQKAWLBgQZX4tVqtEEKInJycav9PLBGjITdu3BAeHh5i9OjRes+/9NJLAoA4ffq0EEKI4uLiKu/dvXu3ACBWrlype06lUgkAQqVS6Z678/yqdOcxW7VqlXBxcRE7duzQ227x4sUCgPjrr7+EEPU7d4gM4S0Vojry8PDA5MmT6/z+devWoXfv3mjUqBGuXbumewwYMAAajQbbt2+v9r2+vr4YMmQI1q5dqzf7Yc2aNbj77rvRrFkzAIC/vz8AYOPGjUZ1/9/pjz/+QFBQEIKCgtClSxesW7cO48ePx/vvv6+33bBhw3S3LgDgxo0b2Lp1K0aMGIGCggLdd7t+/ToGDRqEs2fPIisrCwDwyy+/4O6770ZUVJTu/UFBQUaNaUhKSkJgYCCeffbZKq/JZLIa32utGAGgUaNGGDp0KH766ScUFRUBAIQQ+OGHH9CjRw+0adMGAPR6TCoqKnD9+nW0atUK/v7+OHjwoFGfVZt169ahffv2aNeund55169fPwCASqUCUP9zh+hOTDiI6igsLKxegxjPnj2L3377TdegVz4GDBgAALh69WqN7x85ciQyMzOxe/duAEBaWhoOHDiAkSNH6m3Tq1cvPPbYYwgJCcGoUaOwdu1aoxuQ6OhobN68GX/++Sd27dqFa9euYeXKlVVuJTRv3lzv53PnzkEIgddee63K95s7d67e9zt//jxat25d5bPbtm1ba3xpaWlo27ZtnWZ3WCvGSmPHjkVRURE2btwIANi1axcyMjL0kpaSkhK8/vrrujE9gYGBCAoKQm5uLvLy8kz+joacPXsWJ06cqPKdK5Oeyu9c33OH6E4cw0FUR8bev6+k0Wj0ftZqtbjvvvvw4osvGty+sgGozoMPPghvb2+sXbsW99xzD9auXQsXFxc8+uijejFu374dKpUKP//8M3777TesWbMG/fr1wx9//AG5XF7jZwQGBuoSoJrceSwqG6VZs2Zh0KBBBt/TqlWrWvdrSdaO8YEHHoCfnx8SExMxZswYJCYmQi6XY9SoUbptnn32WSxbtgwzZsxATEwM/Pz8IJPJMGrUqFob+up6dDQajd7/s1arRWRkJBYsWGBwe6VSCaD+5w7RnZhwEJlZo0aNkJubq/dceXk5Ll++rPdcy5YtUVhYaFSDbkiDBg3wwAMPYN26dViwYAHWrFmD3r17o2nTpnrbubi4oH///ujfvz8WLFiAd999F6+88gpUKlWdP7s2lTNo3Nzcav2M8PBwnD17tsrzp0+frvVzWrZsidTUVFRUVMDNzc3gNtU1xNaKsZKHhweGDx+OlStX4sqVK1i3bh369euH0NBQ3Tbr16/HxIkT8fHHH+ueKy0trXI+GWLovANu9c7cPqOpZcuWOHLkCPr371/rbScpzh1yXLylQmRmLVu2rDL+4uuvv67SwzFixAjs3r0bv//+e5V95ObmGlV1cuTIkbh06RK+/fZbHDlyRO92CnBrnMKdunbtCgBVpkCaU3BwMGJjY/HVV19VSbQAICcnR/fvoUOHYs+ePdi7d6/e699//32tnzNs2DBcu3YNn332WZXXKse2VM7suLMxtlaMtxs7diwqKirw5JNPIicnp8oYELlcXqUi6X/+858q544hLVu2xJ49e1BeXq57btOmTcjMzNTbbsSIEcjKysI333xTZR8lJSW6MSZSnTvkuNjDQWRmjz32GKZOnYphw4bhvvvuw5EjR/D7778jMDBQb7sXXngBP/30Ex544AFMmjQJ3bt3R1FREY4dO4b169cjIyOjynvuNHToUDRs2BCzZs2CXC7HsGHD9F5/8803sX37dtx///0IDw/H1atX8cUXX0ChUOBf//qX2b/77T7//HP861//QmRkJB5//HG0aNECV65cwe7du3Hx4kUcOXIEAPDiiy9i1apVGDx4MKZPn66bchoeHo6jR4/W+BkTJkzAypUrMXPmTOzduxe9e/dGUVER/vzzTzz99NOIi4uDl5cXOnTogDVr1qBNmzYICAhAp06d0KlTJ6vEeLs+ffpAoVBg48aN8PLy0k3TrfTAAw9g1apV8PPzQ4cOHbB79278+eefVaYhG/LYY49h/fr1GDx4MEaMGIG0tDR89913aNmypd5248ePx9q1azF16lSoVCr06tULGo0Gp06dwtq1a/H777+jR48ekp475KCknCJDZA+qmxbbsWNHg9trNBoxe/ZsERgYKLy9vcWgQYPEuXPnDE5bLCgoEHPmzBGtWrUS7u7uIjAwUNxzzz3io48+EuXl5UbFN3bsWAFADBgwoMprW7ZsEXFxcaJp06bC3d1dNG3aVIwePbrKVFxDwsPDxf3331/jNpXTYj/88EODr6elpYkJEyaI0NBQ4ebmJsLCwsQDDzwg1q9fr7fd0aNHRZ8+fYSnp6cICwsTb731lliyZEmt02KFuDWV9JVXXhHNmzcXbm5uIjQ0VAwfPlykpaXpttm1a5fo3r27cHd3rzJF1twx1uaFF14QAMSIESOqvHbz5k0xefJkERgYKHx8fMSgQYPEqVOnqpw7hqbFCiHExx9/LMLCwoSHh4fo1auX2L9/v8FjVl5eLt5//33RsWNH4eHhIRo1aiS6d+8u5s2bJ/Ly8oQQ9Tt3iAyRCVHDikJEREREZsAxHERERGRxTDiIiIjI4phwEBERkcUx4SAiIiKLY8JBREREFseEg4iIiCyOhb9wa22BS5cuoWHDhrWW+iUiIqL/EUKgoKAATZs2hYtL9f0YTDgAXLp0SbdgEREREZkuMzMTCoWi2teZcABo2LAhgFsHy9fXV+JoiIiI9O1Lv4EnvjuACrUWD3Vpircf7gQXF9vokc/Pz4dSqdS1pdVhwoH/rSbp6+vLhIOIiGzK8aw8TE8+DY3cEwM7hWDhuG5wldveEMxaVx+2UhxERERkon9yCjFx6V4UlqkR3TwAn425yyaTDWPYZ9REREQO7lJuCcYv2YvrReXoFOaLbyf2gKebXOqw6owJBxERkY25UVSO8UtSkZVbghaBDbB8chQaerpJHVa9MOEgIiKyIYVlakxathdpOUVo4ueJVY9FI9DHQ+qw6o0JBxERkY0ordDgiZX7cfRiHhp5u2FVQhTC/L2kDsssmHAQERHZALVGi+k/HMKutOto4C7H8slRaBVc81RTe8KEg4iISGJCCLyy4Th+P3EF7nIXfD2hB7oo/aUOy6yYcBAREUnsvV9PYc3+TLjIgE9H34VerQKlDsnsmHAQERFJ6MuUNHy1/R8AwHvxnTG4U6jEEVkGEw4iIiKJrN57Ae//dgoA8PLQdhjR03HX9WLCQUREJIFfjl3GKxuOAQCeim2JJ+5tKXFElsWEg4iIyMp2nM3B9B8OQSuA0VHN8OKgtlKHZHFMOIiIiKzo0IWbeHLVAVRoBO6PbIK3H+5U68JnjoCrxRIREf0/jUaDHTt24PLly2jSpAl69+4Nudx865ecuVKAycv3obhcg96tA7FgZBfIbWSZeUtjwkFERAQgOTkZ06dPx8WLF3XPKRQKLFq0CPHx8fXef+aNYoxfkorc4gp0Vfpj8bju8HC138XYTMVbKkRE5PSSk5MxfPhwvWQDALKysjB8+HAkJyfXa/85BWUYvyQVV/LL0DrYB8sm9UQDD+e65mfCQURETk2j0WD69OkQQlR5rfK5GTNmQKPR1Gn/+aUVmLh0LzKuF0PRyAurEqLRqIF7vWK2R0w4iIjIqe3YsaNKz8bthBDIzMzEjh07TN53SbkGjy3fj5OX8xHo445VCdEI9fOsT7h2iwkHERE5tcuXL5t1u0oVGi2mJR7E3owbaOjhihVTotA8sEFdQnQITDiIiMipNWnSxKzbAYBWK/DCuiPYeuoqPFxdsGRST3Rs6lfXEB0CEw4iInJqvXv3hkKhqLYWhkwmg1KpRO/evY3anxACb246iR8PX4KriwxfjuuGqOYB5gzZLjHhICIipyaXy7Fo0SIAqJJ0VP68cOFCo+txLNpyFst3ZQAAPnq0C/q1CzFfsHaMCQcRETm9+Ph4rF+/HmFhYXrPKxQKrF+/3ug6HCt2ZWDhn2cBAG882AEP3xVWyzuch0wYmgfkZPLz8+Hn54e8vDz4+vpKHQ4REUmkPpVGNx7OwvQfDgMApvdvjefua2PBSG2HsW2oc1UdISIiqoFcLkdsbKzJ71Oduorn1x4BAEyMCceMAa3NHJn94y0VIiKietiXcQNTvzsAtVbg4a5NMffBjk6xGJupmHAQERHV0clL+ZiyfB/K1Fr0axeMDx/tAhcnWYzNVEw4iIiI6iDjWhEmLN2LglI1ekY0wudjusFNzma1OjwyREREJsrOK8W4Jam4VliG9k188e3EnvByd56VX+uCCQcREZEJcovLMWFpKi7eLEFEY2+snBIFPy83qcOyeUw4iIiIjFRUpsakZftw5kohQnw9sCohGkENPaQOyy4w4SAiIjJCmVqDqd8dwOHMXPh5uWFVQjSUAd5Sh2U3mHAQERHVQqMVmLnmCHacvQZvdzmWTe6JNiENpQ7LrjDhICIiqoEQAq/+eBw/H7sMN7kMX43vjm7NGkkdlt1hwkFERFSDD38/jdV7L0AmAxaOvAu9WwdJHZJdYsJBRERUjW+2/4MvUtIAAO8+Eon7OzeROCL7xYSDiIjIgLX7M/HOL38DAF4c3Bajo5pJHJF9Y8JBRER0h9+OZ+OlpKMAgCfubYGn+rSUOCL7x4SDiIjoNrvOXcO/Vx+CVgAjeigwZ0g7LsZmBkw4iIiI/t/Ri7l4fOV+lGu0GNQxBO8+Eslkw0yYcBAREQE4d7UQk5btQ1G5Bve0bIxFo+6CKxdjMxtXqQMgIiL7otFosGPHDly+fBlNmjRB7969IZcbv3BZfd9vCVm5JRi/JBU3isrRWeGHryf0gKcbF2MzJyYcRERktOTkZEyfPh0XL17UPadQKLBo0SLEx8db/P2WcL2wDOOXpOJyXilaBjXA8slR8PFg82huMiGEkDoIqeXn58PPzw95eXnw9fWVOhwiIpuUnJyM4cOH485mo3KMw/r162tMGur7fnO4s3ela88YjFu6F8ez8tHUzxPrn7oHTf29LBqDozG2DWXCASYcRES10Wg0iIiI0OuZuJ1MJoNCoUB6errB2yP1fb85VOldkbuh2bj3IAtti8YN3LF2agxaBvlY5LMdmbFtKEfDEBFRrXbs2FFtsgDcWm8kMzMTO3bssMj766uyd0UXg8wFQXGzIQttC21ZMSZGFDLZsDAmHEREVKvLly/Xa7v6vr8+NBoNpk+fftutHBkaD/k3vFvfDaEuR07yW/jolenQaDRm/2z6HyYcRERUqyZNjFtDpLrt6vv++rizd6VRvwT4RA6A0GqQs/E9lF44ZtHeFbqFCQcREdWqd+/eUCgU1RbBkslkUCqV6N27t0XeXx+395r43v0ofHs+DAC4/ssilJzba3A7Mj8mHEREVCu5XI5FixYBQJWkofLnhQsXVjvgs77vr4/KXhOfrkPQqM9EAMCNLd+g6MRWg9uRZTDhICIio8THx2P9+vUICwvTe16hUBg1pbW+76+r3r17Q9krDgEDnwIA5O36AQX7N+pet2TvCv0Pp8WC02KJiExhb5VGt53JwZRlqdAIGQoO/YIbf3yhe82aNUAclbFtKEupERGRSeRyOWJjYyV7vykOnL+BqasOQCNk6BqgwYG//4sbt72uUCiwcOFCJhtWwISDiIgc0qnsfExetg8lFRr0aROEbyb0gPz5+21uHRdnwYSDiIgczoXrxRi/ZC/yS9XoHt4IX47rBnfXW8MWrdW7Qvo4aJSIiBzK1fxSjFuSipyCMrQLbYilE3vC253X11JjwkFERA4jr7gCE5buxYUbxVAGeGHllCj4ebtJHRaBt1SIiMhBlJRrkLBiH05lFyCooQe+S4hGsK+n1GHVibVn8lgDEw4iIrJ75Wotnvr+APafvwlfT1esnBKF8MYNpA6rTqqsaotbs2kWLVpk17NpeEuFiIjsmlYrMGvdEaSczoGnmwuWTuqJ9k3ss6ZSlVVt/19WVhaGDx+O5ORkiSKrPxb+Agt/ERHZGmNvKQgh8PrGE1i15zxcXWT4dmIPxLYNliDi+tNoNIiIiKiSbFSSyWRQKBRIT0+3qdsrxrah7OEgIiKbkpycjIiICPTt2xdjxoxB3759ERERYfDq/pPNZ7Bqz3nIZMCCkV3tNtkAqq5qeychhF2vasuEg4iIbIYptxSW7kzHp1vPAQDejOuEh7o0tWqs5mbsarX2uqotEw4iIrIJGo0G06dPh6E7/ZXPzZgxAxqNBkkHLuLNTScBAM/f1wbj7w63aqyWYOxqtfa6qi1nqRARkU0w9pbCp0kqfHa4HAAwpVdzPNOvlbVCBGC5Kau9e/eGQqFAVlaWwaSrcgyHva5qyx4OIiKyCcbcKvBQdMTnh0qh0QrEdwvDq/e31634ag2mjC8xlVwux6JFiwCgyneq/HnhwoU2NWDUFJImHPPnz0fPnj3RsGFDBAcH4+GHH8bp06f1tomNjYVMJtN7TJ06VW+bCxcu4P7774e3tzeCg4PxwgsvQK1WW/OrEBFRPdV2q8AtuAWCh78OtZBhQPtgvD+sM1xcrJtsWHrKanx8PNavX4+wsDC95xUKBdavX2/XdTgknRY7ePBgjBo1Cj179oRarcbLL7+M48eP4+TJk2jQ4FbBltjYWLRp0wZvvvmm7n3e3t66qTcajQZdu3ZFaGgoPvzwQ1y+fBkTJkzA448/jnfffdeoODgtlohIepXTQg3dUnBt1BShYz+AvIE/oiIaYWVCNDzdrHelb+0pq/ZUadTYNtSm6nDk5OQgODgY27Ztw7333gvgVsLRtWtXLFy40OB7fv31VzzwwAO4dOkSQkJCAACLFy/G7NmzkZOTA3d391o/lwkHEZFtqOxFAP43UFTesDFCx34IV79ghHlr8euLg+Hrad31UVJSUtC3b99at1OpVE63Gq1d1uHIy8sDAAQEBOg9//333yMwMBCdOnXCnDlzUFxcrHtt9+7diIyM1CUbADBo0CDk5+fjxIkTBj+nrKwM+fn5eg8iIpLenbcUXLx8ETziLbj6BSPIU4uNMwdaPdkAHH/KqjXYzCwVrVaLGTNmoFevXujUqZPu+TFjxiA8PBxNmzbF0aNHMXv2bJw+fVp3ryw7O1sv2QCg+zk7O9vgZ82fPx/z5s2z0DchIqL6iI+PR1xcHP5Qbcc7fxXgYokcoX6eWD81BoE+HpLE5OhTVq3BZhKOadOm4fjx49i5c6fe80888YTu35GRkWjSpAn69++PtLQ0tGzZsk6fNWfOHMycOVP3c35+PpRKZd0CJyIis1ML4LsML1wsKUYjbzd8lxAFRSNvyeJx9Cmr1mATt1SeeeYZbNq0CSqVCgqFosZto6OjAQDnzt2qLhcaGoorV67obVP5c2hoqMF9eHh4wNfXV+9BRES2Qa3RYvrqw/jr3HU0cJdj+eQotApuKGlMtU1ZFUJg2LBh2LFjBzQajRQh2jxJEw4hBJ555hls2LABW7duRfPmzWt9z+HDhwH8r9sqJiYGx44dw9WrV3XbbN68Gb6+vujQoYNF4iYiIssQQuCVDcfx24lsuMtd8PWEHuii9Jc6LADVT1l1cbnVlC5cuNCsdTkcjaSzVJ5++mkkJiZi48aNaNu2re55Pz8/eHl5IS0tDYmJiRg6dCgaN26Mo0eP4rnnnoNCocC2bdsA/G9abNOmTfHBBx8gOzsb48ePx2OPPcZpsUREdmb+r3/jq23/wEUGfDG2GwZ3sr0xEZVTVjdu3GhwBmVlD4i9180wll1Mi62uOtyyZcswadIkZGZmYty4cTh+/DiKioqgVCrxyCOP4NVXX9X7UufPn8dTTz2FlJQUNGjQABMnTsR7770HV1fjhqgw4SAikt7ibWl479dTAID3h0ViZM9mEkdUPXtdSt4S7CLhsBVMOIiIpPXD3gt4KfkYAGDOkHZ4sk/dJgVYC+ty/I9d1uEgIiLn8+uxy3h5w61kY2qfljafbACsy1EXTDiIiEgyO89ew/QfDkMrgNFRSswe3Lb2N9kA1uUwHRMOIiKSxKELN/HEqv0o12gxNDIUbz8cadWVX+ujsi5HdfHKZDIolUrW5bgNEw4iIgM0Gg1SUlKwevVqpKSksLaCmZ25UoDJy/ehuFyD3q0D8cnIrpBbceXX+nL0peQtgQkHEdEdkpOTERERgb59+2LMmDGsrWBmmTeKMX5JKnKLK9BV6Y/F47rDw9X+GmZHXkreEjhLBZylQkT/U7la6Z1/Gp2ttoKl5BSUYcRXu5F+rQitg32w9skYNGpQ+6retsyelpK3BE6LNQETDiICWFvB0vJLKzDqqz04eTkfYf5eSHrqHoT6eUodFtUTp8USEZlox44d1SYbwK2y25mZmdixY4cVo3IMpRUaPLZ8P05ezkegjzu+eyyayYaTYcJBRPT/WFvBMio0Wkz7/iD2ZtxAQw9XLJ8cheaBDaQOi6yMCQcR0f9jbQXz02oFXlx/FFtOXYWHqwuWTOqJTmF+UodFEmDCQUT0/1hbwbyEEHhz00lsOJQFuYsMX4zthqjmAVKHRRJhwkFE9P9YW8G8Pt1yDst3ZQAAPn60C/q3D5E2IJIUEw4iotuwtoJ5rNiVgU/+PAMAeOPBDnj4rrBa3kHVcZQidJwWC06LJaKqnL22Qn1sPJyF6T8cBgBM798az93XRtqA7FhycjKmT5+uN3tKoVBg0aJFNpP8sg6HCZhwEEmLjbvjUJ26isdX7odaKzAxJhxvPNTRbtZHsTX2UoSOCYcJmHBYBxsVMsQeruDIOPsybmD8klSUVmgR17UpPhnRFS52tD6KLbGnInQs/EU2hWtTkCGVV3B3/lHNysrC8OHDeX7YkZOX8jFl+T6UVmjRt20QPnq0C5ONenDEInRMOMji2KiQIRqNBtOnT6/SXQxA99yMGTPsdoCcM8m4VoQJS/eioFSNnhGN8MXY7nCTs3mpD0csQsczgiyKjQpVxxGv4JzRlfxSjFuSimuFZWjfxBffTuwJL3feKq0vRyxCx4SDLIqNClXHEa/g7Fldpl7mFpdj/JJUXLxZgvDG3lgxpSf8vNysEK3jc8QidEw4yKLYqFB1HPEKzl7VZYxVcbkak5fvw5krhQjx9cB3CdEIbsjF2MzFEYvQMeEgi2KjQtVxxCs4e1SXMVblai2eXHUAhy7kws/LDSunREMZ4G2tkJ2GoxWh47RYcFqsJVVO7crKyjI4jsOWpnaR9VU2dgD0zg9bqzPgqOoy9VKjFfj3D4fw89HL8HKT4/vHo9GtWSNrhu10bL2kAKfFkk1wxG5BMh9Hu4KzN6aOsRJC4LWNx/Hz0ctwk8vw1fjuTDasQC6XIzY2FqNHj0ZsbKzd/r10lToAcnyVjYqh4k4LFy6060bF1q887EF8fDzi4uJ4HCVg6hirj/44jcTUC5DJgE9GdsW9bYIsGR45GCYcZBWO2KiwQqb5VF7BkXWZMsbq2x3/4HNVGgDgnYcj8UDnppYMjRwQx3CAYzjIdPayxgFRTYwdY/Xh+h2YnXwcAPDCoLaY1reV3j4c6UKiJs70XU3BMRxEFsJiZuQojBljlTD3U8zZcCvZeLx3czwd21K3jTMtWeBM39VSmHAQmYjFzMiR1DRwd/7SJHz/jwe0Ani0uwIvD22vS0ScackCZ/qulsSEg8hELGZGjiY+Ph4ZGRlQqVRITEyESqXCxh2HsfIfL5RrtBjYIQTz4yN1yYYz9fI503e1NA4aJTIRi5mRI7p94O65q4UY8dVuFJVrcE/Lxvh09F1wvW0xNlN6+WxlMHBdx1/Y43e1VezhIDIRK2SSI8vKLcGEJam4UVSOzgo/fD2hBzzd9Btme+vlq8/4C3v7rraMCQeRiVjMjBzV9cIyjF+Sikt5pWgR1ADLJvWEj0fVjnB76uWr7/gLe/quto7TYsFpsVQ3hupwKJVKuy9mRs6poLQCY75JxbGsPDT188T6p+5BU38vg9vay5IFdSndXt0+bP27SonTYokszNBAu/T0dCYbZHdKKzR4YuUBHMvKQ0ADd6x6LLraZAOwn14+c8wos5fvag+YcBDVg6OscUDOS63R4tnVh7D7n+vw8XDFislRaBnkU+v77GEdHHONv7CH72oPOEuFiMhJabUCLyUfw+aTV+Du6oJvJvRApMLP6Pfb+pIF5hx/Yevf1R5wDAc4hoOInI8QAu/8/De+3ZkOuYsMX47thoEdQ6UOy6w4/sI6OIaDiIiq9UVKGr7dmQ4AeH9YZ4dLNgCOv7A1TDiIiJzM96nn8eHvpwEAr97fHsO7KySOyHI4/sJ28JYKeEuFiJzHpqOX8OzqQxACmNa3JV4Y1E7qkKyCK71ajrFtKAeNEhE5iW1ncvDcmsMQAhgT3QyzBraVOiSrub10O0mDt1SIiJzAgfM3MXXVAVRoBB7o3ARvxXWqtjw/kSUw4SAicnCnswswZfk+lFRocG+bICwY0RVyFyYbZF1MOIiIHNiF68UYvyQVeSUV6NbMH4vHdYO7K//0k/XxrCMiclBXC0oxfmkqrhaUoW1IQyyd1BPe7hy6R9JgwkFE5IDySiowYclenL9eDGWAF1YlRMHf213qsMiJMeEgInIwJeUaJCzfh1PZBQhq6IHvEqIR7OspdVjk5Ni3RkTkQCo0Wjz9/QHsP38TDT1dsXJKFMIbN7DoZ7LGBRmDCQcR2TQ2ZsbTagVmrTsC1ekceLq5YNmknmjfxLLFDJOTkzF9+nS9ZeAVCgUWLVrEKp6kh7dUiMhmJScnIyIiAn379sWYMWPQt29fREREIDk5WerQbI4QAm/89wQ2Hr4EVxcZvhzXHT0iAiz6mcnJyRg+fLhesgEAWVlZGD58OP+fSA9Lm4Olzcn2OeNVfmVjduefqMpiVVwHQ9+CzWfw6ZazkMmAhSO7Iq5rWO1vqofKlVjvTDYqcSVW58HVYokchDNe5Ws0GkyfPt3gkuKVz82YMQMajcbaodmkZX+l49MtZwEAbz7U0eLJBgDs2LGj2mQDuPX/lJmZiR07dlg8Fluh0WiQkpKC1atXIyUlhefnHZhwENkwZ+2yZmNmvOSDFzHvvycBADPva4PxMRFW+dzLly+bdTt754wXBqZiwkFOx16uQpz5Kp+NmXH+PHkFL6w/CgCY3CsCz/ZrZbXPbtKkiVm3s2fOemFgKiYc5FTs6SrEma/y7a0xkyKJTf3nOqYlHoRGKxB/Vxheu7+DVRdj6927NxQKRbWfKZPJoFQq0bt3b6vFJAVnvjAwFRMOchr2dhXizFf59tSYSZHEHs/Kw2Mr9qNMrcWA9sF4f3hnuFh5MTa5XI5FixYZbGiBW43twoULHX7AqDNfGJiKCQc5BXu8CrG3q3xzqmzMAFRJOip/toXGTIokNv1aESYt24uCMjWimgfgszHd4Cbnn3KpOPOFgal4lpJTsMerEHu6yreE+Ph4rF+/HmFh+jMuFAqFTUyJlSKJvZxXgnHfpuJaYTk6NvXFtxN7wNNNmqSr8vtXRyaT2VwSbwnOfGFgKiYc5BTs8SrEXq7yLSk+Ph4ZGRlQqVRITEyESqVCenq65MkGYP0k9mZROSYs2Yus3BI0D2yAFVOi4OvpZpZ914UUSbwtDvh29gsDUzDhIKdgr1chtn6Vbw1yuRyxsbEYPXo0YmNjbSbBsmYSW1imxqTl+3D2aiFCfT2xKiEKgT4e9d5vfVg7ibfVAd+8MDAeEw5yCvZ8FWLLV/nOzFpJbJlagydX7ceRzFz4e7thVUIUFI2867VPc7BmEm/rA755YWAcljYHS5s7i8o/WgD07ruzVLZzMVeZ+MrS3llZWQbHcZijtLdGK/BM4kH8ejwb3u5yJD5+N7oq/eu0L3Ozxve//XPsoYS6My5BAJjQhgoJvfvuu6JHjx7Cx8dHBAUFibi4OHHq1Cm9bUpKSsTTTz8tAgICRIMGDUR8fLzIzs7W2+b8+fNi6NChwsvLSwQFBYlZs2aJiooKo+PIy8sTAEReXp5ZvhfZrqSkJKFQKAQA3UOpVIqkpCSpQyMrMPT/r1Ao6vz/n5SUJGQymZDJZHr7rHyuPueVVqsVs9cfEeGzN4nWL/8idp7NqfO+LMWS37+SSqXS23d1D5VKVf8vRHVibBsqacIxaNAgsWzZMnH8+HFx+PBhMXToUNGsWTNRWFio22bq1KlCqVSKLVu2iP3794u7775b3HPPPbrX1Wq16NSpkxgwYIA4dOiQ+OWXX0RgYKCYM2eO0XEw4XAuarVaqFQqkZiYKFQqlVCr1VKHRFZQ2Tje2VDVt3G0VBI7/5e/RfjsTaL5S5vEr8cu1WtflmTpJD4xMdGohCMxMdEsn0emM7YNtalbKjk5OQgODsa2bdtw7733Ii8vD0FBQUhMTNR1hZ86dQrt27fH7t27cffdd+PXX3/FAw88gEuXLiEkJAQAsHjxYsyePRs5OTlwd3ev9XN5S4XIsVm6W97cXelfbUvD/F9PAQDeHxaJkT2b1Xlf1mDJWwkpKSno27dvrdupVCrExsaa5TPJNMa2oa5WjKlWeXl5AICAgAAAwIEDB1BRUYEBAwbotmnXrh2aNWumSzh2796NyMhIXbIBAIMGDcJTTz2FEydO4K677qryOWVlZSgrK9P9nJ+fb6mvREQ2wJQpnHVptCpn0pjDmn0XdMnGS0Pa2XyyAZj3+9+pcsB3bWNFahrw7axjK2yNzcxS0Wq1mDFjBnr16oVOnToBALKzs+Hu7g5/f3+9bUNCQpCdna3b5vZko/L1ytcMmT9/Pvz8/HQPpVJp5m9DRLbEXuqw/Hb8MuYkHwMAPNmnBab2aSlpPLagvtNObXU6rTOymYRj2rRpOH78OH744QeLf9acOXOQl5ene2RmZlr8M4lIOvZQh2Xn2Wv49+rD0ApgVE8lXhrcTrJYbE1dp53a+nRaZ2MTt1SeeeYZbNq0Cdu3b4dCodA9HxoaivLycuTm5ur1cly5cgWhoaG6bfbu3au3vytXruheM8TDwwMeHtIWzSEi6zFHt7wlHc7MxROr9qNco8WQTqF455FIq678ag/i4+MRFxdn9K2R2krPV5Zej4uL4+0VK5G0h0MIgWeeeQYbNmzA1q1b0bx5c73Xu3fvDjc3N2zZskX33OnTp3HhwgXExMQAAGJiYnDs2DFcvXpVt83mzZvh6+uLDh06WOeLEJFNs+VqkGevFGDSsr0oLtfgX60CsXBUV8itvPKrvTCl6qw9rp/k6CRNOKZNm4bvvvsOiYmJaNiwIbKzs5GdnY2SkhIAgJ+fHxISEjBz5kyoVCocOHAAkydPRkxMDO6++24AwMCBA9GhQweMHz8eR44cwe+//45XX30V06ZNYy8GEenYYjXIizeLMX7JXuQWV6CL0h9fje8OD1debZuDvYzbcSaS3lL58ssvAaDK6OZly5Zh0qRJAIBPPvkELi4uGDZsGMrKyjBo0CB88cUXum3lcjk2bdqEp556CjExMWjQoAEmTpyIN99801pfg4jshKnd8pZ0rbAM45fsRXZ+KVoH+2D5pJ5o4GETd7kdgj2M23E2NlWHQyqsw0FE1pRfWoHRX+/BiUv5CPP3QtJT9yDUz1PqsByKtUqvk/FtqM3MUiEicgalFRo8tmI/TlzKR+MG7liVEMVkwwJsedyOs2LCQUQWo9FokJKSgtWrVyMlJQUajUbqkCRVodHimcSD2Jt+Aw09XLFiShRaBPlIHZbDqu+4HZ6/5sVbKuAtFWthtT/nkpycjOnTp+vNFFAoFFi0aJFTrsqr1QrMWncEyYey4OHqgpVTohDdorHUYTmFuvzt4flrPGPbUCYcYMJhDfzldS6VBZfu/PNS2ZUt1awQqQgh8Oamk1j2VwbkLjJ8Na47BnQIqf2NJAmev6ZhwmECJhyWxV9e52LphdLs0adbzmLB5jMAgE9GdsEjdylqeQdJpT7nr7P24nLQKNmE2qr9AcCMGTN4b9SBsOCSvlW7M3TJxtwHOzDZsHF1PX+5ZkvtmHCQRbHxcT62XHDJ2oMANx7Owus/nQAA/Lt/a0zu1byWd5DU6nL+cs0W4zDhIIuy5caHLMNWCy5Z+wpUdfoqnl97BEIAE2PC8dyA1hb5HDIvU89f9uIajwkHWZStNj5kOZULpVW3+JhMJoNSqbTqQmnWvgLdn3EDT313AGqtwENdmmLugx25GJudMPX8ZS+u8ZhwkEXZYuNDlmVrBZesfQX69+V8TFm+D6UVWsS2DcLHI7rAhYux2Q1Tz1/24hqPCQdZlK01PmQdtrRQmjWvQM9fL8L4JXuRX6pGj/BG+HJsd7jJ+WfW3phy/rIX13icFgtOi7UGQ3U4lEolFi5cyCmxDswWpgmuXr0aY8aMqXW7xMREjB49us6fcyW/FMMX70LmjRK0C22INU/GwM/Lrc77I+kZc/7a05otlvp9NLYN5dKEZBW2tEonWY9cLq+yGrS1WeMKNLe4HBOW7EXmjRKEN/bGyoQoJhsOwJjzt7IXd/jw4ZDJZHpJhy314tpC8UX2cIA9HOT4bKGnQSqWvgItLldj3LepOHghF8ENPZD01D1QBnibI3SyI7bci2vp4ousNGoCJhzkyGzhykZqlX9wARi8Aq3rH9xytRYJK/Zhx9lr8PNyw9onY9A2tKFJ+3DmZNDR2OL/pTUq/xrdhgoSeXl5AoDIy8uTOhQis0pKShIymUwA0HvIZDIhk8lEUlKS1CFaTVJSklAoFHrHQalUGn0M1Gq1UKlUIjExUahUKlFWXiGmfX9AhM/eJNq9+qvYn3HDLDEpFAqn+n8hy1KpVFV+/w09VCpVnT/D2DaUPRxgDwc5Jmdc06S2K8y6XoEa7C5/5AW4tOkDN7kMSyb2xL1tgkyK1RnWGLLFK35nY41B0+zhMAF7OMgRWePKxpZYqrfAUC+Rf+9xInz2JtHsxZ/Ea19vMHmfarW6Sqy3P2QymVAqlUKtVtcrdimx98Y22FIPByeIEzkoZypIZKlKooaKhjXsEQe/e0YBAG7+8SWWv/Vvk4uGOXp1Sq4tYjtsqfgiEw4iB+UsBYlqqyQqhMDUqVPx/fffm7xg252JQYNO/RDQ/3EAwM1tK1Bw+Nc6JQaOnAxybRHbYkvFF5lwEDkoW7qysaTaegsAICcnB+PGjTN5wbbbG3yvVtFoPGQ6ACB/7wbk71lncDtjOHIy6Oi9N/bIVir/1jvhyM/Px48//oi///7bHPEQkZnY0pWNJZna2JvSrV/Z4HsoIxEUNxsyFzkKj23GTdUSg9sZy5GTQUfuvbFn8fHxyMjIgEqlQmJiIlQqFdLT0606MNnkhGPEiBH47LPPAAAlJSXo0aMHRowYgc6dOyMpKcnsARJR3dnKlY0lmdrYm9Kt37t3byg634PgYa9B5uqO4jO7cf3X/+her2ti4MjJoCP33ti7ysqpo0ePRmxsrNXPL5MTju3bt+t+uTZs2AAhBHJzc/Hpp5/i7bffNnuARFQ/tnBlY0m19RYYYmy3fsaNEvg+OAcuHt4oPX8EOT99AAgtgPonBo6aDDpy7w3Vj8kJR15eHgICAgAAv/32G4YNGwZvb2/cf//9OHv2rNkDJKL6k/rKxpJq6i2oTU3d+pdySzD+21QUqWVQeGvhtmcZoKnQvR4WFoY33ngDZWVlJg9GreSIyaAj995Q/ZiccCiVSuzevRtFRUX47bffMHDgQADAzZs34enpafYAiYhqU11vQW2uXLmC1atXV0kYbhSVY/ySVFzKK0WLoAbYOHMgMs6d0iUG8+bNAwDMnTsXY8aMMXkw6u0cMRl01N4bqh+TK41+8cUXmD59Onx8fNCsWTMcOnQILi4u+M9//oPk5GSoVCpLxWoxrDRK5BgqK1tmZWXhueeew7Vr1wxOzwRuNfS3JxmV68sMvP8hjPlmD45ezEMTP0+sf+oehPl76bZzhgqh5sJKo87Boou37d+/H5mZmbjvvvvg4+MDAPj555/h7++PXr161T1qiTDhIHI81S3YVh2ZTAbI3dDvjfU4V+CCgAbuWPtkDFoF++i2ccZy8US1MbYNrdO02B49euD+++9HVlYW1Go1AOD++++3y2SDiBxTdd361SUCAjIEPvgCzhW4oIG7HMsn99RLNgDWmCCqD5MTjuLiYiQkJMDb2xsdO3bEhQsXAADPPvss3nvvPbMHSERUV3cOyvzkk0+qHdzZePCz8G4TA6EuxzNd3NBZ4V9lG9aYIKo7kxOOOXPm4MiRI0hJSdEbJDpgwACsWbPGrMEREdXX7YMyQ0JCDG7j33cKfDrfB6HVIOenD+BfftXgdqwxQVR3JiccP/74Iz777DP861//0pvy1LFjR6SlpZk1OCIiczKUCPje/Sj8om4N8rz+66coObun2oSBNSaI6s7khCMnJwfBwcFVni8qKjJ5DjwRkTXdmTD4dBmERn0mAgBubP0WxSe21pgwsMYEUd2ZnHD06NEDP//8s+7nyl+yb7/9FjExMeaLjMhEGo0GKSkpBusqEAH6CYN3u38hYNA0AEDerjUo3L8RQO0JA2tMENWNq6lvePfddzFkyBCcPHkSarUaixYtwsmTJ7Fr1y5s27bNEjES1So5ORnTp0/Xm0FQWVeBDQDdLj4+Hu8sScLiky6QyVxQcOhX5O5YBaVSiYULFxp1vsTHxyMuLo41JohMUKc6HGlpaXjvvfdw5MgRFBYWolu3bpg9ezYiIyMtEaPFsQ6HfWMhJjLFwQs3MfabVJRUaBAVKkdc8E2ENWXCQFRXFi385WiYcNgvFmIiU5zOLsCIr3Yjr6QCvVsHYsnEnnB3rVM5IiL6f8a2oSbfUqmsu1GdZs2ambpLojozpRBTbGys9QIjm5N5oxjjl6Qir6QCdzXzx1fjuzt9ssHS42RNJiccERERNc5G4UA9siYWYiJjXC0oxbglqbhaUIa2IQ2xbFJPeLub/OfPoXDcE1mbyb9xhw4d0vu5oqIChw4dwoIFC/DOO++YLTAiY7AQE9Umr6QCE5fuw/nrxVAGeGFlQhT8vd2lDktS1Y17ysrKwvDhwznuiSzCbGM4fv75Z3z44YdISUkxx+6simM47FflGI6srCyDC3RxDIf12GL3fEm5BhOWpmJfxk0E+nhg/dQYRAQ2kDQmqXHcE5mbRRdvM6Rt27bYt2+fuXZHZBQWYrINycnJiIiIQN++fTFmzBj07dsXERERSE5OliymCo0W0xIPYl/GTTT0dMWqhCinTzYALkBH0jE54cjPz9d75OXl4dSpU3j11VfRunVrS8RIVCOpCjGx0Ngtld3zdzZild3zUiQdWq3ArHVHsPXUVXi6uWDppJ5o34S9lwDHPZF0TB7D4e/vX+VKUggBpVKJH374wWyBEZnC2oWYOODuFo1Gg+nTpxu8nSWEgEwmw4wZMxAXF2e1XiYhBOb99wQ2Hr4EVxcZvhzbHT0jAqzy2faA455IKiaP4bizmqiLiwuCgoLQqlUruLra56hvjuEgU7DQ2P+kpKSgb9++tW6nUqmsNi35k81nsGjLWchkwMKRXRHXNaz2NzkRjnsic7NYHY4+ffrUKzAie2aLV/RSsrXu+WV/pWPRlrMAgHkPdWSyYUDluKfhw4dDJpPpncsc90SWZFTC8dNPPxm9w4ceeqjOwRDZOhYa02dL3fMbDl3EvP+eBAA8N6ANJsREWPwz7VXluCdDtwWNXU+GyFRGJRwPP/ywUTuTyWROO3COnIOtXdFLrXK599q656tb7t1ctvx9BbPWHQUATLonAv/u38qin+cIuAAdWZtRCYdWq7V0HER2wZau6G2BLXTPp/5zHU9/fxAarcAjd4Xh9Qc61FgNmf5HLpc7RU8c2QbnXkiAyESVV/TVNWgymQxKpdLiV/S2RKppyQBw4lIeHluxH2VqLQa0D8YHwzvDxYXJhrPiVHXbVqdKo0VFRdi2bRsuXLiA8vJyvdf+/e9/my04a+EsFTJF5SwVAAav6J1plsrtrF1pNP1aER5dvAvXCssR1TwAK6dEwdONtwOcFaeqS8diy9MfOnQIQ4cORXFxMYqKihAQEIBr167B29sbwcHB+Oeff+odvLUx4SBTGfrjplQq9Qbc2WKpb0eRnVeKYV/uQlZuCTo08cUPT94NX083qcMiiXCqurQslnDExsaiTZs2WLx4Mfz8/HDkyBG4ublh3LhxmD59ul3+pzLhoLqoKaGQ8mrL0ROdm0XlGPHVbpy9WoiIxt5YN/UeBDX0kDoskgjXhpGe0W2oMJGfn584deqU7t8nT54UQgixZ88e0bZtW1N3ZxPy8vIEAJGXlyd1KOQAkpKShEwmEwD0HjKZTMhkMpGUlGTRz1YoFHqfq1AoLPqZ1lRYWiHiPtspwmdvEtHv/CkuXC+SOiSSmEqlqvK7ZuihUqmkDtVhGduGmjxo1M3NDS4ut94WHByMCxcuAAD8/PyQmZlp6u6IHEpthcEAYMaMGRYZzGaLa5qYU5lagydXHcDhzFz4e7thVUIUlAHeUodFEuNUdfthcsJx11136VaF7dOnD15//XV8//33mDFjBjp16mT2AInsibVW4rxzNH55eblkiY41aLQCz605jJ3nrsHbXY7lk6PQOqSh1GGRDeBUdfthdMJR+Yfq3Xff1f3HvfPOO2jUqBGeeuop5OTk4Ouvv7ZMlER2whpXW4aWglcoFA675LgQAq9sOIZfjmXDXe6Cr8f3QFelv9RhkY3gVHX7YfRaKmFhYZg0aRKmTJmCHj16ALh1S+W3336zWHBE9sbSV1vVjcbPyckx6v322K38we+n8cO+TLjIgEWjuuJfrQOlDolsiC0UnyPjGN3DMW3aNKxfvx7t27dH7969sXz5chQXF1syNiK7Y8mrrZrGhxjL3rqVv96ehi9T0gAA7z4SiSGR9hU/WYeUxefIeCZPi01JScGyZcuQlJQEuVyOESNG4LHHHkN0dLSlYrQ4Toslc7JUYTBjl4I3xB6nBq7dl4kXk26tj/LSkHaY2qelxBGRrXP0KeG2ytg21ORBo7GxsVixYgWys7Px8ccf4++//0ZMTAw6duyIBQsW1CtoIkdgqautut4Oscdu5d+OX8ZLybeSjSfvbcFkg4xSuTbM6NGjERsbazfnu7OoU2nzO/3888+YMGECcnNz7XIUPHs47J8tXtmYOyZjezgCAwNx7do13c93VkC1dX+du4bJy/ahXKPFyB5KvDcskouxEdkwY9tQoweN3qm4uBhr167FsmXLsHPnTrRs2RIvvPBCXXdHVGe2uoaCuVfiNHYp+HPnzmHXrl02lXwZ60hmLp5YuR/lGi0GdwzFO490YrJB5CBMvqWya9cuPPbYY2jSpAmmTZuGiIgIqFQqnDlzBi+99JJJ+9q+fTsefPBBNG3aFDKZDD/++KPe65MmTYJMJtN7DB48WG+bGzduYOzYsfD19YW/vz8SEhJQWFho6tciO+Xoxa5uVzkaH0CVRvj22ybu7u522a187moBJi3bi6JyDXq1aoxFo7vCVc4FrYkchdG/zR988IFuhsqxY8fw4YcfIjs7GytWrMC9995bpw8vKipCly5d8Pnnn1e7zeDBg3H58mXdY/Xq1Xqvjx07FidOnMDmzZuxadMmbN++HU888USd4iH7ImVVz8rPt/ZS2I46Gv/izWKM+3YvbhZXoIvCD1+N7wEPV/tIlIjIOEaP4QgKCsK4ceOQkJBgkYqiMpkMGzZswMMPP6x7btKkScjNza3S81Hp77//RocOHbBv3z5dbZDffvsNQ4cOxcWLF9G0aVOD7ysrK0NZWZnu5/z8fCiVSo7hsDPGjmlQqVRmvbUBSH8bx5HGrFwrLMOji3cj/VoRWgX7YN2TMWjUwN0KEROROZh9DMelS5fg5mb95Z9TUlIQHByMRo0aoV+/fnj77bfRuHFjAMDu3bvh7++vSzYAYMCAAXBxcUFqaioeeeQRg/ucP38+5s2bZ5X4yXKkWkOhuuJblbdxrNHTYO7xIfVV1wSsoLQCE5fuRfq1IoT5e2FVQhSTjdvYYmJJVFdG31KRItkYPHgwVq5ciS1btuD999/Htm3bMGTIEF3XdXZ2NoKDg/Xe4+rqioCAAGRnZ1e73zlz5iAvL0/34KJz9kmKNRSkvo1ji+o6jqa0QoPHVuzHiUv5aNzAHasSotDEz8saIdsFQyXsIyIiHGpcEjkXmx6RNWrUKDz00EOIjIzEww8/jE2bNmHfvn1ISUmp1349PDzg6+ur9yD7I8UaCtZanM1e1DUBU2u0eCbxEFLTb6ChhytWTIlCiyAfq8RsD5xpMDQ5D5tOOO7UokULBAYG4ty5cwCA0NBQXL16VW8btVqNGzduIDQ0VIoQyYqMnbVhzi5oR18K29SBsHVJwLRagReTjuLPv6/Aw9UF30zsgU5hfmb7DvaOvWjkqOwq4bh48SKuX7+u6yKPiYlBbm4uDhw4oNtm69at0Gq1dl1qnYxn7VkbjrwUdl268E1NwIQQePvnv5F8MAtyFxk+H9MNd7dobJb4HQV70chRGTVoND8/3+gdmnJ7orCwUNdbAQDp6ek4fPgwAgICEBAQgHnz5mHYsGEIDQ1FWloaXnzxRbRq1QqDBg0CALRv3x6DBw/G448/jsWLF6OiogLPPPMMRo0aVe0MFXI88fHxiIuLs8rgOmOLb9V2G8fWBgPWdSCsqQnYZ1vPYelf6QCAD4d3xoAOIfWM3PE4ei8aOTFhBJlMJlxcXIx6mEKlUgkAVR4TJ04UxcXFYuDAgSIoKEi4ubmJ8PBw8fjjj4vs7Gy9fVy/fl2MHj1a+Pj4CF9fXzF58mRRUFBgUhx5eXkCgMjLyzPpfeSckpKShEwmEzKZTO+8rXwuKSmp1vcrFAq99yoUilrfZylqtbpKPHd+L6VSKdRqdbXvvfNYGHrvyt0ZInz2JhE+e5NYsuMfCb6pfaju7+KdD5VKJXWoREII49tQo+pwbNu2TffvjIwMvPTSS5g0aRJiYmIA3JqeumLFCsyfPx8TJ06sc/IjFa6lQqYyNA3UmDVLqutJqO9KsvVR33omxqyOK28RjRlrDkMI4N/9WmHmwLbmCd4BaTQaRERE1NqLZk8r/5JjM7oNNTWT6devn0hMTKzy/Pfffy/69Olj6u5sAns4qC7UarVQqVQiMTFRqFQqgz0Ad25f154ES0pMTDTqitrQ730lQ702SqVSJCUlCdWpK6LlnJ9F+OxN4rUfjwmtVmvFb2ef6tuLRmRNZu3huJ23tzeOHDmC1q1b6z1/5swZdO3aFcXFxabsziawh4PMpaaxGVJWRq2JueIy9N0PX8zD2G9TUVqhxUNdmmLhyK5wceFibMaoay8akbVZbLVYpVKJb775Bh988IHe899++y2USqXpkRI5iNqqbdrqYEBzDYS9s/rp35fzMXnZPpRWaBHbNggfPdqFyYYJLD0Y2tYGLpPjMznh+OSTTzBs2DD8+uuvuqmne/fuxdmzZ5GUlGT2AInsgTGzPGx1Sm1lPZPhw4dDJpMZHIdhaj2T89eLMGHpXuSXqtE9vBG+HNsd7q7Wn4Vv742qpUrYS70WEDknk2+pAEBmZia+/PJLnDp1CsCt6alTp0612x4O3lKh+qgc5Fdd7YTKHoJz586hZcuWNjsY0Fxd+FfzSzF88W5cuFGMdqENseaJGPh5W39pBDaqhtniwGWyb8a2oXVKOBwNEw6qD1PGQNy4caPWGR1S/rGvb49AXnEFRny1G6evFKBZgDfWT41BsK+nBSM2jI2qYcYmx5wBQ6Ywtg2tUx/njh07MG7cONxzzz3IysoCAKxatQo7d+6sW7REdsyUsRnWroxqqsou/NGjRyM2NtakRqe4XI3Jy/fi9JUCBDf0wHcJ0ZIkGywNXj1WMSUpmZxwJCUlYdCgQfDy8sLBgwdRVlYGAMjLy8O7775r9gCJbJ2pYzPi4+ORkZEBlUqFxMREqFQqpKenS55s1Ee5Woup3x3EwQu58PV0xcqEKDRr7C1JLGxUq2erA5fJOZiccLz99ttYvHgxvvnmG70l63v16oWDBw+aNTgie1CXVWvr05NgazRagZlrD2P7mRx4ucmxbHIU2oVKd2uSjWr1bHXgMjkHkxOO06dP4957763yvJ+fH3Jzc80RE5Fdsdaqtaau5GoNQgi8vvE4Nh29DDe5DIvHd0f38EaSxsRGtXp1SY6JzMXkhCM0NFRvwbVKO3fuRIsWLcwSFJG9sfTYjLqs5GoNCzafwfepFyCTAQtGdEWfNkGSxgOwUa2JtZJjIoNMLWH67rvvig4dOog9e/aIhg0bih07dojvvvtOBAUFiU8//dTU3dkEljYnczG13LkxKstcw0ApdCnLXH+74x/dYmyrdmdIEkN1WBq8ZjWVoicylcVKmwsh8O6772L+/Pm6MuYeHh6YNWsW3nrrLbMlQtbEabFkKfWdZmqr0xiTDlzE8+uOAABmDWyDZ/q1ruUd1sfS4DWz96JoZDssXoejvLwc586dQ2FhITp06AAfH586Bys1JhxkTpV/yDdu3Ijvv/8eOTk5utdMLTxli+uvbD55BVO/OwCNViDhX83x6v3tq719ITU2qkSWZ7G1VKZMmYJFixahYcOG6NChg+75oqIiPPvss1i6dGndIiZyAIauqm93e6lzY5IOW5txsTvtOqYlHoRGKzCsmwKvDLXdZAOwXGlwYzDZIdJn8qDRFStWoKSkpMrzJSUlWLlypVmCIrJHldUta6sBARhfeMqWZlwcu5iHx1fuR7lai/s6hOD9YZFcjK0atjrIl0hKRicc+fn5yMvLgxACBQUFyM/P1z1u3ryJX375BcHBwZaMlchm1VTd8k7ChMJTtjLjIi2nEBOX7UVhmRp3twjAf0bfBVe59RdjswfVJZ6VvVtMOshZGX1Lxd/fHzKZDDKZDG3atKnyukwmw7x588waHJG9qK26pSHG3AaxxEquprqUW4Lx36biRlE5IsP88M2EHvB0c65bA8beHqmtrLpMJsOMGTMQFxfH2yvkdIxOOFQqFYQQ6NevH5KSkhAQEKB7zd3dHeHh4WjatKlFgiSydXUZQ2HsbZDKGh+GVj619IyLG0XlGL8kFZfyStEiqAGWT+6Jhp7WX/lVSqasOmtKWXWpxpYQScXohKNPnz4AgPT0dDRr1symB4qR7XLUgXSmjKGonMpqym2Q+Ph4xMXFWfXYFZapMWnZXqTlFKGJnydWJUSjsY+HxT7PFlW36mx1g39tbZAvkS0xeZbK1q1b4ePjg0cffVTv+XXr1qG4uBgTJ040W3DkWEy5UrQ3lWMtsrKyahzHUZ/bINaccVFaocETK/fj6MU8NPJ2w6qEaIT5e1nls21FXW6P2NIgXyJbY/Kor/nz5yMwMLDK88HBwVwtlqrl6APpaioZfTtbWYa+JmqNFtN/OIRdadfRwF2OFVOi0CrYfuvs1FVdVp21lUG+RLbI5ITjwoULaN68eZXnw8PDceHCBbMERY6ltitFwPhporasuvVUgoKCMGPGDLtYhl4IgZc3HMPvJ67AXe6Cbyb0QGeFv9RhSaIut0e4VglR9Uy+pRIcHIyjR48iIiJC7/kjR46gcePG5oqLHIgzDaSTYqyFuQghMP/XU1i7/yJcZMB/xtyFe1pV7c10FnW9PSLlIN87OeqYKbJPJicco0ePxr///W80bNhQt0z9tm3bMH36dIwaNcrsAZL9c7aBdFJWt6yPL7el4evt/wAA3hvWGYM6hkockbRqG5dT0+BfW0g8HXnMFNknkxOOt956CxkZGejfvz9cXW+9XavVYsKECRzDQQZxIJ3tW733Aj747TQA4JWh7TGih1LiiKRX3xooUiaeps6uIbKGOi/edubMGRw5cgReXl6IjIxEeHi4uWOzGi7eZlmVK57WdqVo7RVP6ZZfjl3GM4kHoRXA07Et8eLgdlKHZFPsbdVZW11hmByXxVeLdSRMOCyv8ooLgMErRV5xSWPH2RxMWb4PFRqBMdHN8M7DnVhjxwB7GgthiysMk2Mz62qxM2fOxFtvvYUGDRpg5syZNW67YMEC0yIlp2BLA+nolkMXbuLJVQdQoRG4P7IJ3opjslEdexqX42xjpsh+GJVwHDp0CBUVFbp/V4d/rKgmtjCQjm45c6UAk5btQ3G5Br1bB2LByC6Qc+VXh8AxU2SreEsFvKVCziXzRjGGL96FK/lluKuZP75LiEYDD5PHj5ON4pgpsjZj21CuL03kRHIKyjB+SSqu5JehTYgPlk3qyWTDwbD4GNkqo/7SmHJ/3d5LVBM5qrySCkxYuhcZ14uhaOSFVQnR8Pd2lzossgCOmSJbZFTC4efnp/u3EAIbNmyAn58fevToAQA4cOAAcnNzeRKT1dnT7AEplZRr8NiKffj7cj4CfTzwXUI0Qnw9pQ6LLIhjpsjWGJVwLFu2TPfv2bNnY8SIEVi8eLHuxNVoNHj66ac5/oGsipUUjVOh0WJa4kHsy7iJhp6uWDklChGBDaQOi6zAnmbXkOMzedBoUFAQdu7cibZt2+o9f/r0adxzzz24fv26WQO0Bg4atT/VVVJkXQ99Wq3AzLWH8ePhS/B0c8GqhGj0jAiQOiwiciAWGzSqVqtx6tSpKs+fOnUKWq3W1N0RmcxZVp+tLyEE3tx0Ej8evgRXFxm+HNudyQYRScbk4emTJ09GQkIC0tLSEBUVBQBITU3Fe++9h8mTJ5s9QKI7SbH6bOVYkaysLOTk5CAoKAhhYWE2fU980ZazWL4rAwDw8Ygu6NsuWNqAiMipmZxwfPTRRwgNDcXHH3+sq1TXpEkTvPDCC3j++efNHiDRnTZu3GjUduaqpGhorEglWx0zsvyvdCz88ywAYN5DHRHXNUziiBwfBzAT1axehb/y8/MBwO7HPXAMh/3QaDQIDQ3FtWvXat3WHGtFVDdW5E7z5s3DK6+8YhMNzI+HsjBjzWEAwIwBrTFjQBtpA3ICHMBMzsyihb/UajX+/PNPrF69WjdI79KlSygsLKxbtERG2rFjh1HJRlBQEHr37l2vz6pprMid5s6di/DwcMnr0Gw9dQXPrzsCAJh0TwSm928taTzOoDIpvbMHrHIpeKnPCSJbYXLCcf78eURGRiIuLg7Tpk1DTk4OAOD999/HrFmzzB4g0e2MvU0yduzYevc21DZW5E5SNzB702/gqe8OQqMVeOSuMLz+QAeub2RhHMBMZDyTE47p06ejR48euHnzJry8vHTPP/LII9iyZYtZgyO6k7ELTsXFxdX7s+o6BkSKBubEpTwkLN+HMrUW/dsF44PhneHCxdgszpQBzETOzuSEY8eOHXj11Vfh7q5fErlysSAiS+rduzcUCkWNV+5KpbLet1OAuq2mKUUDk3GtCBOX7kNBmRpREQH4fGw3uMm5TJI1cCl4IuOZ/FdJq9UavHq7ePEiGjZsaJagiKpT28JUMpnMbAtTVSY3dWGtBiY7rxTjlqTiWmEZOjTxxbeTesDTTfqBq86CS8ETGc/khGPgwIFYuHCh7meZTIbCwkLMnTsXQ4cONWdsRAZVLkwVFqY/1VOhUOgqjGo0GqSkpGD16tVISUmp0y2OyuSmLuMgrNHA5BaXY8LSVFy8WYKIxt5YMSUKvp5uFv9c+p/aetxkMpnZetyI7J3J02IzMzMxePBgCCFw9uxZ9OjRA2fPnkVgYCC2b9+O4GD7Ky7EabH2qbq6B+aeolhTHY47yWQyKBQKpKenW3SKbFGZGmO/TcXhzFyE+Hpg/dR7oAzwttjnUfUqZ6kA0Bs8yjL75CyMbUPrVIdDrVZjzZo1OHLkCAoLC9GtWzeMHTtWbxCpPWHC4TgstcbK7ZVG//zzTyxfvrzKNuZsYGoqIlWm1uCxFfux4+w1+Hu7Ye2TMWgTwtuZUjKUlCqVSi4FT07BIglHRUUF2rVrh02bNqF9+/ZmCdQWMOFwDBqNBhEREdX2RJjS+1Bb1UhLNjA19dDEPfwI/r36EH4+dhne7nJ8/1g07mrWqF6fR+bBSqPkrCzWwxEWFoY///yTCQfZnJSUFPTt27fW7WqrQGrsLRlLNDC19dCM+PBH7MmRw00uw9JJPdG7dVC9Po+IqL6MbUNNXktl2rRpeP/99/Htt9/C1dXktxNZjDmmKFbX4FcW9br9dolcLq9z6XRDyQqAGotI+feZiD05crjIgEWj7qpXssGrcSKyNpMzhn379mHLli34448/EBkZiQYNGui9zjK+JJX6TlGsrWqkTCbDjBkzEBcXV6/GuboelMcff7za20G+UY/A7+5HAQATOrhjaGTdZ8Fw3Q8ikoLJCYe/vz+GDRtmiViI6qVyimJtYziqm6JojWXva+pBmTt3rsH3NIgcgEZ9EwAAN1OWoW2Xuv/+mdKDQ0RkTiYnHMuWLbNEHET1JpfLMXr0aHz44YfVblNTUTBLV400Zt2NO3m1jkHjwc8CAPJSk5CfmoQmTZ6xyOcb24PD2zFEVBdGF/7SarV4//330atXL/Ts2RMvvfQSSkpKLBkbkUmSk5Px0UcfVfv6rFmzarx6t3TVSFMXg/Ns1hlBD70ImYscBUd+R9625fUqImWOdT+Sk5MRERGBvn37YsyYMejbty8iIiJs+laqOYrAEVH9GZ1wvPPOO3j55Zfh4+ODsLAwLFq0CNOmTbNkbERGq20peZlMhh9++KHGxsbSVSNN6RnxaNIGQfGvQubqhuLTu3Dzjy8A1NxDY67Pr247e1yG3R4TJCKHJYzUqlUrsXjxYt3PmzdvFu7u7kKj0Ri7C5uVl5cnAIi8vDypQ6E6UqlUAkCtD5VKVeN+kpKShEwmEzKZTO99lc8lJSVZPMbn5r4vmk3/QYTP3iSCR74jIHcTSqWyXp9tyucbOkZqtVooFIpq3yOTyYRSqRRqtbpeMZpT5f+loVjr+39JRP9jbBtqdA/HhQsX9NZKGTBgAGQyGS5dumRCekNkGeYaf2HMOi01qan73qgelLadscujO2SePmju54L3H2gO1Z9/ID09vd6DOevTg2Nvy7AbM15mxowZvL1CZEVGDxpVq9Xw9PTUe87NzQ0VFRVmD4rIVOYcfxEfH4+4uDiTB0bWNt20cjG44cOHQyaTVVl3w8XbDyGj3kZ2XilaBftg7ZMxCGjgbtT3MkZtnw9Uf8vG3pZht8aMIyIyjdEJhxACkyZNgoeHh+650tJSTJ06Va8WB++NkhQqr96zsrIMXtXWNiX2TqYW9TJ2umllD0qVxCSiFZpN+hgXi10Q5u+FVQlRZk02KlX7+QpFjWXZ7W0ZdntLkIicgdGlzSdPnmzUDu1x2ixLmzsGc67aacrUz7qs4XL7/hsHh2LJOU+kpt9A4wbuWDc1Bi2CfIz+3nVh6tTWyu9YW0JXl1VyLTHN1lxl7omodka3oZYbRmI/OGjUcSQlJVUZ3GjqgEtD+1AoFNXuo66DMdVqtfhzy1Yx+J0NInz2JtHx9V/FsYu59fn6FmWJAbWmHmtjVQ5yNTRotDJmWxvkSmSvzD5o1BK2b9+OBx98EE2bNoVMJsOPP/6o97oQAq+//jqaNGkCLy8vDBgwAGfPntXb5saNGxg7dix8fX3h7++PhIQEFBYWWvFbkLUYU08hPj4eGRkZUKlUSExMhEqlMmnAZV2mfhrbLZ+VlaWL/80330RERHOM/Pgn/J3vBqEuR96mD3AmdYtR+5JCfQfU3smS02wrx6sAqDJItrbxKkRkIdbJfwz75ZdfxCuvvCKSk5MFALFhwwa919977z3h5+cnfvzxR3HkyBHx0EMPiebNm4uSkhLdNoMHDxZdunQRe/bsETt27BCtWrUSo0ePNikO9nDYPlOvhNVqtVCpVCIxMVGoVCqjrmTrOvXT2B6OoKAgvZ8b9XtMhM/eJJq9sFF4tYqym+madTm2hvZhjWm25ujxIqKaGduG2swtlTsTDq1WK0JDQ8WHH36oey43N1d4eHiI1atXCyGEOHnypAAg9u3bp9vm119/FTKZTGRlZRn92Uw4bJup9RTq2k1fn1sjNXXfG3r4xowQ4bM3ifDZm0SDjv2crqvfXHVTjGGOBImIqmcXt1Rqkp6ejuzsbAwYMED3nJ+fH6Kjo7F7924AwO7du+Hv748ePXrothkwYABcXFyQmppa7b7LysqQn5+v9yDbZGo9hfp009d1ZkNN3feG+HQdgkb3TgAA3NjyNYpObNW9JmysnoWlWHMWSeWMo9GjRyM2Npa3UYgkYrMJR3Z2NgAgJCRE7/mQkBDda9nZ2QgODtZ73dXVFQEBAbptDJk/fz78/Px0D6VSaeboyVxMqadQ32JP9Zn6Wd34hqCgIL2fvdvfi4CBTwEAcnf9gIL9Pxn8DEefrmlv02yJqP5sNuGwpDlz5iAvL0/3yMzMlDokqoYpV8L1rYZZ37VUDA1Y/eSTT3SvezbvhsD7Z0Imc0HBwU3I2/FdtbE6ekNr6XVriMj22GzCERoaCgC4cuWK3vNXrlzRvRYaGoqrV6/qva5Wq3Hjxg3dNoZ4eHjA19dX70G2yZQr4fp205tjZsOd3feVPR4eYe0Q9MjLkMldUXRyG25s/srg+52loeUsEuNwpVtyJDabcDRv3hyhoaHYsuV/0wTz8/ORmpqKmJgYAEBMTAxyc3Nx4MAB3TZbt26FVqtFdHS01WMm89JoNNBoNAgICKh2m9sbaHN005t76mfv3r2h6BSF4OFvwMXNEyX/7Me1nz/BrTGRVb8L4DwNrbmPtaPhSrfkcCw9erUmBQUF4tChQ+LQoUMCgFiwYIE4dOiQOH/+vBDi1rRYf39/sXHjRnH06FERFxdncFrsXXfdJVJTU8XOnTtF69atOS3WARiaaXLn485ZKuYs9mSumQ3nrxWJTq/+V4TP3iRCx34gZK4e1X4fZ52uyVkkVXGlW7IndjEttrqpcRMnThRC3Joa+9prr4mQkBDh4eEh+vfvL06fPq23j+vXr4vRo0cLHx8f4evrKyZPniwKCgpMioMJh22p7o+tMQ20JZeXN9WVvBLR+/2tInz2JnH3Gz8JRYs2Vabqzps3jw0t6bFWjRIiczG2DTV6LRVHxrVUbEdt65IAQOPGjbFmzZpqpzgaWrVVqVTWuDiZueUVV2Dk17txKrsAzQK8MLMzUHT9MnJychAUFISwsDCzrBnibCyx7oqt4TowZG+MbUONXi2WyBpqm2kCANevX4dcLq+2oanr8vLmUlyuxpQV+3AquwAN3QTSlz2PR04d1r1euWS9ozWUlmYokaw8lo403oMr3ZKjYsJBNsVcf2xNXV7eXMrVWjz13UEcOH8TXnKBM189i/KcDL1t7lyynmpXWdDtzg5ZRzyWrFFCjspmZ6mQc7LnP7YarcDz645g25kceLq5oPCXD6skG4BxRcjof+pb0M3esEYJOSomHGRT7PWPrRACb/x0Av89cgluchme7uyGi4e317i9M5QwN4f6FnSzN6xRQo6KCQfZFHv9Y/vJ5jNYtec8ZDJgwYiuCFLnGPU+3oevnTOOaWCNEnJETDjI5tjbH9ulO9Px6dZzAIC34jrhwS5N7frWkK1x1mNpqFR+enq6zZ3/RMbitFhwWqy1mDql0R6mQCYduIjn1x0BAMwa2AbP9GsN4H/Te7OysgyOPZDJZFAoFEhPT7e572RreCyJbJuxbSh7OMgqTC3TbA/JxuaTV/Bi0lEAwJRezTGtbyvda+a4NcR1NG6x19tsdcX/d3JYlqs9Zj9YadSyTC3TbKisuUKhsKlyzrvTronWr/wiwmdvEjPXHBYajdbgdoa+izElzG31GEhZhryux9Ke2Or/O1FN7KK0ua1gwmE5ppZptoc1JI5dzBUdX/9NhM/eJBKW7xMVak2N25vaSNvqMbCFxtCR112x1f93otqwtLkJOIbDckwp09y7d+8ay5rbwr36f3IK8eji3bheVI7o5gFYMSUKnm51i8XQbSMANnkMqiu8VXlLwxYH89qT2kr628K5T1QdjuEgm2DKlEZbr7dwKbcE45fsxfWicnQK88W3E3vUOdmobkzLO++8Y3PHwNkKb0nB1s99InNgwkEWZcqURluut3CjqBzjl6QiK7cELQIbYPnkKDT0dKvTvip7C+5sYLKysjB37lyj9mHNY8DG0PJs+dwnMhcmHGSQuUbKm1I51FbrLRSWqTF52V6k5RShiZ8nVj0WjUAfjzrty5jeAmNY8xiwMbQ8Wz33icyJCQdVYeoU1pqYMqXRFsual6k1eHLVfhy5mIdG3m5YlRCFMH+vOu/PmNVwayLFMWBjaHm2eO4TmRsTDtJTU3f/8OHD65R0GFs51NbqLag1WkxffRh/nbuOBu5yLJ8chVbBDeu1T1N6AWzhGABsDK3B1s59Iouw5FQZe8FpsbcYO4W1rKysTlMTjZ3SaAv1FrRarXhx3RERPnuTaP3yL2Ln2Ryz7FelUlV7fG9/zJs3T/JjcLvKKZt3TtvklE3zsoVzn8hUnBZrAk6LvcXYKayBgYG4du2a7meFQoFFixaZdVqk1JVG5//yN77a/g9cZMAXY7tjcKdQs+zXlDLdAGyq2mpycjKmT5+u1/ulVCqxcOFCTok1I6nPfSJTGduGMuEAE45Kq1evxpgxY0x+n6PVYvgyJQ3v/3YKAPDBsM4Y0VNp1v1X3rYC9AeK2sNxZGNIRHdiwmECJhy3GNvDYYijFCZavfcC5iQfAwC8PLQdnri3ZZVtzNHosreAiBwFEw4TMOG4pbbufmOoVCrExsaaNzAr+eXYZTyTeBBaATwV2xKzB7erso2hRKGut5TYW0BEjsDYNtTVijGRjascKT98+HDIZLI6JR32Woth59lrmPHDYWgFMDqqGV4c1LbKNtWV966cwWPqrRC5XG63yRkRkak4LZb0VDeFNSgoyKj3W6oWgyWX7D504SaeWLUf5Rot7o9sgrcf7lRlaiLLexMR1Q8TDqoiPj4eGRkZUKlUSExMhEqlwsWLFyWrxWDOQmR3OnOlAJOX70NxuQa9WwdiwcgukLtU/Y4s701EVD+8pUIGGerur+52iyULE5n7NsbtMm8UY/ySVOQWV6Cr0h+Lx3WHh6vh+Fnem4ioftjDQUYztmKouVjyNkZOQRnGL0nFlfwytA72wbJJPdHAo/r8m+W9iYjqhz0cZJL4+HjExcVZZXaFKbcxTBl8mV9agYlL9yLjejEUjbywKiEajRq41/ieyvLeNc3gkcvlyMnJMToOIiJnwoSDTGat2RWWuI1RUq7BY8v34+TlfAT6uGNVQjRC/Txrfd/tM3iqo9FoMHLkSMjlctbSICK6A2+pkM0y922MCo0W0xIPYm/GDTT0cMWKKVFoHtjA6Hji4+Oxdu3aWntzOFuFiKgqJhxks8y5SqlWK/Di+qPYeuoqPFxdsGRST3Rs6mdyTIGBgTUmE5ytQkRkGG+pkNmZq4JmTYXITJkZI4TAm5tOYsOhLLi6yPDluG6Iah5gcjwAZ6tYAyuwEjkm9nCQWZm7ZoY5ZsZ8uuUclu/KAAB89GgX9GsXUqdYAMeerWLJ4mrGsmTNFSKSFtdSAddSMZfqamaYYxXUyqverKws5OTkICgoCGFhYbVe/a7YlYG5P50AALzxYAdM6tW8Tp9/exzGLi9vT1fl5lwjpj4xWOr8ISLLMboNFSTy8vIEAJGXlyd1KHZLrVYLhUIhABh8yGQyoVQqhVqtrvNnJCUlVfkMhUIhkpKSDG7/46GLInz2JhE+e5NY8MfpOn+uoThkMpmQyWRVvqNMJtPFo1arhUqlEomJiUKlUtXru1tS5fcx9H92+/exJGucP0RkGca2oUw4BBMOc1CpVNU2Frc/VCpVnfZvaqO49e8rouWcn0X47E3i9R+PCa1Wa4ZvqR/PnQ2kUqnUxWFqciQVW2noLX3+EJHlGNuGcgwHmYUlB1OaWnF0X8YNTP3uANRagYe7NsXcBztWO9OlrgytN5Oeno74+HjdrYE7i5ZVlmO3pfEItrJGDAfjEjk+zlIhs7DkYEpTGsXgNt0wZfk+lKm16NcuGB8+2gUuBhZjMwdDBdBqS45kMhlmzJiBuLg4mxjjYSsNvSMPxiWiW9jDQWZhzpoZdzK2sTuWkY0JS/eioFSNnhGN8PmYbnCTW/cUt5UeA2PZSkNvyfOHiGwDEw4yi8qaGQCqNBr1XU3WmMZO7hOAFed9ca2wDO2b+OLbiT3h5W79HgRb6TEwlq009JY8f4jINjDhILOx1GqytTWKcq+GCBv7Hq6VCEQ09sbKKVHw83Kr02fVl630GBjLlhp6a69GTETWxTocYB0Oc7NEpcjKgZgA9MZHuLh7InjE2/AIa4cQXw+sn3oPlAHe9fqs+rDXOh2G6nAolUosXLjQ6g09K40S2Rdj21AOGiW7UHn1q9coyl2hGPUmZE3awc/LDasSoiVNNgDzlWO3tvj4eMTFxdlEQ2+t1YiJyLrYwwH2cJiTpStW6iqOXrqMH682wr5sDbzd5fjusWh0a9ao3vs3F1vqMSAisiRj21AmHGDCYS7WKk0thMDLG45j9d4LcJPLsHRST/RuHaS3jS10y9tCDERElsaEwwRMOOqvcuxCdVNCzTl24YPfTuGLlDTIZMBno7vh/s76AzBtYV0QIiJnYWwbylkqZBbWqj/xzfZ/8EVKGgDg3UciDSYb9lLlk4jImTDhILOwRv2Jtfsz8c4vfwMAXhzcFqOjmum9bmoJdCIish4mHGQWlq4/8euxS5i9/igAYHCEG574V0SVbeytyicRkTNhwkFmYcmKle8v34CpK/dCACg8+ge+emoQmjdvXuX2iL1V+SQiciZMOMgsLFWxctHKDfj8qBoyuRuKT+/C9d8+A2B4TIa9VfkkInImTDjIbMxdmvpMdh4WHCyDi7s3SjKOIOe/HwJCC8DwmAxbWReEiIiq4rRYcFqsuZmj/kRWbgkeXKjCjVKBsstncOWHVyDKSwxuq1KpdJUpqyuBbu5aIEREdAtLm5Nk6lOaWqPR4Oct2/HWXwW4USZHxfVMXF33RrXJBqA/JsNgCXTc6mVhlU8iIukw4SCbkZycjOnPz4a6zzPwCG0Fdf5VXFnzGrQl+TW+784xGba0LggREd3ChINsQnJyMoaPHI3g4W/AM7QVNEW5uLLmNWgKrlX7nsrqpYbGZHABMCIi28JBoyQ5jUaD6TOeQ+CDL8AzvDO0ZcW4um4u1Deyqn2PLa+8SkREVbGHgyS3bft2lEQ+Ap82MRDqclxNehPlV9JqfE9tYzK4cBoRkW1hwkGSEkJg6cGb8Im8D0KrQc7G91CWedzgtq+++io6dOhQawLBxduIiGwPEw6S1BcpadiZ4wEAuP7LIpSc21vttv379691XEbltNg7Z3tXFgrjtFgiImlwDAdJ5rs95/Hh76cBANoD61B8UmVwO2MLdnHxNiIi28WEgyTx3yOX8NrGW7dOnunbCgufeghA/cqic/E2IiLbxYSDrG7bmRzMXHsYQgBjo5vh+YFtzFIWnYu3ERHZLo7hIKs6cP4Gpq46gAqNwAOdm+DNuE66Xoz6Fuzi4m1ERLbLpns43njjDchkMr1Hu3btdK+XlpZi2rRpaNy4MXx8fDBs2DBcuXJFwoipJqey8zF52T6UVGjQp00QFozoCrmL/i2UyoJdo0ePRmxsrElTWbl4GxGR7bLphAMAOnbsiMuXL+seO3fu1L323HPP4b///S/WrVuHbdu24dKlS5yBYKMuXC/GhCV7kV+qRvfwRvhyXDe4u5r39JPL5Vi0aBGA+o0FISIi87P5hMPV1RWhoaG6R2BgIAAgLy8PS5YswYIFC9CvXz90794dy5Ytw65du7Bnzx6Jo6bbXS0oxfilqbhaUIZ2oQ2xdGJPeLtb5m6eOcaCEBGR+dn8GI6zZ8+iadOm8PT0RExMDObPn49mzZrhwIEDqKiowIABA3TbtmvXDs2aNcPu3btx9913V7vPsrIylJWV6X7Oz695cTCqu7ySCkxYshfnrxdDGeCFlVOi4OftZtHP5OJtRES2x6YTjujoaCxfvhxt27bF5cuXMW/ePPTu3RvHjx9HdnY23N3d4e/vr/eekJAQZGdn17jf+fPnY968eRaMnACgpFyDhOX7cCq7AEENPfBdQjSCfT2t8tlcvI2IyLbYdMIxZMgQ3b87d+6M6OhohIeHY+3atfDy8qrzfufMmYOZM2fqfs7Pz4dSqaxXrKSvXK3FU98fwP7zN+Hr6YqVU6IQ3riB1GEREZFEbH4Mx+38/f3Rpk0bnDt3DqGhoSgvL0dubq7eNleuXEFoaGiN+/Hw8ICvr6/eg8xHqxWYte4IUk7nwNPNBUsn9UT7JjzGRETOzK4SjsLCQqSlpaFJkybo3r073NzcsGXLFt3rp0+fxoULFxATEyNhlM5NCIG5P53AT0cuwdVFhsXjuqNHRIDUYRERkcRs+pbKrFmz8OCDDyI8PByXLl3C3LlzIZfLMXr0aPj5+SEhIQEzZ85EQEAAfH198eyzzyImJqbGAaNkWZ/8eRar9pyHTAYsGNkVsW2DpQ6JiIhsgE0nHBcvXsTo0aNx/fp1BAUF4V//+hf27NmDoKAgAMAnn3wCFxcXDBs2DGVlZRg0aBC++OILiaN2Xsv+SsenW84CAN6M64SHujSVOCIiIrIVMmFoaU0nk5+fDz8/P+Tl5XE8Rx0lH7yImWuPAACev68Nnu3fWuKIiIjIGoxtQ+1qDAfZpj9PXsEL648CAKb0ao5n+rWSOCIiIrI1TDioXlL/uY5piQeh0QrEdwvDq/e3r3YtEyIicl5MOKjOjmfl4bEV+1Gm1mJA+2C8P6wzXFyYbBARUVVMOKhO/skpxMSle1FQpkZU8wB8NqYb3OQ8nYiIyDC2EGSyy3klGL9kL64XlaNjU198O7EHPN24TgkREVWPCQeZ5GZROcYv2Yus3BK0CGyAFVOi4Otp2cXYiIjI/jHhIKMVlqkxafk+nLtaiCZ+nliZEIVAHw+pwyIiIjvAhIOMUqbW4MlV+3EkMxeNvN2wKiEKikbeUodFRER2ggkH1UqjFZjxw2H8de46GrjLsXxyFFoFN5Q6LCIisiNMOKhGQgi8nHwMvx7PhrvcBV9P6IEuSn+pwyIiIjvDhINq9N5vp7BmfyZcZMCno7uiV6tAqUMiIiI7xISDqrV4Wxq+2vYPAGB+fCQGd2oicURERGSvmHCQQT/svYD3fj0FAJgzpB1G9mwmcURERGTPmHBQFb8eu4yXNxwDAEzt0xJP9mkpcURERGTvmHCQnp1nr2H6D4ehFcDoKCVmD24rdUhEROQAmHCQzuHMXDyxaj/KNVoMjQzF2w9HcuVXIiIyCyYcBAA4e6UAk5btRXG5Br1bB+KTkV0h58qvRERkJkw4CBdvFmP8kr3ILa5AV6U/Fo/rDg9XLsZGRETmw4TDyeUUlGH8kr3Izi9F62AfLJvUEw08XKUOi4iIHAwTDieWX1qBiUv3Iv1aEcL8vbAqIRqNGrhLHRYRETkgJhxOqrRCg8eW78fJy/kI9HHHd49FI9TPU+qwiIjIQTHhcEIVGi2mfX8QezNuoKGHK5ZPjkLzwAZSh0VERA6MCYeT0WoFXlx/FFtOXYWHqwuWTOqJTmF+UodFREQOjgmHExFC4K2fT2LDoSzIXWT4Ymw3RDUPkDosIiJyAkw4nMh/tp7Dsr8yAAAfP9oF/duHSBsQERE5DSYcTmLV7gws2HwGAPDGgx3w8F1hEkdERETOhAmHE9h4OAuv/3QCADC9f2tM6tVc4oiIiMjZMOFwcKpTV/H82iMQApgYE44ZA1pLHRIRETkhJhwObF/GDTz1/QGotQJxXZti7oMduRgbERFJggmHgzp5KR9Tlu9DaYUWfdsG4aNHu8CFi7EREZFEmHA4oIxrRZiwdC8KStXoGdEIX4ztDjc5/6uJiEg6bIUczJX8UoxbkoprhWVo38QX307sCS93rvxKRETSYsLhQHKLyzFhyV5cvFmC8MbeWDGlJ/y83KQOi4iIiAmHoyguV2PK8n04faUAIb4e+C4hGsENuRgbERHZBiYcDqBcrcWTqw7g4IVc+Hm5YeWUaCgDvKUOi4iISIcJh53TaAWeW3sYO85eg5ebHMsm90Tb0IZSh0VERKSHCYcdE0LgtY3H8fPRy3CTy/DV+O7o1qyR1GERERFVwYTDjn30x2kkpl6ATAZ8MrIr7m0TJHVIREREBjHhsFPf7vgHn6vSAADvPByJBzo3lTgiIiKi6jHhsEPr9mfi7Z//BgC8MKgtxkQ3kzgiIiKimjHhsDN/nMjGS8nHAACP926Op2NbShwRERFR7Zhw2JHdadfxzOpD0GgFHu2uwMtD23MxNiIisgtMOOzEsYt5eHzlfpSrtRjYIQTz4yOZbBARkd1gwmEHzl0txMRle1FYpsY9LRvj09F3wZWLsRERkR1hq2XjsnJLMGFJKm4UlaOzwg9fT+gBTzcuxkZERPaFCYcNu15YhvFLUnEprxQtghpg2aSe8PFwlTosIiIikzHhsFEFpRWYtGwf/skpQlM/T3yXEI3GPh5Sh0VERFQnTDhsUGmFBk+sPIBjWXkIaOCOVY9Fo6m/l9RhERER1RkTDhuj1mjx79WHsPuf6/DxcMWKyVFoGeQjdVhERET1woTDhgghMCf5GP44eQXuri74ZkIPRCr8pA6LiIio3phw2AghBN795W+sO3ARchcZPht9F2JaNpY6LCIiIrNgwmEjvkhJwzc70gEA7w/rjIEdQyWOiIiIyHyYcNiA71PP48PfTwMAXr2/PYZ3V0gcERERkXkx4ZDYpqOX8OqPxwEA0/q2xGO9W0gcERERkfkx4ZDQtjM5eG7NYQgBjIluhlkD20odEhERkUUw4ZDIgfM3MXXVAVRoBB7o3ARvxXXiYmxEROSwmHBI4HR2AaYs34eSCg3ubROEBSO6Qu7CZIOIiBwXEw4ry7xRjPFLUpFXUoFuzfyxeFw3uLvyv4GIiBwbWzorulpQinFLUnG1oAxtQxpi6aSe8HbnYmxEROT4mHBYSV5JBSYu3Yfz14uhDPDCqoQo+Hu7Sx0WERGRVTDhsIKScg0Slu/D35fzEdTQA98lRCPY11PqsIiIiKyGCYeFVWi0ePr7A9h//iYaerpi5ZQohDduIHVYREREVsWEw4K0WoFZ645AdToHnm4uWDapJ9o38ZU6LCIiIqtzmITj888/R0REBDw9PREdHY29e/dKGo8QAm/89wQ2Hr4EVxcZvhzXHT0iAiSNiYiISCoOkXCsWbMGM2fOxNy5c3Hw4EF06dIFgwYNwtWrVyWL6ZM/z2Ll7vOQyYCPR3RB37bBksVCREQkNYdIOBYsWIDHH38ckydPRocOHbB48WJ4e3tj6dKlksSz7K90fLrlLADgzYc6Iq5rmCRxEBER2Qq7TzjKy8tx4MABDBgwQPeci4sLBgwYgN27dxt8T1lZGfLz8/Ue5rLh0EXM++9JAMDM+9pgfEyE2fZNRERkr+w+4bh27Ro0Gg1CQkL0ng8JCUF2drbB98yfPx9+fn66h1KpNEssW/6+glnrjgIAJveKwLP9Wpllv0RERPbO7hOOupgzZw7y8vJ0j8zMzHrvM/Wf63j6+4PQaAXi7wrDa/d34GJsRERE/8/u62oHBgZCLpfjypUres9fuXIFoaGhBt/j4eEBDw8Ps8VwPCsPj63YjzK1FgPaB+P94Z3hwsXYiIiIdOy+h8Pd3R3du3fHli1bdM9ptVps2bIFMTExFv/89GtFmLRsLwrK1IhqHoDPxnSDm9zuDysREZFZ2X0PBwDMnDkTEydORI8ePRAVFYWFCxeiqKgIkydPtujnXs4rwbhvU3GtsBwdm/ri24k94Okmt+hnEhER2SOHSDhGjhyJnJwcvP7668jOzkbXrl3x22+/VRlIak43i8oxYcleZOWWoHlgA6yYEgVfTzeLfR4REZE9kwkhhNRBSC0/Px9+fn7Iy8uDr2/tpceLytQY820qjmTmItTXE+ufioGikbcVIiUiIrItxrahHGxgojK1Bk+uOoAjmbnw93bDqoQoJhtERES1YMJhAo1W4Lk1h7Hz3DV4u8uxfHIUWoc0lDosIiIim8eEw0hCCLyy4Rh+OZYNd7kLvpnQA12V/lKHRUREZBeYcBjp/d9O44d9mXCRAZ+O7operQKlDomIiMhuMOEwwlfb0rB4WxoAYH58JAZ3aiJxRERERPaFCUct1uy7gPm/ngIAvDSkHUb2bCZxRERERPaHCUcNfjt+GXOSjwEAnuzTAlP7tJQ4IiIiIvvEhKMaO89ew79XH4ZWAKN6KvHS4HZSh0RERGS3mHAYcDgzF0+s2o9yjRZDOoXinUciufIrERFRPTDhuMO5qwWYvGwviss1+FerQCwc1RVyrvxKRERULw6xloq5ZN0sxuTvT+BmcQW6KP3x1fju8HDlYmxERET1xYTjNk+sOoDsQqB1sA+WT+qJBh48PERERObAFhW3qogCQPqlawgLDsAXj7aHXFOK/PxSiSMjIiKybfn5+QD+15ZWh6vFArh48SKUSqXUYRAREdmtzMxMKBSKal9nwgFAq9Xi0qVLaNiwYb1no+Tn50OpVCIzM9Oope6dAY+JPh6PqnhM9PF46OPxqMqWjokQAgUFBWjatClcXKqfi8JbKgBcXFxqzMrqwtfXV/KTwNbwmOjj8aiKx0Qfj4c+Ho+qbOWY+Pn51boNp8USERGRxTHhICIiIotjwmFmHh4emDt3Ljw8PKQOxWbwmOjj8aiKx0Qfj4c+Ho+q7PGYcNAoERERWRx7OIiIiMjimHAQERGRxTHhICIiIotjwkFEREQWx4TDzD7//HNERETA09MT0dHR2Lt3r9QhWcUbb7wBmUym92jXrp3u9dLSUkybNg2NGzeGj48Phg0bhitXrkgYsflt374dDz74IJo2bQqZTIYff/xR73UhBF5//XU0adIEXl5eGDBgAM6ePau3zY0bNzB27Fj4+vrC398fCQkJKCwstOK3MJ/ajsekSZOqnDODBw/W28aRjsf8+fPRs2dPNGzYEMHBwXj44Ydx+vRpvW2M+T25cOEC7r//fnh7eyM4OBgvvPAC1Gq1Nb+KWRhzPGJjY6ucI1OnTtXbxlGOBwB8+eWX6Ny5s66YV0xMDH799Vfd6/Z+fjDhMKM1a9Zg5syZmDt3Lg4ePIguXbpg0KBBuHr1qtShWUXHjh1x+fJl3WPnzp2615577jn897//xbp167Bt2zZcunQJ8fHxEkZrfkVFRejSpQs+//xzg69/8MEH+PTTT7F48WKkpqaiQYMGGDRoEEpL/7dI4NixY3HixAls3rwZmzZtwvbt2/HEE09Y6yuYVW3HAwAGDx6sd86sXr1a73VHOh7btm3DtGnTsGfPHmzevBkVFRUYOHAgioqKdNvU9nui0Whw//33o7y8HLt27cKKFSuwfPlyvP7661J8pXox5ngAwOOPP653jnzwwQe61xzpeACAQqHAe++9hwMHDmD//v3o168f4uLicOLECQAOcH4IMpuoqCgxbdo03c8ajUY0bdpUzJ8/X8KorGPu3LmiS5cuBl/Lzc0Vbm5uYt26dbrn/v77bwFA7N6920oRWhcAsWHDBt3PWq1WhIaGig8//FD3XG5urvDw8BCrV68WQghx8uRJAUDs27dPt82vv/4qZDKZyMrKslrslnDn8RBCiIkTJ4q4uLhq3+PIx0MIIa5evSoAiG3btgkhjPs9+eWXX4SLi4vIzs7WbfPll18KX19fUVZWZt0vYGZ3Hg8hhOjTp4+YPn16te9x5ONRqVGjRuLbb791iPODPRxmUl5ejgMHDmDAgAG651xcXDBgwADs3r1bwsis5+zZs2jatClatGiBsWPH4sKFCwCAAwcOoKKiQu/YtGvXDs2aNXOaY5Oeno7s7Gy9Y+Dn54fo6GjdMdi9ezf8/f3Ro0cP3TYDBgyAi4sLUlNTrR6zNaSkpCA4OBht27bFU089hevXr+tec/TjkZeXBwAICAgAYNzvye7duxEZGYmQkBDdNoMGDUJ+fr7uKthe3Xk8Kn3//fcIDAxEp06dMGfOHBQXF+tec+TjodFo8MMPP6CoqAgxMTEOcX5w8TYzuXbtGjQajd5/NACEhITg1KlTEkVlPdHR0Vi+fDnatm2Ly5cvY968eejduzeOHz+O7OxsuLu7w9/fX+89ISEhyM7OliZgK6v8nobOj8rXsrOzERwcrPe6q6srAgICHPI4DR48GPHx8WjevDnS0tLw8ssvY8iQIdi9ezfkcrlDHw+tVosZM2agV69e6NSpEwAY9XuSnZ1t8ByqfM1eGToeADBmzBiEh4ejadOmOHr0KGbPno3Tp08jOTkZgGMej2PHjiEmJgalpaXw8fHBhg0b0KFDBxw+fNjuzw8mHGQWQ4YM0f27c+fOiI6ORnh4ONauXQsvLy8JIyNbNWrUKN2/IyMj0blzZ7Rs2RIpKSno37+/hJFZ3rRp03D8+HG9cU7OrLrjcft4ncjISDRp0gT9+/dHWloaWrZsae0wraJt27Y4fPgw8vLysH79ekycOBHbtm2TOiyz4C0VMwkMDIRcLq8yYvjKlSsIDQ2VKCrp+Pv7o02bNjh37hxCQ0NRXl6O3NxcvW2c6dhUfs+azo/Q0NAqA4zVajVu3LjhFMepRYsWCAwMxLlz5wA47vF45plnsGnTJqhUKigUCt3zxvyehIaGGjyHKl+zR9UdD0Oio6MBQO8ccbTj4e7ujlatWqF79+6YP38+unTpgkWLFjnE+cGEw0zc3d3RvXt3bNmyRfecVqvFli1bEBMTI2Fk0igsLERaWhqaNGmC7t27w83NTe/YnD59GhcuXHCaY9O8eXOEhobqHYP8/HykpqbqjkFMTAxyc3Nx4MAB3TZbt26FVqvV/aF1ZBcvXsT169fRpEkTAI53PIQQeOaZZ7BhwwZs3boVzZs313vdmN+TmJgYHDt2TC8R27x5M3x9fdGhQwfrfBEzqe14GHL48GEA0DtHHOV4VEer1aKsrMwxzg+pR606kh9++EF4eHiI5cuXi5MnT4onnnhC+Pv7640YdlTPP/+8SElJEenp6eKvv/4SAwYMEIGBgeLq1atCCCGmTp0qmjVrJrZu3Sr2798vYmJiRExMjMRRm1dBQYE4dOiQOHTokAAgFixYIA4dOiTOnz8vhBDivffeE/7+/mLjxo3i6NGjIi4uTjRv3lyUlJTo9jF48GBx1113idTUVLFz507RunVrMXr0aKm+Ur3UdDwKCgrErFmzxO7du0V6err4888/Rbdu3UTr1q1FaWmpbh+OdDyeeuop4efnJ1JSUsTly5d1j+LiYt02tf2eqNVq0alTJzFw4EBx+PBh8dtvv4mgoCAxZ84cKb5SvdR2PM6dOyfefPNNsX//fpGeni42btwoWrRoIe69917dPhzpeAghxEsvvSS2bdsm0tPTxdGjR8VLL70kZDKZ+OOPP4QQ9n9+MOEws//85z+iWbNmwt3dXURFRYk9e/ZIHZJVjBw5UjRp0kS4u7uLsLAwMXLkSHHu3Dnd6yUlJeLpp58WjRo1Et7e3uKRRx4Rly9fljBi81OpVAJAlcfEiROFELemxr722msiJCREeHh4iP79+4vTp0/r7eP69eti9OjRwsfHR/j6+orJkyeLgoICCb5N/dV0PIqLi8XAgQNFUFCQcHNzE+Hh4eLxxx+vkpw70vEwdCwAiGXLlum2Meb3JCMjQwwZMkR4eXmJwMBA8fzzz4uKigorf5v6q+14XLhwQdx7770iICBAeHh4iFatWokXXnhB5OXl6e3HUY6HEEJMmTJFhIeHC3d3dxEUFCT69++vSzaEsP/zg8vTExERkcVxDAcRERFZHBMOIiIisjgmHERERGRxTDiIiIjI4phwEBERkcUx4SAiIiKLY8JBREREFseEg4iIiCyOCQcROaSIiAgsXLhQ6jCI6P8x4SAiAIBMJqvx8cYbb1gljsjISEydOtXga6tWrYKHhweuXbtmlViIyHyYcBARAODy5cu6x8KFC+Hr66v33KxZs3TbCiGgVqstEkdCQgJ++OEHlJSUVHlt2bJleOihhxAYGGiRzyYiy2HCQUQAgNDQUN3Dz88PMplM9/OpU6fQsGFD/Prrr+jevTs8PDywc+dOTJo0CQ8//LDefmbMmIHY2Fjdz1qtFvPnz0fz5s3h5eWFLl26YP369dXGMW7cOJSUlCApKUnv+fT0dKSkpCAhIQFpaWmIi4tDSEgIfHx80LNnT/z555/V7jMjIwMymUy3vDkA5ObmQiaTISUlRffc8ePHMWTIEPj4+CAkJATjx4/X601Zv349IiMj4eXlhcaNG2PAgAEoKiqq+cASEQAmHERkgpdeegnvvfce/v77b3Tu3Nmo98yfPx8rV67E4sWLceLECTz33HMYN24ctm3bZnD7wMBAxMXFYenSpXrPL1++HAqFAgMHDkRhYSGGDh2KLVu24NChQxg8eDAefPBBXLhwoc7fLTc3F/369cNdd92F/fv347fffsOVK1cwYsQIALd6gEaPHo0pU6bg77//RkpKCuLj48H1L4mM4yp1AERkP958803cd999Rm9fVlaGd999F3/++SdiYmIAAC1atMDOnTvx1VdfoU+fPgbfl5CQgCFDhiA9PR3NmzeHEAIrVqzAxIkT4eLigi5duqBLly667d966y1s2LABP/30E5555pk6fbfPPvsMd911F959913dc0uXLoVSqcSZM2dQWFgItVqN+Ph4hIeHA7g13oSIjMMeDiIyWo8ePUza/ty5cyguLsZ9990HHx8f3WPlypVIS0ur9n333XcfFAoFli1bBgDYsmULLly4gMmTJwMACgsLMWvWLLRv3x7+/v7w8fHB33//Xa8ejiNHjkClUunF2a5dOwBAWloaunTpgv79+yMyMhKPPvoovvnmG9y8ebPOn0fkbNjDQURGa9Cggd7PLi4uVW4pVFRU6P5dWFgIAPj5558RFhamt52Hh0e1n+Pi4oJJkyZhxYoVeOONN7Bs2TL07dsXLVq0AADMmjULmzdvxkcffYRWrVrBy8sLw4cPR3l5ebX7A6AX6+1xVsb64IMP4v3336/y/iZNmkAul2Pz5s3YtWsX/vjjD/znP//BK6+8gtTUVDRv3rza70JEt7CHg4jqLCgoCJcvX9Z77vaBmR06dICHhwcuXLiAVq1a6T2USmWN+548eTIyMzORnJyMDRs2ICEhQffaX3/9hUmTJuGRRx5BZGQkQkNDkZGRUWOcAPRivT1OAOjWrRtOnDiBiIiIKrFWJloymQy9evXCvHnzcOjQIbi7u2PDhg01fg8iuoUJBxHVWb9+/bB//36sXLkSZ8+exdy5c3H8+HHd6w0bNsSsWbPw3HPPYcWKFUhLS8PBgwfxn//8BytWrKhx382bN0e/fv3wxBNPwMPDA/Hx8brXWrdujeTkZBw+fBhHjhzBmDFjoNVqq92Xl5cX7r77bt2A123btuHVV1/V22batGm4ceMGRo8ejX379iEtLQ2///47Jk+eDI1Gg9TUVLz77rvYv38/Lly4gOTkZOTk5KB9+/Z1PHpEzoUJBxHV2aBBg/Daa6/hxRdfRM+ePVFQUIAJEybobfPWW2/htddew/z589G+fXsMHjwYP//8s1G3IRISEnDz5k2MGTMGnp6euucXLFiARo0a4Z577sGDDz6IQYMGoVu3bjXua+nSpVCr1ejevTtmzJiBt99+W+/1pk2b4q+//oJGo8HAgQMRGRmJGTNmwN/fHy4uLvD19cX27dsxdOhQtGnTBq+++io+/vhjDBkyxIQjRuS8ZIJzuoiIiMjC2MNBREREFseEg4iIiCyOCQcRERFZHBMOIiIisjgmHERERGRxTDiIiIjI4phwEBERkcUx4SAiIiKLY8JBREREFseEg4iIiCyOCQcRERFZ3P8BcihnF4yFfXMAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regr_catboost = CatBoostRegressor(random_state=42)\n",
    "\n",
    "# Train the model using the training sets\n",
    "regr_catboost.fit(diabetes_X_train, diabetes_y_train)\n",
    "\n",
    "# Make predictions using the testing set\n",
    "diabetes_y_pred = regr_catboost.predict(diabetes_X_test)\n",
    "\n",
    "print(f\"Mean absolute error: {mean_absolute_error(diabetes_y_test, diabetes_y_pred):.2f}\")\n",
    "\n",
    "\n",
    "print(f\"Mean absolute percentage error: {mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred):.2%}\" )\n",
    "\n",
    "print(\"Root Mean squared error: %.2f\" % root_mean_squared_error(diabetes_y_test, diabetes_y_pred))\n",
    "# The coefficient of determination: 1 is perfect prediction\n",
    "print(\"Coefficient of determination: %.2f\" % r2_score(diabetes_y_test, diabetes_y_pred))\n",
    "\n",
    "results_list.append(\n",
    "    {'model': 'CatBoost',\n",
    "     'RMSE': root_mean_squared_error(diabetes_y_test, diabetes_y_pred),\n",
    "     'MAPE': mean_absolute_percentage_error(diabetes_y_test, diabetes_y_pred)}\n",
    ")\n",
    "\n",
    "plot_actual_vs_predicted(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0oAAAHWCAYAAABJ4Xn8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA9dUlEQVR4nO3deVwVZf//8fcR5IAooIIiqYBrrmhW3qalFkpuaVlptxma3W2WuWRJd26ZopmmlUp1l6gtlt1l3ZpbLpmllVuLmVsqlgtmCoKKCNfvj36eb2cAlyMwR3w9H495xFxzzcznnAuEdzNzHYcxxggAAAAA4FLK7gIAAAAAwNsQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAMCLjBo1Sg6Ho1jO1aZNG7Vp08a1vmrVKjkcDn344YfFcv4+ffooKiqqWM7lqYyMDD3wwAMKDw+Xw+HQwIED7S4pD4fDoVGjRp23X3F8b539Hlq1alWRngcAigNBCQCKSHJyshwOh2vx9/dXRESE4uLi9PLLL+v48eOFcp79+/dr1KhR2rx5c6EcrzB5c20XYty4cUpOTtYjjzyiOXPmqHfv3gX2jYqKchvvwMBAXX/99Zo9e3YxVgwAKCy+dhcAACXdc889p+joaGVnZ+vgwYNatWqVBg4cqMmTJ+vTTz9V48aNXX2fffZZDRs27KKOv3//fo0ePVpRUVFq0qTJBe+3dOnSizqPJ85V2xtvvKHc3Nwir+FSrFixQv/4xz80cuTIC+rfpEkTDRkyRJJ04MAB/ec//1F8fLyysrL0r3/9q0hqPHnypHx9+XUOAIWNf1kBoIh16NBB1157rWs9ISFBK1asUOfOnXXbbbdp69atCggIkCT5+voW+R+9J06cUJkyZeTn51ek5zmf0qVL23r+C5Gamqr69etfcP+rrrpK9957r2u9T58+qlGjhl566aUiC0r+/v5FclwAuNJx6x0A2ODmm2/W8OHDtXfvXr399tuu9vyeI1m2bJlatWqlkJAQlS1bVnXr1tUzzzwj6a9nQq677jpJUt++fV23fSUnJ0v66zmkhg0basOGDbrppptUpkwZ177WZ5TOysnJ0TPPPKPw8HAFBgbqtttu0759+9z6REVFqU+fPnn2/fsxz1dbfs8oZWZmasiQIapWrZqcTqfq1q2rF198UcYYt34Oh0OPPfaY5s+fr4YNG8rpdKpBgwZavHhx/m+4RWpqqvr166fKlSvL399fMTExmjVrlmv72Wdtdu/erYULF7pq37NnzwUd/6ywsDBdffXV2rVrl1t7bm6upkyZogYNGsjf31+VK1fWQw89pKNHj7r1W79+veLi4hQaGqqAgABFR0fr/vvvz/NeWJ9RWrNmja677jr5+/urZs2aeu211/LUtmfPHrfxONcx9+7dq0cffVR169ZVQECAKlasqLvuuuuC3o8dO3aoe/fuCg8Pl7+/v6pWraqePXsqLS3tvPsCgJ24ogQANundu7eeeeYZLV26tMCrDVu2bFHnzp3VuHFjPffcc3I6ndq5c6e++uorSVK9evX03HPPacSIEXrwwQd14403SpJuuOEG1zGOHDmiDh06qGfPnrr33ntVuXLlc9Y1duxYORwOPf3000pNTdWUKVMUGxurzZs3u658XYgLqe3vjDG67bbbtHLlSvXr109NmjTRkiVLNHToUP3+++966aWX3PqvWbNGH330kR599FGVK1dOL7/8srp3766UlBRVrFixwLpOnjypNm3aaOfOnXrssccUHR2tefPmqU+fPjp27JieeOIJ1atXT3PmzNGgQYNUtWpV1+10YWFhF/z6JenMmTP67bffVL58ebf2hx56SMnJyerbt68GDBig3bt369VXX9WmTZv01VdfqXTp0kpNTVX79u0VFhamYcOGKSQkRHv27NFHH310znP++OOPrv1GjRqlM2fOaOTIkecd93P57rvv9PXXX6tnz56qWrWq9uzZoxkzZqhNmzb6+eefVaZMmXz3O336tOLi4pSVlaXHH39c4eHh+v3337VgwQIdO3ZMwcHBHtcEAEXOAACKxMyZM40k89133xXYJzg42DRt2tS1PnLkSPP3f5pfeuklI8kcPny4wGN89913RpKZOXNmnm2tW7c2kkxSUlK+21q3bu1aX7lypZFkrrrqKpOenu5q/+CDD4wkM3XqVFdbZGSkiY+PP+8xz1VbfHy8iYyMdK3Pnz/fSDLPP/+8W78777zTOBwOs3PnTlebJOPn5+fW9v333xtJ5pVXXslzrr+bMmWKkWTefvttV9vp06dNixYtTNmyZd1ee2RkpOnUqdM5j/f3vu3btzeHDx82hw8fNj/++KPp3bu3kWT69+/v6vfll18aSeadd95x23/x4sVu7R9//PF5v3+M+eu9GDlypGu9W7duxt/f3+zdu9fV9vPPPxsfHx+3763du3cXODbWY544cSJPn7Vr1xpJZvbs2a62s99DK1euNMYYs2nTJiPJzJs375yvAQC8EbfeAYCNypYte87Z70JCQiRJn3zyiccTHzidTvXt2/eC+993330qV66ca/3OO+9UlSpV9Nlnn3l0/gv12WefycfHRwMGDHBrHzJkiIwxWrRokVt7bGysatas6Vpv3LixgoKC9Ouvv573POHh4brnnntcbaVLl9aAAQOUkZGhL774wuPXsHTpUoWFhSksLEyNGjXSnDlz1LdvX02cONHVZ968eQoODla7du30xx9/uJZmzZqpbNmyWrlypaT/G/sFCxYoOzv7gs6fk5OjJUuWqFu3bqpevbqrvV69eoqLi/P4df39SmJ2draOHDmiWrVqKSQkRBs3bixwv7NXjJYsWaITJ054fH4AsANBCQBslJGR4RZKrHr06KGWLVvqgQceUOXKldWzZ0998MEHFxWarrrqqouauKF27dpu6w6HQ7Vq1bro53Mu1t69exUREZHn/ahXr55r+9/9PQicVb58+TzP+eR3ntq1a6tUKfdfgQWd52I0b95cy5Yt0+LFi/Xiiy8qJCRER48edXv/d+zYobS0NFWqVMkVqs4uGRkZSk1NlSS1bt1a3bt31+jRoxUaGqquXbtq5syZysrKKvD8hw8f1smTJ/OMoSTVrVvX49d18uRJjRgxwvXsWGhoqMLCwnTs2LFzPmsUHR2twYMH6z//+Y9CQ0MVFxenadOm8XwSgMsCzygBgE1+++03paWlqVatWgX2CQgI0OrVq7Vy5UotXLhQixcv1vvvv6+bb75ZS5culY+Pz3nPczHPFV2ogj64NCcn54JqKgwFncdYJn4oTqGhoYqNjZUkxcXF6eqrr1bnzp01depUDR48WNJfEzlUqlRJ77zzTr7HOPsc1NkP/123bp3+97//acmSJbr//vs1adIkrVu3TmXLlr2kWs81hlaPP/64Zs6cqYEDB6pFixYKDg6Ww+FQz549zxvaJ02apD59+uiTTz7R0qVLNWDAACUmJmrdunWqWrXqJb0GAChKXFECAJvMmTNHks57S1SpUqV0yy23aPLkyfr55581duxYrVixwnWLVkF/8Hpqx44dbuvGGO3cudNthrry5cvr2LFjefa1Xo25mNoiIyO1f//+PLci/vLLL67thSEyMlI7duzI8wd+YZ9Hkjp16qTWrVtr3LhxyszMlCTVrFlTR44cUcuWLRUbG5tniYmJcTvGP/7xD40dO1br16/XO++8oy1btmju3Ln5ni8sLEwBAQF5xlCStm3b5rZ+doIJ6zjmd0Xtww8/VHx8vCZNmqQ777xT7dq1U6tWrfL9HshPo0aN9Oyzz2r16tX68ssv9fvvvyspKemC9gUAuxCUAMAGK1as0JgxYxQdHa1evXoV2O/PP//M03b2g1vP3oIVGBgoKe8fvJ6aPXu2W1j58MMPdeDAAXXo0MHVVrNmTa1bt06nT592tS1YsCDPNOIXU1vHjh2Vk5OjV1991a39pZdeksPhcDv/pejYsaMOHjyo999/39V25swZvfLKKypbtqxat25dKOc56+mnn9aRI0f0xhtvSJLuvvtu5eTkaMyYMXn6njlzxvVeHT16NM/VMevYW/n4+CguLk7z589XSkqKq33r1q1asmSJW9+goCCFhoZq9erVbu3Tp0/P97jWWl555ZV8rz79XXp6us6cOePW1qhRI5UqVeqctxACgDfg1jsAKGKLFi3SL7/8ojNnzujQoUNasWKFli1bpsjISH366afn/MDQ5557TqtXr1anTp0UGRmp1NRUTZ8+XVWrVlWrVq0k/RVaQkJClJSUpHLlyikwMFDNmzdXdHS0R/VWqFBBrVq1Ut++fXXo0CFNmTJFtWrVcpvC/IEHHtCHH36oW2+9VXfffbd27dqlt99+221yhYutrUuXLmrbtq3+/e9/a8+ePYqJidHSpUv1ySefaODAgXmO7akHH3xQr732mvr06aMNGzYoKipKH374ob766itNmTLlnM+MeaJDhw5q2LChJk+erP79+6t169Z66KGHlJiYqM2bN6t9+/YqXbq0duzYoXnz5mnq1Km68847NWvWLE2fPl233367atasqePHj+uNN95QUFCQOnbsWOD5Ro8ercWLF+vGG2/Uo48+6gqBDRo00A8//ODW94EHHtD48eP1wAMP6Nprr9Xq1au1ffv2PMfs3Lmz5syZo+DgYNWvX19r167V559/fs5p2KW//ofAY489prvuukt16tTRmTNnNGfOHPn4+Kh79+6evaEAUFxsnXMPAEqws9ODn138/PxMeHi4adeunZk6darbNNRnWacHX758uenatauJiIgwfn5+JiIiwtxzzz1m+/btbvt98sknpn79+sbX19dtyufWrVubBg0a5FtfQdODv/feeyYhIcFUqlTJBAQEmE6dOrlNNX3WpEmTzFVXXWWcTqdp2bKlWb9+fZ5jnqs26/Tgxhhz/PhxM2jQIBMREWFKly5tateubSZOnGhyc3Pd+sky5fZZBU1bbnXo0CHTt29fExoaavz8/EyjRo3ynSb7YqcHL6hvcnJynqm4X3/9ddOsWTMTEBBgypUrZxo1amSeeuops3//fmOMMRs3bjT33HOPqV69unE6naZSpUqmc+fOZv369W7HlmUqb2OM+eKLL0yzZs2Mn5+fqVGjhklKSsrzvWXMX9N+9+vXzwQHB5ty5cqZu+++26SmpuY55tGjR13vV9myZU1cXJz55Zdf8rzf1unBf/31V3P//febmjVrGn9/f1OhQgXTtm1b8/nnn1/QewoAdnIYY+NTrwAAAADghXhGCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFiX+A2dzc3O1f/9+lStXTg6Hw+5yAAAAANjEGKPjx48rIiJCpUqd+5pRiQ9K+/fvV7Vq1ewuAwAAAICX2Ldvn6pWrXrOPiU+KJUrV07SX29GUFCQzdUAAFBMMjOliIi/vt6/XwoMtLceAPAC6enpqlatmisjnEuJD0pnb7cLCgoiKAEArhw+Pv/3dVAQQQkA/uZCHslhMgcAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsbA1Kq1evVpcuXRQRESGHw6H58+fn6bN161bddtttCg4OVmBgoK677jqlpKQUf7EAAAAArhi2BqXMzEzFxMRo2rRp+W7ftWuXWrVqpauvvlqrVq3SDz/8oOHDh8vf37+YKwUAAABwJXEYY4zdRUiSw+HQxx9/rG7durnaevbsqdKlS2vOnDkeHzc9PV3BwcFKS0tTUFBQIVQKAMBlIDNTKlv2r68zMqTAQHvrAQAvcDHZwGufUcrNzdXChQtVp04dxcXFqVKlSmrevHm+t+f9XVZWltLT090WAAAAALgYvnYXUJDU1FRlZGRo/Pjxev755zVhwgQtXrxYd9xxh1auXKnWrVvnu19iYqJGjx5dzNUCAPITNWyh3SW42TO+k90lAAAuE159RUmSunbtqkGDBqlJkyYaNmyYOnfurKSkpAL3S0hIUFpammvZt29fcZUMAAAAoITw2itKoaGh8vX1Vf369d3a69WrpzVr1hS4n9PplNPpLOryAAAAAJRgXntFyc/PT9ddd522bdvm1r59+3ZFRkbaVBUAAACAK4GtV5QyMjK0c+dO1/ru3bu1efNmVahQQdWrV9fQoUPVo0cP3XTTTWrbtq0WL16s//3vf1q1apV9RQMAAAAo8WwNSuvXr1fbtm1d64MHD5YkxcfHKzk5WbfffruSkpKUmJioAQMGqG7duvrvf/+rVq1a2VUyAAAAgCuArUGpTZs2Ot/HON1///26//77i6kiAAAAAPDiZ5QAAAAAwC4EJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwMLX7gIAALgSRQ1bWKTHDzh9Slv//9f1hi/WST//AvvuGd+pSGsBgMsRV5QAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFrYGpdWrV6tLly6KiIiQw+HQ/PnzC+z78MMPy+FwaMqUKcVWHwAAAIArk61BKTMzUzExMZo2bdo5+3388cdat26dIiIiiqkyAAAAAFcyXztP3qFDB3Xo0OGcfX7//Xc9/vjjWrJkiTp16lRMlQEAAAC4ktkalM4nNzdXvXv31tChQ9WgQYML2icrK0tZWVmu9fT09KIqDwAAAEAJ5dVBacKECfL19dWAAQMueJ/ExESNHj26CKsCAFyuooYttLsEAMBlwmtnvduwYYOmTp2q5ORkORyOC94vISFBaWlprmXfvn1FWCUAAACAkshrg9KXX36p1NRUVa9eXb6+vvL19dXevXs1ZMgQRUVFFbif0+lUUFCQ2wIAAAAAF8Nrb73r3bu3YmNj3dri4uLUu3dv9e3b16aqAAAAAFwJbA1KGRkZ2rlzp2t99+7d2rx5sypUqKDq1aurYsWKbv1Lly6t8PBw1a1bt7hLBQAAAHAFsTUorV+/Xm3btnWtDx48WJIUHx+v5ORkm6oCAAAAcKWzNSi1adNGxpgL7r9nz56iKwYAAAAA/j+vncwBAAAAAOxCUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWNgalFavXq0uXbooIiJCDodD8+fPd23Lzs7W008/rUaNGikwMFARERG67777tH//fvsKBgAAAHBFsDUoZWZmKiYmRtOmTcuz7cSJE9q4caOGDx+ujRs36qOPPtK2bdt022232VApAAAAgCuJr50n79Chgzp06JDvtuDgYC1btsyt7dVXX9X111+vlJQUVa9evThKBAAAAHAFsjUoXay0tDQ5HA6FhIQU2CcrK0tZWVmu9fT09GKoDAAAAEBJctlM5nDq1Ck9/fTTuueeexQUFFRgv8TERAUHB7uWatWqFWOVAAAAAEqCyyIoZWdn6+6775YxRjNmzDhn34SEBKWlpbmWffv2FVOVAAAAAEoKr7/17mxI2rt3r1asWHHOq0mS5HQ65XQ6i6k6AAAAACWRVwelsyFpx44dWrlypSpWrGh3SQAAAACuALYGpYyMDO3cudO1vnv3bm3evFkVKlRQlSpVdOedd2rjxo1asGCBcnJydPDgQUlShQoV5OfnZ1fZAAAAAEo4W4PS+vXr1bZtW9f64MGDJUnx8fEaNWqUPv30U0lSkyZN3PZbuXKl2rRpU1xlAgAAALjC2BqU2rRpI2NMgdvPtQ0AAAAAisplMesdAAAAABQnghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwMLWoLR69Wp16dJFERERcjgcmj9/vtt2Y4xGjBihKlWqKCAgQLGxsdqxY4c9xQIAAAC4YtgalDIzMxUTE6Np06blu/2FF17Qyy+/rKSkJH3zzTcKDAxUXFycTp06VcyVAgAAALiS+Np58g4dOqhDhw75bjPGaMqUKXr22WfVtWtXSdLs2bNVuXJlzZ8/Xz179izOUgEAAABcQbz2GaXdu3fr4MGDio2NdbUFBwerefPmWrt2bYH7ZWVlKT093W0BAAAAgIth6xWlczl48KAkqXLlym7tlStXdm3LT2JiokaPHl2ktQEAUJJEDVtodwkue8Z3srsEr8U4AcXLa68oeSohIUFpaWmuZd++fXaXBAAAAOAy47VBKTw8XJJ06NAht/ZDhw65tuXH6XQqKCjIbQEAAACAi+G1QSk6Olrh4eFavny5qy09PV3ffPONWrRoYWNlAAAAAEo6j55R+vXXX1WjRo1LPnlGRoZ27tzpWt+9e7c2b96sChUqqHr16ho4cKCef/551a5dW9HR0Ro+fLgiIiLUrVu3Sz43AAAAABTEo6BUq1YttW7dWv369dOdd94pf39/j06+fv16tW3b1rU+ePBgSVJ8fLySk5P11FNPKTMzUw8++KCOHTumVq1aafHixR6fDwAAAAAuhEe33m3cuFGNGzfW4MGDFR4eroceekjffvvtRR+nTZs2MsbkWZKTkyVJDodDzz33nA4ePKhTp07p888/V506dTwpGQAAAAAumEdBqUmTJpo6dar279+vt956SwcOHFCrVq3UsGFDTZ48WYcPHy7sOgEAAACg2FzSZA6+vr664447NG/ePE2YMEE7d+7Uk08+qWrVqum+++7TgQMHCqtOAAAAACg2lxSU1q9fr0cffVRVqlTR5MmT9eSTT2rXrl1atmyZ9u/fr65duxZWnQAAAABQbDyazGHy5MmaOXOmtm3bpo4dO2r27Nnq2LGjSpX6K3dFR0crOTlZUVFRhVkrAAAAABQLj4LSjBkzdP/996tPnz6qUqVKvn0qVaqkN99885KKAwAAAAA7eBSUduzYcd4+fn5+io+P9+TwAAAAAGArj55RmjlzpubNm5enfd68eZo1a9YlFwUAAAAAdvIoKCUmJio0NDRPe6VKlTRu3LhLLgoAAAAA7ORRUEpJSVF0dHSe9sjISKWkpFxyUQAAAABgJ4+CUqVKlfTDDz/kaf/+++9VsWLFSy4KAAAAAOzkUVC65557NGDAAK1cuVI5OTnKycnRihUr9MQTT6hnz56FXSMAAAAAFCuPZr0bM2aM9uzZo1tuuUW+vn8dIjc3V/fddx/PKAEAAAC47HkUlPz8/PT+++9rzJgx+v777xUQEKBGjRopMjKysOsDAAAAgGLnUVA6q06dOqpTp05h1QIAAAAAXsGjoJSTk6Pk5GQtX75cqampys3Nddu+YsWKQikOAAAAAOzgUVB64oknlJycrE6dOqlhw4ZyOByFXRcAAAAA2MajoDR37lx98MEH6tixY2HXAwAAAAC282h6cD8/P9WqVauwawEAAAAAr+BRUBoyZIimTp0qY0xh1wMAAAAAtvPo1rs1a9Zo5cqVWrRokRo0aKDSpUu7bf/oo48KpTgAAAAAsINHQSkkJES33357YdcCAAAAAF7Bo6A0c+bMwq4DAAAAALyGR88oSdKZM2f0+eef67XXXtPx48clSfv371dGRkahFQcAAAAAdvDoitLevXt16623KiUlRVlZWWrXrp3KlSunCRMmKCsrS0lJSYVdJwAAAAAUG4+uKD3xxBO69tprdfToUQUEBLjab7/9di1fvrzQigMAAAAAO3h0RenLL7/U119/LT8/P7f2qKgo/f7774VSGAAAAADYxaMrSrm5ucrJycnT/ttvv6lcuXKXXBQAAAAA2MmjoNS+fXtNmTLFte5wOJSRkaGRI0eqY8eOhVUbAAAAANjCo1vvJk2apLi4ONWvX1+nTp3SP//5T+3YsUOhoaF67733CrtGAAAAAChWHgWlqlWr6vvvv9fcuXP1ww8/KCMjQ/369VOvXr3cJncAAAAAgMuRR0FJknx9fXXvvfcWZi0AAAAA4BU8CkqzZ88+5/b77rvPo2IAAAAAwBt4FJSeeOIJt/Xs7GydOHFCfn5+KlOmDEEJAAAAwGXNo1nvjh496rZkZGRo27ZtatWqFZM5AAAAALjseRSU8lO7dm2NHz8+z9WmS5GTk6Phw4crOjpaAQEBqlmzpsaMGSNjTKGdAwAAAACsPJ7MId+D+fpq//79hXa8CRMmaMaMGZo1a5YaNGig9evXq2/fvgoODtaAAQMK7TwAAAAA8HceBaVPP/3Ubd0YowMHDujVV19Vy5YtC6UwSfr666/VtWtXderUSZIUFRWl9957T99++22hnQMAAAAArDwKSt26dXNbdzgcCgsL080336xJkyYVRl2SpBtuuEGvv/66tm/frjp16uj777/XmjVrNHny5AL3ycrKUlZWlms9PT290OoBAAAAcGXwKCjl5uYWdh35GjZsmNLT03X11VfLx8dHOTk5Gjt2rHr16lXgPomJiRo9enSx1AcAAACgZCq0yRyKwgcffKB33nlH7777rjZu3KhZs2bpxRdf1KxZswrcJyEhQWlpaa5l3759xVgxAAAAgJLAoytKgwcPvuC+57pN7nyGDh2qYcOGqWfPnpKkRo0aae/evUpMTFR8fHy++zidTjmdTo/PCQAAAAAeBaVNmzZp06ZNys7OVt26dSVJ27dvl4+Pj6655hpXP4fDcUnFnThxQqVKuV/08vHxKbZb/wAAAABcmTwKSl26dFG5cuU0a9YslS9fXtJfH0Lbt29f3XjjjRoyZEihFNelSxeNHTtW1atXV4MGDbRp0yZNnjxZ999/f6EcHwAAAADy41FQmjRpkpYuXeoKSZJUvnx5Pf/882rfvn2hBaVXXnlFw4cP16OPPqrU1FRFRETooYce0ogRIwrl+AAAAACQH4+CUnp6ug4fPpyn/fDhwzp+/PglF3VWuXLlNGXKFE2ZMqXQjgkAAAAA5+PRrHe33367+vbtq48++ki//fabfvvtN/33v/9Vv379dMcddxR2jQAAAABQrDy6opSUlKQnn3xS//znP5Wdnf3XgXx91a9fP02cOLFQCwQAAACA4uZRUCpTpoymT5+uiRMnateuXZKkmjVrKjAwsFCLAwAAAAA7XNIHzh44cEAHDhxQ7dq1FRgYKGNMYdUFAAAAALbxKCgdOXJEt9xyi+rUqaOOHTvqwIEDkqR+/foV2ox3AAAAAGAXj4LSoEGDVLp0aaWkpKhMmTKu9h49emjx4sWFVhwAAAAA2MGjZ5SWLl2qJUuWqGrVqm7ttWvX1t69ewulMAAAAACwi0dXlDIzM92uJJ31559/yul0XnJRAAAAAGAnj4LSjTfeqNmzZ7vWHQ6HcnNz9cILL6ht27aFVhwAAAAA2MGjW+9eeOEF3XLLLVq/fr1Onz6tp556Slu2bNGff/6pr776qrBrBAAAAIBi5dEVpYYNG2r79u1q1aqVunbtqszMTN1xxx3atGmTatasWdg1AgAAAECxuugrStnZ2br11luVlJSkf//730VREwAAAADY6qKvKJUuXVo//PBDUdQCAAAAAF7Bo1vv7r33Xr355puFXQsAAAAAeAWPJnM4c+aM3nrrLX3++edq1qyZAgMD3bZPnjy5UIoDAAAAADtcVFD69ddfFRUVpZ9++knXXHONJGn79u1ufRwOR+FVBwAAAAA2uKigVLt2bR04cEArV66UJPXo0UMvv/yyKleuXCTFAQAuXtSwhXaXAHjMm75/94zvZHcJAGx0Uc8oGWPc1hctWqTMzMxCLQgAAAAA7ObRZA5nWYMTAAAAAJQEFxWUHA5HnmeQeCYJAAAAQElzUc8oGWPUp08fOZ1OSdKpU6f08MMP55n17qOPPiq8CgEAAACgmF1UUIqPj3dbv/feewu1GAAAAADwBhcVlGbOnFlUdQAAAACA17ikyRwAAAAAoCQiKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGDh9UHp999/17333quKFSsqICBAjRo10vr16+0uCwAAAEAJ5mt3Aedy9OhRtWzZUm3bttWiRYsUFhamHTt2qHz58naXBgAAAKAE8+qgNGHCBFWrVk0zZ850tUVHR9tYEQAAAIArgVffevfpp5/q2muv1V133aVKlSqpadOmeuONN865T1ZWltLT090WAAAAALgYXh2Ufv31V82YMUO1a9fWkiVL9Mgjj2jAgAGaNWtWgfskJiYqODjYtVSrVq0YKwYAAABQEnh1UMrNzdU111yjcePGqWnTpnrwwQf1r3/9S0lJSQXuk5CQoLS0NNeyb9++YqwYAAAAQEng1UGpSpUqql+/vltbvXr1lJKSUuA+TqdTQUFBbgsAAAAAXAyvDkotW7bUtm3b3Nq2b9+uyMhImyoCAAAAcCXw6qA0aNAgrVu3TuPGjdPOnTv17rvv6vXXX1f//v3tLg0AAABACebVQem6667Txx9/rPfee08NGzbUmDFjNGXKFPXq1cvu0gAAAACUYF79OUqS1LlzZ3Xu3NnuMgAAAABcQbz6ihIAAAAA2IGgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWPjaXQAAAIA3ihq20O4SvJY3vTd7xneyuwSUUFxRAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFhcVkFp/PjxcjgcGjhwoN2lAAAAACjBLpug9N133+m1115T48aN7S4FAAAAQAl3WQSljIwM9erVS2+88YbKly9vdzkAAAAASrjLIij1799fnTp1Umxs7Hn7ZmVlKT093W0BAAAAgIvha3cB5zN37lxt3LhR33333QX1T0xM1OjRo4u4KgDeIGrYQrtLAADYzJt+F+wZ38nuElCIvPqK0r59+/TEE0/onXfekb+//wXtk5CQoLS0NNeyb9++Iq4SAAAAQEnj1VeUNmzYoNTUVF1zzTWutpycHK1evVqvvvqqsrKy5OPj47aP0+mU0+ks7lIBAAAAlCBeHZRuueUW/fjjj25tffv21dVXX62nn346T0gCAAAAgMLg1UGpXLlyatiwoVtbYGCgKlasmKcdAAAAAAqLVz+jBAAAAAB28OorSvlZtWqV3SUAAAAAKOG4ogQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALghIAAAAAWBCUAAAAAMCCoAQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAICF1welxMREXXfddSpXrpwqVaqkbt26adu2bXaXBQAAAKAE8/qg9MUXX6h///5at26dli1bpuzsbLVv316ZmZl2lwYAAACghPK1u4DzWbx4sdt6cnKyKlWqpA0bNuimm26yqSoAAAAAJZnXByWrtLQ0SVKFChXy3Z6VlaWsrCzXenp6erHUBQAAAKDkuKyCUm5urgYOHKiWLVuqYcOG+fZJTEzU6NGji7kyFIaoYQvtLsFlz/hOdpfglbxpjAAA8Db8nizY5fi3ldc/o/R3/fv3108//aS5c+cW2CchIUFpaWmuZd++fcVYIQAAAICS4LK5ovTYY49pwYIFWr16tapWrVpgP6fTKafTWYyVAQAAAChpvD4oGWP0+OOP6+OPP9aqVasUHR1td0kAAAAASjivD0r9+/fXu+++q08++UTlypXTwYMHJUnBwcEKCAiwuToAAAAAJZHXP6M0Y8YMpaWlqU2bNqpSpYpref/99+0uDQAAAEAJ5fVXlIwxdpcAAAAA4Arj9VeUAAAAAKC4EZQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwICgBAAAAgAVBCQAAAAAsCEoAAAAAYEFQAgAAAAALX7sLuNJEDVtodwkue8Z3srsEr8U4AQAAXNm4ogQAAAAAFgQlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACwuCyC0rRp0xQVFSV/f381b95c3377rd0lAQAAACjBvD4ovf/++xo8eLBGjhypjRs3KiYmRnFxcUpNTbW7NAAAAAAllNcHpcmTJ+tf//qX+vbtq/r16yspKUllypTRW2+9ZXdpAAAAAEooX7sLOJfTp09rw4YNSkhIcLWVKlVKsbGxWrt2bb77ZGVlKSsry7WelpYmSUpPTy/aYi9QbtYJu0tw8Zb35Cxvem+8iTeNE2MEXD5yTp/S2X89crJOKNfk2loPgCubt/w9c7YOY8x5+3p1UPrjjz+Uk5OjypUru7VXrlxZv/zyS777JCYmavTo0Xnaq1WrViQ1Xs6Cp9hdAS4E4wTAU8Fnv5h+n51lAIDX/T1z/PhxBQcHn7OPVwclTyQkJGjw4MGu9dzcXP3555+qWLGiHA6HjZXBKj09XdWqVdO+ffsUFBRkdzm4BIxlycJ4lhyMZcnCeJYcjKV9jDE6fvy4IiIiztvXq4NSaGiofHx8dOjQIbf2Q4cOKTw8PN99nE6nnE6nW1tISEhRlYhCEBQUxD8SJQRjWbIwniUHY1myMJ4lB2Npj/NdSTrLqydz8PPzU7NmzbR8+XJXW25urpYvX64WLVrYWBkAAACAksyrryhJ0uDBgxUfH69rr71W119/vaZMmaLMzEz17dvX7tIAAAAAlFBeH5R69Oihw4cPa8SIETp48KCaNGmixYsX55ngAZcfp9OpkSNH5rlVEpcfxrJkYTxLDsayZGE8Sw7G8vLgMBcyNx4AAAAAXEG8+hklAAAAALADQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISitzYsWN1ww03qEyZMgV++K/D4cizzJ07163PqlWrdM0118jpdKpWrVpKTk4u+uKRx4WMZ0pKijp16qQyZcqoUqVKGjp0qM6cOePWh/H0TlFRUXl+FsePH+/W54cfftCNN94of39/VatWTS+88IJN1eJ8pk2bpqioKPn7+6t58+b69ttv7S4J5zFq1Kg8P4NXX321a/upU6fUv39/VaxYUWXLllX37t116NAhGyvGWatXr1aXLl0UEREhh8Oh+fPnu203xmjEiBGqUqWKAgICFBsbqx07drj1+fPPP9WrVy8FBQUpJCRE/fr1U0ZGRjG+CvwdQQlF7vTp07rrrrv0yCOPnLPfzJkzdeDAAdfSrVs317bdu3erU6dOatu2rTZv3qyBAwfqgQce0JIlS4q4elidbzxzcnLUqVMnnT59Wl9//bVmzZql5ORkjRgxwtWH8fRuzz33nNvP4uOPP+7alp6ervbt2ysyMlIbNmzQxIkTNWrUKL3++us2Voz8vP/++xo8eLBGjhypjRs3KiYmRnFxcUpNTbW7NJxHgwYN3H4G16xZ49o2aNAg/e9//9O8efP0xRdfaP/+/brjjjtsrBZnZWZmKiYmRtOmTct3+wsvvKCXX35ZSUlJ+uabbxQYGKi4uDidOnXK1adXr17asmWLli1bpgULFmj16tV68MEHi+slwMoAxWTmzJkmODg4322SzMcff1zgvk899ZRp0KCBW1uPHj1MXFxcIVaIi1HQeH722WemVKlS5uDBg662GTNmmKCgIJOVlWWMYTy9WWRkpHnppZcK3D59+nRTvnx511gaY8zTTz9t6tatWwzV4WJcf/31pn///q71nJwcExERYRITE22sCuczcuRIExMTk++2Y8eOmdKlS5t58+a52rZu3WokmbVr1xZThbgQ1r9rcnNzTXh4uJk4caKr7dixY8bpdJr33nvPGGPMzz//bCSZ7777ztVn0aJFxuFwmN9//73Yasf/4YoSvEb//v0VGhqq66+/Xm+99ZbM3z7ia+3atYqNjXXrHxcXp7Vr1xZ3mTiPtWvXqlGjRm4fCh0XF6f09HRt2bLF1Yfx9F7jx49XxYoV1bRpU02cONHttsm1a9fqpptukp+fn6stLi5O27Zt09GjR+0oF/k4ffq0NmzY4PZzVqpUKcXGxvJzdhnYsWOHIiIiVKNGDfXq1UspKSmSpA0bNig7O9ttXK+++mpVr16dcfVyu3fv1sGDB93GLjg4WM2bN3eN3dq1axUSEqJrr73W1Sc2NlalSpXSN998U+w1Q/K1uwBA+utWn5tvvlllypTR0qVL9eijjyojI0MDBgyQJB08eNDtD29Jqly5stLT03Xy5EkFBATYUTbyUdBYnd12rj6Mp/0GDBiga665RhUqVNDXX3+thIQEHThwQJMnT5b019hFR0e77fP38S1fvnyx14y8/vjjD+Xk5OT7c/bLL7/YVBUuRPPmzZWcnKy6devqwIEDGj16tG688Ub99NNPOnjwoPz8/PI8H1q5cmXXv6/wTmfHJ7+fyb//bqxUqZLbdl9fX1WoUIHxtQlBCR4ZNmyYJkyYcM4+W7dudXsA9VyGDx/u+rpp06bKzMzUxIkTXUEJRauwxxPe5WLGd/Dgwa62xo0by8/PTw899JASExPldDqLulTgitehQwfX140bN1bz5s0VGRmpDz74gP+JBBQzghI8MmTIEPXp0+ecfWrUqOHx8Zs3b64xY8YoKytLTqdT4eHheWb1OXTokIKCgvjFUQgKczzDw8PzzKx1duzCw8Nd/2U8i8+ljG/z5s115swZ7dmzR3Xr1i1w7KT/G1/YLzQ0VD4+PvmOFeN0eQkJCVGdOnW0c+dOtWvXTqdPn9axY8fcrioxrt7v7PgcOnRIVapUcbUfOnRITZo0cfWxTrZy5swZ/fnnn4yvTQhK8EhYWJjCwsKK7PibN29W+fLlXf8Hu0WLFvrss8/c+ixbtkwtWrQoshquJIU5ni1atNDYsWOVmprquoVg2bJlCgoKUv369V19GM/icynju3nzZpUqVco1li1atNC///1vZWdnq3Tp0pL+Gru6dety250X8fPzU7NmzbR8+XLXDKK5ublavny5HnvsMXuLw0XJyMjQrl271Lt3bzVr1kylS5fW8uXL1b17d0nStm3blJKSwr+fXi46Olrh4eFavny5Kxilp6frm2++cc0i26JFCx07dkwbNmxQs2bNJEkrVqxQbm6umjdvblfpVza7Z5NAybd3716zadMmM3r0aFO2bFmzadMms2nTJnP8+HFjjDGffvqpeeONN8yPP/5oduzYYaZPn27KlCljRowY4TrGr7/+asqUKWOGDh1qtm7daqZNm2Z8fHzM4sWL7XpZV6zzjeeZM2dMw4YNTfv27c3mzZvN4sWLTVhYmElISHAdg/H0Tl9//bV56aWXzObNm82uXbvM22+/bcLCwsx9993n6nPs2DFTuXJl07t3b/PTTz+ZuXPnmjJlypjXXnvNxsqRn7lz5xqn02mSk5PNzz//bB588EETEhLiNiMlvM+QIUPMqlWrzO7du81XX31lYmNjTWhoqElNTTXGGPPwww+b6tWrmxUrVpj169ebFi1amBYtWthcNYwx5vjx467fiZLM5MmTzaZNm8zevXuNMcaMHz/ehISEmE8++cT88MMPpmvXriY6OtqcPHnSdYxbb73VNG3a1HzzzTdmzZo1pnbt2uaee+6x6yVd8QhKKHLx8fFGUp5l5cqVxpi/pr5s0qSJKVu2rAkMDDQxMTEmKSnJ5OTkuB1n5cqVpkmTJsbPz8/UqFHDzJw5s/hfDM47nsYYs2fPHtOhQwcTEBBgQkNDzZAhQ0x2drbbcRhP77NhwwbTvHlzExwcbPz9/U29evXMuHHjzKlTp9z6ff/996ZVq1bG6XSaq666yowfP96minE+r7zyiqlevbrx8/Mz119/vVm3bp3dJeE8evToYapUqWL8/PzMVVddZXr06GF27tzp2n7y5Enz6KOPmvLly5syZcqY22+/3Rw4cMDGinHWypUr8/39GB8fb4z5a4rw4cOHm8qVKxun02luueUWs23bNrdjHDlyxNxzzz2mbNmyJigoyPTt29f1PyJR/BzG/G0OZgAAAACA+BwlAAAAALAgKAEAAACABUEJAAAAACwISgAAAABgQVACAAAAAAuCEgAAAABYEJQAAAAAwIKgBAAAAAAWBCUAQIm0Z88eORwObd68ucA+q1atksPh0LFjxwr13A6HQ/Pnzy/UYwIAihdBCQBgmz59+sjhcMjhcKh06dKKjo7WU089pVOnTl3ysatVq6YDBw6oYcOGhVApAOBK42t3AQCAK9utt96qmTNnKjs7Wxs2bFB8fLwcDocmTJhwScf18fFReHh4IVUJALjScEUJAGArp9Op8PBwVatWTd26dVNsbKyWLVsmScrNzVViYqKio6MVEBCgmJgYffjhh659jx49ql69eiksLEwBAQGqXbu2Zs6cKSn/W+8+++wz1alTRwEBAWrbtq327NnjVsuoUaPUpEkTt7YpU6YoKirKtf7dd9+pXbt2Cg0NVXBwsFq3bq2NGzcW6nsCALAfQQkA4DV++uknff311/Lz85MkJSYmavbs2UpKStKWLVs0aNAg3Xvvvfriiy8kScOHD9fPP/+sRYsWaevWrZoxY4ZCQ0PzPfa+fft0xx13qEuXLtq8ebMeeOABDRs27KJrPH78uOLj47VmzRqtW7dOtWvXVseOHXX8+HHPXzgAwOtw6x0AwFYLFixQ2bJldebMGWVlZalUqVJ69dVXlZWVpXHjxunzzz9XixYtJEk1atTQmjVr9Nprr6l169ZKSUlR06ZNde2110qS25UfqxkzZqhmzZqaNGmSJKlu3br68ccfL/oWv5tvvtlt/fXXX1dISIi++OILde7c+aKOBQDwXgQlAICt2rZtqxkzZigzM1MvvfSSfH191b17d23ZskUnTpxQu3bt3PqfPn1aTZs2lSQ98sgj6t69uzZu3Kj27durW7duuuGGG/I9z9atW9W8eXO3trMB7GIcOnRIzz77rFatWqXU1FTl5OToxIkTSklJuehjAQC8F0EJAGCrwMBA1apVS5L01ltvKSYmRm+++aZrtrqFCxfqqquuctvH6XRKkjp06KC9e/fqs88+07Jly3TLLbeof//+evHFFz2qpVSpUjLGuLVlZ2e7rcfHx+vIkSOaOnWqIiMj5XQ61aJFC50+fdqjcwIAvBNBCQDgNUqVKqVnnnlGgwcP1vbt2+V0OpWSkqLWrVsXuE9YWJji4+MVHx+vG2+8UUOHDs03KNWrV0+ffvqpW9u6devyHOvgwYMyxsjhcEhSns9h+uqrrzR9+nR17NhR0l/PPv3xxx+evFwAgBdjMgcAgFe566675OPjo9dee01PPvmkBg0apFmzZmnXrl3auHGjXnnlFc2aNUuSNGLECH3yySfauXOntmzZogULFqhevXr5Hvfhhx/Wjh07NHToUG3btk3vvvuukpOT3fq0adNGhw8f1gsvvKBdu3Zp2rRpWrRokVuf2rVra86cOdq6dau++eYb9erVSwEBAUXyXgAA7ENQAgB4FV9fXz322GN64YUXlJCQoOHDhysxMVH16tXTrbfeqoULFyo6OlqS5Ofnp4SEBDVu3Fg33XSTfHx8NHfu3HyPW716df33v//V/PnzFRMTo6SkJI0bN86tT7169TR9+nRNmzZNMTEx+vbbb/Xkk0+69XnzzTd19OhRXXPNNerdu7cGDBigSpUqFc2bAQCwjcNYb8YGAAAAgCscV5QAAAAAwIKgBAAAAAAWBCUAAAAAsCAoAQAAAIAFQQkAAAAALAhKAAAAAGBBUAIAAAAAC4ISAAAAAFgQlAAAAADAgqAEAAAAABYEJQAAAACw+H8hx4F0c/JTHgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_residuals(diabetes_y_test, diabetes_y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Compare Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_df = pd.DataFrame(results_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "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>model</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>MAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CatBoost</td>\n",
       "      <td>55.137472</td>\n",
       "      <td>0.393867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>XGBoost</td>\n",
       "      <td>58.011463</td>\n",
       "      <td>0.398371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>54.959277</td>\n",
       "      <td>0.403435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Light GBM</td>\n",
       "      <td>57.971124</td>\n",
       "      <td>0.407147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Multiple Linear Regression</td>\n",
       "      <td>54.253023</td>\n",
       "      <td>0.407386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Linear Regression</td>\n",
       "      <td>62.004712</td>\n",
       "      <td>0.488760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Decision Tree</td>\n",
       "      <td>76.000948</td>\n",
       "      <td>0.506753</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        model       RMSE      MAPE\n",
       "6                    CatBoost  55.137472  0.393867\n",
       "4                     XGBoost  58.011463  0.398371\n",
       "3               Random Forest  54.959277  0.403435\n",
       "5                   Light GBM  57.971124  0.407147\n",
       "1  Multiple Linear Regression  54.253023  0.407386\n",
       "0           Linear Regression  62.004712  0.488760\n",
       "2               Decision Tree  76.000948  0.506753"
      ]
     },
     "execution_count": 311,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.sort_values('MAPE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's remind ourselves - what's the scale of the output feature?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "diabetes_y_test.hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 313,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "diabetes_y_train.hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting Residuals\n",
    "\n",
    "Plotting the residuals (errors) can help us to understand whether the model is consistently over or under estimating."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ml",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}