{"cells":[{"metadata":{"_uuid":"08c0c8baf241a57ba91214df7f43329f106fe0de"},"cell_type":"markdown","source":"## 1. About this notebook\n\nIn this notebook I try a few different models to predict the time to failure during earthquake simulations (LANL competition). Every model is implemented trough a Scikit-learn pipeline and compared using cross-validation scores and visualizations.\n\nFor more details about LANL competition you can check my [previous kernel](https://www.kaggle.com/jsaguiar/seismic-data-exploration)."},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"import os\nimport time\nimport random\nimport numpy as np\nimport pandas as pd\nfrom tqdm import tqdm_notebook\nfrom boruta import BorutaPy\n# Visualizations\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nimport plotly.graph_objs as go\nfrom plotly.offline import init_notebook_mode, iplot\n# Sklearn utilities\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.metrics import mean_absolute_error\nfrom sklearn.model_selection import GridSearchCV, KFold, RandomizedSearchCV\nfrom sklearn.feature_selection import RFECV, SelectFromModel\n# Models\nfrom sklearn.linear_model import LinearRegression, Ridge\nfrom sklearn.tree import DecisionTreeRegressor\nfrom sklearn.svm import NuSVR, SVR\nfrom sklearn.kernel_ridge import KernelRidge\nfrom sklearn.ensemble import AdaBoostRegressor\nfrom sklearn.ensemble import RandomForestRegressor, ExtraTreesRegressor\nimport lightgbm as lgb\nimport warnings\nwarnings.simplefilter(action='ignore', category=FutureWarning)\nwarnings.simplefilter(action='ignore', category=UserWarning)\nsns.set()\ninit_notebook_mode(connected=True)","execution_count":1,"outputs":[{"output_type":"display_data","data":{"text/html":"        <script type=\"text/javascript\">\n        window.PlotlyConfig = {MathJaxConfig: 'local'};\n        if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n        if (typeof require !== 'undefined') {\n        require.undef(\"plotly\");\n        requirejs.config({\n            paths: {\n                'plotly': ['https://cdn.plot.ly/plotly-latest.min']\n            }\n        });\n        require(['plotly'], function(Plotly) {\n            window._Plotly = Plotly;\n        });\n        }\n        </script>\n        "},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"This competition has a single feature (seismic signal) and we must predict the time to failure:"},{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"data_type = {'acoustic_data': np.int16, 'time_to_failure': np.float32}\ntrain = pd.read_csv('../input/train.csv', dtype=data_type)\ntrain.head()","execution_count":2,"outputs":[{"output_type":"execute_result","execution_count":2,"data":{"text/plain":"   acoustic_data  time_to_failure\n0             12           1.4691\n1              6           1.4691\n2              8           1.4691\n3              5           1.4691\n4              8           1.4691","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>acoustic_data</th>\n      <th>time_to_failure</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>12</td>\n      <td>1.4691</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>6</td>\n      <td>1.4691</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>8</td>\n      <td>1.4691</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>5</td>\n      <td>1.4691</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>8</td>\n      <td>1.4691</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{"_uuid":"ab25676aa4a93cb6213347bfd3dcd15d9c767fe7"},"cell_type":"markdown","source":"## 2. Feature Engineering\n\nThe idea is to group the acoustic data in chunks and extract the following features:\n\n* Aggregations: min, max, mean and std\n* Absolute features: max, mean and std\n* Quantile features\n* Trend features\n* Rolling features\n* Ratios"},{"metadata":{"trusted":true,"_uuid":"1772c3a5aa5dbe3d672031eb64e6bac596206588"},"cell_type":"code","source":"def add_trend_feature(arr, abs_values=False):\n    \"\"\"Fit a univariate linear regression and return the coefficient.\"\"\"\n    idx = np.array(range(len(arr)))\n    if abs_values:\n        arr = np.abs(arr)\n    lr = LinearRegression()\n    lr.fit(idx.reshape(-1, 1), arr)\n    return lr.coef_[0]\n\ndef extract_features_from_segment(x):\n    \"\"\"Returns a dictionary with the features for the given segment of acoustic data.\"\"\"\n    features = {}\n    \n    features['ave'] = x.values.mean()\n    features['std'] = x.values.std()\n    features['max'] = x.values.max()\n    features['min'] = x.values.min()\n    features['q90'] = np.quantile(x.values, 0.90)\n    features['q95'] = np.quantile(x.values, 0.95)\n    features['q99'] = np.quantile(x.values, 0.99)\n    features['q05'] = np.quantile(x.values, 0.05)\n    features['q10'] = np.quantile(x.values, 0.10)\n    features['q01'] = np.quantile(x.values, 0.01)\n    features['std_to_mean'] = features['std'] / features['ave']\n    \n    features['abs_max'] = np.abs(x.values).max()\n    features['abs_mean'] = np.abs(x.values).mean()\n    features['abs_std'] = np.abs(x.values).std()\n    features['trend'] = add_trend_feature(x.values)\n    features['abs_trend'] = add_trend_feature(x.values, abs_values=True)\n    \n    # New features - rolling features\n    for w in [10, 50, 100, 1000]:\n        x_roll_abs_mean = x.abs().rolling(w).mean().dropna().values\n        x_roll_mean = x.rolling(w).mean().dropna().values\n        x_roll_std = x.rolling(w).std().dropna().values\n        x_roll_min = x.rolling(w).min().dropna().values\n        x_roll_max = x.rolling(w).max().dropna().values\n        \n        features['ave_roll_std_' + str(w)] = x_roll_std.mean()\n        features['std_roll_std_' + str(w)] = x_roll_std.std()\n        features['max_roll_std_' + str(w)] = x_roll_std.max()\n        features['min_roll_std_' + str(w)] = x_roll_std.min()\n        features['q01_roll_std_' + str(w)] = np.quantile(x_roll_std, 0.01)\n        features['q05_roll_std_' + str(w)] = np.quantile(x_roll_std, 0.05)\n        features['q10_roll_std_' + str(w)] = np.quantile(x_roll_std, 0.10)\n        features['q95_roll_std_' + str(w)] = np.quantile(x_roll_std, 0.95)\n        features['q99_roll_std_' + str(w)] = np.quantile(x_roll_std, 0.99)\n        \n        features['ave_roll_mean_' + str(w)] = x_roll_mean.mean()\n        features['std_roll_mean_' + str(w)] = x_roll_mean.std()\n        features['max_roll_mean_' + str(w)] = x_roll_mean.max()\n        features['min_roll_mean_' + str(w)] = x_roll_mean.min()\n        features['q05_roll_mean_' + str(w)] = np.quantile(x_roll_mean, 0.05)\n        features['q95_roll_mean_' + str(w)] = np.quantile(x_roll_mean, 0.95)\n        \n        features['ave_roll_abs_mean_' + str(w)] = x_roll_abs_mean.mean()\n        features['std_roll_abs_mean_' + str(w)] = x_roll_abs_mean.std()\n        features['q05_roll_abs_mean_' + str(w)] = np.quantile(x_roll_abs_mean, 0.05)\n        features['q95_roll_abs_mean_' + str(w)] = np.quantile(x_roll_abs_mean, 0.95)\n        \n        features['std_roll_min_' + str(w)] = x_roll_min.std()\n        features['max_roll_min_' + str(w)] = x_roll_min.max()\n        features['q05_roll_min_' + str(w)] = np.quantile(x_roll_min, 0.05)\n        features['q95_roll_min_' + str(w)] = np.quantile(x_roll_min, 0.95)\n\n        features['std_roll_max_' + str(w)] = x_roll_max.std()\n        features['min_roll_max_' + str(w)] = x_roll_max.min()\n        features['q05_roll_max_' + str(w)] = np.quantile(x_roll_max, 0.05)\n        features['q95_roll_max_' + str(w)] = np.quantile(x_roll_max, 0.95)\n    return features","execution_count":3,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Functions for extracting features and creating dataframes. Make train also returns one Series with the target variable and another one with the earthquake number for each segment. "},{"metadata":{"trusted":true},"cell_type":"code","source":"def make_train(train_data, size=150000, skip=150000):\n    num_segments = int(np.floor((train_data.shape[0] - size) / skip)) + 1\n    features_list = []\n    target_list = []\n    quake_num = []\n    quake_count = 0\n    \n    for index in tqdm_notebook(range(num_segments)):\n        seg = train_data.iloc[index*skip:index*skip + size]\n        \n        target_list.append(seg.time_to_failure.values[-1])\n        features_list.append(extract_features_from_segment(seg.acoustic_data))\n        \n        # From which quake does the segment come from\n        quake_num.append(quake_count)\n        if any(seg.time_to_failure.diff() > 5):\n            quake_count += 1\n    return pd.DataFrame(features_list), pd.Series(target_list), pd.Series(quake_num)\n\n\ndef make_test():\n    submission = pd.read_csv('../input/sample_submission.csv', index_col='seg_id')\n    X_test = pd.DataFrame(index=submission.index, dtype=np.float64)\n    features_list = []\n    \n    for seg_id in tqdm_notebook(submission.index):\n        seg = pd.read_csv('../input/test/' + seg_id + '.csv')\n        features_list.append(extract_features_from_segment(seg.acoustic_data))\n    return pd.DataFrame(features_list)","execution_count":4,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"ff26be4db60a010ddf7e81554456bbd19e068f96"},"cell_type":"code","source":"X_train, target, quake = make_train(train, skip=150000)\nprint(\"Train shape:\", X_train.shape)\nX_train.head(3)","execution_count":5,"outputs":[{"output_type":"display_data","data":{"text/plain":"HBox(children=(IntProgress(value=0, max=4194), HTML(value='')))","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"55af3bdbecd544d5aed1d91c917edd3b"}},"metadata":{}},{"output_type":"stream","text":"\nTrain shape: (4194, 124)\n","name":"stdout"},{"output_type":"execute_result","execution_count":5,"data":{"text/plain":"   abs_max  abs_mean      ...       std_to_mean         trend\n0      104  5.576567      ...          1.044425 -3.268300e-06\n1      181  5.734167      ...          1.394229  9.090424e-07\n2      140  6.152647      ...          1.420060  3.962182e-06\n\n[3 rows x 124 columns]","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>abs_max</th>\n      <th>abs_mean</th>\n      <th>abs_std</th>\n      <th>abs_trend</th>\n      <th>ave</th>\n      <th>ave_roll_abs_mean_10</th>\n      <th>ave_roll_abs_mean_100</th>\n      <th>ave_roll_abs_mean_1000</th>\n      <th>ave_roll_abs_mean_50</th>\n      <th>ave_roll_mean_10</th>\n      <th>ave_roll_mean_100</th>\n      <th>ave_roll_mean_1000</th>\n      <th>ave_roll_mean_50</th>\n      <th>ave_roll_std_10</th>\n      <th>ave_roll_std_100</th>\n      <th>ave_roll_std_1000</th>\n      <th>ave_roll_std_50</th>\n      <th>max</th>\n      <th>max_roll_mean_10</th>\n      <th>max_roll_mean_100</th>\n      <th>max_roll_mean_1000</th>\n      <th>max_roll_mean_50</th>\n      <th>max_roll_min_10</th>\n      <th>max_roll_min_100</th>\n      <th>max_roll_min_1000</th>\n      <th>max_roll_min_50</th>\n      <th>max_roll_std_10</th>\n      <th>max_roll_std_100</th>\n      <th>max_roll_std_1000</th>\n      <th>max_roll_std_50</th>\n      <th>min</th>\n      <th>min_roll_max_10</th>\n      <th>min_roll_max_100</th>\n      <th>min_roll_max_1000</th>\n      <th>min_roll_max_50</th>\n      <th>min_roll_mean_10</th>\n      <th>min_roll_mean_100</th>\n      <th>min_roll_mean_1000</th>\n      <th>min_roll_mean_50</th>\n      <th>min_roll_std_10</th>\n      <th>...</th>\n      <th>q95_roll_mean_10</th>\n      <th>q95_roll_mean_100</th>\n      <th>q95_roll_mean_1000</th>\n      <th>q95_roll_mean_50</th>\n      <th>q95_roll_min_10</th>\n      <th>q95_roll_min_100</th>\n      <th>q95_roll_min_1000</th>\n      <th>q95_roll_min_50</th>\n      <th>q95_roll_std_10</th>\n      <th>q95_roll_std_100</th>\n      <th>q95_roll_std_1000</th>\n      <th>q95_roll_std_50</th>\n      <th>q99</th>\n      <th>q99_roll_std_10</th>\n      <th>q99_roll_std_100</th>\n      <th>q99_roll_std_1000</th>\n      <th>q99_roll_std_50</th>\n      <th>std</th>\n      <th>std_roll_abs_mean_10</th>\n      <th>std_roll_abs_mean_100</th>\n      <th>std_roll_abs_mean_1000</th>\n      <th>std_roll_abs_mean_50</th>\n      <th>std_roll_max_10</th>\n      <th>std_roll_max_100</th>\n      <th>std_roll_max_1000</th>\n      <th>std_roll_max_50</th>\n      <th>std_roll_mean_10</th>\n      <th>std_roll_mean_100</th>\n      <th>std_roll_mean_1000</th>\n      <th>std_roll_mean_50</th>\n      <th>std_roll_min_10</th>\n      <th>std_roll_min_100</th>\n      <th>std_roll_min_1000</th>\n      <th>std_roll_min_50</th>\n      <th>std_roll_std_10</th>\n      <th>std_roll_std_100</th>\n      <th>std_roll_std_1000</th>\n      <th>std_roll_std_50</th>\n      <th>std_to_mean</th>\n      <th>trend</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>104</td>\n      <td>5.576567</td>\n      <td>4.333310</td>\n      <td>-0.000011</td>\n      <td>4.884113</td>\n      <td>5.576566</td>\n      <td>5.576655</td>\n      <td>5.579118</td>\n      <td>5.576561</td>\n      <td>4.884088</td>\n      <td>4.883864</td>\n      <td>4.883418</td>\n      <td>4.883969</td>\n      <td>3.507118</td>\n      <td>4.050450</td>\n      <td>4.288590</td>\n      <td>4.011743</td>\n      <td>104</td>\n      <td>68.5</td>\n      <td>10.04</td>\n      <td>5.629</td>\n      <td>12.82</td>\n      <td>32.0</td>\n      <td>1.0</td>\n      <td>-2.0</td>\n      <td>3.0</td>\n      <td>78.313047</td>\n      <td>52.335255</td>\n      <td>31.029445</td>\n      <td>61.404926</td>\n      <td>-98</td>\n      <td>-18.0</td>\n      <td>8.0</td>\n      <td>12.0</td>\n      <td>7.0</td>\n      <td>-60.0</td>\n      <td>1.04</td>\n      <td>3.896</td>\n      <td>-3.10</td>\n      <td>0.421637</td>\n      <td>...</td>\n      <td>8.1</td>\n      <td>5.58</td>\n      <td>5.338</td>\n      <td>5.80</td>\n      <td>4.0</td>\n      <td>-1.0</td>\n      <td>-3.0</td>\n      <td>0.0</td>\n      <td>7.226494</td>\n      <td>8.195903</td>\n      <td>8.185756</td>\n      <td>8.310922</td>\n      <td>18.0</td>\n      <td>15.034220</td>\n      <td>16.948797</td>\n      <td>15.055998</td>\n      <td>17.430339</td>\n      <td>5.101089</td>\n      <td>2.897799</td>\n      <td>2.193084</td>\n      <td>1.612370</td>\n      <td>2.302065</td>\n      <td>4.568265</td>\n      <td>6.594138</td>\n      <td>11.064159</td>\n      <td>5.891685</td>\n      <td>2.801800</td>\n      <td>0.452294</td>\n      <td>0.295715</td>\n      <td>0.606039</td>\n      <td>4.526280</td>\n      <td>6.522304</td>\n      <td>11.248606</td>\n      <td>5.787738</td>\n      <td>2.809071</td>\n      <td>3.111524</td>\n      <td>2.769772</td>\n      <td>3.176148</td>\n      <td>1.044425</td>\n      <td>-3.268300e-06</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>181</td>\n      <td>5.734167</td>\n      <td>5.732758</td>\n      <td>-0.000005</td>\n      <td>4.725767</td>\n      <td>5.734187</td>\n      <td>5.734653</td>\n      <td>5.739510</td>\n      <td>5.734444</td>\n      <td>4.725732</td>\n      <td>4.725623</td>\n      <td>4.724876</td>\n      <td>4.725729</td>\n      <td>3.761435</td>\n      <td>4.436359</td>\n      <td>4.843486</td>\n      <td>4.379248</td>\n      <td>181</td>\n      <td>145.1</td>\n      <td>16.67</td>\n      <td>5.667</td>\n      <td>28.26</td>\n      <td>91.0</td>\n      <td>1.0</td>\n      <td>-2.0</td>\n      <td>2.0</td>\n      <td>122.978273</td>\n      <td>87.972617</td>\n      <td>38.643217</td>\n      <td>94.730087</td>\n      <td>-154</td>\n      <td>-99.0</td>\n      <td>8.0</td>\n      <td>11.0</td>\n      <td>7.0</td>\n      <td>-128.0</td>\n      <td>-4.09</td>\n      <td>3.412</td>\n      <td>-13.30</td>\n      <td>0.421637</td>\n      <td>...</td>\n      <td>8.2</td>\n      <td>5.39</td>\n      <td>5.066</td>\n      <td>5.64</td>\n      <td>4.0</td>\n      <td>-1.0</td>\n      <td>-3.0</td>\n      <td>0.0</td>\n      <td>8.265726</td>\n      <td>9.829922</td>\n      <td>10.544982</td>\n      <td>9.882948</td>\n      <td>21.0</td>\n      <td>18.772705</td>\n      <td>23.457270</td>\n      <td>33.704332</td>\n      <td>23.233021</td>\n      <td>6.588802</td>\n      <td>4.403190</td>\n      <td>3.563252</td>\n      <td>2.399157</td>\n      <td>3.752625</td>\n      <td>6.615549</td>\n      <td>10.727877</td>\n      <td>21.604075</td>\n      <td>9.299945</td>\n      <td>3.924070</td>\n      <td>0.496220</td>\n      <td>0.231587</td>\n      <td>0.764507</td>\n      <td>6.398675</td>\n      <td>10.162552</td>\n      <td>21.096014</td>\n      <td>8.823748</td>\n      <td>4.120785</td>\n      <td>4.893431</td>\n      <td>4.492905</td>\n      <td>4.953472</td>\n      <td>1.394229</td>\n      <td>9.090424e-07</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>140</td>\n      <td>6.152647</td>\n      <td>5.895925</td>\n      <td>0.000010</td>\n      <td>4.906393</td>\n      <td>6.152557</td>\n      <td>6.153045</td>\n      <td>6.159850</td>\n      <td>6.152678</td>\n      <td>4.906229</td>\n      <td>4.906088</td>\n      <td>4.905840</td>\n      <td>4.906072</td>\n      <td>4.080841</td>\n      <td>4.917334</td>\n      <td>5.423013</td>\n      <td>4.849219</td>\n      <td>140</td>\n      <td>94.8</td>\n      <td>12.38</td>\n      <td>5.957</td>\n      <td>17.36</td>\n      <td>60.0</td>\n      <td>2.0</td>\n      <td>-2.0</td>\n      <td>2.0</td>\n      <td>96.247424</td>\n      <td>53.353832</td>\n      <td>35.326896</td>\n      <td>65.431379</td>\n      <td>-106</td>\n      <td>-33.0</td>\n      <td>8.0</td>\n      <td>11.0</td>\n      <td>7.0</td>\n      <td>-72.9</td>\n      <td>-1.55</td>\n      <td>4.055</td>\n      <td>-7.72</td>\n      <td>0.516398</td>\n      <td>...</td>\n      <td>9.1</td>\n      <td>5.66</td>\n      <td>5.344</td>\n      <td>5.94</td>\n      <td>4.0</td>\n      <td>-1.0</td>\n      <td>-3.0</td>\n      <td>0.0</td>\n      <td>10.211649</td>\n      <td>13.485267</td>\n      <td>14.845834</td>\n      <td>13.127670</td>\n      <td>26.0</td>\n      <td>23.865014</td>\n      <td>28.598375</td>\n      <td>23.928873</td>\n      <td>29.448105</td>\n      <td>6.967374</td>\n      <td>4.429093</td>\n      <td>3.516817</td>\n      <td>2.414414</td>\n      <td>3.699345</td>\n      <td>6.789628</td>\n      <td>10.729778</td>\n      <td>19.898844</td>\n      <td>9.372244</td>\n      <td>4.179729</td>\n      <td>0.530151</td>\n      <td>0.267012</td>\n      <td>0.811309</td>\n      <td>6.641086</td>\n      <td>10.535572</td>\n      <td>19.047827</td>\n      <td>9.125804</td>\n      <td>4.227960</td>\n      <td>4.959233</td>\n      <td>4.402140</td>\n      <td>5.035827</td>\n      <td>1.420060</td>\n      <td>3.962182e-06</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"## 3. Cross-validation strategy\n\nSince there are only four thousand samples, it is better to use cross-validation instead of a simple split for validation. This can be implemented trough a generator, so it is easy to change our strategy. I am using KFold, but you can try a [Group KFold](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GroupKFold.html) with the quake series or [Stratified KFold](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.StratifiedKFold.html) if you prefer."},{"metadata":{"trusted":true},"cell_type":"code","source":"def fold_generator(x, y, groups=None, num_folds=10, shuffle=True, seed=2019):\n    folds = KFold(num_folds, shuffle=shuffle, random_state=seed)\n    for train_index, test_index in folds.split(x, y, groups):\n        yield train_index, test_index","execution_count":6,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## 4. Feature Selection\n\nOne option for selecting features is the [Boruta package](https://www.jstatsoft.org/article/view/v036i11) with a Random Forests model."},{"metadata":{"trusted":true},"cell_type":"code","source":"rf = RandomForestRegressor(n_jobs=-1, max_depth=5)\n# define Boruta feature selection method\nselector = BorutaPy(rf, n_estimators='auto', verbose=0,\n                    random_state=1, max_iter=500)\n# find all relevant features\nselector.fit(X_train.values, target.values)\n# transform values\nX_filtered = selector.transform(X_train.values)\nX_train = pd.DataFrame(X_filtered, columns=X_train.columns[selector.support_])\nX_train.head()","execution_count":7,"outputs":[{"output_type":"execute_result","execution_count":7,"data":{"text/plain":"   min_roll_mean_1000        ...         std_roll_mean_50\n0               3.896        ...                 0.606039\n1               3.412        ...                 0.764507\n2               4.055        ...                 0.811309\n3               3.722        ...                 0.959834\n4               3.918        ...                 0.903958\n\n[5 rows x 13 columns]","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>min_roll_mean_1000</th>\n      <th>q05_roll_abs_mean_1000</th>\n      <th>q05_roll_mean_1000</th>\n      <th>q05_roll_std_10</th>\n      <th>q05_roll_std_100</th>\n      <th>q05_roll_std_1000</th>\n      <th>q05_roll_std_50</th>\n      <th>q10_roll_std_10</th>\n      <th>q10_roll_std_100</th>\n      <th>q10_roll_std_1000</th>\n      <th>q10_roll_std_50</th>\n      <th>q95_roll_mean_1000</th>\n      <th>std_roll_mean_50</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>3.896</td>\n      <td>4.589</td>\n      <td>4.379</td>\n      <td>1.636392</td>\n      <td>2.475639</td>\n      <td>2.706474</td>\n      <td>2.331768</td>\n      <td>1.873796</td>\n      <td>2.576114</td>\n      <td>2.770139</td>\n      <td>2.465766</td>\n      <td>5.338</td>\n      <td>0.606039</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>3.412</td>\n      <td>4.552</td>\n      <td>4.345</td>\n      <td>1.646545</td>\n      <td>2.475965</td>\n      <td>2.674879</td>\n      <td>2.345991</td>\n      <td>1.873796</td>\n      <td>2.576820</td>\n      <td>2.729840</td>\n      <td>2.478973</td>\n      <td>5.066</td>\n      <td>0.764507</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>4.055</td>\n      <td>4.717</td>\n      <td>4.446</td>\n      <td>1.686548</td>\n      <td>2.538591</td>\n      <td>2.761534</td>\n      <td>2.404163</td>\n      <td>1.911951</td>\n      <td>2.645274</td>\n      <td>2.829290</td>\n      <td>2.540207</td>\n      <td>5.344</td>\n      <td>0.811309</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3.722</td>\n      <td>4.658</td>\n      <td>4.433</td>\n      <td>1.649916</td>\n      <td>2.496442</td>\n      <td>2.716991</td>\n      <td>2.368501</td>\n      <td>1.888562</td>\n      <td>2.595470</td>\n      <td>2.773216</td>\n      <td>2.492519</td>\n      <td>5.317</td>\n      <td>0.959834</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>3.918</td>\n      <td>4.714</td>\n      <td>4.543</td>\n      <td>1.646545</td>\n      <td>2.491521</td>\n      <td>2.719174</td>\n      <td>2.365052</td>\n      <td>1.885618</td>\n      <td>2.596404</td>\n      <td>2.792436</td>\n      <td>2.497672</td>\n      <td>5.306</td>\n      <td>0.903958</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"## 5. Making a Pipeline\n\nWe will be using a Sklearn Pipeline to perform hyperparameter search and to make predictions. The advantage of using a pipeline is that we are not leaking information from the training to the validation set.\n\nThe feature selection could also be moved to this pipeline, but it would take too long to perform the grid search."},{"metadata":{"trusted":true},"cell_type":"code","source":"def make_pipeline(estimator):\n    pipeline = Pipeline([\n        # Each item is a tuple with a name and a transformer or estimator\n        ('scaler', StandardScaler()),\n        ('model', estimator)\n    ])\n    return pipeline","execution_count":8,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"79675d6eca47c3039e729c937b1498eea8a54b60"},"cell_type":"markdown","source":"The next cell has two functions: one for searching the best hyperparameters and another for making predictions and ploting."},{"metadata":{"trusted":true,"_uuid":"c2f28f7269af8bd5ba9fe1bec70da271bc93d3fe","_kg_hide-input":false},"cell_type":"code","source":"def search_cv(x, y, pipeline, grid, max_iter=None, num_folds=10, shuffle=True):\n    \"\"\"Search hyperparameters and returns a estimator with the best combination found.\"\"\"\n    t0 = time.time()\n    \n    cv = fold_generator(x, y, num_folds=num_folds)\n    if max_iter is None:\n        search = GridSearchCV(pipeline, grid, cv=cv,\n                              scoring='neg_mean_absolute_error')\n    else:\n        search = RandomizedSearchCV(pipeline, grid, n_iter=max_iter, cv=cv,\n                                    scoring='neg_mean_absolute_error')\n    search.fit(x, y)\n    \n    t0 = time.time() - t0\n    print(\"Best CV score: {:.4f}, time: {:.1f}s\".format(-search.best_score_, t0))\n    print(search.best_params_)\n    return search.best_estimator_\n\n\ndef make_predictions(x, y, pipeline, num_folds=10, shuffle=True, test=None, plot=True):\n    \"\"\"Train, make predictions (oof and test data) and plot.\"\"\"\n    if test is not None:\n        sub_prediction = np.zeros(test.shape[0])\n        \n    oof_prediction = np.zeros(x.shape[0])\n    for tr_idx, val_idx in fold_generator(x, y, num_folds=num_folds):\n        pipeline.fit(x.iloc[tr_idx], y.iloc[tr_idx])\n        oof_prediction[val_idx] = pipeline.predict(x.iloc[val_idx])\n\n        if test is not None:\n            sub_prediction += pipeline.predict(test) / num_folds\n    \n    if plot:\n        plot_predictions(y, oof_prediction)\n    if test is None:\n        return oof_prediction\n    else:\n        return oof_prediction, sub_prediction","execution_count":9,"outputs":[]},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"def plot_predictions(y, oof_predictions):\n    \"\"\"Plot out-of-fold predictions vs actual values.\"\"\"\n    fig, axis = plt.subplots(1, 2, figsize=(14, 6))\n    ax1, ax2 = axis\n    ax1.set_xlabel('actual')\n    ax1.set_ylabel('predicted')\n    ax1.set_ylim([-5, 20])\n    ax2.set_xlabel('train index')\n    ax2.set_ylabel('time to failure')\n    ax2.set_ylim([-2, 18])\n    ax1.scatter(y, oof_predictions, color='brown')\n    ax1.plot([(0, 0), (20, 20)], [(0, 0), (20, 20)], color='blue')\n    ax2.plot(y, color='blue', label='y_train')\n    ax2.plot(oof_predictions, color='orange')","execution_count":10,"outputs":[]},{"metadata":{"_uuid":"d5255831db3f42ce8c31fcb18363ed7920f2d410"},"cell_type":"markdown","source":"## 6. Testing Models\n\nThe predicted values in the following plots are using a out-of-fold scheme.\n\n### Ridge Regression\n\nThe first model is a linear regression with L2 regularization.\n\n"},{"metadata":{"trusted":true,"_uuid":"38dd94339081930175c7ca60ed983fa802bd0e0a"},"cell_type":"code","source":"grid = {'model__alpha': np.concatenate([np.linspace(0.001, 1, 200),\n                                        np.linspace(1, 100, 500)])}\n\n\nridge_pipe = make_pipeline(Ridge(random_state=2019))\nridge_pipe = search_cv(X_train, target, ridge_pipe, grid)\nridge_oof = make_predictions(X_train, target, ridge_pipe)","execution_count":11,"outputs":[{"output_type":"stream","text":"Best CV score: 2.1485, time: 79.1s\n{'model__alpha': 9.332665330661323}\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x432 with 2 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAA0wAAAF/CAYAAAB67EmdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xl8HHX9+PHXzO7m3KRNQ5om3aUtlH7kKjdyiAjSUkCBooioKCIqqPAVRQXxp6CACMipCILggaKiWA4xGOQqYLlpLYWhBVqatEnTM00212bm98dmk02y987uzCbv5+NRSHZnZ9+Znd39vOfz+bw/mmVZCCGEEEIIIYQYT3c6ACGEEEIIIYRwK0mYhBBCCCGEECIBSZiEEEIIIYQQIgFJmIQQQgghhBAiAUmYhBBCCCGEECIBSZiEEEIIIYQQIgFJmIQQQgghhBAiAUmYhBBCCCGEECIBbyGfTClVC/wB2B3oB1YDXzUMo0MpdRhwB1AOrAU+ZxjGpkLGJ4QQQgghhBCxCt3DZAHXGoahDMPYF3gHuEYppQP3Al83DGMe8AxwTYFjE0IIIYQQQohRCpowGYax1TCMp2JuWgbMAg4Ceg3DeHbo9tuBTxUyNiGEEEIIIYQYS7Msy5EnHupV+jfwENAKnGMYxkkx94eAgGEYW9PYXSlwCLARGMxDuEIIIVLzAA3AS0Cfw7G4jXxPCSGE87L6niroHKYxbgW6gF8Ai3Pc1yHA0pwjEkIIYYejgGdTbjW5yPeUEEK4R0bfU44kTEqp64E9gI8bhmEqpd4nMjQvev8ugJlm7xJErtixbVs3ppldj1ltrZ8tW7qyemwhuD0+cH+MEl/u3B6jxJe5Bx/0cuWVpZgmfO5z/Vx5ZWnWMeq6Rk1NJQx9JotRJvz3lBvJMcucHLPMyTHLnFPHLNvvqYInTEqpq4nMWTrJMIxoV9grQLlS6kND85jOA+7PYLeDAKZpZf1FFH28m7k9PnB/jBJf7tweo8SXvosuKuWPf/Th8cAdd/Rw8smDQKkdMcqQs/EmxfeUG8kxy5wcs8zJMcucw8cso++pQpcV3xu4FHgbeF4pBfCeYRiLlVJnAXcopcoYKiteyNiEEGKyCIXgpJMqeOMND1OmmDQ1dbP77k5HJYQQQrhTQRMmwzDeALQE9z0P7FvIeIQQYrJZs0bjhBMq2LFDZ999B/nnP0OUlTkdVf4NDQX/BDAb2NcwjJVKqdnAkpjNpgLVhmFMi/P4y4GvARuGbnrOMIyv5zNmIYQQ7uBk0QchhBAF9OCDXs47r4zBwch8pRtumFSF7JYANxNTeMEwjLXA/tHflVI3kfx78feGYVycrwCFEEK4kyRMQggxCfzgB6X8+tc+dB1uuaWXT3867HRIBRVd529oKPg4SqkS4LPA8QUMSwghRBGQhEkIISaw/n449dRyXn7Zi99v8fDDIfbe23Q6LDc6GWg1DOPVJNt8Wim1EGgDfmQYxn8LE5oQQggnScIkhBAT1Pr1sHBhJVu26MydO0hTU4jqaqejcq1zgLuT3H87cJVhGANKqQXAg0qpPQ3D2JLJk9TW+nOJkbq6qpwePxnJMcucHLPMyTHLXDEdM0mYhBBiAnr8cQ9f+EI5AwOwePEAd9zR63RIrqWUmgkcDZyVaBvDMNpifm5WSq0H9gGezuS5tmzpyrqUbl1dFR0dO7N67GQlxyxzcswyJ8csc04dM13XsrpwpechFiGEEA762c98fOYz5YTDcPXVfZIspfYF4J/JeouGkqroz/sTqbZn5D80IYQQTpMeJiGEmCBME04/vZylSz2Ul8MDD4Q46CCZrwSglLoFOA2YATyulNpiGMbeQ3efDVwY5zGPAj80DONl4Gql1EFEFjvsB86K7XUSQggxcUnCJIQQE0B7OyxYUElbm04waPL4493U1DgdlXsYhnEhcZKiofvmJbj9xJifv5Cn0IQQQricDMkTQogi99xzOgcf7KetTWPhwgFeekmSJSGEEMIukjAJIUQRu/VWH6edVkF/P3z/+33ce28vunyyCyGEELaRIXlCCFGETBPOPruMpiYvJSVw330hjjpK5isJIYQQdpOESQghisy2bZH1ldat06mvN2lu7mbGDKejEkIIISYmGbghhBBF5LXXdPbf38+6dTpHHhnmtdckWRJCCCHySRImIYQoEvfc42PRogp6euD//q+Pf/yjB6+MExBCCCHySr5qhRCiCHzta2X87W9evF64++4eFi0adDokIYQQYlKQhEkIIVysqwsWLarg7bc9TJtm8thj3cya5XRUQgghxOQhQ/KEEMKlVq3S2W8/P2+/7eHAA8OsWCHJkhBCCFFokjAJIYQL3X+/l2OPrWDnTvjSl/ppauqhpMTpqIQQQojJR4bkCSGEy3z3u6X89rc+PB64/fZeTjst7HRIQgghxKQlCZMQQrhEby98/OMVLF/uobra4tFHu5k3z3I6LCGEEGJSk4RJCCFc4L33YNGiSrZt09lrr0EefTRERYXTUQkhhBBC5jAJIYTD/vlPD0ce6WfbNo0zzhjgqackWRJCCCHcQnqYhBDCQVdcUcIvf1mCrsP11/fy+c/LfCUhhBDCTSRhEkIIB4TDsHhxOS+84KWy0uLBB0PMn286HZYQQgghxpCESQghCmzDBli4sJJNm3TmzIksRjt1qtNRCSGEECIemcMkhBAF9OSTOoce6mfTJo2PfWyA//5XkiUhRHyvvKJz550+p8MQYtKThEkIIQrkyivhjDMqGBiAK67o4+67e9HlU1gIkcAJJ1Ry2WVlDA46HYkQk5sMyRNCiDwzTfjMZ8p54gkoK4O//jXEYYfJfCUhRHJer0U4rNHertHYKGuyCeEUubYphBB5tGULHHxwJU884SUYhFde6ZJkSQiRlkAgkiStXy/NNSGcJO9AIYTIkxdf1DngAD8tLRrHHBNm7Vqoq3M6KiFEsQgGIxdXWlo0hyMRYnKThEkIIfLgjjt8fPzjFfT2wne+08df/tIj85WEEBmJ9jC1tMiHhxBOkjlMQghhI9OEL3+5jIcf9uLzwb33hjjmGBmCJ4TIXEVFdEie9DAJ4SRJmIQQwiadnbBwYQXvvuuhrs6kubmbxkanoxJCFDvpYRLCWfIOFEIIG6xYobPffn7efdfDBz8YZvlySZaEEPaQOUxCOEsSJiGEyNG993pZuLCC7m44//w+Hn64B6/03wshbNLSomNJVXEhHCNf6UIIkYP/+79S7rvPh8cDd9/dw8c+JitMCiHsFQppbN2qUVsrWZMQTpCESQghshAKwYknVrBqlYepU03+9a9udt/d6aiEEBNVS4skTEI4RYbkCSFEht5+W2P+/EpWrfIwf/4gK1ZIsiSEyC9ZvFYI58i7TwghMvCPf3g5+uhKOjs1Pv/5fh5/PERZmdNRCSEmOin8IIRzZEieEEKk6bLLSrnzTh+6Dr/4RS+f+lTY6ZCEEBOYZUFNjUVfn5QWF8JJkjAJIUQK/f1w8snlvPqqF7/f4pFHQuy1lyxGW0yUUtcDnwBmA/sahrFy6Pa1QO/QP4DvGYbxWJzHVwD3AAcBYeBiwzAeyXvgYtLTNItg0JLFa4VwkCRMQgiRxPr1sHBhJVu26MybN0hTUwi/3+moRBaWADcDS+Pc98loApXExUCnYRhzlVJ7AEuVUnMNw+iyO1AhxgoELOlhEsJB8u4TQogE/v1vD4cd5mfLFo3TThvg2WclWSpWhmE8axjG+hx2cQZwx9C+VgMvAyfYEZsQqQQCpiRMQjhIepiEECKOq68u4aabStA0uOaaPs45Z8DpkET+/FEppQHPAt83DGN7nG12BdbF/P4+ECxEcEIEgxbbtml0dSEXbYRwgCRMQggRIxyGT32qnGef9VJebvHAAyEOOkjmK01gRxmGsV4pVQrcBPwC+Fy+nqy2NrfWbl1dlU2RTB7FfMzKy0HTYK+9SgEIhaqYMyf/z1vMx8wpcswyV0zHTBImIYQY0t4OCxZU0tams+uuJs3N3dTUOB2VyKfoMD3DMPqUUrcBDyXY9H1gFtAx9PuuwJOZPt+WLV2YZnaLj9bVVdHRsTOrx05WxX7MenpKsSwv1dU9QCUrVoSYPn0wr89Z7MfMCXLMMufUMdN1LasLVzIgVgghgKVLdQ4+2E9bm8bxxw/w4ouSLE10SqlKpdSUoZ814NPA6wk2vx/46tC2ewCHAE2FiFOIYDCSZMvitUI4Q955QohJ75ZbfHzykxX098P/+399/OEPvejy6TihKKVuUUq1AAHgcaXUG0A98JRSagWwEpgHfC3mMa8rpRqHfr0OmKqUWgM8AnzFMAy5pCwKor7ewuezZPFakZUBmYKbMxmSJ4SYtEwTvvCFMh57zEtpKfzpTyGOOkrmK01EhmFcCFwY564Dkjxm/5ifu4HT8xCaECnpOjQ2Smlxkbmnn/Zw+ukVPPZYNwccIN9v2ZJ3nhBiUtq2DQ49tJLHHvMxY4bFyy93SbIkhHCtYNCUIXkiYy+95AHgX/+SPpJcyDtPCDHpvPKKzv77+3n/fZ2jjgrz6qvd1Nc7HZUQQiQWWbxWhuSJzAQCkQuBkmznRo6eEGJSuftuHyeeWEFPD3zzm338/e89eOXCmxDC5QIBk/Z2jf5+pyMRxaSmJlIwRJLt3EgzQQgxaZx3XhkPPODF54N77ulh4cL8lucVQohcWDEV6INBE8vSaG3VmDMnu9L0YvLRhvIkmf+WG0mYhBATXlcXHH98BatXe6itNfn3v7sJBp2OSggh0hcIRHsKdObMkYs9IjOtrZIw5UKOnhBiQlu1Smf+fD+rV3s46KAwy5dLsiSEKB7RHoLoXBQZWiVE4UnCJISYsP76Vy/HHltBVxd8+cv9/OtfPZSUOB2VEEJkbuZMC02zZPK+yFp3t9MRFK+CD8lTSl0PfAKYDexrGMbKodvXAr1D/wC+ZxjGY4WOTwgxMVx8cSm//70PjwfuuKOXxYvDTockhOutXavR2GjJhQUXKimJLGArc1FEtlpadJSS5TOy4cQcpiXAzcDSOPd9MppACSFENnp74aSTKvjf/zxUV5s8+miIefNkgrQQqfT1waGH+jn11AF+/eve1A8QBSelxUUuWlo0lHI6iuJU8MsUhmE8axjG+kI/rxBi4lu9GubPr+R///Ow116DrFjRLcmSEGkKD3XCLlniczYQkZAsXityIedO9tx25P6olFqhlLpNKTXV6WCEEMXjkUc87LUXbN+uceaZ/Tz1VIiKCqejEkII+wQCJhs2aJgyqkpkYf166Z3MlpvKih9lGMZ6pVQpcBPwC+BzmeygttafUwB1dVU5PT7f3B4fuD9GiS93bozxO9+B668HXYc779Q499wSwJ2TMNx4/MYqhhiFmIwCAYuBAY32do2GBuk9F6nFruUl89+y55qEKTpMzzCMPqXUbcBDme5jy5YuTDO7D5C6uio6OnZm9dhCcHt84P4YJb7cuS3GcBgWLy7nhRe8VFZaPPOMRjC4k44OpyOLz23HL55cYtR1LecLV8IdLGuknLVwj2Aw0rW0fr0kTCIzVVVSYTEXrjhySqlKpdSUoZ814NPA685GJYRwsw0bYL/9KnnhBS+77TbI8uVdHHig01EJMTFs2+Z0BCKe2MVrhchEIGBKwZAcFPwdp5S6RSnVAgSAx5VSbwD1wFNKqRXASmAe8LVCxyaEKA5PPqlzyCF+Ojo0Tj55gOefD1Fd7XRUQkwc0iB3p+jitdJTIDIVDFq0t2v09zsdSXEq+JA8wzAuBC6Mc9cBhY5FCFF8rrvOx3XXlaJp8JOf9PHVrw44HZIQE8769Trz50tlAbfx+6GmxpLJ+yJjgYCJZXlpbdWYM0eGc2bKNXOYhBAiGdOEM88s58knPZSVwd/+FuLQQ6VBJ0Q+yNAdd7DitGsjQ6ukh0lkJto72dKiM2fOoMPRFB9JmIQQrtfRAQsXVtLaqhMImDQ3d1Nb63RUQkxc0iB3r0DA5J135PURmQkGo/Pf5GJINuQdJ4RwtWXLdA46yE9rq8ZHPxrm5ZclWRIi32TIl3sFgxYtLXrc3ichEmlsNNE0qZSXLTlqQgjX+tWvfJxySgW9vXDJJX3cd18PunxqCZF30sPkHmPLuweDJqGQxtatktSK9JWUQH29Je/tLMmQPCGE65gmnHtuGY884qWkBO69N8RHPiLzlYTIp9ELXEpj3K1GSotr1NZKN5NIXyBgyXs7S5JmCiFcZft2OPzwSh55xMf06RYvvtglyZIQBVRTY7F1q053t9ORiHhGFq+VJpzITDBoynmTJTlqQgjXWLFC54AD/Lz3ns5hh4V5/fVuGhudjkqIySW2mpZwn5HXR3oKRGYCAZMNGzRMuQaZMfk0FEK4wu9/72Xhwgq6u+HrX+/joYd68MqgYSEKLtqDIQ1yd6qpgYoKmYsi0hM71DYQsBgY0Ghvl/d2pqQ5IoRw3AUXlPGXv3jxeuE3v+nhpJNkjQghnBItPxwZuiPvRbfRtOjQKmn0ivRFzxuIVMFsaJD5b5mQhEkI4ZhQCE44oYI33/RQU2PS1NTNnDlORyXE5DZ9uoXPJ5PD3SwyeV96mERmRgqG6LLwe4bk3SaEcMTbb2vMn+/nzTc97LffIMuXS7IkhBvoukVjozTI3SwQMOX1ERmT+YnZkyMmhCi4Bx7wcvTRlXR2wtln99PcHKKszOmohBBRUk3L3YJBi23bNLq6nI5EFBO/P1IFU4ZzZk4+DYUQBXXppaWcd14ZlgW//GUv117b53RIQogxZL0Wd5OeApEt6Z3MjsxhEkIURH8/nHxyOa++6qWqyuLhh0PstZeMoRbCjQIBk7Y2L/39UFLidDSTl5VgXn5safEPfKCAAYmiFwiYvPuuJEyZkiMmhMi7detg/vxKXn3Vi1KDLF/eJcmSEC4WDJpYlsaGDdLL5EajKxkKkb5g0GL9ej1hMi7ik3eaECKvmpo8HH64n61bNT75yQGWLg3h9zsdlRAimdhqWsJ96uulkqHITiBgEgppbNvmdCTFRYbkCSHy5qqrSrj55hI0DX72sz6++MUBp0MSk5hS6nrgE8BsYF/DMFYqpWqBPwC7A/3AauCrhmF0xHn8b4HjgM1DN91vGMZVBQi9oDRt9JAvt2tr01i61MPpp4edDqVgdB2pZCiyEnsxZNo0GemRLkmYhBC2C4fh9NPLee45L+XlFkuWhDjgAPlgFo5bAtwMLI25zQKuNQzjKQCl1HXANcCXEuzjGsMwfpHPIN1g5kwLTbOKYsjXeeeV8fzzXo48sovGxskzzkgqGYpsjCxeqzN/vnwvp0sSJiGErdraYMGCStrbdWbNMvn3v7upqXE6KiHAMIxnAZRSsbdtBZ6K2WwZcH5BA3OhkpLIsK9iaJD39UV6wd5/X6excdDhaAonELB46in3vz7CXUZ6mNzfe+wm8k4TQtjm6ad1Dj7YT3u7xgknDPDCC5IsieKhlNKJJEsPJdnsW0qp/ymlliil9ixQaI4oltLiI1fM3R+rnQIBk/Z2jf5+pyMRbmZZo98X06ZZVFTIcM5MSQ+TEMIWN93k4+qrS9E0+OEP+/jGN2S+kig6twJdQKIhd5cBGw3DMJVSnwealFK7GYaRdrdGbW1uFU/q6qpyenwy0cWj/f4y6urKmDsXXnwxv89ph3nzIv/ftq2currx97s9/mTKyyPzleL9DXvtFSk73ttbxcyZ9j5vMR8zp7j1mE2ZEvn/tGmVw++PWbNg06YS6uqcXTPArccsHkmYhBA5MU0466wympu9lJbCn/8c4sgjZVy0KC5DBSH2AD5uGEbcE9gwjNaYn3+vlLoRCADr0n2eLVu6MM3s5tnU1VXR0bEzq8emo6sLoIqurl46Ogaoqyth/foS2tu70F18MbqkpAQoxTD66egYvRB2vo9ZvvX0lGKaXjo6usfdN2WKB6hg+fIQ1dX2DUUs9mPmBDcfsx07vEA527Z109ER+WhraCjnnXc0OjpCjsXl1DHTdS2rC1cu/ggUQrjdtm1wyCGVNDf7aGiwePnlLkmWRNFRSl0NHAScahhGX5LtZsb8fDwwCLQm2r7YBQIWAwMa7e3FMdStGOZb2amYKhkKdwkETDlvMiQ9TEKIrLz0ks4nP1lBTw98+MNh/vrXHldfhRZCKXULcBowA3hcKbUF+BRwKfA28PxQQYj3DMNYPPSY14ETDcPYAPxOKVUPmEAncLJhGBO2lnXs3KCGBvdXn5tsDcBiqmQo3CUYtNi6Vae7GyornY6mOEjCJITI2G9+4+P73y/FsuCii/q59FKZdSzczzCMC4EL49yVsKVtGMb+MT8fl4+43Cp2vZZDD3V/z/Fkm8QerWQ42f5ukbuR3kkdpdz/3nYDSZiEEGkzTfjKV8pYssSLzwe/+10Pxx03ecr4CjGZxDaqikFvr0ZXF/hzq6tRVIqlkqFwl9jS4jGrLIgkiuNTUAjhuK4u2HtvWLLER22txbJlXZIsCTGB+f1QU2MVVbnuiTY8zUoxElIWrxXZiF28VqRHjpQQIqU33tCZP9/PW2/BIYeEWb68m2DQ6aiEEPkWmRxePE2FydbbEgiYbNigYcqoKpGB+noLr1d6JzNRPJ+CQghH/PnPXj760Qq6uuCb34R//rOHEmeXbhBC5EG0N0OLaUMVWzWtyXbFvNgqGQp38HigsVHmv2VCjpQQIqFvfauUCy8sQ9Pgzjt7ufFGpyMSQhRSMBipwpZqaJhbFFNyZ4fYSoZCZEKGc2ZGjpQQYpzeXvjoRyu4994SpkyxWLq0m1NOmbDVk4UQCQQCJqGQxrZtTkeSWkNDcQ0ftENsJUMhMiEFQzIj7zAhxCjvvAP77lvJ//7nYZ99Blm+vJu5c4vk8rIQwlbF1CAPBCbmmkRakjZtsVUyFIWXqHc4EDBpa9Pol1VB0iLvMCHEsIce8vChD/nZsUPjs5/t54knQlRUOB2VEMIpxVRNKxgsrvlWdijGSobCHYJBE8vS2LBBzp10uP8TUAhRED/6UQnnnluOZcHNN/dy4419TockhHBY7HotbhcImLS36/RNso+uYqtkKNyhWHqPe3th3TrnP3/cfZSEEHkXDsNJJ5Xzq1+VUlkJjz8e4swzZb6SEG5lWRSsjPS0aRYVFcVRTSvaAGxtdb5xVUjFVslQuMPIcE53nzvf+lYZhxziJxRyNg73fwIKIfKmtRXmz6/kpZe8zJ07yPLlXeyzjyzoIYSbzZ3r54QTCjNWVtOi1bTc3aiCkeGDxZDc2anYKhkKd5g5M3LCuH247csvewBobXU2TncfJSFE3vznPzqHHupn82aNU04Z4NlnQ1RXOx2VECKVnTs1XnvNU7Dni1TTcn9zoZiGD9qpmCoZ5iIcTlzAQGSutBTq690/nNMtpfPdfZSEEHlx7bU+zjyzgnAYrryyjzvv7EWXTwMhRBxuH/IVbUTPnGmiaROzUl4yxTIXJRe9vdDYWMW118qq6XYqhtLibjm/J+67SwgxjmnC6aeXc/31pZSVwSOPhPjKVwacDksIkYXe3sI8TzBosXWrTnd3YZ4vWz4fzJhRHL1hdiqmSobZipa+/vnPS50NZIIphsVrGxrcMdfK3UdJCGGbjg448MBKnn7aSyBg8dprXRxyiMxXEqJY5asc8Nh1f4pprZ9iuGKeiXSGoE2GoYjJ1qIS2QsETDZs0ApWRCYbPl/k/04ndu7/9BNC5Oz553UOPNDPhg0axx03wMsvd1Nb63RUQohc2N2ASLbAJRRHg7wYrpjbrZgqGQp3CQQs+vs1Nm1y/3vb6c8feXcJMcHddpuPxYsr6OuDSy/t409/kvlKQhSz8vL8j+nX+tqHfw4Gi6OaFoxcMR8cdDqSwtG0yN/t9KR4UXzcUlAhHU5fEHD/p58QIiumCWefXcbll5fi88H994e46CKZryREsWtoiCYw+WnkqIqH2OWZPfBtfRqA+noLn684hroFAhbhsEZ7u/tjtVOxVDK0g9vn0hUTtxRUSIfTF0Lcf4SEEBnbvh0OO6ySRx/1MX26xcsvd3H00YUbpLyxuYmlZ5xK8zGHs/SMU9nY3FSw5xZiovP58tvICZT/FwBv52sA6Do0NhZHg3wyFECIx+2VDO1k93m4YoXOVVeVTNiS5cn+rmJ6v1iWRlubc+e4+4+QECIjr7+us//+ftau1TniiDCvv97NjBmFe/6NzU2suv4aetvbwLLobW9j1fXXSNIkhM3y1kC2hvYb09IKBk3ef9/9TYaJWAAhnYIHxVLJ0A52v7af+1w5N99cOuHXsYp3Hvn9MHWqu3uPLQtu++L5rLp2T0cTO/d/+gkh0nbLJQbHLywnFILFu/6JX339H3i9hY1hzV23Y/aNrnds9vWy5q7bCxuIEBNUNI/JV4+PNdw0GEmYiqX6XDFV9LPTZPq77W4019REzvPW1ol/7OKJ9E7m528vW38Xes/6nPdz/nG3s+fMtxz9DCpwU0qIiWNjcxNr7rqd3k3tlE2vZ+6559GwYJFj8XzptC08/OxBeAhzYeCbHOx/ilXXlwEUNK7eTe0Z3S6EyE50TL/HY/eetaH/jgzjDQRM2tu99PdDiYvXDq2shGnTJl8BhNieNaUcDiYPYoeV2d1oDgQs3nwzkojtu6+L62vnSSBgsnat/QlT2fpfU/XWxVh6OZs/as/3v5MXBCRhEiJDG5ubMG69kYHOHcO3RYedRRUykerqghNPrOCtt6rwe7bxkzlnML2kFRjp2Un0/PlI+sqm10eG48W5XQhhn3A4MqZ/5kx7J19YxB+SZ1kara0ac+a4cbKHReCZakKzv0UgcO2k6GmJNXouysQuEWj3a1tMZfPzIRi0WLpUx7LsXe+q6q2LAdDMHtv26eSFEEmYxIQULxGo+8zpWT3eV1WNhUV45068VVUMhnqwwuOrzZl9vRi33shgX9/wkLTYRCofSdNbb+mcdFIFO3dq7Fa2gh/NPguvHh61TaKenehcI7tjnXvueaP2C6CXljH33POy3qcQYjS/36KrS2P9ep2ZM+1uIEcbJaOH5EGksTpnjo3PN9jNtOcPYcf+f2awan5Wu7As8OiRmMrX3kQgcA1r1kyuhKm+3sLrzc+wSe/2FyjZ8iSh3S+G+g/SAAAgAElEQVSxfd/ZsHtI3rRpxVM2Px8CAZPubo3t26GmxuloknPyQsjkPDvEhJao6MB7jzyS1eMHOncQ7uwEyyLc2Rk3WYoa6NxRsPk7f/+7l2OOqWDnTjj77H6uO/yicckSJO7ZeevWG/ISa8OCRex18SWU1c8ATaOsfgZ7XRz5opXKeULYI9qjkI8G8odqrh36KTZhys/zVay7DU9vC9OWfSin/UQTJg2TYDBS0W+iVj2Lx+PJXyXDmpcWUPnu1WC5o+fK7nMw2qsyWXuYiqm0uMxhEmKMTIaKjd023BOKmwgsv+kmdjunJ+V+4xUtyJXd83e+971S7rnHh8cDv/pVL5/4RJiNzen37GxsbookgXmKtWHBolHHNV+9WUJMVjNnRuZd5LWRE5NxzJxpoWmW/VfhrfEXebIRTZggktyFQhpbt2pMn27L7gHwv/VtPF0GOw5O7+JbPN5t/yU85UDQS+0LbEgwaOa1l8S/6gK69r4tb/tPKwa/RVublpe5dMWQMORD7HDOfM3h0vo2YZXm/maMXgixc+hguibn2SFcLZOy1PG2TZQIhDZuTLrf6NpB8ebf5Mqu+Tv9/bBwYQX33FNCVRU8+WRoKFlqGkn09MjbOtqzEy8hSdaLZOdcozdvvJbmY49g5VWXS+U8IWxUUWFRW5vv4gYjjaeSksiwL7sblZ6edbbsx+sZSbzyVVq8fP2dlGx7JuvHe7oMal4+Hr+Rn6Ft+a5kWL7h3rztO13RuXQbNtj/d0oPU/7+/l2emWvLfnp7LTZvduZ1kh4m4TrJylLn0huk6Xrc/a786Y/ZvnIFG5oetb1nCeybv7NuHSxcWMm2bTof+MAgTU0hKirG995gRho5uxx2RMLem2S9SNFYc50H9uaN19Ly4ANJtxkbh9sqDwr3UEppwLnAmcAuhmHMV0p9GJhhGMZfnY2u8A6c0QTzd6elZV7enkMbHD1ZOx8N8rKNf7JlPx5tpIfJrYtxagNbAfDuXJmX/QcCJm1t7q9kmItoRbuWFp3Zs+0dIrhlS2Qdq8pKW3ebM61/C/rAVgYr98jL/mtrLcrL89B7nAfzd11BS8se1NUVfrytJEzCdRL18MRr5GfSG2SZCbqaTTNlwz4XiXp50rWxuYk//nQ5P1/5IwbROH6f1/hK3cU8d1IkqRjs6Ymb6LU8+AAtDz5AWf2McYmHr6p6VJW/KE95OQ0LFiUcQldVXY7/g0enFXfLw0tSbhPbm5XPYXvDiVh7W6QHzjTjHhfhaj8GFgA3AdGuyRbgRmDSJUw/OvpkOBrmXW7PkDZg3JwfbbBr1O/BoMmrr9pew9wWP//st4d/dmXVMzOMb8fLQ79k2tizwDJBS96g3XXXkd6X2bMn5gSufLy2sed9a6vOvHnuKi0+7bkD0MPb6drjx/TM/mZW+yj19XJUWz3ddbfRX7941H2aFl2LyUXvlwR0LbJm1AEHFP41cn86mWfRYVh/2mcfmYjuAsmO/9ihYm/eeG2CLcFXPSVSdMBh0RiyLXawsbmJ/3dRF9eu/AkmOufMuILPm58bNawwXuITK96QRivBF/ZgT8/o4X0xovPAUsUb/VtJlKAO0bw+wj2h4eNi3HpjXobtjRq2CcNxJRvqKVzpbOBjhmH8mZEW53vAbo5F5ALTtBX5K25gjX4PBwImGzZoqd7aWYv2wGTjnI/cM/xzTU1kyGK+5qSUbP53xo+pfOcq/KsvA8C348WMHtv0zcOpe3zq+Ix2jHxM3tf6O2zblx2iJfTt7g3xl+3ksLn/tb+gxMA26pqrKWu5O+t96OHtAPhX/zDrfcysacVrdeNf/aO490d6j208pjZ+KMXuStecW2NtUidMmcyVEYWRrHEcO6xtY3NT0l6haBLhra62L7gs7HLYEVmfY+EwfPoru/HQpi9SqvXw4zmf5qPT7s8qDrOvl5VXXU7zsUfQ/JHDEs7zAkYnF2OE2hL36I19P6VihQeGqw8mS/xynVOWbNhmbEIWm+zJxRNX8gDRLo/oCeaPuW3SiH17HTL7WbZssa8B4fWMVAEtb71n1H2BgMXAgEZ7e34aLFNePc2W/WhatABCfuL0hN7J+DG+bUuzfr4DZr0U2cfWJ5Jul4/eF70/+yQ2qnTDffi2PJXTPiwL5u+6nPPnHMGcYKftyfDO31Tz3yuOYGurvXOY9d4WAMrX32XrfjNhWZFEIyL+uWF3D5Nvy+O27StWVWXYseIckzphSjZXRjgj2dya2KFTb916Q+p9JSkAUShtTz6e1TnW1gb77GHxZvdBTPe9z617HMNu5atyDyiNS8NmX2/CEjQVMxL32uWjumBUbAKTaWKTqupf76Z2uXhSHP4F3KCUKoXhOU0/AR52NCqHHbr7i7Y2dAZ+n3jyy8jcoPwkIt6d/7NnR9ag/VfMY2j9mzN+TKa9SlFHzfzdyPOGk18byEslQxvKkVW/8VWmvnpyzvv52ae/R2PZq5x06NK8DR/r2pz5a5vMtGVH2rq/bOl65L3r6Xkv7v3BoDU8h8sO+ZqnN2NG2LGhgwVPmJRS1yul3lNKWUqpfWJun6eU+q9S6u2h/+dndluMRA2pfFRJE+lJVqEttnHsdCKUroSlu9vbEjbGn35a5+CD/WztruLgqmZ+vvsJVHoLfBHdssAzeoqjXlpG44c/nDBZsbt0eqxoAvPmjddmnNikqvrnraqSiyfF4SKgAdgBTCHSszQL+F46D871u0cp5VFK/VIp9Y5Sao1S6tzc/6Ts/O4Lxw3/fNaH7i3YZO38r9dizzCe0ra/D10xz0+cle9dl5f9xnPO3ufH/Jb8gld+Khm6Z16LNRTLt479ft7O+c2b7JsTGNsV7O3KTwKRrkN2eynp/dGLIa2tNh3XOIm2p8vIebdqVvukGpK3BPgwMLaW6O3ALw3DmAf8Ergj34H4qhIP15Iry86Ye+55aF5f3Psm2lX/eH/LjTf6OP30Cgb64TPTr+Oi4DejVcILb3Dki6OsfgaNi07kvQcfTJis2FmOPB6zr5eWh5dknNikqlCooSUuNCIXT1xhqDdpF+B0YFfgMGB3wzAWG4axM83d5Prd81lgLrAHcDhwuVJqdiZ/h10On/PUqN8LdcV1ZMiXuwanjCtWYfYSDFps26bRVeQDNk1GimxoaSwca38lQxclTFYkljlTV+RtLt3WzYkXps+YmZ8RF5mapr3OvV87K+k2hSgt7gmtznkflx/7ST4S/F3qDfOg4J96hmE8axjG+tjblFLTgQOB+4Zuug84UClVl89YEk18BzBuvTGfTy0SaFiwCE9FecL7J9JV/2hJ8+ZjDufp00/l2EPa+elPSynxmVw252xO2sWZD4V4djnsCDYve57B3sTJytxzz0MvLctvIAm+IVMN5Uw2l21gZyfJstKJkqAXM8MwLOB/gGkYxibDMF4yDCOjbNaG754zgDsNwzANw+ggkoClX2c/jwqVwPj9UFNj5fEKb3Y9TB+o+ueY3VjDyd06e5Z5Gi9vlTZGM62YqoRW6gzB7sVrrRSV+ZIy+5j27HzbYjGtkVgCU99l0yb7z8PP7neVjXtzR7JZrz+fchv7S/HH+dvTSPjTcdicJ9mZ7mUyG7nlMlEQaDUMYxBg6P8bhm7Pm3CSI56q8pjIn1TD7Xrb2xwv5pCRZGPATZPOgSmc99zvefLleqb52rlp7kfZs/zlxI9xQMuDDyTthVl6xqmsvPoKPKWl6GWJE958ide7tbG5iadOPj5lkYuy6fVJ53ZNlAR9AngNsHvRoUy+e3ZldO/U+wm2K7hCDlHJ51C3dHpQ4vnkzC+P3s9gV/4TJpuGD6Z8lpgkIdrgrGuupvKt78Td3v5KhtmfW3pvC56etXYFgmmOHIuPHfCIbed9Q/nI3Lma8nYG7OpksmH+ly2s1HHU11t4vfld+DjVkNJMOLFm1IRah6m21p/R9r7qagZ2JE6M6uqqcg3JVm6LJ55cY3zvkUfS2q68ro6dRTKPCcuioqGB0MaN4+56O7QfP133G/qtMvatfJbvBr+KXujPWE3L+WppNJmKXmjwlJejl5QkfX9lw1tRgWWao3q6PGVlHPCti4bPvbq6Kt575BFWXXsVZopvPk9ZGYGPHM2av/wl4Ta9m9ptfe9NhvdxnjwFNCmlfgusJ6bFahhG9jV7CyzT76mx4r02bW0+6uriD2XOhC/OLsY+3+67w9tv5+8cyWa/oTGfmf4Kjf33j6w+um4dnHii/bHW7VIJevZNqHT/zp6Y69rVVSUw9LiK9XdQcdT4izl77gn9/TA4WEV9DqOkh+Prqop/ezp2jj7XczlnfL6ROUwAmmbR2VlJnQ3jkL695yHDPzdO3UBXXxWNjZnvZ9zfN1iS/P4sZLOP1vJSiJmalWgfwSB0dJRSV1eabXgjOsbvY4p/5PxNFUussQsJ11ZtoauzxJbPvEy4JWFaD8xUSnkMwxhUSnmAxqHb07ZlSxemmUHDL8mm3upqOjoc6PNLoK6uylXxxGNHjK/dkN5QyJ3vZF7W1TG6HjdZatryGe5tvxQLjcW73MYnp99W+NBKy2hcdKLtC/cO9vRgDoTRy8oxe3ts2284FBp11U4vK2fPb38P/wePpqNj5/A5+NoNN6ZMltB1Bnt7WXN/8lLtZdPrbXvvTfT3sa5rOScESRxJZN2lsSsnW0C2CVMm3z3vEykyEZ09PbbHKS0Zf0/FiL42Y9uIa9dadHTkPllnxw6YOua2sefC9Oml/PvfPjZt6rLlAvrYvyWbc69yzOHs6urDW7sTn8/PunWafe/fmJ87OnaAnriiYLLHRh6fXkwVMUPyujavo6dq5PWPt4+pUz1ABa+/3o3Pl90V/djPAL2nm9os4gbQQ11ZP3asHTtG5jABePRB3nijj46O/qz3GRX72tRVb+a55SH8/sx6O+N+bg72jjlnsvv7c91HqCcMMblFon00NpazZg10dOT+nV3e3c/Yb4LOzm76Yp473e8as3d06b4T9mvirfYz6OjIbtpCtt9TrhiSZxjGJuB14Myhm84EXhsaJ543AzsT91DMOOa4hPeJ/MlnpTXHjBkbYZpwS8vP+UP799EJ853gVx1JlgD2uvgSNi9LPb45nlRD76zwQPxkyZvjdZqY3jCzt4ftK1eM2ySt8yj6uiQZu6KXlqUsGiEKwzCMYxL8OzaHfWby3XM/8GWllD40x+lU4G/ZPredtm/PX3EDz5hS34GASSiksW1bfp4vum5NTvsY2IyuQ2OjlbcheZ7u3Ct+pSN2SJ5/zeUpt7e/kuHobNTTvSbtR2pjRi5o/VtyimRkLSHwV+SvvLR9Q1xzH7apDWy3IY70/p58luIHsp7D9PHG8cNPP1Dxj1yjyZgTZcVvUUq1AAHgcaXUG0N3nQdcoJR6G7hg6Pe88lYl7grMthEpcpPvSmtOC4Ur+N66Jl7oXES1dys3zF3E/lXPORbPW7fekHUluGx6jsrqZ6AnqIKYrZaHl4x/HjvOI11nr4svGbX+l3DOUKIS91+aj8/4u0cp9ahS6uChX/8AvAusBpYBPzYMI/6iJg7I15j+sevI5Lu0+NQXc79YWbHuViAykT1fCVPVm9/KaHurQAUAonO3bDsfxiQ9U1/6aCYPHvWbZvblFMpRHxhZ/LemJn8LmNq134q1N+e8j8rVl+ceSNoJk0lbm2bfHK4xzJLsxk/69JDNkWSn4EPyDMO4ELgwzu1vAR8sZCxakpNIygk7Y+6557Hq+mvytgCqk9b2zOMn6/5Ar+ln/3kd3HnZU7z3q14GHJyKlfN6VhnOf8rL+ypOD9Hcc8/jjZ9dhRXO7pNfLy2TZMl9wiS+ZOtJcPuwbL57DMM4MebnQeD8eNu5QUuLxp575raPdN7KsdW05s+3v66zp29Dxo/Rtfjv80DA4plnco0okcz+9p5dv0bF+7/M4nlGvyhVbyQ/BaOVDO3rfRmT9AzkMi81+x4XywIt5vE1U/PXw2TXfive/WnO+9DM3IfHWWP7Rsxw3Pl3waCJaWps2KAxa1auvWPjj6HlqYyzXTp7KkyBlVTcMofJEcmG5InC2tjcxJq7bqd3Uzu+qmr00pKiWZw2HU9tW8xdG6/AQueEab/l85W3UuK9hI889NjI397eFilvnY/FJfLFstC8vqwTE1toGkvPOJXeTe1UzJjBbud8dTjRefOGnzHYk9kXTln9DOaee54kS+4zZ8zvDcAlwMMOxOI6kR6F3Mr2lnYlX9wSCrNeS6bKPPHnQQQCJhs3RooglKQ/3Sg9aZT4HkVLmdPHf5g2urFYtuGPKR9jZyVDbezfmUGZ8XEJeKbHbIxBc+QYTpsaZv16PZJI2Xwq2tU7Z09DP/f2wNjeTW1wJ5ZeM2672N7jWbPsKQE+6nmz/Ft0zf5YsuGKOUxOSbZwrSicjc1NrLr+muEFUQc6dzAYsq9QgNN+veEK7tz4EzQsLpx5EZ+bcd249aSGG/XFlCwB6LqzydKQ6LkT2riRlVddzpMnLwTg2H89mVEJ+rL6GRz1lyWSLLmQYRjrxvxbBnwB+J7TsTmtpMSeHoWS3tTFdKZNs6ioyPNcB5sEgyaWBa2tuR8bbWDrqN99na9kuIfsGs9eLfOiBpGEKV/zcNJ/3cfPc8vt+629c2So9cdmXUV3t8Z2O6b4xHhr8wfzd26HM59oqPeOLxiVqfE9TIl6ZKO9x3m6GJLlHKZ9pj5kcyDZSdrDpJTaLZ2dGIbxrj3hFFayhWsh0pCXhlNqsb1DZdPrM746v+au28cNwXNDIzxXvWYZV7x3L+/37UmlvoMr5pxBQ+lI8a3e9rbhZNH1QxC9Poj3mqSZ4Hmrq+3vMdR1tJISrN7xxy7c2cmq668B4AMXfCvtYywFHopONeMLkE06jY32JDCWnrqcsKZFGlaFXPspG6Ub/kQgcBYQuWI+Z05uV6n9ceYsaf0dWOnOy8hy6YZSb+bzN4JBi2eeyU/vSyY9ZZ6xLUzLxLNzJYP+vbMKrD880k0Y7XlradGpqbHvQuOm8H60tkbWsUqynnlWvN1vE55yYEaPsWNI3tjhcXq4k8HS6eO2mjnTxvmJ8V7fHHsYnZbqqKwhMsE1+v/VCX4vSqkacLJgZWpje4d629tYdf01bGxuSnsfE7Ey3sa+WVzw9hO837cns0pX8Yt5HxmVLEXFSxZdKYcENnDKaRzz0L9tX9DW6/fHTZaizL5eVl59BRCpBlhWPyPlPuUCiXsppf6glPp9zL+/Aa8A9zodW0HFaXgHg6Ytw4gsLb1xa/mupqX3vJ/zPko3PTR8xdyO3hYtPH7YnzaYSWO2cPMwAgHTxt6XsYUbsm/Al2x+jGnLjqA0jWGF8cQb4mZ3sZOqSpP+fo2OjnxcEMj8HLC03IskjVr8GKh49+q425WWQn29Pb2TGzeOf13sqIDppKRnmmEYumEYHsMwdOBc4M/AB4Cyof//CfhS3qPMlxSXDyZiQ95u8Rr8Y4ebpTLRKuMt23E8333nIUJmNcdM/StX7346JXr8YRWTobhIy4MP8NTJx9u6HhOkWbDCslh5zZUAHPWXJUmTpnQSKuGoNcA7Mf+WAZ8xDOMCR6MqNGv8xYsD562xpZFjeqaktZ29Q77G0/vs+VycOdNC02xqVOuFXSQzF7ZWMsyhV2Ds/CfvUCl2b9fKrPa336zxS0jY3dMZ8K/My34BPKEsBmPFmzNmDuDtfC2TnYz+LUnhjkDAsuX9Em+Zg+pVX8OzM7vX3g0yOSo/Ac41DGO1YRj9hmGsBr4KXJmf0Aog1XCicf3JYqxESWUmyeZEGgb1h43f5dbWn2Oh8ZXGH3Bu4xVOh+QKA525VFbK0WAY49bIgsjJzsve9jaWnnFqRr2jonAMw7hizL/rDcN43Om4Cs4cf/Fl39mraW/X6cutajOmnl4vcDBosXWrTnd36m2TSjinwZ6hOyUl0NBgT+JQ2vFonFszaVSP9C6EK3MsZ5hCbCXD3OXSMxb/dax4/zZKNmVWq6Wk581xt122+BrbE/c6bRmQn7L51Suz6F8Ym7BaFpWrf0DNC0fj6X47rV1UaKPnQWlJisMEg/YUDKlMUBDPv/qHOe/bKZlkBDowG4g9a2eRRjlX10pVkWwCzKPJt7Lp9XF7STLpNWpYsIiVV11uY1SFFza9XLnut6zuOYAyvYsfzj6LWWXpfZi5ieb1YVkWDIadDsVWA5072NjclPB8jYoOKQUZnucGSqlz0tnOMIy78x2LW2jW+Pfmh2beC3yc1laN3Xazf+iXb+szDEz78PDvI0PddJTKIbmJ87cAaFlODh+zFwBmzcrjJHZNg8EQ3q5VhKccnGLjmNfF9olFo9lbyTCH8ylOYz9qyvLP0rEg/Xmt+uD4ba/85KVc9p/9gSPHPyAn9vSy2GHse6F87Q34hnqXtP4tkEal7oN9Y5KUJPPpAgGTRx/12jCHK/65V7LlcbzbXyA8taCrCNkik8NxI/CEUupqpdT5Sqmrgf8M3V6ciq0imQvNPfc89NKyUbfppWUZ9RoV+xX9zf31XLD6CVb3HEBDybvcOveY5MmS3TNJczUUT1n9DPb+3mXsc8kPIsPTNA1Pub3zjpy05q7b456vY2U6pFTk1Vlp/PucY9G5hCr9K2DHVfH4DanKNT8Z9bttDfJEDbcEiVQ2IglT/j5zq1d+lZoXj0Xr35x8wyyLPmQjWsnQlr87p7jHtrGy39egZ1rc27dvtf/C9i7T+l1UNn/0MSzbcC++7cts3WesQMCyZw5Xkofr/R257dshafcwGYZxnVLqf8DpwAHARuAcwzCKt7VbbGveuFD0KnwuVfLeuvWGfIWXd6/vPIob1t/CID4Or/4nX2v8bup8yLLcde6Z5nCSG33dGhYsYmNz0/D8n6KR5Lj2bmoffb4m62mS+YuuYBjGMU7H4D6RRudvXr+eL+1/8ah7cm/kpdegHT3kK5feoEQJk32fjbNmwd/+pjE4CB7bx8Now3NJtMFuLHZJsu3I3+rtWmV3IKOj0uybZ6aZOYzzHNfDlP25kujM3LIl610mNHvXPlpaKuzfcTbGHMNR88Ky7alM8jqMvLc16utz6V1MdqdbktHMZDRJZyg5Kt4EaawUDVa90l+gQIpbw4JFNCxYRF1dFR0d8RcQjDV2kdpiXaD2/k1fZ8nm89GwOKv+ahbV/im9B1pWQa82psPs68W49cZRiW+4J1R0Q/NKqqvp37Ej7vGNrrsWPV+fPHlhwnNvohUiKVZKKc0wDGvo54SXIgzDcMnVhwKyxpQK1vM5jGj0+6m+3sLrtWPtp0Qvm32fj7NmQTis0d6u0diYj8/d6DFIvm97FjEds88kZc3tqmQ49aUF2T94bGN/MNdJb+Pt3BmZS5dozkw2ZgX7WLXGxh3mZMx7ZNQcxuzef8mGvMYWDDn44Fw+VifeR3La7yalVKlS6iql1LtKqR1Dty1USn0jf+HlWYrs3OzOfJExkVy8RWqLjWnCT9feyZLN51Oi9fLD2Z9LP1lysYHOHaPKwxdjItu/YweBkxfHvW+gu3t4+OfG5qakf99EKkRS5GI/IMLAwJh/0dsmkaGG95ivrxkzcm8gJ27Uj278eDz2rP00toraMFt6mEbmMEHuw/LCFXPH3ebduWKkHZHyIpj9CVP1ii8mvM+2HqY4VRnTfuyY3sey9r+P3iCDC4eV2x9LeJ/dBRqCgX7Wr9fzcl0zs+p2jDtGnr7W3INI8v6yq2BIssISU5afmdO+nZLpHKZ9gM8y8s5/Azjf7qAKxmVX+d1gY3MTS884leZjDs9LxbCiWXcoge3hWi5c8x9Who5gF+8Gbt7jOOZVLHc6LDHEW15Oy4MPxL9zMDw8NynZHCVPebkUfHCPvWN+ngPsNuZf9LZJxxqTMdnRQE74lRingWXL2k8JG2729jBB7sMV+2uPG3fblNfPiPktRcx5aG8kmwsSDFps2WJDJcNcpEx800+MKzub495e699C63p7ezNmNobp6tLYkYfruTUvHJ3R9skLoGR7Tic+Xn4/TJ1qQ+9xzGs/wMQYrZXJkLzFwFzDMLrVUFkcwzBalVIz8xOaKLRo7080ocmkYtjG5iaeu/sOQm1to+YxxQ6/S1WhzO1WdR/Mte//mgGrhAP8T/KtwDdcV79hsguHQknvj85NSjZHabB/gI3NTZI0uYBhGOtjfl7nZCyukaDhHQiYvPxyrkthxN+3PrA1zvNZPPNMfopM2GkkYcox1oQjUtIbkpeXvzXJ/KJoJcPWVp1585waHpVGEpl2uzz+hr8//wu8snMZ8LNMAktqZuMgXzn2DsrfeBWOvN62/WbDSjYSKus5TMmH2kcuvuT2fhnQI0U6jr+miYVfOJpvNxR/AalMPl37x26vlKoD8jDlrkA0TXqZYiRbhDZZ43FjcxNv/OwqrKEy7L3tbay86vJxpcKLOVl6ZPPZ3Lfp2wB8qu4mTqm7y+GIcuOrjixQWTRDIm0qkhGdm5Q0eR/qiZKEyX2UUicDRwO7ENOCMgzj844F5ZQxc5h2DQ7y0EO+vBQ38PSsxdNlMOhXw7cFAiZtbV76+yPrHWUjHI65Cl19EL7OVwC75vtE9uH3Q02NlXtp8RS9YZoDbQktRbUziPSszZtXqIhGSxZfRCaf6YmP79ySf9BvY8LU2GByx5fOgxB04GzClFx65/Q7g59id89fh3/3dK9Oun0gYLJunT0XQ1a27MNeLRo05Lg7F8jkiNwP/E4pNQdAKdUA/AL4cz4CKwhJlkbJdhFa49Ybh5OlYqd5R6/mbppww/s3c9+mi/FqA3x/1peKPlnC40VdcFFRlQzXS0rj3h445bRICfR09hFT7j5VeXGpkuc+SqkfAXcQ+d46ncjFuuOB7U7G5RaBgEk4rNHWlp8KVJ7QmlG/B4MmlqWxYUMOz2eOfAdv/+CTI7fbMIeptOOfwz/bccU8YRI3HGvy9kR5y29yev5M2fGBcJEAACAASURBVLt47Wi+rU+nt2Gq19GmaoimaW9brqHBjnXACiG9994Oc49Rv+uDyefnB4MW77+f2xyuaLJsWZpr1rTKVSZ/xfeB94D/AVOB1cAG4Md5iKsgUja08ry4nNskqgwWe3u8OU5F00uRBs070onaFa7m2+/8i1e6jmOqdxM3zV3I3pUvOhidPfa55Ac0LFhUVEmB2dsDmj7yntR1Aqecxp4XfTetv8NbXc1eF18yqmz6XhdfknBNLKmS50rnAAsMw7gI6B/6/8eJLKg+6YydwxQMRIbZ5NQ4iWkhde7z6zF3jp0zNVJNK3v5n8MENhVASFigwrnGtadnbcL77KtkGGGWTB/+uWrlV9N9VI73p/ksNr8E06YVR8JUsvXJ1BslkmLx2u5uje05XYqyhv6ruWhNq9yk/UlnGEa/YRgXGYbhB+qBqqHfcyjS76yUlbAmWQ9UvKvumtdHuCdE8zGH8+TJC3njZ1eNqqRWdOv0pGD29gCwJrQ3F67+D5sGdmWvihe4ee5x1PhSLExYZFIt4Oo6lomvqpoFTy1jwRPPs+dF340UJUlxYSNwymkc89C/xw2xa1iwiH0u/WHOCy+LgplqGMbKoZ/7lVI+wzBeJDJEbxJJNIcp0sizq3HSV3/a6Bu0sQmTacPz5fk7dmiOTzAYqeiX21d6qkV23dVeiFYytOvqfmyC7unbkN6DUg2jdmkPk6dIFlatXJNef4WmRY5Pvz6yTlj5ulsTbm/LxRArOiRWs72KoVMyKSs+POvTMIyOmLUxNuUjsEJoWLBo0vUiJRO96l5WPwM0DV/1FCzLipRfHvr/uKF3RbZOTzqat36Ky9f+mT6rnJN3uYPLZp+DV584awoYt97Ixuam4eSwmAx07qD5I4ex9IxTefPGayNFSRJ8KXurq9nnssvZ86LvJtzf2HO+rH7GqJ4o4SrvKKWiVfNWAucrpc4CtjkYk2PG9jDNbIh8FufWOIlpeOrJJybNnBnZNrcerZH3bjjmq8S3/YXs9xlDC0fWBQwGTUIhja1bc/i+T5BtDQ/Vc+EF1mDQntLiQKSHP2PJvzczmquWZFPThP7+xPdnqubFj9i3s1zZclpFdrKj9KjhW/yrf5BwazuGc0aH5DU0WGzcaFMvp3eqLfvJViZFH3xjb1BK+QDb184upMDJixOXIZ6Eoot6Aiw949QJNdwuHbe1XsNzOz6GhzD/F/wmB1U95XRIthvo3DGuIEex6W1vS/y+1XX2ufSHaSc9see8cLUfALVDP18C/AnwA19zLCJHxG9B1W6/h9ra7+RU3CCTBmxpKdTX5zo3aKRB3damDc8Lr1h7A917XJ7DfqMixyK2AEJtbbYtUAeq4KXDHAB9XPMMiPzdS5fa1UQbf15pAzvwdq1koObIBI8oTA8TwIYNGrNnuy9pzVVXF9Qk28Cy0rjwHzkum8s/Tl3PP4Zv1frasUrHDz+Pfb9kL7KPxkYL86Xx+6lrroYT/wfMSrmn97sPYuXq6Sw6ZFkO8eQuZcKklFpK5C8vU0o9M+buAPB8PgIThTeRSoBnqtcs54fv3kdr/x5Uebby4zmfZnqJDQvEieTyUanShmp6wh2UUt8wDOMXQ7++bRjGGoChoXjjVxKdTCyN7t0uofLdyNIPfuMSAoGL8zj8ZXyjJxDIcY5MzHu/pUXngOz3lGD/kW6r2Cvm++2X7eeDMwnT+p17E6x6I+H9VW9eyM69fxX3vmglw4EB8MXPqdKmh0dfPK1rrsbSy9HMHjYf04LlrR7/IBvXYUqmfsomnm/RmT3b/rlHoRBUVOS2j+5wLZXe7ApKDw6StFuirPX39Aa+kHQfGhaDps7WstEXBku2PsXA1MMwy0cnLbW1FuXluQ7nHEqYki081PoQTL8g5Z40LExLd7wXN52jcRdwN5EV1X8T8+8uIovWnpb4oe7X8vASp0NwhegaTLHzkyaL9b1z+cbbT9Lavwe7l63gF3sc445kSdcn/pDRPH0Arrr+GtsXXRaOuCrm51cdi8JNou8ZDUzftFF35V7cYPT7sb/mw0m3znnx2pgGdc5lv+Moa7sfsG++1WDJ+EJRev+m4fsz4d2+jJL2f6TczrQ8PLby41ha/OvbJZseTvjYYNDENHOsZDgcyPjp6po5NKzbTFQlt3A9TL72f6beKAVLG59VtrbmfgHiyfe/OOp3rT/95ClVolvS8UjKfWhEKlpaY5r81Su/TO2z+6L3bhwVk6bZ8Fky9Dk1I1k58bS//6PbOXsxNGUPk2EYvwNQSi0zDOOt/IdUYHI1Goi/BtNk8Oz2j3HHhqsx0VlQ80fObrja6ZCGVQRnYfb2TKrk1S7prB8misK7SqmfA28APqXUOfE2Mgzj7sKG5QYaA7XHjrolELB44gk9vVE6aRis3AO2RQeWjG/cBAImjz7qxTQTFpxMYeT7t6VFj9TfHVL+3o30zLkom53G7D7SyK+pgYoKK/dJ7MkOaoYXf2peWghAx4LFqZ8ajYSLCoc7Ez4udvL+rFm59b507/Fj/G9/P6PHaDYmRJ21Z1K5M3FVuF36HwNy+7zXrPGJ3/r1GnvsEWfjNPSFyyj19vLb/17Ax3YbWc+pbMMf6Zl9YbZhjhIv5vEsLEsDLX5XVe1ShYXO5gUjZfEivce5zGGKDsnLehejWJbz66ZmcjS+ppQ6IvYGpdQRSqmbbI6psCb6Ffw44pUGL6YS03a5Z+MP+NWGawCLrzd+x1XJEkBo3XsTOlnSvD681XGGcdhkMp7TE9AZwBTgTCLzaM+K8+9zjkXniJFGw2DlPMKVew7/Hgya9PRobNmS7ffa6AaJpcesfxan8RsIWPT3a2zalOXzDe3zC7f/dtzVbP+aH6W9mx0D8Vtl/jVXQKgVTYv2huX6fa/RN/2UBPflqzGXfVGJaM+aHb13pm8aOz9wY6aPyvH+mC09/qT3d3XnZyjq7ju+T0Wa1ejGWr9jHg++cjJvrJk+6nb/6h+g96Y7iiX5a1ey5Ym09mIxvodp9LOMfi3s6q0uKYW6OnsS59gYu+dcbMs+M5FJ0YczgbERvgIsAb5pW0QFppeWFWW1sGxFh95Fe5N629tYdf01eKuqItXwJoF+s4Qfr/0D7/XuQ7neyRWzP8PMsvecDmtS8VVPQV0QuXocez5GaT4f1kBuiyHLWkrFzzCMt4FzAZRS/zEM46MOh+QaljXUmIkZqhU7WXuXXTJvYI9rk+sxJffjrDc0MjdIY8aMbBKGyGNMS6fFhopccS0JwILOnK+YR2Pt3O8PkQnrCe7PlP/Ni+jaM3kiMvxaZyhaydCueW29gXOoeiv9Xj8rRZKnhTuxSurS3VnSu7u6ck8KwxW74w29M+q2g/03w3sQmvvDjPena4MMmh7Wt4xvaut9GzHLkk3wsY821MNkZVCjLRi02LJFp7sbKiuzedboOkw6wWCikvzp9XpGy6LHXrTpmfWNbILKSSbvIivO9p4M9+E6k20YWryhd2ZfL9bAQPGty5OF9v4A33j7Cd7r3Ydg6dv8Yt4xkiwVSLTM94KnlvGRhx4brk43tqx34JTT0PTcKjvJWkoTjyRLEeMr2Y38Hgya6NogXS2JiwQkN7aHaaS0uK/z5XFb57xey1ADyDT1nK5mRxOmcPluCbeJXDHPNWFKHGPJtqVZ7bW85TdJ78+o9PYYI5UMs+0BjH3uzMd5poq99rlMynykOA4DXRnsqzA8msmg6aE7ND5h8huXFjCSoSF5GSRM0d7JbOdwjVzE0IYvrMSLK10WGqFZX4+5pfCpRybPuBS4UimlAwz9//Kh24tWqqvQE23ieKJhSoM9PTQuOjHbgehF4aXOY7l4zSN0m1M5aso/uGb3xZTpkythTipfw1N1nX0uuzzu4rEQKet91F+WsODJ/3LUX5awednz6V/I0HUCp5zGPpddLmspiUllZB2mkUZHIGBy+ScuZ3HpB/HszCZpGtOAiem9qlg7fvR9ruu1aDE9TLlMrte0oQZZgsIIELlivm2bRleW7erwgEVff+LPSL9xCd7O17LbeQoWWuoS3QkEArlUOxt5zoGpRyTZLpHCzTk5df8/uW5KujaUMMXj2/FC3EIaeWENVZnL4Ds+99Li0ddei/Tubg3EjSsd0c+J0NwfYulDJQuzWhcsN5k84/8BxwEblVIvAhuABUDqmoAuluoq9Jq7bi9QJIWRLEHcvOx59rk0827nYnBf+7e4qeUWLHS+NONyzpuZeNG2SSvBpObAKaehlWXZ+6jpeP1+Vl59xfB8uVQymntkmmxoehSAo/6yhM+sXMlRf1kiyZKYuJI0Mmq9b3LM3pEiDXrfhqyf4ultl0eeKsEk8Si/H6ZOzaG0+FAPU1WVRU9P9hdsdFIP7RmplJddQ+ullzx0dCTvCdP6N6e1L9NTlfbz5tLDBNHFa3NvXJoVc3LeR07SOAxZz6UbYpYmK+mWOV0bTNqwr/tPHb4tTyXdxybvwpzjiFbJy+SaaO6L144MyQsETE75ebyK1JksXDz6AlGy+Vj5kvYzGobRAhwInApcN/T/g4ZuL1qpGlYTbeJ4sgSxd1M7DQsW8f/ZO/M4N8r6j39mcmyy2fs+EnpQOi2UUg6hIBWKVCtWqMglihdV61EukasKBQERyvET1KKIiieCyCUUltNCqQiF1tIy7dJr2z2avc9sjnl+f0wmmWTuyeTY3Xm/XoXN5JnneTKZJM/3+X6/n6//nAmtFJ9ClKPxkz2/x7M9l6KIGsMtMy7CGVWP53taBQvlcPJCDBQFZ1kZXGXlOPDUEyAhM544CqBpPjcuLlWvR+7baO6RoIhnYzO1kHqYqjedhFNnp5dLNNBj3BhrG497E9IMJrncnUxC3YQcl6rqzBY/CQ+TCplKi4+N8XPdvVstcd6oTLIOKGI6hwngX/fBg1R+vC+Wqppp92VW3OLAyDEAgJHZ1go/Oegoijz8Z+i1HvlNWlef+ueVA68r/siGS5QbxbR+n4nII62PhgYCpzPzzRCK4kPyNu89XrGNjs6kh/Ig2GboW4plWY5l2bdYln2MZdlNLMsWmAPUeiZb4njjkqWKymST7bX2hOtw2a5X8OHYx9Dg3ov7j1iMmd7t+Z5WQUOiEUSHhuA/+/PgxsOIDA5onyQDr4BXCsSiKcf1GDezVqyUzadTU9SbbBsbNvIwDFPJMMxXGIa5Pv7/Ku2zJhtpiweFRWlmngmpoIQSmalp8UuIygzfRT3hakLiudkd8xIfb7iYfa1R35zE30bfGwIKY4FvmhpXUDIMBk3MW+beipRJ846oaL/kWLwD42NmQCaetKc3n41o2QILZwM46QgohwslJQQDg5l5ar667hHFFmVbNQrXmjC6HQ6gqcl8OCeV8ARRifC+p0Kvp7UylsOUb1SvBMMwO0R/tzEMs1/uX/anmT8mWuK4nGR4OnNWXSW7IK1ZeAp23HsnDjz1RC6mmlX+N7QQV7W+gIFYDT5W+gLumvlZ+JyFlxRakBCCA089kZEgCiFEUXVRy7iRE4IQcqA89dLCkQDgKs2ePLlNYcAwzMkAPgKwEsB8AN8G0Bo/PuVILoCs3LdMz2GSWSJw4ZSHgQC/qDLjTBDq9FRVKSxFdO5A6wnJq68ncLnM75j7fHxYk+oCUu9FMFCfSFh4Dh9xq+5zxIiVDI0jYzCVnyQ5VqRYPDe3Hqbhzr2mek6osFmMg4oixrkQCHBwRjqyMgYAFHU/r9HCuIcJyHQzRLimdOIe3LzvxLQm+j4HqRsMydyoXKO1fSTe0phitS54CjUXoqNlPVofWofQoS546uoThp2cZDiQ+joalyxF/7atEsOoff1zk0I18J/Bb+Px4CoABBfX34nPVivvzNhkiViUFxCRiQPR48kUFPTSmbViJT742W0g0VTJ8cjICDpa1qP24vPNz9mm0LkPwHdZlv2bcIBhmAsB/BzAx/I2qwJlbMyIiLAAEf031SsiUMJeh+G59yQe+/0cRkYo9PfzBWKNwX8/FBcDJSXSRSsVHQBxaXeqJySPpvkdc7NeCJeLgMQo/vzjlFqpLbzlFn3xuY21gfMGlM8kFODw6p1qCmIlwxNOMG9cJ1JcZfPalBav/Nh/7HwZlzRkX+RyWfU3AWgZD/JkEvaohIOOIEacvHczMmSqDz3eSEK5tFqYen1+P8Ebb5hVrOXiI1MoLQXKy+OfPSZ1XnoR5k8cXlBcCAVnMLEs+4bo73Rf2qTBU98gWyA0m0U1M0GplhJd5JaVDG99aJ1k8dm9aaOk34luLHEcsLZtHbaMnAoXNY7rDluBOb7sqBYVIp76Bix69Em0LD457xWxAQAcx9c5E91Xmcp9Ny5Zig/vvwfRwbQaTbEoWh9ah/m2wTSZmQ3g72nHHgeQcQIbwzDTwdcUFKgAUMaybFVauzUAvgte9AgA3mRZVqx1mwOUZcXFBIM0GpTX4KpQ8fyASOUiyXPOoS0pj4VQtwMHaFRWGl2Qi2u1yJzL6avFRok8TOHKRYoS33zxWrMhRsgoJC+FtJ318s1no+/j8r9VVog+AGZDEZNjd3dTqK0lCmGaSnV2rEzQ174OsyvfRDBLfZvBQUcR5Zzw+zmMjpq/bziOQk2NymdLw2CiEFfJM4jfz6Gz04lIBHBp2WQyYwp/CX1JNyuM5/z1n/Ai3MHnTG8gZIKqwcQwjK7yxizLTmhpNX7X+laQaGq+RWx0DB0t6wvOy6RUS0nJ4JELgZpsOR+D0Qr8aPdj6Ik2ocrZgVtnno9yZ1++p5VTBEPEU1cvuwGQazz1DZi1YqXEE2r085TuTTUb6mcz4dkF4CIAfxEdOx98mF5GsCy7F0AigYFhmPug/Pv4CMuyuS8zn4YQYhOpOh3OkZ2S50t6HodzsA7RsmMM9ZoC7cLotMtRvO//FM8QL8iPPtqgwUSEXWg64QkRQ3HjupZUYg/TwPHPoPalipTn6VAHOGcp/H4PXnsts1wSdQ+VPg8TRVLDGh1jezVGNr/YzljJEMCz730W7hplg8nT8SjGpl8hc6a2wURFh0CcOlQDs7wJmI0cGQoEILyxcD7zV1N9CMIoisVfoa1mibhKnlECAQKOo9DeTmHaNKPXX/zeEwQCHPbtS7sPtv8MbtcChOs+q91bfP6xEgZjJYxG6+ygFZIn3p/yAPgCgP8C2AfgMAAnAvhHdqaWOxqXLMXOX9yHcH9q4iKJRmS9M/nG6MJQLgSqUBbVVsCOHIOf7n8YEVKE+b438MPAtydzOSlF2Pvvxbbbb4bDrAS4lTicCeMok8+PnDdVickmWmIj4QoAzzIMcxn436DpAI4AsMzKQRiGcQP4EoBPW9mvdaQuXIZn3wFv268lrWbgr8B//orgEvkNBtURVBdXqeNnVq8lWdxSULFLHSosPSZDquiDdB7VG/gF1vRAP7q6yhAOA263pJkGmYk+qEERlRwsmfwazlUFOtKru3/zSob82Bt3noJZc2kceywHzlUhaeUcVqr5FZ+7irR26f9WYPDYR03MTWFE4/V1swoBpWrsaEHF849kPx+JRurvLWU6JC8pxT9tmnaeYMqYJPnZBgj8foI335TOs3zLFzW/o7KVY2YU1avMsuzXhX/gX/UXWZb9OMuyF7Mseyr43b5JQXhAXg2sEHetlRaGrrJyiZiDUghUzUIzRegKj+d7LsFP9v0JEVKEL9T8AtdOm5rGEgBe0Y4QxMbG8j0VuHw+SzYa5LypcmQa6mdT+LAsuxHA4QAeAPAugPsBzIoft5KzARxkWXazwvMXMQyzlWGYF/MrOBFfANHaSnb60SHfK9rpr20pw4z9F6K42KSalihky+8n+PeHqSGAFKfPYKLFCyqV1fJFzLUghMLBg8YXjxTFL1xjsSytxhWkoXkvReqYQ0f+0lDXppP3E+8PlRCNCPm/Yfx8lXAw57Be5Vp9i+Z+JcE+FTINe9TqV9XY0QEhlKwHVjySRg+mPGiZCIYQmZC8oSHzn51CUMkz8k37GfC7bmKeBvA766aTPyiHQxKSBwC002DgZg6QS3ynnC4wq64EAMUQqERo0yTwLHEccP/Be/D20KfgRAQ/CHwP80utXjdNXmiPFwTEZH0lbSJDxne15VDbsPDUN2QU6mczsWAY5ucsy14G4E9px+9jWVYuHsgs3wDwsMJz6wDcxrJshGGYJQCeYhhmLsuyPXo7r64uyWhy1VU+AIDT5UBtrXYok542AgM9fF6At9iTPO9g6iacy0mn9FnU/RymTaMQDLpRW2vMbTMe4ccr8hThqBlF+PyKf6LnwZrE81UVbqBS//wdaXNLp7miDQAwPFyC2lpDU0W305HYpR/2nYGSkVckbcrLPIDS+A71BV9tx1rg2J9JjndQFEClvq7y8tT8Da33ePZs4M03gZqaUt3el9raUiDGr38IodDT40FtrQeIFSm3T2OgnW9b7FPON3E41N8zgcEO+XHTGRkpxezZupom6KApgFCq86iuLtXcjE0/f4wioCgHFizwAdLbBQDg27MWvpN+qrjxMeZxAiFg7lzlzxZNq1/DAbcDZJRCVZUPaFOZf3kUcCdFVkrjXfb1eQ1/XsIeNzAGVFQUo7YWOOoolXE13v+u+E1r5LssGxgxmFoBfA+8IpHAd2BB/Hi+6WhZL2ssAQAX0bfDlWtIWjyv8FgpBCo9tGkiMxotxo/3/h2d4Rkod3TjlhkXoMZdeJ5Ao9BeL7gceYe48Lisgp1VWBUepxQ6Kghc2EwpvgbgMpnjl4AP18sYhmGaAZwW71MCy7Kdor9bGIZpAzAPgG5RpJ6eYXCcuR3t2tpS9PQMoxpAJMwhGOSVt9TWMkIbPYQGx1AOYHQ0nDjPNxpBsahNJBJFf9q4jY1RfPQRhWBwVPdYAMD1DaMewFgoivqyEcS41FyMwbbNGI8ertmPMI9YlENfcEjxejgd/O/8tm1jOPpo+d98JaLRaEJ18KnhZ/DZ419GxeazU9oMDIwhXCR/vSujnOqCa2ygA8My7xUhHAghCIpe18DAKMpFbbTe4+pqF4aGPGhtHUKFNKJOQm1tKd9nbCwx5q5dEQSDIYCLyl5fuTmEh/j7aXgkCsTXuoeGm1BX0p5oE+MIenXco+GhEPTIcG3dOga/39h7S+K/hUGVe2f79mHU1yt/bhPXTEQJ+NdHUeqvr2/3vxEtlxf6DIUiIIRCRYXyZ4sjQI/KNYyE+T56e0dU5zH6n+sxMueulGN1dT6wbBTB4LjqudJ58+0HBkYRDMZQVkYD8Mm21bp/OcLpaqcXmqZMbVwZMZhWAPgnwzDXADgIoBlAFMC5hkctMLQKaRYarQ+tkxQEFVTClHbZ9YY2FTp7xubgJ3sfwTjxYbb3Xaye9jU46clRP9lZVIRwrsLpslz23arwuFkrVkoMfTv8bmrBMIwQA+QU/S0wE0C3hcN9FcC/lDxGDMM0syx7MP73AvB5VKyF4+sgi/H8Mon1ROKSkLbx+zls2WIiNFCU5xAIEERjqX2UbbsUwUbrlC89RRwoylz4oDgPpK2NRuRMmWhMnTlXRpHmnxi7B4RwrrY2GhUVxr/7U+pPaQoMpJwZ/2/yejtMh8zzff3t0LM45/in4G37jWwrszlmWiFf5KM/A/UXG+4VgI40Ae2cQZ8PgOISTuM1k2RIXv+xT6DiPfllu1xoolBnzQwcR0HYZVAPKdQmG7LvRtH9Dcey7HsMwxwBYCGAJgAdAN5iWVaf7mcBM9FC1JTClNTClybaa5Tjtb7P46GOm0FA46yqh/GlhrvzPSVLSRcdmaj4zznXsvA4oZ9MlfZsJjSCt8eNVM8PAdAF3sixiq8hzYvFMMxzAG5kWfYdALczDHM8gBiAMIBLxF6nXJLdmH79og8Av6jq6aExMhJf2OkmmeNSW0tAOcyFwHOEBk1x4Ny8fyDmnS6rPEfTBPX1ZmsxJRedBw7QAC0NESveczfC9csVzzeDFfk1Qi7KgQNGlQyTYyeumQFFBSKTw+R0mn09/HkD0cMwXneOosFk1hjW4uiRlQjCmMEk5L3pmYEa2n2oz5+iOHAcDYoCIjVnGhonEOCwZYuZWkypc6qpIfB68/cZsALT2aIsy/6bYRgfwzBulmXV/XyFjkKBTQCoPO6EHE9GSkfLerz58IMY7eyEp64eztJSWWllpTCojpb12Z5i1vl1+y14vf9c0IjhMv8VOLHspXxPaepCUaoSr3OvvMbS4TJV2rOZ2LAsuxgAGIa5lWXZH2V5LEn2A8uyZ4n+ttI4Mwn/2cuOEpiMtHfaZ92VVocJSCa1HzxIY/Zs/QtySiQrTtNAbb0b6969HyuPX2Vk0ugNz0BN0UcYmC8UKVcupOr3m5PYFjxMdXXKAgr0uPWh4RSVuXGc9DAZ7Scp+jAwQGFoKJnXogcKMh4mR/o9ZnRO6u2j/bvBB0AZI2sejIz7tcJY0Ge4OUekznK/n8NzzznBcXo8ZemjJsekKKC5mcMzO7+Hz83+hbGOkO0NIn3ofvkMwxwNYCeA3wD4bfzwaVBOjp04qIQnnXDPAzmcSCodLevx0tLTse22NRjt6AAIQairE7HRMcCRautSTheiY6NoWXwyNly4PGEk7bj3Tmy7bU0eZm8NIc6D6z/6B17v/wJ89ADuPHyZbSzlEVd5OebdcBM89Q2yzysdt7HJlGwbSxMNMws8H3sdalvUMkGSC+Qk2gaQeWnxpMHE98Phode/Y7APnj+/eTGIWyMznZCMitcKEtGKHqqMrFjlc+VC8qK+uZo9+nbdjNqWMtSUdsPrNR9aJQ5FNHhm/L/ikLy0/GvdC2F9hoM3LK1JZqTv/uOfMXG+PPo9I8rXQDDUKQq46/2XFVppeJh0yoq7e1+THPP7CcJhCsGgOWM7va+hocwKKOcTIzP/FfiwhDkAhDC81wGcavmscozSIs9ZpifFMDsIho6cihmJRgAuxs+PonjFs2iE9zrFjaptt63Bm1/9Ig489UQetrztegAAIABJREFUZm8NHePTsGrnq9g/PgfTPdvwwOzFaCxSkXjJMw5v7itP55rIwABaH1qHmoWn6Jawt7GxsRLp4mH4iJ/oOrN4v1456uTiKNRwgfTp6HDKQ3HxWkMIHqb4Yi4Q4LB/vxmjI/WacCqGk9/Pob2dMpHGmZSIVvLURKo+oXuORsdNPUQwLir2Wfnm8XAM75A08+7nN3yr35xvTlo8zbuodn7NS1VwDryjcL6ah8nglDQMrO7uzHKYIgriC2agRJ6dQ1EjBaRTET4f7kq/UgOtHkx7aMxLi0uNtEDAvLR4IeQwGfl2OwpJOVcCAPFQvAm/Spy1YiUopzQ6MTY6lpdwto6W9dqGDiHgxsPwn/15cCF5oYDRfXuyMLvcsHFgKa756GmMcqU4o+JR3DbzQrjpwlQsFIhlSaK70Ah1daJ9/XNoWnoWv9lAUfDUN+DIq6+zQ+dsbHKEeIlEHCqyzUPbTPbKEys9CqPTUkPkal9tSnlcX0/gdBoPdSNpHohAgCAYzMwLAkClkCcfkheJUOjqMjZXYZfe7yc4dIiG3Nd9zMconj84kHkNmvG6z8k+7xzdheI9axXPp2ND8VBEc9dWCMNTM4gpEkXl22ekHZV6mGgHwfJ7/pk8T7chqa/d4BCfS2caytpSMsJ9GSs6TGVMfTmDDQ2ijYzGLxqZhWmDI+k9NhyPl/CMifsaHctOjlkuMDLzvQCOFx9gGOZE8HLjE5rGJUvhKpFKDJJoJKcKeh0t67HhwuW6Q+i48RAOPDP5pJUf6bgOvzi4FgCFlU3X49KmW/I9JX1o7vJMHrjxELo3bcSiR5/EklffwqJHn7SNJRubXGDwe4Yel9Gk0OgjfTc6Vqxe2MbhAJqajId8iXOYgMwLfGpDMt4xF+bY3i5zPqcsZ93ba27BKr9YlDsm07/ofTbjYRLG9pUQFBWZMbhkDCYKeO/g6Slt3F1PAySmPpfEa1F/DfP823DwoMH7UFz4mFYxmIjB+5MiCYPBo+pa0Cf60NiUbDc070GMNX1F3zQyMJhMe49l8Ps5cCpFjNWYUDlMAH4M4F8Mw9wMwM0wzPUAHgMwKeLKwwMDssfVlOeMIhhE6XlGwnPb195hXM0uy/LQuSTKOXHT7j/jhb5L4KFHcPvMc7Gowrp44oInO1ncWcPKz4aNjR4YhpnDMMyPGYb5hejx/HzPKz/o/L6Q9bYo/W4oGFI6vpsCARMhX0IOE0l6mMwgMSqUDEJCzO+YQ8hhUl5AOgf+q3iuy6Xx2tSucXyxGyk7DgDAeeRDs9zB50FFRbVqSFLEWKxkqJ+4wAgoNDcb9yAmRR9E5xGCBpGD0jG2F+Vbvwzv3vt0zEQ7NOtnX7zOlKiHHoPCs/lSQ32KDRWPx+zve/K+aWxSaaaBWUOlpASoqDAnlJJu5AQCZg23wtiM1n0FWZZ9FsBS8DXiXgcwDcC5LMu+mKW55RSnQv6Jw+ORPa5m/Ci1TxhE8Tyj7WvvSJw3WeokmSUYbsT3d72K1tACNLlb8YvZpyPgmfDOS2Pk00NFUZi3eg3mrV4DyqkvJMGq4rQ2NnpgGOZ8AP8GL4ElyIuXALgnb5PKC0a/J9Q9DyktFXfxtRc5pkK+RHWY+D7MbwASUDrsOpIYw+hchcW/YNQdOECj59QPUtoU9bQoni8T9W9oXAAYm34lehduRLT8eEk7x8hOlL9/IUp2XC577kw/X1rMqPcF4K8t76HK3MMEEDQ1Se+/ktabdfWVTu/CNyXHzIpTaFHa+w+D/SaNBnWFOfWQPMHIqCznDeC+8XQVQC3RBy6jHCAz772cZ9TvNz+PCZPDxDCMg2GYPwDYzrLsd1mW/SzLsitZln03y/PLGdFR+SrKsbExdW+QjPEjh5xBxI2HEiF/U3m3fvPQJ/CD1ucxFKvEKWXP4K5Z58BD56iAa4ZQHg9olzvf08gcQhL34lHXrtbcUbYFHmzywC0AlrAsuxJ8HSQA2ALAfDb1BEbvAsKzT07CVyMkL61vKjYsaVOxKVXgwO/n0NlJIWKoMmPqgrqxkYCms7dx5BjZhRLvCCoriemQvKYmDjTNn895A7rPNl1/SFzLh6IRK53HzyZN2IKO8AaRI3RQtptLy/lFtiFPgciw5tUFDeaoJdxCqUvN5mYTi18iyOknzw1XnopY6dEYnrUmpakV3hArSAn1U1NBVPUuJj+PghHSMZieD5U90QfAXDgnIP0eaWggpgJpJlQOE8uyMQCfgh590UmIHm+Q2PiR7UOj2OxU3a1/7ND3cXfbL8HBga813Irv+a/L95SMEYmAixS2GIVehPscAObdcJNEBU/AFniwyRN1ALbG/yai/xfGr2nOMPZy5b0e8n0QhePurqckx1xD76c8DgQ4cBwln9ujgODREgwml4s3moyj7xxHuBPl719k2ltCCAWnk1/4GT0/XU7b0MgyxvFY4FspjylunG8rU1BXjJlcFC2xC5Uz4/+lUo41NWdwLVL6ShqSYpwytcLUyNaCPDV3SO1zoS36QFEA5w1g7YYH8b2/pXq66Ki0LqfyPDSISTerAwE+P9FYEIy0sdMJPLX9m0Y6EfU2QTxMce4FcDPDMNZKiEwQ9HiD1LxEigYRRaHl9IVTzsMU5WjcvvchPNm9Em4qhDXTL8aSqr/le1qGITH1RNWJhnCfNy5ZiiOvvi5FBW/e6jW4+IMPbIEHm3zxLpKheAIXAXg7D3PJO3oXEETuZ14zed344sRcbpAg+pAcz2xYnt4Fobv3dfM75qLQQaXzHUP/Q9Xrs+Hq22i4fzkUF/OUA5yrSjS51PBGOYwrGaaKRgC82EW46nRdZyfnnhyTUG40mfEwqYpfpN5zAdfrxnvPUsiXvs+pPtEHAPjf6NewdafRDXb9BhPFSS1iv5/D8DAFhVR/5VFlxox5Z+I/+89Ma6jxmacKY0/MSFTtKgANAK5iGCYI/k6lABCWZVX0EicPYm+QrDgDIWg5fSH/N02jcsFxGDt4AKFDXXCVlvHFZmNpKjqCaMMUUljrj1bjR7sfQ1+0HrWuNtw64wKUONV3SGxyR6irEx0t69G4ZKltGNkUEpcBeJFhmEsB+BiGeQHAbPDRD1MGiiR3nPUgv1BSFkVQGFVzHGFBbSRsiwgeJlHIlmB4pRAdBhw+xRdt1EMQCBC8/jq/Y673Ooqb+f0E77zjkG1XtenjAIDij27DwAn/MjQvJfQsuhNhkyovyLiSYdLgEXK32tpozDztadS0VIDSCjoS7ieKAueuBR0OgjiK+fe428A0oPEep3mYBgxLuGdr/ZXsVzXsToX01x0IEPT28uIdUm1n5XlkFpKXfO8rKvRuaMhfU7+fQJqur339J0wOU5wvAzgTwKfjf18i+v/UgBBsuHC5bNFOCRyHvs3vJPKcIoMDQCwKyuOZcGpoVvLByIm4fFcL+qJ1OK7kFdxz+FLbWCpAtt1+C149+1O6RU1sbLINy7IfApgD4Bfg1Vl/B+BolmV35XVieULvAoLIqmMpLVD0STfL0dxs3sMkXvAGAhwef/u8xGMq3IPaV5tQvOcu1Z6MLKj8fg6joxT6+sDvbuuRi6aSu/SBAF/8Vj3AIO0aa+6SK81f7bzkdRM8A1RUWQYvE89auliGprEUPzM+O/R8ohWj06/C4LGPo9mfeU0qoV/+f6nG6+AQUP1yAyr+c7r+fnXeP66e10BphMAlZkcJfgVA7TPl6pMKVyjNzZxoCQHH0Yml53iN2kao9H4zJS1O5IVYAgEOofG0fiaIw8DIFX8LwCcBPATgufj/zwTwnyzMq2CRFO00CAmFMO+Gm7Iws8Lnme6v4/Z9v0WUuHBh7b34wWGrNJRjbPIG4RAdHEyImmy7bQ1aTl+IJ8880zaebPIGy7KjLMv+nWXZu1iW/RvLslI1gkmPscVFjMh5QozVYdJDURFQX29sQZ6swyReEBL8fdP5iceO0AG+/y6NYu4GEIcPVr8+A1X/Vi44m5irKKzJ7yeIRil0dlq5+an8nigu5mU2X10Dyksy40qGIknruCCHMeEHUQ4TRWHkiDWI+WahutpAFzJzAcUHRxHaE38m9TURQoHmRuEa3GxmIFUqNp+N6tem626vxxArZX+o1kPKo+S9a+BzBi7lMzY8526V4eTU7YyPqeZhSh+CDqunpBSK6IORkLxfAWDAh0XsAy8rfgN4iddvWD+13OLwehEb06fMJi7a2bL4ZMPWsd7CtJMFjgPuPXA/Ng8vhpMK45rDVuIo35RMO5jwjHZ0JIQh7HA9m1zCMMxhAG4CcCzSolFYllWvrDop0bd44ThpO4pwCqVPtTxP6vj9xovX8t2n7qCXH52sVlL5n0UAAOfwdrUODA0n3jGn3X16J5lYdIrPn3nUgyj74NsyzQ2Kc3T9E8NH3i85TlmYv8ErGToRifACG3ohhEoIcpgzuNJUF2Vrg+mFQqRiIUZmXosx/zeFDlNaFDnHDfZo7BpTRLlAsRjakvcuNZwuxdujc89eKvpgbF7V1QRer/HPtpyx6PdzQJp9VL1hLnoXvoVY6VEqneU/MsuIwbQcwOEsy/bHH29nGOY/AFoxwQ2mjpb1uo0lgVBXJ28sUdSEcSfmg+FoGX605+8IRgKodHbh1pnno8LZk+9pTS1o2tICx2JhCBubHPIYgA8B3AhgYtQdKAA4zkhInoC5xUkgwGHLFvncHv19EKiZRkoY8Yql7JjPNDAG4UOMxOdHFpym61ytRTkdHYRjaCucQ//DeNOXUs5Tfm3G3qejpu0Bx81BezuFadN0rFtIunfDaEifnEoeTKUlEPFcKBqjh69OPOSKGlPa+qsOGO8/SypsREdInkYHKYZHfT1Jind8TNxOLSFPfw4THemDY+BthGs/kzjG3/Pm3vt0AgEO/e9I+3GM7lIxmApjjW3EYOoEUAygX3TMC6DD0hnlATU5cFUIsY0lFVpHj8Zt+36HMPHiyOJNuPawb8JJT0ll+rzhKisHs+pKbF97h6WFkaeaqqNNQTAHwMksy9pfIkhd4BGHcvp3jKOkaynF3y0hSd/cnPx+Ds895wTHaRXqTBtPRHMzZ8pgMkJVFUFxcdxbImMwUZEBlG77FoaO/DlIEa9IJjZ4mpuTeSScp8n0PAjlTPFWVG06FQAQFBlM8Yby5xt8o2Y0HAQwBwcO0Jg2Tb/Ca1IdkOC//9VvECeumQXeATnFPYFw7bKUx8aNn2yKVgt5VmZHSBV5cTh48Q6Jp4/EEqGKcn2kenuU51K18TgAQM8ndoIrSrqwzBSmlvMwNTeTFCNCd18TTFb8jwDWMwzzTYZhPsMwzLfA5zI9wjDMGcK/7Ewzu9iLP+tp6b0Ia/b+BWHiwTk1D2L19EttY8ki/Oeci3mr16RIfvvPOReUUxpjERnhE4DTJcL955yrLVyiwlStG2aTV54BoG87f1IjNTLGGy9UbE2BoLc3fbGRHQ+T308QDlMIBo2dL14MFRcDXq+x34p07024apF6+/iOuVI+jqfjzyjqfh7Fe8S5HslFp88HVFerF3J1978JKqKlw6xwfooIhdp7Zew6N9TyjlnjRXuNiF2IIAoepgyQ7SvNGPF6jW9kyy3uh2f/1HA/og5THzpKM+gqdW58EeF0g0k5TNBQHSaBNHlxqzxMHg/gdssHBCsxEXOYhCDdG9KOr4z/A/grZMDBnQrDMHsBhOL/AOBalmVfMNufXhRlwm1M8YsDd2Dj4DI4EMUVgctwXOm/8z2lSUX3po2omDcfix59EgAfUtr60DqQaETaOBbFh/ffg8VPvygJoauYNx+tD61D6FAXPHX1iI6N8kIPGtBFHsxasVKznY2NxVwGYCPDMB8hLQqeZdkJHRZuhpQFEKW861/qHcaBAxSqq8WLDnWDJH1RqnfBEghwWLH4N4h85Abqz9V1jhwlPuOba+LrMXr4aoSaLgFxlaNGJkGfivTB7/ca2jFPD43Ts+NesuNyDM3/vTBD2V7lSVtsK7UzmAtUV8avc/S/bqnggCB2oS99RvCOSOf/4MAAvl1ernMeor50tCzRr7fN96mUa5SRorEgmc/3MTL7J+A8zfB9dKupfsT4/QT//nfqe0iRCAiUNkGNy4pTXOp6IhAg6Onh5cx9Ph3nq4xZ5DFQ6iDxbP49TLoNJpZlZ2RzIiLOY1l2W47GAgDMWrESO+6+AzFjJaxt0ghxXvx499/QHp6FUkcvfjLjAtS6J3zEZsER6upMCC8A0Ay3iw4OJuoqiUmvs9Sy+GTFPjz1DQh1dYKi6ZQiznYek00O+R2AGIAdmMo5TCbCwNvaaBxzjA4jJMMQc7+f4DcrvgWMAEFoG0wkUVMqdVwzBlMKlANcsfKSpXj3nfD770F99HmliSkcThWn2LlT3fCgw4c05knLrxMJlzCC1Y1VY4vIyj03oK7uG4alxcUFewH+flqg80zx+WKa/E5gyNA04mjP/YIFvzLcq+yC3JKUC75f4izD6MxrTBhM0rkJ4h0pqclcWGUGMcQ48aaK9usq+fAqDBz/TMqYAHDwII3Zs/V9PpW8Wh7DwS0Tz8M0aWlcshSh1h1offTRfE9lwtIWmoWb9/4JY1wpZnm24MfTvwInrU9JxsY43HgI2356C5wlJbpyk/SINCh5Wj31DZi1YmWKYSY22myjySZHnAGgiWVZU8usyYf+Ra9kgay5EDQv+oB4vr1jeAdiJXN1npk6ns+owWRYjYwgECC49dPLDZ0jxu8neOUVvvhtzN0AR9jKKJXkWBRlIpxKhUDAiNqZtGgqEL+fFJxDtS1lCH6yB6BdovOl8w8EOKQnq5W9dwEGF/xV1WOaDU9DdkK+DPapINog92oDAQ6EUBgdpSA4eygupDgijRiiMWPLfXfv6ymPxUIns3Vpkiq/fo/HmpDJXFOIVXD+zDDMVoZhfskwTEWuBm3/tx02ZpYN/Z/D9bufwBhXgk9X/hE3z7zYNpZyAcfpCqED9OXpzVqxUpLXJITftT60TmKYiT1NNjY5YCsAUxVcJiUG1g9n+r4JOtSeeFy27VLZdiRDWXFxKFTVWyfpOEO+X8MGEwwuqAiX2DE3cJJE3nlsjEJPD4W+k99SmlTiT/lFuZ4cJrWQPPnjVLhbvn28AK3ukDxh/vFrKxa7UMPb9qDQQfy/0nk2NEivR1H3etDjCr9VWRbYkr1/1ELyovpKwOk38JTuR6nBLBgvIyOi46o5TFFEucz8I2aK1yq9do9sSJ4yVkrrZ0KheZgWsSzbxjBMEYD7ADwA4Mt6T66uNhi4KmK0085hMsNv22/EK/0XgEYM322+GqeU20VNC5HihgbU1qonndZefD5Ky7zYct99GO3sRHFDA4654grMWLYM226/Wfac0KEuzX5zTaHNJ51Cnx9QsHN8BcCLDMP8DtIcpofzM6UcE+6HY2yv4dNObvgTwh8cTDx297yifkKOd3PTF1a+YgsXSEdeB2y/I+UQBY5fdGrtN4kWzELivHBIvONes0DJjtd4HYo5SJmJPtS8PhPBJdIXR4cP4bBAGM8/X2xAyTD5/hQXAzU1cbELlZI5JTtvQKj5K6CIsofJYVh9XrmvbMG5lPdnivf/CqMzVQrOGjXwRGGY0q6kIXkAMDICCGlLvtZb4Ol8DAPHPo5wzadS2tNUekiecRoaRHLmulDzMBkfvxA8TAVlMLEs2xb//zjDML8E8LSR83t6hsFx5r5onV4voqOjps6dioQ5N27e+0fsDc1DMT2INdMvRrNnT76nZSMDXeTBzG98G8GgdiRTyUmn4eN/TRUiCwaHlMP16up19ZsramtLC2o+6RT6/IDM5kjTVEYbVxqcCuAggE+lHScApobB9Nx8lI+28X8bXUCQNFEYEpMs0CiLd/Hp0T2quURKFBfHktJPOjAcUkU4fsf8A0MnpSzaxDvuCxbIewe0jFsCSn75r9fDZMJ4uOrYZbg//CqCQQr19VrXTV5wQJeHiguLPJZZVslTwd31FIirApEqNYFN+euQbnjoOSf9ef0Lffl7yEmGQFEkxdnV3Mz3PTxCJQwmT+djAIDy986TGMsUoojGnJIFP+eqwvCctSj7n7ZmjsMBHHygEe90fQHAHZrtAY0cprTawu7e1xGu+5ys0VgoKnkFE5LHMIyPYZjy+N8UgIsAvJ+r8W1jST8d4wF8f+er2Buah8OKPsT9sxfbxlK2oCg4y8pUm7jKyiWS4eLHR159XcZ5RmrhejY2uYBl2cUK/yZkOQtTCMaSBbi7X5Q5aq0MdAl7rerzSgshX7H+GkECinOuWSg55A4+p8NYUB9D2OVX3XHXkgdX8DBRosWzlaIPADDd8xoAo9Liqa/bqJfBkvvJpDFfvvUSVLz7Oe3uRXPs+cRO9CzaDisMPd2vnUgNJircg2ncn1FV0pdyvKgIqK/n8Mf3fqyrawocOEJLIgyJoxjjDefpmx+AutIunDXrl/oaq7xfHhnZd++B36J4773K3U0klbwcUA/gHwzDOAA4wKcDfje/U7JJ5+3BM3H/gbvBwYHTKv6BbzXdmO8pTW4IATeurH4DAMyqKyUGkSA1HjrUZYminXDu7ocfxGhnJzx19Zi1YqUt+GCTVRiGoViWJfG/FTf4pmIx2/QFxMjhq+EY3Q1Px19l26fXzqFiIyq9p/ZNTO6tOgff09kydTxnBjXiJDSfjUjpAriGkvuvjvF2aUgYFwVoYUkkXdCl51FUVAA+n/Finmm9yh9O9zApeSlMyl7TVAwHDtD4eNMfEalcBM57mMLs5D1ML79MI1xxMtz9CrlbwvlpOVD60GqbfdEHoWArFTFTYlXAqIEn/QqjFXPR+Pdha2sAOFP6HD22F5x3eups8hDSpjSmU8Hy8LXegtHpP5C5rwvDw1QwBhPLsrsBHJvvedgo8+fOH+C53q+DAodLG27EGVVP5HtKUwJuPMQHm3PSL1RXWbmssZSuaLfttjX48P57MGfVVaaNnMYlSzH/4vMLPqTMZlIxAEBwsUYh/eWk4scyC9CfBIzOvBbgoooGU3icwCt6XLz7TozXn5vm5ZBfmIxNvxLu9zcanpMjrC42I6yn05dHXLF8OUfPgT8g5P+qsUlQFIirUrNZ7ctVGJr7fwj5v67QIjUkj6KEAqLKC1FHuJOXe6bdSpNTHCs5jvWLxV994zuoH5qPsg9WIeZuQO9pOxWmIQ0rE8Qudh/2T9TO60b1G/MURhE+msiwnlFiMhb0odK77OLegjpMOvtwDbwDOtyN8YYv6GofCHB4/335r73qN+YjUn4C+o9/FnAU65uuDLUtZehZxILzNKYcj0QAl0vrbJOeURIBKOnnpRBymAomJM+mcIlyNG7e8wie6/0GiqhR/GTGBbaxlGs4DpQz9RuKcrrArLpS0lRO0Q7g6zFtX3sHOlpsYQ6bCYM4tXwG+MLo4n/CsSmH7EKMVt4DHR9P8+KMfAiXkvhD2gKXc6qHBYMYD6GLn2joeOmOVaZGGZ0hTc53DEnLPZbuuByuvjcht6CjIK9WpuVh8hz8Q+J8KQqvPz2Uz+LF4iWn/hFn1V0FALrk0NML9gJAW0eJomcKAFx9b8BaoQYrjS+FvtOPqnyenEPbULblywAnUzA+BX3zrXj3cyj7n5KxLsXv59DeTiFU93nZ510D78A1sDn+SOlzpj03x7A00a+9Xd9rIqBMvF2F4U2SwzaY4lB65WKmGD3hOly261XsHDseDe49eOCI0zHD+2G+pzUpcXg8yvlKFAUSTf1iJgoxwmoS4rYUuM1EQhACinM+y7L70v8B0LclO8FxjLRmdP74uPQYxaVvrCgsVrKySBWPmtp/uPqTlvYfqTpVcszXuka2rWN4BxQX0GkLQLFEd8+pW2XPKf3wB/DtklcZjfmOUJhxvKBvpBflRUFlL4VKnoizTzlczusOgdNlhMmF5OmTly7feomoFyvvn+zci7JzdPikx+IUHXoSRYeehmNEYT2UZRl0v58gHKYQGtdpvMi+39pz9HT8VWIU6gtDNfn6CQE93omqN45BcetPANiiDwUHkQl3mupsHToFV7W+gIFYNU4sfQF3zVyGYqctjgFAIoCQeYc0ZpxzDuasukq+b7kv31gU225bgw0XLk/xGnnq6lWH0lOTycamAFFKmPxRTmeRJ0q2fz/tiLGFY6DoDelBiWfIpEdAYXFIKM24HVmi5SdgQ8/V8k9y0noz5hZURhfe8vk8/f0UhochyRkRU7z3btnjHK2gKBn3MFVvUNHuBkBUCrxqCW64Heq5sSnjpNWfApJiF4PzfqN6Jk9hh+SphT1q38PqYZVKoWTUlzJbcwrvw+ioyrXVudExeJTyJqqn8zEU77kr5Vhn2zCoqHYNSOUwOrV5cfC0/RqOsT3w7bkLjhE+XLQQTCbbYIpT3Nio3WgK8cSh7+Bnbb9GDE58uf5nuDxwle6aDRMeLWU6msaRV18H/znnWjcmx2HPU08BAI68+rqEyp2eix7q6kwJtZNTtBOjZVDZ2BQSDMOcwTDMGQAcDMMsFh7H/60AMCWS6lxDW3S16z3lXd19lu64AlSkV3JcuttudsGrsShU2YUvKZX/7qt9uQqu3tdlusquF0wuJM9YMU/paw2FixTG4vsVhDmUXtvo4dfpGDcTpHMuLwdKSpKhiOONFyqeLYg+GPIwKS7yrVVwlPSueP9kulS3RhI+/bIIoZGjo5kvzGIlc1Wfp8PBlMcrq+tR86pf9RzTXiHCwdsmMsKJXOpqfpgqS2BNjrniinxPoSDgOOBn+x7EP7q/Bxc1jh9P+zI+U/3HfE8rtxCC6NAQaI9X+pzDiXnX34jGJUsx98prLB02FuLD5RqXLMWiR5/Eklff0u3WF4faNS5ZiiOvvg6usnJJO1sK3GYC8tv4Pw/4ekvC44cAfAOAucSWCYZeI4aohBGlQ0d6UsPFNL5vIuUnKs5O/rDehY70tZSUKS9P3D2v6uzX2JjqbeRymHTTeNwGAAAgAElEQVRIi8uwmV4LAOgfb0DvyW9LnvftvEHWk5bOeONFis+JVQHNI/WS6BG7SJ6dhWKzOU7+H51+lckz5e/9nkUfoufEjXC7MzMCBGN9ZFRb7yaXIW1VG46G78NrIPd50QMdHQQdHZAcLwTRh4JRycs3M5Ytw1vXqruwJzsD0Ur8ePff0RNtQrWzHbfOPB9lzkxkNScwhIALjUkOU2nbPJ76BtmCrgkU1O2USA+XUyoYK3tuVyc6WtajccnSxD+xvLgtBW4zEWFZdgYAMAzzCMuyX8n3fPJHlhYMsoINmYer8b2YDzsqK6dVitfmXnaYogg4kmrEBQJxAYQ2GoB+4YvySjfQA4yMAMUyeUyezsd5BcM4+a9BY1zsgsc6r1DGi36ZQs16GJ31I7gG3oa79zWFFsZC8jhPE+BpShSfNUtJCVBRQdA9WA5I90ZlZmO1CqA8jtA+FLetA9zfMldwWfJ9ZEY4IjvYHiYRsh6FKcKHI8fi8l0voSfaiGN8G3DfrCVT11hSgUQjKXlDmt4aGXU7NcThch0t6xEbkxptaqSr4Im9VYsefdI2lmwmLFPbWIIkJsfUYkSuNRHnsmgt4hQMIx35DKr9yUzZ1SwVakicJbOCMr4DbfA6xYt/iqmrI3C7iS4Pk3jKFXGV85ERSnEe7mABqJkqeAjFYhdqJOowWVi41qzxVf7u2erdqz5rbExXz2soOvRcvF/5cwXvpAQuCt+uG0GFezRzkPx+DqMj1hY2tg5tg3D91s9IjlW/kZ63R8V7y7/VZBtMcfY8+6ysR2Eq8HzPJfjJvkcQIUU4r/Z+XDNt5dTJVzKJkDfUv22rchU2AKBpHHXtatnwuHQcnmS43I5778S229YgMpjqmnaWlcF/zrmKOUq2Cp6NjTkYhtnLMMyHDMO8H//3aZk2xQzDPMowTGu87bJ8zFUdYzvXno6/iU5VSFSn+e8bzlWNgQV/l/RR83qmyu7SxVCsWtlgsgIqNmysPaQeJpoGmpqS3hY1gYAK98HE36T5szjQF8DjH1wBpUWtt/2RZPs8hyPJyakLYhcAEFIMDYyr/RlwEWgvjM1dC3ffBskxenQ3QGKWh6xVbD4bZdsuBaA8WyGkLh13TwuK996Hkg9/oDmO389hRK3+tEXQoXY4RnbJPlfU/he4gy/IPqd13z69/duaYxcFn0Wh5DDZIXlxttx3X76nkHM4Dvj5wXvw36FPwYkIrg58B0eXbsr3tCYM3HgIB57SqEfFcYnwOIFEmFxXZyJkz1PfgGOvuhIlJ52Gjpb1iv06vcWYe+U1qJg3H9tuWyPbxlbBs7ExzXksy0oL9CS5GsAgy7KzGIY5AsAGhmFmsSxrbPVtijQPk8JihDhLLR8rWjofQ3PuxXjD5xUXTvzg+VGbNVPcVZ/BJMrdAQeOoyWb/nw+D28wRao+AXfPy5q9Enctzv7tHpSVEYDS3qhVMyKCZxxCCXsNvAd/r9mPceSvq1jsYu5cDtESeTU/iijGVCriHNkFEmpDtPwEXXMxCz22F9VvLsDIjKsRBrKWG6XsYZJ/PcW77wAgJ/kvJRAgGBvR85kzVu8snaLu51HU/bzsc2Uf8Ju8wSWpXmYqYSzLnBQ/WFenPb6v9Rb0oynnuWty2H6EOKMdHfmeQk4ZiZbgh7v/hf8OfRrljm7cM+vTk9JYqjzuhKzXEFHDU9+Ajpb12HDhcrQsPhkbLlwOAHyY3GubsOSVjVjy2iYsevRJzFjGb1areYgEY6hxyVJeSU9uTFsFz8YmW1wI4EEAYFl2F4B3AEjjSrKCznorzlJEyo4zOYZyHaZQ4FIQV5X66ZxMsScz45nAaMiOa/A9Q+0pipMdg8/niUtsH/173f3pDWvTxOEBoeXV9jJHPgxOr9hFXfd94Dhj70vFu2eh8u0zlBtk8HvuGGYTf9Pj/G+pu/d1ZHIfVm06GRWbPqH4vNGQPCP3pd/P4ZYnblCf34aj0eDYBEIoUxsLagh1kuQghNKUFa+t1TeOXYfJJm/sGZuDVbteQWd4OuZ4/4ufH3EGqt2H8j2trNC3+Z2sF5BTY7y3B9tuW8N7kwiRSIDLoeYhEhtDcvLhtgqejU1G/JlhmK0Mw/ySYZgKmecPA7BP9Hg/gEBuppYKTSnvLIfrMosUVDM+iFPusvBQREbZLTaG8nfPgVNlIWg8PyEfm2AEHCddMgUCHLq6aIyPA8SlI/s+jt9PcPAgpevn6fjDtDYzM7setS1loNKko9X6F8Qu9u/XXkJmnHtCONBj+6Bl1Iwe9h3Nror3xSOJuHGUsNck+gfU5xnzTlPt14wioXANFdFxY/j9BHuDMxSfp8KH4AjtU3w+8b6aXCP50uozpaLdp16Die8t/x4mOyRvivFK73l4uPNGENBYVv1bfLH+nnxPaVJDIhHJMSHPSEmAQU0ZT2wMCefbKng2NpawiGXZNoZhigDcB+ABAF+2epDqaoVipVqkJZbWlR5Cba1C+F3NTcDcrwDPzNLVdS21A6g5EQMHeU+Fz1ek3HftCcBbModrS4GwVCmuduwFoPdVuHcOAJ9JrRE11s0LLfmKVcaTwVdcBJ+o/RBFAFCKfRjpu7TUA0T5xV6x143i+LkdNJ/DVFtbiiKRQ+fII/n/h0Kl8KuXpkmZz9y5QCgEEKI9t2MC7wBqr2GfW9/AKtT0PwocdX3KHDHMS9Q7HHTKNayuBtxuoLfXg9paDxBU9nARQqG0NN7OAInxPrgd2LIao43fBACUlHjl38/aXwLdXwNePEmxT0/7n+E5/U/A7n8AcQPe5aRBa9w/WHg38I/fq0+Yi8qeT9MO2ePHHAO89chCnHyEvDFcVOREUWWyREBlZbHEwJg/X31K5VtTdXKqqkr4PsLNAADHYcv5uVHG3hs5asMb4jWTeIqKnCCgUFEhnTeKeNOjrlafoUZpvT85wjaYphDrDt6KDQPLQSOKy/1X4GNlr+R7Ssagaf5fVLs+Rbbw1DfwHqAMvVZiL1K69HfNwlPQvv45cOOpMcz+c86VGEPp+VE2NjbmYFm2Lf7/cYZhfgngaZlm+wFMAyBsxx8GwFBRoJ6eYXCc8e+PakJSQkLCEQ7BoFrN3jro3sB98ST0nfgaRofGUA5gdDSs2ncNKEmYTDA4BCrcj5r0xm/xi7ZIlEN/ep9Do/ACGBkdlx1Paf7jwW0YFLUvIvxXsmwftaUIBod0Xwtu87UI+b+BYgCjY2GMxPskXAwcoREMDqUYTOXlDgDFeP/9UZSXx3SNEwwOoaJCOG8EEnURhXOU8IWiKNbRhypbbkCw7vsAkteMHhtGNYBojEjGb272YefOGILBELwj41DaBiCEwtBQCMGgdPNQ7VoJ45W3vQQ3gNjQbgDA0LD8vQIAjhEnNIJGEQwOwTM4CmHpHYlEQVTuHx6X9vvKhRHs4fPRxG1jMfnPaVERsKN9rqLBFA6NYbhvJPF6+vpGEQymbkj4fBSgeOWl9PYOIxgkABygPrELxFUNBIfg7B9Epe5eFHj9sykPw+MReAD098vMm25CMYDK+gqgW7tr7ffHGDRNmdq4skPypgAhzoPrPvonNgx8Hj66H2tnLZt4xhLAq1Rk2VhylpUpPuepb8CsFSstyRES+uhoWY/ta+9ICdlrX/8cmpaexecoURQ89Q2Yt3qN5YVybWxseBiG8TEMUx7/mwJwEQC5OJvHAHw73u4IAB8DkCP953TRB2t/vulwchNHM/xFJmfGMbIrZYdZinXyx0WHngRi2ZEHo2NDKN73f5LjFMXJhuQl83mMvR9C0r8leUxWhSulbwSqbAym5GBRynvvVoVS6cljifmOwFjzVzXbpc7JIqGSsLEyLC4XUOxVfk3unpc0+6iuJvCq9CEm/X0gRfUAzb9vVLxIcrji47r60otSDtPIrJsxMP8RVDCLdfVj5zAVGK5y/bHHE4mDoRlYtfNVtI3PxgzPNjww+wzUuw/ke1oFi9Mrv09HF3lQs/CUpHGTIUJoXetD6ySeJG48hO5NG+36STY2uaMewGsMw2wFsA3AbADfBYC4xHhTvN1dACoYhmkF8CyAb7Esa822pybpi49sxPTrXHzJyGcX77kLlGwRXB46OiA5JqzHzbwSikvWj8rFgooCL/qQrjnQ1ERA0wRtbcZeRVJpLv+5GQJULP1WVpCZh6AOyB8f81+q2KdpSXSJ4qK+mk7RsuO1+xa9iRThACvEEJ5s5msnceGUw2oGY0mJ/mLHclCUSj0nIwgbHbSVQWcq19ThQbh+OZqb9fVU5i4M5V87JC/OCTfcgLeuuy6vAgFWs7H/M/hV+x3g4MAnK/6GbzQpK5rY8CgZQ01Lz0L3po0S48YMDq83YQApCTzY0uA2NrmDZdndAI5VeG6B6O8RAOfnal6pFM7CGpTcXiul6mFyjO2VOWr+99a360aEa5chXPvpeE/ZvT5ydZgA3lPQ0EAMe4rKy4HSUv68oeUPoHT79zOanTUoSNUrqAMeOkQjFAI8Hg9GZt4A3+7bpecSqZEpMMzcgRL2OtnnKt7+JPpPkot21XqtGvdUmkGjZhQaxTm0BRWbl6d1r9yvz5f5elNJntwIkcpTEGo4DyOH/wjVby7QPkEnWtdUrYRlOl/42GOI4u4MZ5QZtocpzoxlyxSLgU5Eft9xA37RfhcACt9pus42ljKke9NGy4yYWChpdCmF99nS4DY2NimkrTqzYiDo3jCUju3p+CtApHkquoY18Vq8B/+A8vfzZLumwYenGX8NQlhbqPkrIHTGWUhZIF5LR+YZwbPR3h5/VsEqUntvxw77ruJzrsFUgRCrNrNrX65B2QdiRT1imYeSDrUbal/i0/IOiTxhCpdRqQBuOp86ukX5SdqNoaMfBlecaQFqMSp1mExQ6TMW8pgNbINJBBfSLiBX6EQ5J27c/Ve09H0JXnoYt888F6dWPJvvaU14Qoe64CzVr9DiqW/QVSfJlga3sbHRAy2Rfc5eSJ6WARMtl6/zREl27zOj+/T9OltqL3h7FrGabaRQoMf2wdW3UXUMvhYTv5wannWz7t4DgWQo39DcTHbPrS8YTI/tRwnLq+bJ3Q+CLLZQtDcr9yM3Llpxa8t/m0KHrLhuZKwDtX4jRdMzHlJTnjxvkIKQArcS22CaRBwKN+P7u17FR6H5aHbvwgOzFyPgac33tGShPJ6EqEG6XG4h4qmrB2Xgwx/q6kRsbAxwpPqc042hxiVLceTV16UIPBx59XV2zpKNjY0qVoQQiXH1/wdUYuGt3vfQkQ8oTEo9J8M5uCX9BNX2xKVc80nAs/9Bvq3G9eA8jZp9yVH9xtGoeGep6hiBAIf2dgqxGDA240rdfacKJ2QgDZ6FwrVl/7sURd3KeiYSsQvZMM3MqNyYlAinuVH+D4vve0dov4U5cMYMpgPl1+Dh176e0YiW5DBlgWmR36OpskOz3Xu9heEl1kPhr1RziKts4go/vDt0Oq5u/ReGYpX4ePkzuHPWcnjowvWYESEsjRBe/a6AEYycyNCgofMigwOgKIpX3lMxhhqXLMWiR5/Exdu22QIPNjY2uiAW/3wX770HNV36vBzEIR8+Jlu4VoQ7+C+FZ5QXlR+FPqv4HACUsj/Mu4qW308QjVLo7DS2mPf7OQwOUhgYADLJ5yIqKnXj1UswMvN6xeeVSRq/coZiuthFNrwJzrHdiTC34rF3AFgX4iVAxUZQ5u6yZP5UzNiaqzngxDPvfU7xeXePShhdHCtymPLJa5Hf53sKurENJhHMKv07Q4XEo4cuwz1tD4CDA19ruAXfbZZPoiw0rFCbs5pEGF3c6yU2chTzilQ8ZCQagdNbbKvd2djYFDzuyD4AerxX8t95Ff81+P2mY633ju/vmm1KXN1ZWbAXHXom8Xe1+yPFdknFO2NLqtSwNvMLX65IOeeVK6rHeJ260SkPpfA3j1TsQiGHKUOPkHPkw9T+CjjMq/RD6RpS7fU3N3Oqxr6SIIYYvTlMhYreYs+FgG0wiejftjXfUzBElKNx296H8XT3t1FEjWHN9IuwpEr7x8VGAZrmpbxf24Qlr2zEktc2pRg5SvlG866/UTFfCeANw46WHJVqsbGxmRJYHZJnBZQQNqVA8Z67FE5Ufi16czTKvNaruztC+xJ/O+mwouJbsqaScQ9T8jzzBlPIv0LlWQqUAcEEemwv8BcKroH/arZNFbsolPvR3HXM3udJTSUPKCvLTFq8vl7/67XaO2cFgQCHs+9+Kt/T0IVtMIk48PQ/8z0F3fRFanB560vYPnoSal1t+PkRn8Ss4g/yPa38QdOYt3pNZt8IGqGBavlGWgp629feoWo0dbSsx5NnnomWxSdjw4XLbQPLxsZGlezutGco3azUq0aOkxzNzfp20M894R+G+7YKYY5a0uJb9h+D3lOS6m8pxWszUYGjaISrP6n0JKK+2bq7kuaZKSMWu1Dsz2F1sfkCXPVnQHV1Zuc7HNbMI180NxN8cOCofE9DF7bBJGaC1GD6YOREXNH6IvqjdTi+5CXcc/hSlDiN5ddMOjgO29feYUjJLh01L5GAkG+UHmKnJQPOjYfQ+tA62ec6WtZj+9o7MNrRARCCUFenpoFlY2Mz1cniwjFnW9Hav7nFOtW2i4vyl7Pr8wHV1ZxmEdot+49FzHdE4nFtLUFRETEcyicHoRXKolAU4PAiUqpdX4eK9MLIfSUWu1C6Z9zOiOrt1HfSBgzNWat7TO2NguzcuwMLHjV1ntZ8q6v1bQhQVhTX1UHvKe/kZBwBjweoqpoYa2/bYJpgPB28FLfv+y2ixIWL6u7GVYddPhFE5nICNx4CBUoSNkc5XbzwAqCYb5SplLdcuF46Sl6o1ofWSQriqhlYNjY2NvmEUK58TyEvqC1+9Xhb0g0HmuZ32DMNyQMAQiuo7BnYCK5+42g4h96XdqHwusViF5HKRbrHERMtOwaRytNMnStPdkLywrWfyUq/fc6Pm+o3W8QMeCP1oGfvpbHRNpgmHJSrcH8EOA5Yu/8XeDR4JVxUBKunfR2fq/ldvqdVcESGBiVhc0dduxqLn34RF3/wQSI3ad7qNYalvDta1mPDhctlw+ZSwvUUUPJCKRlSVhXKtbGxmXzkNSTPWYL+Y5/IeBTP4BsZ91Eo6CteK10YJqXFs7tojFScpN0IgG+P1NujJqcO8KIV0fLjEa463dzkDHk0J1dIXkVDDf7x9rn5nkZeaWiYGMIVylqUUxCH14toxFyl8mwyHC3Dj/Y8hmDEj0pnJ26deQEqnD35npZ5aBrOkhJEB3WEEVKUoR0yT109Gpcs1TR+9LQRI4TNCZ4gIWxO6EvcZ3pbQN2D5amrl1UM1Arzs7GxsckXkZozMTT35yjdcZmh8yrePgP9J74CACjr/ks2ppYX/H6CV16hDUf2BwIcXnjBiqWYkiHBT2hk9u0gzjL4lIQ3TGBW7EKK/vP5jQLrjcvsbUCo9+v3E/z0t9fjCydmvgExUWlsnBgGk+1hEhEdsl5lJ1N2jh6DVbteQTDSjHnFG/HzWZ+c2MYSAHAcYqPSoq5yuErLdOUWCYS6OrMimmAkbM5oMVol9b1MQgRtbGwmN9n0MOntO1x9huG+XQPSHIlClooWo6SSB/CGz9gYhZ4eo0p5BN3dNCLhzIyAaPnH1BvQLsRK5prqW+n90St2MZXRurcDAQ7bDszL0WwKk4liMNkeJhFKO/354sWei/BI12oQUFhesw7n1ylUV7cci3dwZLxEJBqBq6wcDq+XDz1T2JaLDA6AWXWlxGOjhpz3BwD2PPss3rvnXoQOdcFTV49ZK1bq9jIphs3FJcPlitHq7Vtot/vhBzHa2Wl4bjY2NlMPQrK5SNW56KesmcPEMJfUEXtblMwSucR9QVq8r48gE8G0scO+C+IsQ+n276U9k71QP6nYRfZzUShKfX0SqTjFVL/ZkhXX6tXv5xDjFPLPpgiNDVYrKWYHe1tARCHt6D9w4Gf4Q9ePQCOKHwa+kz1jiaJk4oet+9JzlZUrPhcZGkwozml5kbTyg9JJ9/50tKzH2zfdxBvEJpTo1MLjrFC0a1yyFMtfeskucGtjY6OLwvDKFMIcCoNMi9f29Spfyxino0+KQqghW7kwOsUuTCsNGw3JUyZWMgfBM/tNzsN6tOZbUQF4vFN7Kd5YbxtME45CKFwb4rz4YevTeGtwGcocvbh71llYULohK2N56ht4QyCLcuoERNHY8NTVJ4QU1Dx7rQ+tS8h5z1u9RlONTkDsFWp9aB1iIfNKdGoqeLainY2NzeRC7wLWvMHkHBSrsU0Mw0vNC5FahFYeuWeE8zZ3LgVx+GTPi3E6i+1QRZJDzqHM6zOqqwOKxS7k1xK5ksQWjWj4jGxtQGh5rigK8Pu1+ynEorNW0dCQWfHeXGEbTCIOPPNkXsffF5qN7+18De3hw3GE9z3cf8QZqHV3ZG280KGurCuxRQcHUbPwFFljw9vsx/a1d2iGQYrnqEeNTkBsqGWqRCeMq2eONjY2NllHRwhRzDMtu3PIICSv8j+fsHAi+aeiAvD5eG9L9+n7dZ/X2EhA0wTs/iZ0nyH/e8/pDb+kpVkWzuHkRnDMa/39IHiY+H1Xk5LeVqvkTTDrQgjnnIzoeStKfBMjh8k2mMRw+XvT/t1/DlbvfhwhzoellX/AmhlfhpPOrpvSU1efEyW29vXPofwoaVJj3+Z3dOUlpc9R8DapGU3poglqXi69NC5ZqjimrWhnY2OTS/TsiI/XnZW1vo20S8cxssvUeflG7fVSFB+W19ZGgbgqdPfpcvFGk3oon3kDgCLJ3XtNYQglVIxzs2IXqUwsA8cIej4jgpfRLDH3RF9/2AbTxCNPFWAfar8JD7bfBgoEq5p/gEsa78z+oA4nZq1YqavgqhKusvKEEhzlUe6DGw+h7/3NpsZQU4tTmruzrEyiSjdrxUo4PJkr0dmKdjY2NoVBASwyzXqYuPGUh4WRj5U5WsVrlULTtGo46fYw5YEUafEshvcnyU/onOl+dRlMmV23cM2nAAA/fvxW1XaF6ngTG/WFTOF+CvOA/3PLczpemHPjho8ew6v9F6CYHsSdh5+NheUv5GRsKv7J0Qpxo4s8qDzuBNnjzKorE6INReUaO2pmvHc0jaalZykKIMjJd89bvQaLn35RVrXuxJtvNlysVs+YZvqxsbGxyQdhZ5OOVtnNYaLHC0eN1ihqi85kEVpjaBla+TaY1Bb9KcVrK0x6sCaJwWwW4RpmSkefNZ6mkRlXW9KPbsjE8DDZsuIi5l55DQ48lZviYR3jAdy0528Y4SowrWgHbpzxZXhofbLZ2lBwlpWqFoYl0UhCTEEsgd3Rsh6tD62TSG93tKwHe/+9iAwOAADoolQZTM0cHoMFaAEAHIf29c+hYt58VaNJr7EyY9kylJx0mrE5ZDimjY2NTSExOuencG/7qkar7BpMFe+lqrlp7Xx3n74XoefPg98rreGUS7S8EH4/QX8/heFhoNZAv4EAhyefdCKqEIWfb4NJXSUvKXYROuMSFO+9z0T3BWAwZcnDpOcz4vdzQEbCfgVw/TJiYuRw5ftTWFBYXexUif8MLME1Hz2LEa4cp1U8jtsPP89CYwkACBY//aKmopyc2IKQHyQnbx0bT4ZRRAcHsX3tHdhx753YcOFyTWNIaR60x6sqPW6lAt2eZ5/FhguXo2XxyVkpbmtjY2NT6FBOcyHY8p3JL9S6T9uNmCdgpCPVZ4mrCq9w/zTQX34wKy3u9xPEYhQ6OuSvQ3ZrbmWGWOwi5tEh91ao6LA5ek9513i/OgwxQVq+UKByElqZJFp2LNjoV3I6phkK91OYB9j77836GH/q/CF+fvBeEFD4ZuOP8a2mm7I2VkLZTSU3S6/R0PrQOolAAzcewoGnntBUufOfc66iuAMXGksxxOQQisNmQqZ1mGxsbGwKCfN5P3rOy8zDRFxV6D11m+4Z6aGpqbAWlXJoSYtTCjvpyfPkf6t1y4pnCb1iF3B4TY6g/17OljNKTw5TzHcEOEep5WPX1Wnf27l1wuX4s0bROFB7W27HNIFtMIkQws2yQZSjsWbPH/F879fgoUdw64zzcHpldnbMxB6bxiVLMe/6GxXb6vXemJHNdpaVYd7qNZh75TXKKnI0rUspL1PjJtM6TDY2NjY2qRBaWvcngcUrvKZmS7szhdaiWvAUmC1e29Ym338mIXkkB8s8bbEL9fOJs8ziGRnnqGZrDXwBPRsbVumNFRcTPNF6swU95X5zoqk5v5sCerANphzQHa7Hql2vYtfYcWhw78H9sxZjundn1sarX/zJREHYlsUnqxoFeg0hQ7LZFIUlr21KEV9QUpfTKwaRqXGTaR0mGxsbm8Iiex4m3d4rR7HpMYyOV1NjqMu8UFdH4HYTRQ+TyxmWPa7lYcpEwS1WPDPlce/CjYb70M7dMid2kejfVQlCuUyfbwWnz31FZ0tjxkQuFSArKwnWvXGDBT1ZZzDp3TepbiixbMxsYRtMIpxl1u9yvD/0cVzVuh6DsWosLHsOd81chmLnqGxbV7U1vwhdr76cLAgbDz9TQq8hZEh+nBBJjpCSupyeArQCmRg3VtRhsrGxsckX0RJpLTtTiFYwW6dvVWqkuztCubUbWQBNF0Ziu9oCkKb50EEl4+G8j/1d9rjXC9TUKEuLZ5LDlO69iZUav4+0Fv1isQuzRCo/bv7kHKIUVlkIVFWqy9PrJsc5TEBh6H5oYRtMIuasusrS/h4/9F3c1fYgODhxSf1Pscr/Q1XXa6SnW/E5Kn6iq6ycN+xU7q7I4ICuMDcj9YPkDB7/OecqGlFyOUJyghJG6hdlYtxYVYfJxsbGJh8Mzn/E+EkaC599nXUmZyNC7rdoIqx+TKDHW8Dn8xhfWgUCysVrM1PJy/57YVbsYiqQrfpOclRWaRVAtskEW1ZcROOSpTj4/LPo25yZdCnHAT/b/2tsGz0FbiqE6xOH6xYAACAASURBVKdditnFW8x3SFH44v/+h2BwKOXwhguXawoupOOpb5BIhutFTk67Yt58XoZcZh5CGJ3aGI1LluLD++9RlUAHMjduGpcsRWmZF+/dc6/p129jY2OTL1TzhUxy4KATqJYZK9cyxXoMrAlihPn9BK+8QmPMfym8B35r4DwO27fL53FwhIbpDA9Lrpt2SB6gLHYxqTDsfcltSN7wsBXjJV8j56wAHc1I83zSYBtMaZxwzwPYce+dpusxDUQr8aPdj6E32ohq50HcOvMClDn7eV+9meKtUPaszFqxEtvX3pHiTaKLPKCL3LIGiKe+AYsefdLUHJTqMwn/WhafLPtFkh5GJ9fPnFVXSV4H5XTBUexFdGjIMuPGqjpMNjY2Nrkn/fs184XR/jZ5g6kwmRiLcb+fQ1eXCyHvfBjRjPP7CVpa5F9jjDhgPsMnFx4mkdiFETX5CUnhhuRVVlo1N76f4SN+grFp30ftS5UW9TuxsX13aXS0rEfXqy+bOnfHyPG4fNdL6I02YIHvddw361O8sQSYNpboIg9qFp6CJ888U1I/SCkvaM6qq2QFFsx6aDpa1ktyotLD7fTkCCn1A0DyOo66djUWP/2ibD0oGxsbmymHRUnxBBTGa5aid6QKbQeU+iw848QKr9d4zWcAAGNNXzY3B101dfjf+j6Dm/KBAIdQKBt1mDK/blrXXkvsQu8oE4P8iD7ocRRWWWYwJUa1uL+Jje1hEiEs6PXk/6Tzr+6v4i+HrgZA4fza/8Py2t/IN6QobZdu/JPhqatHzcJT0L7+ucScBCOjf9tWdG/amPDUzLvhJolRIecRMoNSDSZxuN2sFSux4+47JNLdNQtP0dWPbRTZ2NjYKMN5Gi3ra/DYv+Os1cXwemKyzxtZ5MW8M+Ec2ZHhjHKzMBs85o+gokMg7mp42/+UlTH8fv73vb/PaPFa5U3VzPJgrBABUO9DS+wiHwwe9SuUffCdLPScnZA8zvH/7d15mFxVmfjxb1XvW7qTdCfd6eokQMJhCTvIooAokYig0RERdHBUeAaYwRFEQZBNxYmI6Ig6qBkFRPkxjIqoEM3MyAA6jCKish0WWbLTWTrdWXqrur8/7q3uW9V1q+5adbv6/TxPnnTdusupW7eqznvPOe9pI5keKr1iwUOYxwithSmnjipBU5YETDaFKvSlZDLwLxu+yuNDp1KbGOXy1MUc0vaY8wYu+r/Wtc3izff/EjDHKTlNGJtlb6nJBh2Fxhv55SYld8/yFQy/+Cwv3nNPzjob1zxAx7JDzecltbcQQlSYWQHq68vwxBOFR8Z4GfYytOzbzP6/E8MoWHFhjMVJ1mPU+++D6CZwyQY+O3YAHuY4zQZaBY8bpNJaprFf+ckuxlsOnAik3RRhuPdD1G//H4bnv5vGLc5zVLp9OSMLPsDI5h/TsG2tuw2qQEsrNDWFETRl95FAAqZJ8bkdEANeK+67x1v5xEsP8vjQctprt/LVJW8rHiy5ZJ9A122ZopyE1W1K7o0PP1y0XHVthdO2Oy0XQghRmJtKdKbBedqGVCrDxo1OVQD3laRMrYeowImrFpR4VNxKVdgXLDBIJAy2b/dW3mxXvkLSGe8pH8ab9i29kkturjVz8trJ9cY6jvd0jJHu99K/fJCxOW/xXD43xltUODvymPShnFnyEpif63968DGGDrglwJ6s1xhCsD1NcrW4IgGTTaLBfRaiv+49iEte+DWvjy3kgKbf8bUlb2F2nXNacK+y44O8pNKOqqXGadLZ/DFRezYXztiXLZfh0JTttFwIIcpBKTVXKfWAUkorpf6ilPqxUqqrwHq3K6XWK6WetP5dXYnyurV30ccYXFY4U1sqZTA2Fn0GtfBUvua18ujSSZPq66G722BgwFv1qr0d2trC+y3cddDXrb/KU+k1k10k2XLEQwwuW+2/pmyM+9uu9I6LPpvJuCvvSM9ZYRRmqgCRRdoKjjMN3aRSBv/zl6MZmxNSgqtqingCkoDJxhgZcbXef20/i2tf/n+MGE2cMXc11+zzYWqT/pI6OHnqnz/L2lOOZ3zvHhK17gb7RjUJq1Nyifwuf83dhe9mZss1PlS4f67TciGEKBMDuElrrbTWhwAvAasc1l2ltT7c+ndj+YqYy1U3rWTt1ApeYrJLXqB9uzCkbnK3oqvDVb7i1t5cfPqLrFTKYGCH9+DHPo5p11J/l9b243/P4CHfnXgcxnvpJdnFK4NHMdLzvgBHC7cuNSGkyViHDrx14u/hBR8ofdgyXLd7F3+MgSN/ymjXClKp7OS1YSTgqPxnLmvnnsr3RJKAyc7FB+pfN9zIdzdfR5IMl6b+kXPmfyWasmQyYBiMDw5iGAb1HR1FV496EtZCk87mO+zjHy/aEuW2a59bm9au4ZGzV07JHiiEEF5orbdrrR+yLXoMWFSh4pRFdszMXbuen/qkp25Ezr+bwwvd/ibFp2IWhr6+DNt3eH9N2fTcAMO9fzvxt5duXelWxUj3ewmz0uu2Sx4wmfgh4a966SYIcc1D64jrwCY5OfR/dM6bS++3HF3yEjWMzT0FMK+hbduS7Nkb4LhGeF3yqokETHZJ59MxnGnkipd+wqM7V9JaM8CXlpzB0bMeiux4OdLj1DY1sezq6x1bmxasOL3iWeb2OeOMoi1Rbrv2ubFp7RqeWvX5nBTlT636vARNQohAlFJJ4CLgfodVLrO67d2nlDqwjEULiVkJyrZm6HULA+6vTF2qp1HlLZXKsHPAe3mdMuX5e+nlrfROnbzW53FrmsMpUJ5068FFn/cT2Ix0v4+Rzrf5LZInbt/G7PuwZUuQ6r0MkyhEsuTZpM5cWXDC2vXD+3L9Kz9gb2YW+zb+hWsWn0d9ctT1futmtZMeGZkywexBl18J4CqV+Z7Nm+lZvoLnbr2F8cGxKc9vfey3rssTpWLZ+bLLw0h3/tytt0A6r69zepznbr2l4oGjEGJauxXYBXy9wHNXA5u01hml1HnAGqXUvlrrwvm5C5g7tzWUQtbX1dLV5T3hQkdHC3S10dUFc+bA1q0NUNMI6cnfoNa2Jrq6XE67OjT19XgtV0tLY+lt0qXHGDvtw895KnaMUpXXAw+EPz5VeKViZTnggMm/Oztzz6vn15A23z+/14ldTU3pfbS3m5X67dub6OoCXp18v9rarGUhaGtrpKursfSKAPWTVdyGk78P9zq/BoOE+/PUfSp0nUjXvFnwlvvg352DvNpal+e/xEU1e3aLq3N4yCHm/9v2Lp3y3Ny5re7eh3Xmtdc6axatAa+djo7mosd0fc4T4X6O/ZCAyaZj2aGsv/8nOV3zfjNwOrdtXEWGJKfOvpsP93ze0z6TDY2oSy4FJgOF2rY2EiR46gs30DhvPgtWnM76n91XdHLb7Pggp/E+0yU1t5d055vWrnEMrsYHC/cld1ouhBClKKVuBpYCZ2qtp3wha6032P6+Uyn1FSAFvOr2GNu27SKT8XcH117vGBtL09/vbvynfbuBgT2M1Zjb9fY288ILBkMX3Ejbc5+YWGfX0DD9/e4G39fsHmJO3jJ7udzUz3btHin9WjJjJfdVaB9dXW2Fl7soVyFbt5Y+5x0dzlntir3O2bMnq2Rbtw7R6XK7QuoGdtMBjI5l2Jm3rdfXPj7u7lrr7m5B6zT9/cO07B0jG0YMDe11fT2VKt/Q0DD9/VNvGhcya3ScbNjWP2D4un4KOsS6sW6tX2y/Y+MZV/udaxTv8rVjx276+0uP72ptTQCt/PnpBG9bkPvctm27aG8v/d2T6LmM5r3j7J51FvQP+f6sgPl9099f+H5S/mez6HEM758BJ8lkwteNK+mSZ/Pi6ttygqXbN13FNzeag1YvXvApz8GSvUtadgzQsquuIzMyaqYOt7qSbVzzQNFgKdnQyGEf/7i5z5DHAcVVdhJhe5e7Z25eJV3uhBCRUEp9ATgKWKm1LpgBSCnVa/v7NCANbCi07nTgNEDc20B16b6Tzz6n0njrQR62CzPhQZhd8tztIye1uM8xTJVSzvTfUeruNqitzU3x7pVR28bu/T8HycKtuqOzT3K9r7B6hJYjeUYp0+uKjli2lWY8U8s1f72btTs+QFNyF/+870re2PGAp301zu8umByh0OS4RbvjJZMcdPmV7HPGGUC444DizOk8TczpNKu94HY1TS67kQghhEUpdTDwaWAB8FsrZfhPrOeeVEpl79XeYY1f+hPwGeCdWuuo8iBHrq/PYN26ZIHKSOUrJ1MVL9OWnfN873ls1lFsPy68bu2pVIa9o+ZvkVHT4mG7yUDLSLrsduaovEkfIH/y2ulVvYwqYHK739Gut4dyvJoacy4w+yTCYTNqoq9nPdx/aeTH8CpWXfKUUvsDdwBzgW3AeVrrF8p1/MZ583ltXQ3XvHwPu9KzSTU8zw37nEtjcq/nfTl1kfPSdS47zskedIU5DijOSp0/dcmlPL3qcxjp3Kbe9OgYm9auqbrzIYSIjtb6aRxqllrrw21/n1q2QkVm8mWmUhn27EmwZw8ETdo73rwftXtemrJ8dPabGOk+m7ZnL3FVJudViq+TztT4qqIbiXoGjv21jy2dtbTAg8+ew73Pv8by9x9CxxMrXW3X1WVrratpYuCIH9Pxx/cELE35gt++vgz3319L2vWIPu8SiWhaNMNuwUg39lEzvM71focO+jq7l1zL3Efct0g6mWw5jkr019Szmw7m0Lc/SMfj4QSSYYjbLYDbgG9orfcHvgF8q5wHf+3wG7j8xV+wK93Bm9rv44v7vbtgsJRsaCT1rvdQO8v5JyZw1zmrZcle8c+m0X7qCzcAsOyq6xxTfE93pc5fz/IV1LQUuHOXHp9ohRJCiKpywGXB95GwB0xm5XNgRxhd8gpXJ3Ye/QDDqQ+5LVLV6E0lue2RKzE8ZH3LT5ybaewDou0utvXkqUHuVO675I2PJ9iyJRGTLnke0oqHfo6T3kqQrCfTmArlyGbXyKnnP6zP2XDqw+HsKM/gsm9P/L19W4Kx2W+M5Dh+xeGKBkApNQ84ErjbWnQ3cGSh2daj8OlPw6VfeSMZavhI9w1c1FtgAndbquwDL/0Up9z/K5Zdfb2nLnKdx53grkCGkRMIvfzzn8+oMT1uuh5O9wQYQgjhyZFfZqTrHUA4I4eyk43uyAuYPFWsjIyPjXxI1LBLfZGdh/0glN1l6syUCmMdbwhlf/km7vKHNGFqFDK1HRj14VWxstfTunXJmARMhb08cAh/+71fsf2EP0R3EOvzEFbLlZePVyqVYfPm6D6P4y0H0L98kIGjfh7qfkd63j/x944doe46FHG6ovuADdn0rNb/G63lkbrkkkZWrYLmZvjsPmfz1jn3FlwvO2krMDFh6ourb2PBitMd5x6y27R2jZngwYX8FpY/ffWrRcf0VJue5SuKzukEMycBhhBCTAqvAp5tYdqxI78qEF4Lk7s9uDve3oUXkW5e4vs4duNt5rw8e/b5RIk1/Zm4yx8oYArjvQ6+j8ba3a7Wm5y8dmoiES+MhHOWwTDU1Sa495FTGWuaTL0dflKBcAMmL/r6DDKZCI87Eb1Fd4yBgdyPTlRdMb2I1RimoPzOb5FOw5FHwnc/+Uue/twzBddJJJN0dbXx8s9/zrNfXkV62AxehrdsZsMv7qeu1Tx2MpmgbVZTwXzxv/nut0rOtwRQ09jIEZddmrOPPZs3F1x3+PUtFc9NbxfqfBfnnsWh557l+PwRl13K7667buK9gMLnLqryRSHu5YP4l1HKF9x0KKNwb8/Cf6T5tanTSs2ZY9DcbLBjIJGT09es5LmsoEy0MJXn/mu6aXHIe4ym3H195viwwUGD2T62Hx4G9+kiotVU5y5g6u3NTl6bhEP9V6Yz9fOoGdnke/tSauuTjIwk6O9PkL29GnaXvKiDvmLCzbZYiHmu0k3RtWeMj0N/f4JsKpdEDLJxxilgWgf0KqVqtNZppVQNZsaidW534Hd+i29+06wg/PgtX3Jcp/fMlfT3D/HHW76SU0EHMMbHGR0YAGDPpk3837XXMTS4d0ork1PQA2ZWPXsSh9ZjT87JOd/c3c2eTVO/QBrnzQ8tN31QTvNdRKX12JM58BNXTkmAkX/uKlU+r+JePoh/GaV8wQUpo9/5LUS00s372R5NVgwTCbNiv317kKAhWzkr0510hwxd/u/kR1PubGtLfz8s8rH9hg0J9p8fdUpwd/Wl2hp38x61tMDcuRnWrQvWwhT1tVRTVw+YLWHLrGVhBkxjzQdAjZWSuwLpyrNdI6NjvqZM02J2Hnon7X8+L4IjmKnRDw59z/7FJmDSWr+ulHoSOAe4y/r/j1rr/nKVodjYlwMv/VTJdbKyXeXyA6bGefPNMUh5sinIizns4x/n/669LqeFqhLpxItNJlsJXibCFUKIauG/gpe7XSplTBnDhJcWphC65FU2jXk0d66zldb1r8/l6Hrv269bl2T/+VHfVXe3/5qE+8z52a6IRiI3MPcmmuthvHkptXteYH3nZwGrJczKyRFm17mtTe+kc3Rt6Pt1a8GCqK+byddUagxckKGN69cnoSm7n8q3MMVpDBPAhcAlSqnngUusx2XjOCZmfnfJdfIVCqyCzKG0zxlnlBzTEzWZTFYIISrM6tgfVo6FVCpYC1PC6pJnxHiQfyVku0X9Zd0hnrYbaDmNT/7wppwsZ5FVul2Or/JyfKfJkGMhYbYRdC4w53E0W8JMQVqYdhz7cM7jwcFERVM/NjbCvHkZXt55BGPtR4d/gDK8tkTCyHl/4iBW33Ba6+e01sdqrfe3/tflPP6S8y8kUVs3ZfnY4M6JoKBQ0FNIocDKTSKDYnqWr+DEe+6bSD5R7paVUpPJCiGEKA9PleicCk7udn19Bvf85kz/+zbK3CVvmujogJaWwumdi9lzzL3c8uDlkc6js/2EJ3IeD/ecW3wDD8HEZEpr/+Xfvf+NZGrC6NqbXwYzQGxpTTBrlhFaUDo+6/Ccx4OD2Mb0VeZz0ddn8KG7f8fAG/47gr1Pvqax9mMj2D80NWVYvz7JSOdpgDnPWqXFpkteXBjpqU3P6b17efqLNwJTJ46tbWsjvWcvxvhkH99irUbTuQtZGJPxCiGEiI9UKsPmgR5e7LiZJQOXW0t9ZMkLcNfZ6939ka530ND/C9/HK4fs+DCvd8nr6qCnxzBTcwfqLui8baauw/W65rPuX0M22cXevUnfSStGut/DSPd76Fo7da5LL5eZMWW8W/Z1JqyWsBDbDA7/Ijx5BQBDQwnoTlpHLH9acTA/13/+c7hBhlHTQiK9m5zvh2Q0YUSycTbr1yfZddCtNDy8P+lM5cOVWLUwVdqLq29zbKI2xscmWlLsLT2n3P8rDr7i6op2lSsXSeMthBDVJdt1bGDAX3XASJq9Moy6OaGVqZTxNm/d3IqLrgXAaQLR0ttlclqYwp9UNbu/3PqOkWwIvOdssoudQ5WvXu464Ja8JZMBU1+fkdclL2B5D/rUxJ9ml7zKvv5UymDDhgSZSPI/RN9qtmH8JNatS2Ak6qwjyhimWCnVUuL0fKW7ypVLkDFYQgghwhCw4pCY2iUPYHvOXEzuK0Tp1kMYUjcxuGx1sHIF5D+omDyfW09+mW0nPjvx+LGtHwlUJr+tGPmBVugD3q1rYLISWmr/3lqYAHburHz10qifm7cgwhYmm507k9QNPAbAwo6yjiyZ0NeXYXTUTJ0eujKMYVrQ4+9mQ5TiVZoKq2ub2vxrV6mWlE1r13Dfqaey9pTjeeTslRVLshB0DJYQQgh/1q9P8Oyz9iXhVFrmzzeoqzMYGPC5v0SC4YUXTq2cettJxOu7Y9TPJdPYO/G4sTlYi0vhDISl9fVl2LgxQTodVcW08Ngeu8dfe5PtWW9JHwAGB+NXvUw37wOYXfVSqYzZdc4SZiveoG2/+835U2j79SIbuEaROKEcmf8WLMgwOJhgKDvDRQyy5FW+U2CMGEXusiRq6yrSkpLNTJdNtpDNTAdUJFCZzmOwhBBiurr55nr+9Cf4wy1BKw65lZ1k0kxDPLAjAfsE3HXZhFlhc95Xa6sxOc2UD37nw0mlDNLpBFu35swn7E3RDHjJkuu868s/Y8O/zLZWc3++s8kuBnbWQPl6aboytOzfGN7xKJmmhRMtq1lhBkxDg6Htyrds18j165MTMyAHbhiytdBFrbfXPNaGjTWx+VqK3y2AChofcp6s8eArrq5IoCCZ6YQQQrS3w0svRbPvvr5MTpe8CmZEjpXW1mDBaba1xe92mzebb0SgynyhNzNv2Vj7MQCkm/adWDYwNDVjsNvD9fVlclos43I9GXXtjM57BzD1vQmz1WR83P++tp/wBNve8NvAZZhsYYqgml/iDf3U/zzpdlVHPdZcUhs35ncfrRwJmGzimLxAMtMJIYTo68swPAyjI963zdRPziVYqGJodh0LUrqAvNaqylQDb20NNmI+vxXD63bZgCl8ufsdTl3A9hOemAicAPaM+M1xZ15PAzsrnwa6mChbmHJ522+6ZQmZ9mWBj9raCh0dRkTp6Yvv8y8v9QU+Qnby3Q0b4nMdScBkU6zLXaVadCQznRBCiOwd8b3D3lsdRrveXnLf9vEc8Te1rFGUviZgXW3ePIP6+smKud58gKvtenvN9/r1LSG8qgLd7iYnGZ5MCZ9uWYJT8gevSSdSqYzvrIvl0tlp0NQ0+brKMS4nCD/3CKJLbFG8MBs3Bj9mZ6dBQ4PBhg3mvkJPfOJDvK/oMivW5a5SLTqSmU4IIUR2TMLIsPm/pwpUIlF0gsm+vgyZoGmVAwleWY1jhTc7Psyr5mbo7Myw2QqYfL22ohdI4bTigTMwWlIpgz174129TCQmA1OIroWpktdlfnr6ctkQQsCUTCbo7TUmgi8JmGKoblZ74eUlMuhFJZuZrrmnRzLTCSHEDJUdk7B3b4BKNBSsSGeDsekjfsGRE3viBy+V8lTKCNYlr2jSB6eAaZK99cWrvr4MmUw01cswe2Par/swA6Z6W3LFSo69MeeaKv8YpuHhcI6ZSmVCCb7CIlny8jhlyiuWQS9qPctXcOi5Z9Hf75yUQgghRPVqb4dZs2B4GGj1swfn37BUKhPhGI4IxCWLgAv2SrmXu+SpVIYtYXTJK3iuCu/XXrm3t754lUpl+Myj53HzBz7pex92g3vbmNXkr/4zcPQaanc+XvA5v1kMS5k1Kx43IFKpDLt2hflZcZclL6xWtb6+DI89Ep/PenxCt5hwypRXLIOeEEIIEbVFi2B4OOgErVO37+01KtzlJT6VIjujxn/igyx7NjYvrzKVMnh9S5mraLZWqSCtjn19Bv2D8wIXJ1Nj3hlYetkLvvcxNvsE9i7+WMHnclqYQrwGZ9k6JFW2S17In2lj3PwvWR/ufgtKkEoZ9PdL0ofYkiQLQggh4sgMmMIPbOrrYVZbPO6KuzHSdXpZjrN73ysC78Nvl7y+vgzDIxFVtq1WJ6PGoanygMsCtb7kJ7vw2yC4/cRn2XrySwwTPPgqxB7MhtnCmtvCVLmAKewWtIQxZv1RPGAK61zm3myo/PeTBEx5Oo87wdNyIYQQohzsLUxhd6HrmF2+il26KXcqSq8V6nTrQcELUXSMjyWUFiaDS79/i6/tgimyfbKBXWoVA8f8qvA2HYcEOr7fZBf5jLp2jPou3/NZleI37Xsp8emSF245MrVW01kidzTP6Jw3s7f37yYezwppyH9fnxGrZC4SMOXZ+ljhCcOclgshhBDlsGgRpNPRVMZmzy5fJW/3flfnPA6jUhTXMVipVIb/evqtvrYLp5tk4fOyd+HFpFuW5i2d7LYZNEgJs3UjqsDG/hrDzBKZ0yUvpOvSTyvd3LlGoOQd+Qbe8N8MHfDlKYXZedT97DroaxOPwwiWSeReg5IlL4ZkolghhBBxtGhRkK2LVzhmz47mLn5BieBVDyNRF7AM5QmwFizwNz4sJ+AoezCYyAlS/Bw9zNaNqFqYurv9JeQoJS4tTIlEuOcu3bI/w30XlFwvrGP29BiMZ8zP+UPPnRrKPoOQgCmPU/pwGcMkhBCikgIFTMZk60Eh9i55cbibW8q2k//KSGeA6TXcdMkLQX09zJ3rfbv2dmgOsXXAHXvSh2CV3jAr6lEFTLUR5YkOq0taGHKzNJbnmL294eynrg7mdDaw9LLn+ei//SCcnQYgAZPNprVrGNu9e8ryRG2dTBQrhBCionIDpvDmYQKYPdvf7sLh/bUYde2MzTkpgrKEb/5874FPIgHd3WVs9QOMZKP5R7Iup/WFoC1kAUXVJc8uzKQCLc3RZN/zI6pgs5gFveF1JU2lMry4ZSnDY00h7DMYCZhsXlx9G6THpyyvaW6SiWJFjk1r1/DI2StZe8rxPHL2SjatXVPpIgkhqtz8+eaA+iiUcwxTWPYuvJgvPXZ3pYtRUjbw8Vp59hNoZfkJAHYv/Rx7Fl8KC8+ithZ+9MQHfR9/OnTJA7jm4YfC32nC8YFr6abFYZSkLMFmvlRfiPuyrqM4TL0mAZON0zglmYNJ2G1au4Znbl7F8JbNYBgMb9nMMzevkqBJCBGpRAIaG6dn0ocdx6y1PQqp9pNIsrvhSKDUCK3K8hv4BAmYJrk/10ZdB7uX3gBJs69ac7alxMcYqnC75EX37tZ2LATi1w3VSIbTolKJFqZUKC1MpqgmF/ZDAiYbmYNJuPHi6tvIjAznLMuMDJstlEIIEaHGBr9bFq/ENDZGewt3vONYdu9zOQCZ+s7cJwPcPu7pMV+X52xkZbxlPd9nFSKcLnn+K6/NzdYfPoIJe6a0oKe6qyu6YKanx/w/3IBp8gUPjzcXWc9Zpt7HwLcCogw2nfSGMG7KmOiSF59AVgImG5mDSbghmRSFEJXS1OS3El086cNI52k+9+venn0/zc4j/oOxOSeHts9swORZmZI+gP/Ap3u+v658frfJ19Li/xzVF5/b1JMoY9tu7bxxcQAAF3hJREFUv9dPCbvHzYDnR0/9Uyj783sOKtFC09lZeh23KtFC5kQCJhuZg0m4IS2RQohKabTG5fuvEBfezmgow/dXso7RzreFusuenmgrVLv3+VTgfVS2S55/2RamGAwfiUyv1RIWZtKHdPO+kDTTYY9lfDYJhxTQ5yTvKJNEiBFuJcZgOZGAycax5WDLZhmfIiYsOf9Ckg2NOcuSDY2SSVEIETnfY5jiU++YKsA8Qz0RZ5Lbs+QzvP6WbYH24TfwCdIlL4wAoNlfb7Jppbsn/H2OdL+PmhqfXUUnhPOBrakJZTcehRAwJSaz5MVFRFnop6fGefPNgfwFPHPzKgDJloeZ9ODF1bcx/PoWGufNZ8n5F86o85J9rTP5HAhRbZRS+wN3AHOBbcB5WusX8tapAb4GrMCs0azSWq8uZzkbGw1Il/OI8dYUtFLv4m540DvmTc3+tg8nEYf/sre0GLArhCLEWIPVABTqGKZEgpok1ue0mtvnotfUBJ2dGUZHK38epYXJplDLQZYM6jdJhjhTz/IVnHjPfSz/9f9y4j33SbAkxPR3G/ANrfX+wDeAbxVY5wPAEmApcDxwvVJqcdlKyGSXPJHL9538soxl8neMRLKylcSZ0MIU1QCppNWy47frbH4LYRmH3E2RbkyxZ+HF7jcI+ZzGpVueBEw2PctXcNDlVzo+L4P6JUOcEKL6KKXmAUcC2Ul97gaOVEp15a16NvAdrXVGa90P3AecVb6SQkN98eQNzvxuF71KT+5ZLv67Z1VGuuvNAPx1+7LKFqQMwujCuP24R9l+/O9D2l/u9tu2Ve7a2X7iM+xWq8p81MnXG5dueRIw5elZvoLG+d0Fn5NB/ZIhTghRlfqADVrrNID1/0Zrud1C4FXb49cKrBOpbNehffbxWSFzcff3iCPKW0E57LAgfQwDVkzLmF58Yl4j18z1W1t9vMYQmiRGFpzDGd/bzOvjRwTaT0tL4KJMCD/NtPn+14WQ1S/ddijpVmU9Msu5bFk4WRw7O4O/7qZwpnbyZPHi4N8lRx+dDmU/QckYpgKWnH8hz9y8KqclRQb1m5zGeUkwKYQQpc2d2xpo+/p682f7ve9tgK429xvWmhXD2bNbYE7x7U46KVgNt8tLuYC3vrUFanz2Ndxlns8FvQnqHY5bsDz1Zp+pjvamqefxzBfN7dqs5ZnJgM7rawOgzjyfixbh8T0zt0ulkt62Axg1+9PV19f5KnN2m9vvbiOZhDlz6jzvI+vcc5uoDam2+e53h9xPcNis57XPMryf4zw559mKw6+9rgka/Vwzue0Zb3lL8Kjz4IODffe4ZT8Py5YVP6bTtdnV2Qa15nt9zTVw7bUAwd6foCRgcpBsqJ8ImOpmtaMuuVTGqSDBpBCiKq0DepVSNVrrtJXcYYG13O41YBHwe+txfotTSdu27SKT8Xe3uKurjdHRceqBgZ17GKsdcr1tx3iGOmDHjj2Mpwtvl+1/2N/vfr9+t7f3dezfuguSY76Omdy7i7lATTJT8LhdXW0Fl7ePps3zOLCHsZr85+eZ/w1by410oHNTs2s3c4DxcYMdHravGcpul/G0HUD9zj20AyOj4wx63Db/nKXT0N/vaRfmfqz/d+zwdz0V2pffa9NJYnQ3nUAmk2FbgH3nn7O5mTRJYOu23Rj13vfbMZbGHqIGed1Rnbtix3FzzEKfzYnttu6Cmmiy2ySTCV83rqRLXp5sUoPxwcGJZemRkQqWKF6y47wa53dDIkHj/G4OuvxKCSaFENOW1vp14EngHGvROcAfrXFKdvcCFyilktb4ppXAf5SvpNUjUz+v0kUwlbFLnnfxGOxezYy6Oexd8EF2HvGjsPds/V/ZtOIiPNLClEff+hXHpAYSFJh6lq+QcyGEqDYXAncopa4FdgDnASilHgCu1Vo/DnwfOBbIphv/rNb65UoU1rt4VcC2n/AHOh8Kc/hXnAMfEVuJBLsO/mb4+82OQfIZkGfq5oZYGBEGCZhsNq1dw9jgzoLPSVIDIYSoXlrr5zCDofzlp9v+TgMXlbNczvwGCPEILIy6dtujeJTJWaXKF+S48QqQZ55gLUxDy25jdPOPaHvuE+EVSQQiXfJsiqXGDiupwaa1a3jk7JWsPeV4Hjl75Yybv0gIIUQQPuf0qeRELlGq1tcFSNAznWXfO3/VbKNuDsN9F4RXnGknfjdRpIXJplgrUhhJDbLjo7Jd/rKTvgLSxU0IIUTkjFiO2QmhTLF8XZbAQZ2f1xbj8zETGFYa7Dhfl8ITaWGycWpFqp01K5SARiZ9FUIIURlV3lpR1S1NYrrJTlxrSDW7asg7abPk/AtJNuTOBZFsaOSASy4LZf8y6Wth0k1RCCG8mt5jmCrPS4AV8JxVpJWhugLIkc4VpOu7K10M13Yt/bz5R00FZoutBjFsmZMueTbZVqQXV9/G8OtbaJw3nyXnXxhadzmZ9HUq6aYohBBeRFcR3nHMWmoH/xjZ/p1Jl7xCQhl3Fufz4sHgEf9e6SJ4Mtz3UYb7PlrpYogQSQtTGTm1YM3kSV+lm6IQQvjhtSJcuvI93nEswwtnyu9RBQKJKglehJiJpIXJJurWjqhbsKYj6aYohBDuZRpTABi13meqN1VbpX0adT3z2GIUzwQdQpRD/K59CZhsirV2hBXUyKSvuUp1U9y0do0EmEIIYRk64GZG557KePvRHreMcWAR98CgQuWr2lTwQkxD0iXPxrG1o0CFXoSjWDfFbIvf8JbNYBgTLX6SFEIIMWPVtDDS/TcBdhDz4KRsKhCM+Ay8/LU0SbAlRJgkYLIplnxBKunR6Fm+goMuv5LG+d2QSNA4v5uDLr+SnuUrZHyTEEIIEYgEyGI6it91K13ybDqPO4H1P/1xwefC7JYncjl1U5TxTUIIMRNUonIUvwrZVNJKJERcSAuTzdbHfuv4nFTSy8+pxW8mp2EXQghfZDxMnul0PqZDcCdErkzd3EoXIVQSMNkUC4qkkl5+koZdCCHCYgUIcU+w4Nl0Cnz8mgmvUVSTradsYNuJTwfYQ/y+p6RLno1TxjZAKukVIGnYhRAibPGriIRTJq/7KOd5kIBHzCxGbVulixA6CZhslpx/Yc48TFmpd71HKukVImnYhRAiDNVeaff6+qbT+YhjkCvEzCIBk420aAghhBAiXBLwCOFJDLsOS8CUR1o0hBBCiDIKpXIUvwrWpOnUmiWEKESSPgghhBAzRpwDi0oo5/nweiwJtISICwmYhBBCiKpXnZXvROB06eU8L36PFSSokwBZTEfxu24lYBJCCCGq3URgEb+KSBiMKn1dpuoMdoU7mbrZlS6CQMYwCSGEEDNHDAdThyFRkdabOB9LVIOtJ70IyYZKF0MgAZMQQgghRIxJoDVTGQ3zKl2EConfNS9d8oQQQggxrVV3lzwfAo/tEkLYxaKFSSl1O3AqsNVadK/W+sbKlUgIIYSoJlKBnnZCCXokkBQiDLEImCyrtNZfr3QhRLxsWrtGJhIWQojQVFsFejoEgkHLWG3vmRAlxHCsZZwCJiFybFq7hmduXkVmZBiA4S2beebmVQASNAkhhCfTIbAIIIYVrEnVnaFQiJkgTgHTZUqpvwdeAj6ttX7Ww7Y1AMlksC+joNtHLe7lg3DLuPFnP6Fp7pyCy3tPe7uvfcb9HMa9fBD/Mkr5gvNbRtt2NaEVpnpU9nequRcS4yRr6zGiugZbFgEeyuh1/QKStfXmfhoXOO6n4PKmbhhdRLK20d3xA5Q1Wdtgbt/U62n7ye26PR83UddsnZcuf2UO4xoJ4f2dTmL7Osv1Pvg4zpR1y1BWv79TCaMMAwOVUk8ACx2eng90A5u01hml1HnA54B9tdZpl4d4E/BI8JIKIYQIwYnAo5UuRMzI75QQQsSHp9+psgRMXimltgFHaq1fdblJA3AMsAlwG2QJIYQIVw3QA/weGKlwWeJGfqeEEKLyfP1OxaJLnlKqV2u9wfr7NMwfkw0edjGC3M0UQog4eKnSBYgp+Z0SQoh48Pw7FYuACbhDKTUfyACDwDu11uMVLpMQQgghhBBihotllzwhhBBCCCGEiINkpQsghBBCCCGEEHElAZMQQgghhBBCOJCASQghhBBCCCEcSMAkhBBCCCGEEA4kYBJCCCGEEEIIB3FJK14WSqn9gTuAucA24Dyt9Qt569QAXwNWAAawSmu9ukzlmwt8H9gPGAVeAP5ea92ft97twKnAVmvRvVrrG8tUxleAYesfwBVa61/mrdMMfA84ChgHLtda/7wMZVsM3Gdb1AHM0lrPyVvveuBiYKO16Dda63+IqEw3A38DLAYO0Vo/ZS0veS1a60V+PRYqo9tr0dr+diK8Houcw1cocS1a60V+PTqcw8W4uB6t7a8nomuy2HuplDoO+BbQBLwCfFBr/XqBfVTkMy1Mbr8vqp2f71O/z1ULv59/v89VC6XUfcA+mNPd7AIu0Vo/KddaaUqp64DrmfwtrIrrbKa1MN0GfENrvT/wDcw3It8HgCXAUuB44Hqr4lMOBnCT1lpprQ/BnFhrlcO6q7TWh1v/yhIs2bzXduwpFVTgcmBQa70EOBNYrZRqjbpQWutXbOU6HLOy+kOH1e+0rRtJsGS5DzgJeDVvuZtrEcpzPRYqo5drEaK9Hp3OIZS+FqE81+OUMnq8HiG6a7Lge6mUSgJ3Af9gXYcP4/weV+QzLSa4/b6odn6+T/0+Vy08f/79PldlPqS1PkxrfQRwM/Bda7lca0UopY4EjsP6jFbTdTZjAial1DzgSOBua9HdwJFKqa68Vc8GvqO1zlh30+8DzipHGbXW27XWD9kWPQYsKsexQ3Y21peBdQflceDt5SyAUqoeM9j4bql1o6S1flRrvc6+zMO1CGW4HguVMU7XYqHyeRT59ViqjJW8Hou8l0cBw1rrR63ltwHvc9hNxT/TM5XH74uq5vX71O9zUb+OcvL5+ff7XNXQWu+0PWwHMnKtFaeUasAMBi+yLa6a62zGBExAH7BBa50GsP7faC23W0ju3avXCqwTOSu6vgi432GVy5RSf1FK3aeUOrCMRQP4gVLqz0qpbyqlOgo8H4dz+E7M9/sJh+ffb72GXymlji9nwXB/LUIMzqWLaxEqdz2WuhYhBueQ0tcjlOGazHsvc86L1norkFRKTekymL8uFfpenKG8fF/MRMXOj9/nqpKHz7/f56qKUmq1Uuo14EbgQ8i1Vspngbu01q/YllXNdTaTAqbp5lbMfrNfL/Dc1cASq3n9x8Aaa6xLOZyotT4MOAZIOJQvDj6C893824B9tNaHAl8Cfmr18xaFFbsWoXLX43S5FqH49QjluyZLvZdCiOoln38PtNbna60XAldhfi8LB9ZNvqOBb1a6LFGZSQHTOqA3W5Gz/l9gLbd7jdyuRwsLrBMpa2DrUuBsrXUm/3mt9Ybscq31nUArkCpH2bLdIbTWI5gfjDcWWK2i51Ap1QucDPyg0PNa681a6zHr77VW2ZaVq3y4vxah8uey6LUIlbseXV6LUPlzWPR6hPJckwXey5zzopTqBDJa6+0FNq/49+IM5uX7YiYqdn78Pld1PH7+/T5XlbTW3wdOAdYj15qTk4EDgZethEwp4JeYY7Cr4jqbMQGTNjNrPAmcYy06B/ijnpr1617gAqVU0upfuhL4j3KVUyn1Bcy+myutimChdXptf58GpIENZShbi1Kq3fo7Abwf85zmuxf4e2u9pZgtAGuiLp/Nh4BfaK23FXoy7/wdjplxSZenaJ6uRajg9ejmWrTWK/v16OFahJhfjxD9NenwXv4BaFJKvcl6fCHmuSqk0udwxvL4fTHjFDs/fp8rX+nLw8fn3+9zVUEp1aqU6rM9PhPYDsi15kBrvUprvUBrvVhrvRgzuDwNs2WuKq6zhGEYlTx+WSmlDsBM6zgb2IGZ1lErpR4ArtVaP25F/l8H3mZt9kWt9bfLVL6DgaeA54G91uKXtdbvVko9CZyutd6olPpPYD5mustB4JNa68fKUL59gR8BNda/Z4CPaa035ZWvBbgdOAKz8vwprfVPoy6frZzPW+VaY1tmf4/vwPzxSGOmWb1Oa/1ARGX5GvAeoBsz7fY2rfXBTtdigbJGfj0WKiPm4MqC16K1TdmuR4fynYnDtVigfJFfj07vs/XclOvRWl6Wa7LE98oJmMkcGplM27rF2i42n+mZrtj3xUzi8/vU13PVIsDn39dz1UApNR/4KdCC+X23HXMqhSfkWnPHamU6Q5tpxaviOptRAZMQQgghhBBCeDFjuuQJIYQQQgghhFcSMAkhhBBCCCGEAwmYhBBCCCGEEMKBBExCCCGEEEII4UACJiGEEEIIIYRwIAGTEFVCKWUopZZUuhxCCCHiTSl1m1LqGp/bPqiU+pDPba9XSt3lZ1shKqm20gUQYqay5ik4X2v9nxUuihBCiGkijN8OrfWFAbZ9u99thZiupIVJCCGEEKJKKKXkZrgQIZMPlRAhUEpdCVwAzAPWAVdrrX9iPXcBcBmQsp77IHApsBD4mVIqDXwW+B1wl9Y6ZdvvK1h3EpVSbwD+BTgQc8b2HwGXaa1Hy/EahRBCVJZS6vtM/e34d+Bl4HzgOuAV4CSl1L3AiUAT8CfgIq3109Z+bgfWa60/o5R6M3AX8BXgCiANXKW1/p5DGR7C/K1arZT6O+u4jwEfBQaAi7XWD1rr7gPcDhxpraPz9nUccAtwEPAq8E9a64eUUnOAP1tl/plSqhV4Evis1vpOn6dPCN+khUmIcLyE+cPUDtwA3KWU6lFKnQVcD5wHzALeCWzTWv8t8Bpwpta6VWt9k4tjpDEDrU7geOCtwMVhvxAhhBDxVOK342TMG2qnWY8fBJZi3sh7AvhBkV13Y/5+9WIGPt9QSs12WaxjMQOhTuAm4N+UUgnruR8Cf7Ce+xwwMfZJKdUL/AL4PDAHuBz4kVKqS2u9HfgI8B2l1DzMYO5JCZZEpUgLkxAh0Frfa3t4j1Lq08AbMO+83aS1/r313IsBjvEH28NXlFLfwvyB/KrffQohhKga12utd2cfaK2/m/1bKXU9sEMp1a613llg2zHM1ptx4AGl1C5AYbYKlfKq1vo71nHuAL4JzFdK1QPHAKdqrUeAh5VSP7Nt90HgAa31A9bjtUqpx4HTgTu01r+yWsn+CzOgOtRFWYSIhARMQoRAKXUeZre7xdaiVsw7an2YrU9hHGN/zK4LRwPNmJ/fPxTdSAghxEyxLvuHUqoGuBE4C+gCMtZTnUChgGmbFSxl7cH8HXNjc/YPrfUepRRM/gbusAdxmN3u+qy/FwFnKaXOtD1fB/za9vjbwD8CX9Bab3NZHiFCJ13yhAhIKbUI+A7ml/pcrXUH8BSQwPwB289hUyPv8W7MQCi73xrMH7qsfwWeA5ZqrWcBV1nHEEIIMXPk/3YUWn4u8C7gVMyudout5eX8zdgEzFZKtdiWLbT9vQ74vta6w/avRWu9CiZ+A78N3AlcLNNmiEqSFiYhgmvB/KHqB1BKfRhYZj23GrhFKfUoZh/y/YAxrfWrwBZgX9t+ngcalVLvAH6FGRA12J5vAwaBXUqpA4CLsscUQggxY+T/dhTSBowA2zBvxH0h6kLl01q/anWxu0EpdRVmN/UzgfutVe4Cfq+UOg34T8zWpeOAF7XW6zF/Aw3MsUxXAHcqpU7UWqfL/FKEkBYmIYLSWj8DfBn4X8wfskOA31jP3YvZLeKHwBBwH2ZfbIB/Bj6jlBpQSl1u9Su/GDPI2oDZ4rTedqjLMe8aDmG2aN0T7SsTQggRQzm/HQ7r3InZ/W0D8AzuxiJF4VzMpBDbMTP4TSRt0Fqvw2wFuwrz5t864JNAUil1FGY39/OsAOmLmMHTlWUtvRCWhGE4tewKIYQQQgghxMwmLUxCCCGEEEII4UACJiGEEEIIIYRwIAGTEEIIIYQQQjiQgEkIIYQQQgghHEjAJIQQQgghhBAOJGASQgghhBBCCAcSMAkhhBBCCCGEAwmYhBBCCCGEEMKBBExCCCGEEEII4eD/AzPkcw04SrO6AAAAAElFTkSuQmCC\n"},"metadata":{}}]},{"metadata":{"_uuid":"37fc4cff531998c5bfae84772795e8c01e607fbe"},"cell_type":"markdown","source":"There are some negative predictions when using a linear model. We can try to change negative values for zeros:"},{"metadata":{"trusted":true,"_uuid":"97bbc24d95919791d48766abe1f23f11eea78597"},"cell_type":"code","source":"ridge_oof[ridge_oof < 0] = 0\nprint(\"Mean error: {:.4f}\".format(mean_absolute_error(target, ridge_oof)))","execution_count":12,"outputs":[{"output_type":"stream","text":"Mean error: 2.1309\n","name":"stdout"}]},{"metadata":{"_uuid":"7c9f936d7866fea46f5d83401b0e1882291c4d63"},"cell_type":"markdown","source":"### Kernel Ridge\n\nThis model combines regularized linear regression with a given kernel (radial basis in this case)."},{"metadata":{"trusted":true,"_uuid":"f0b7e94dda285927518b226b716c2966e9def90f"},"cell_type":"code","source":"grid = {'model__gamma': np.linspace(1e-8, 1, 100),\n        'model__alpha': np.linspace(1e-6, 2, 200)}\nkr_pipe = make_pipeline(KernelRidge(kernel='rbf'))\nkr_pipe = search_cv(X_train, target, kr_pipe, grid, max_iter=60)\nkr_oof = make_predictions(X_train, target, kr_pipe)","execution_count":13,"outputs":[{"output_type":"stream","text":"Best CV score: 2.0336, time: 1421.1s\n{'model__gamma': 0.15151515999999998, 'model__alpha': 1.678392120603015}\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x432 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"_uuid":"7738cdc47cc603b645b35aa8d931985888fa6d77"},"cell_type":"markdown","source":"### SVM\nSupport vector machine with radial basis function kernel."},{"metadata":{"trusted":true,"_uuid":"9bb9c43c3a5de20552fbb2a6c7d91e81ddb55e1e"},"cell_type":"code","source":"grid = {'model__epsilon': np.linspace(0.001, 1, 100),\n        'model__C': np.linspace(0.01, 10, 100)}\nsvm_pipe = make_pipeline(SVR(kernel='rbf', gamma='scale'))\nsvm_pipe = search_cv(X_train, target, svm_pipe, grid, max_iter=60)\nsvm_oof = make_predictions(X_train, target, svm_pipe)","execution_count":14,"outputs":[{"output_type":"stream","text":"Best CV score: 2.0108, time: 969.3s\n{'model__epsilon': 0.29363636363636364, 'model__C': 4.449999999999999}\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x432 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"_uuid":"7cc3b80df2183655437dfad7607ecd45acfd02f7"},"cell_type":"markdown","source":"### Random Forests\n\nThis regressor fits several decision trees with a different subset of the original data for each tree. Predictions are the average between trees."},{"metadata":{"trusted":true,"_uuid":"f293eae9eac71f20a299b44270010eaab1f94a60"},"cell_type":"code","source":"grid = {\n    'model__max_depth': [4, 6, 8, 10, 12],\n    'model__max_features': ['auto', 'sqrt', 'log2'],\n    'model__min_samples_leaf': [2, 4, 8, 12, 14, 16, 20],\n    'model__min_samples_split': [2, 4, 6, 8, 12, 16, 20],\n}\nrf_pipe = make_pipeline(RandomForestRegressor(criterion='mae', n_estimators=40))\nrf_pipe = search_cv(X_train, target, rf_pipe, grid, max_iter=10)\nrf_oof = make_predictions(X_train, target, rf_pipe)","execution_count":15,"outputs":[{"output_type":"stream","text":"Best CV score: 2.0150, time: 1200.2s\n{'model__min_samples_split': 8, 'model__min_samples_leaf': 14, 'model__max_features': 'auto', 'model__max_depth': 6}\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x432 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"_uuid":"44aa21f6b7a41bcb47aef88d83c7892614700dc4"},"cell_type":"markdown","source":"### Ada Boost\n\nAdaBoost is a sequential ensemble model which begins by fitting a base estimator on the original dataset and then fits additional copies on the same dataset. At each iteration (estimator), the weights of instances are adjusted according to the error of the last prediction. It's similar to the next model, but gradient boosting fits additional estimators on the current pseudo-error and not on the original target."},{"metadata":{"trusted":true,"_uuid":"146acb052cc6bf6bb448b2244b52ec8f92036ab2"},"cell_type":"code","source":"grid = {'model__learning_rate': np.linspace(1e-5, 0.9, 100)}\nbase = Ridge(alpha=2)\n\nada_pipe = make_pipeline(AdaBoostRegressor(base_estimator=base, n_estimators=200))\nada_pipe = search_cv(X_train, target, ada_pipe, grid)\nada_oof = make_predictions(X_train, target, ada_pipe)","execution_count":17,"outputs":[{"output_type":"stream","text":"Best CV score: 2.1471, time: 78.5s\n{'model__learning_rate': 1e-05}\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x432 with 2 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAA0wAAAF/CAYAAAB67EmdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xl8HHX9+PHXTHZzJ20a0hzdpRRKP4Kl3AgqAkprAUFAEeHrwRdR8ADBHyqIByAgInJ6gCCooHh9sUDFQhCKRazcLeWYUihtkzRpmp45Nslm5vfHZpNNsvfO7Mwm7+fjUdjs7nzms7Ozu5/3fD6f90ezLAshhBBCCCGEEBPpbldACCGEEEIIIbxKAiYhhBBCCCGESEACJiGEEEIIIYRIQAImIYQQQgghhEhAAiYhhBBCCCGESEACJiGEEEIIIYRIQAImIYQQQgghhEhAAiYhhBBCCCGESMCXz50ppWqB+4B9gAHgLeB8wzA6lVJHAncCZcC7wGcMw9iSz/oJIYQQQgghRKx89zBZwA2GYSjDMA4A3gauV0rpwP3AVw3DmAf8C7g+z3UTQgghhBBCiDHyGjAZhrHNMIzlMXetBGYDhwIhwzCeGb7/DuBT+aybEEIIIYQQQoynWZblyo6He5UeBx4GWoFzDcM4KebxXiBgGMa2NIorAQ4HNgNDDlRXCCFEakVAI/A80O9yXbxGfqeEEMJ9Wf1O5XUO0zi3A93Az4DTcizrcGBFzjUSQghhh6OBZ1I+a2qR3ykhhPCOjH6nXAmYlFI3AvsCJxuGYSqlNhIZmhd9fA/ATLN3CSJX7Ni+vQfTzK7HrLa2kq6u7qy2zQev1w+8X0epX+68XkepX+YeesjHNdeUYJrwmc8McM01JVnXUdc1amoqYPg7WYwx6X+nvEiOWebkmGVOjlnm3Dpm2f5O5T1gUkpdR2TO0kmGYUS7wl4EypRSHxyex3QB8JcMih0CME0r6x+i6PZe5vX6gffrKPXLndfrKPVL3yWXlPD73/spKoI77+zjlFOGgBI76ihDziaaEr9TXiTHLHNyzDInxyxzLh+zjH6n8p1W/L3A5cBa4FmlFMB6wzBOU0p9FrhTKVXKcFrxfNZNCCGmit5eOOmkcl57rYhp00yWLethn33crpUQQgjhTXkNmAzDeA3QEjz2LHBAPusjhBBTzbp1GiecUM7OnToHHDDE3//eS2mp27Vy3vBQ8E8AewEHGIaxRim1F7Ak5mnTgWrDMGbE2f5K4CtA2/Bd/zYM46tO1lkIIYQ3uJn0QQghRB499JCPCy4oZWgoMl/pppumVCK7JcCtxCReMAzjXeCg6N9KqVtI/rv4O8MwLnWqgkIIIbxJAiYhhJgCvvvdEn71Kz+6DrfdFuLTnw67XaW8iq7zNzwUfAKlVDHwP8BH81gtIYQQBUACJiGEmMQGBuDUU8t44QUflZUWjzzSy3vfa7pdLS86BWg1DOOlJM/5tFJqEdAO/MAwjP/kp2pCCCHcJAGTEEJMUps2waJFFXR16cydO8SyZb1UV7tdK886F7gnyeN3ANcahjGolFoIPKSU2s8wjK5MdlJbW5lLHamrq8pp+6lIjlnm5JhlTo5Z5grpmEnAJIQQk9ATTxTx+c+XMTgIp502yJ13htyukmcppWYBxwCfTfQcwzDaY243K6U2AfOBpzPZV1dXd9apdOvqqujs3J3VtlOVHLPMyTHLnByzzLl1zHRdy+rCle5AXYQQQrjoxz/2c/bZZYTDcN11/RIspfZ54O/JeouGg6ro7YOIZNsznK+aEEIIt0kPkxBCTBKmCWecUcaKFUWUlcGDD/Zy6KEyXwlAKXUbcDrQADyhlOoyDOO9ww+fA1wUZ5tHge8bhvECcJ1S6lAiix0OAJ+N7XUSQggxeUnAJIQQk0BHByxcWEF7u04waPLEEz3U1LhdK+8wDOMi4gRFw4/NS3D/iTG3P+9Q1YQQQnicDMkTQogC9+9/6xx2WCXt7RqLFg3y/PMSLAkhhBB2kYBJCCEK2O23+zn99HIGBuA73+nn/vtD6PLNLoQQQthGhuQJIUQBMk0455xSli3zUVwMDzzQy9FHy3wlIYQQwm4SMAkhRIHZvj2yvtKGDTr19SbNzT00NLhdKyGEEGJykoEbQghRQF5+WeeggyrZsEHnAx8I8/LLEiwJIYQQTpKASQghCsS99/pZvLicvj74+tf7+dvf+vDJOAEhhBDCUfJTK4QQBeArXynlr3/14fPBPff0sXjxkNtVEkIIIaYECZiEEMLDurth8eJy1q4tYsYMk8ce62H2bLdrJYQQQkwdMiRPCCE86vXXdQ48sJK1a4s45JAwq1dLsCSEEELkmwRMQgjhQX/5i48Pf7ic3bvhC18YYNmyPoqL3a6VEEIIMfXIkDwhhPCYb32rhN/8xk9REdxxR4jTTw+7XSUhhBBiypKASQghPCIUgpNPLmfVqiKqqy0efbSHefMst6slhBBCTGkSMAkhhAesXw+LF1ewfbvO/vsP8eijvZSXu10rIYQQQsgcJiGEcNnf/17EBz5QyfbtGmeeOcjy5RIsCSGEEF4hPUxCCOGiq64q5uc/L0bX4cYbQ3zuczJfSQghhPASCZiEEMIF4TCcdloZ//2vj4oKi4ce6mXBAtPtagkhhBBiHAmYhBAiz9raYNGiCrZs0ZkzJ7IY7fTpbtdKCCGEEPHIHCYhhMijp57SOeKISrZs0fjYxwb5z38kWBJCxPfiizp33eV3uxpCTHkSMAkhRJ5ccw2ceWY5g4Nw1VX93HNPCF2+hYUQCZxwQgVXXFHK0JDbNRFiapMheUII4TDThLPPLuPJJ6G0FP78516OPFLmKwkhkvP5LMJhjY4OjaYmWZNNCLfItU0hhHBQVxccdlgFTz7pIxiEF1/slmBJCJGWQCASJG3aJM01Idwkn0AhhHDIc8/pHHxwJS0tGscdF+bdd6Guzu1aCSEKRTAYubjS0qK5XBMhpjYJmIQQwgF33unn5JPLCYXgm9/s509/6pP5SkKIjER7mFpa5MtDCDfJHCYhhLCRacIXv1jKI4/48Pvh/vt7Oe44GYInhMhceXl0SJ70MAnhJgmYhBDCJrt2waJF5bzzThF1dSbNzT00NbldKyFEoZMeJiHcJZ9AIYSwwerVOgceWMk77xTxvveFWbVKgiUhhD1kDpMQ7pKASQghcnT//T4WLSqnpwe+/OV+HnmkD5/03wshbNLSomNJVnEhXCM/6UIIkYOvf72EBx7wU1QE99zTx8c+JitMCiHs1dursW2bRm2tRE1CuEECJiGEyEJvL5x4Yjmvv17E9Okm//hHD/vs43athBCTVUuLBExCuEWG5AkhRIbWrtVYsKCC118vYsGCIVavlmBJCOEsWbxWCPfIp08IITLwt7/5OOaYCnbt0vjc5wZ44oleSkvdrpUQYrKTxA9CuEeG5AkhRJquuKKEu+7yo+vws5+F+NSnwm5XSQgxiVkW1NRY9PdLanEh3CQBkxBCpDAwAKecUsZLL/morLRYurSX/feXxWgLiVLqRuATwF7AAYZhrBm+/10gNPwP4NuGYTwWZ/ty4F7gUCAMXGoYxlLHKy6mPE2zCAYtWbxWCBdJwCSEEEls2gSLFlXQ1aUzb94Qy5b1Ulnpdq1EFpYAtwIr4jz2yWgAlcSlwC7DMOYqpfYFViil5hqG0W13RYUYLxCwpIdJCBfJp08IIRJ4/PEijjyykq4ujdNPH+SZZyRYKlSGYTxjGMamHIo4E7hzuKy3gBeAE+yomxCpBAKmBExCuEh6mIQQIo7rrivmlluK0TS4/vp+zj130O0qCef8XimlAc8A3zEMY0ec5+wJbIj5eyMQzEflhAgGLbZv1+juRi7aCOECCZiEECJGOAyf+lQZzzzjo6zM4sEHezn0UJmvNIkdbRjGJqVUCXAL8DPgM07trLY2t9ZuXV2VTTWZOgr5mJWVgabB/vuXANDbW8WcOc7vt5CPmVvkmGWukI6ZBExCCDGsowMWLqygvV1nzz1Nmpt7qKlxu1bCSdFheoZh9CulfgE8nOCpG4HZQOfw33sCT2W6v66ubkwzu8VH6+qq6OzcndW2U1WhH7O+vhIsy0d1dR9QwerVvcycOeToPgv9mLlBjlnm3Dpmuq5ldeFKBsQKIQSwYoXOYYdV0t6u8dGPDvLccxIsTXZKqQql1LTh2xrwaeCVBE//C3D+8HP3BQ4HluWjnkIEg5EgWxavFcId8skTQkx5t93m55OfLGdgAL73vX7uuy+ELt+Ok4pS6jalVAsQAJ5QSr0G1APLlVKrgTXAPOArMdu8opRqGv7zJ8B0pdQ6YCnwJcMw5JKyyIv6egu/35LFa0VWBmUKbs5kSJ4QYsoyTfj850t57DEfJSXwhz/0cvTRMl9pMjIM4yLgojgPHZxkm4NibvcAZzhQNSFS0nVoapLU4iJzTz9dxBlnlPPYYz0cfLD8vmVLPnlCiClp+3Y44ogKHnvMT0ODxQsvdEuwJITwrGDQlCF5ImPPP18EwD/+IX0kuZBPnhBiynnxRZ2DDqpk40ado48O89JLPdTXu10rIYRILLJ4rQzJE5kJBCIXAiXYzo0cPSHElHLPPX5OPLGcvj64+OJ+/u//+vDJhTchhMcFAiYdHRoDA27XRBSSmppIwhAJtnMjzQQhxJRxwQWlPPigD78f7r23j0WLnE3PK4QQubBiMtAHgyaWpdHaqjFnTnap6cXUow3HSTL/LTcSMAkhJr3ubvjoR8t5660iamtNHn+8h2DQ7VoJIUT6AoFoT4HOnDlysUdkprVVAqZcyNETQkxqr7+us2BBJW+9VcShh4ZZtUqCJSFE4Yj2EETnosjQKiHyTwImIcSk9ec/+/jwh8vp7oYvfnGAf/yjj+Jit2slhBCZmzXLQtMsmbwvstbT43YNClfeh+QppW4EPgHsBRxgGMaa4fvfBULD/wC+bRjGY/munxBicrj00hJ+9zs/RUVw550hTjst7HaVhPC8d9/VaGqy5MKCBxUXRxawlbkoIlstLTpKyfIZ2XBjDtMS4FZgRZzHPhkNoIQQIhuhEJx0UjmvvlpEdbXJo4/2Mm+eTJAWIpX+fjjiiEpOPXWQX/0qlHoDkXeSWlzkoqVFQym3a1GY8n6ZwjCMZwzD2JTv/QohJr+33oIFCyp49dUi9t9/iNWreyRYEiJN4eFO2CVL/O5WRCQki9eKXMi5kz2vHbnfK6VWK6V+oZSa7nZlhBCFY+nSIvbfH3bs0DjrrAGWL++lvNztWgkhhH0CAZO2Ng1TRlWJLGzaJL2T2fJSWvGjDcPYpJQqAW4BfgZ8JpMCamsrc6pAXV1VTts7zev1A+/XUeqXOy/W8ZvfhBtvBF2Hu+7SOO+8YsCbkzC8ePzGK4Q6CjEVBQIWg4MaHR0ajY3Sey5Si13LS+a/Zc8zAVN0mJ5hGP1KqV8AD2daRldXN6aZ3RdIXV0VnZ27s9o2H7xeP/B+HaV+ufNaHcNhOO20Mv77Xx8VFRb/+pdGMLibzk63axaf145fPLnUUde1nC9cCW+wrNF01sI7gsFI19KmTRIwicxUVUmGxVx44sgppSqUUtOGb2vAp4FX3K2VEMLL2trgwAMr+O9/fey99xCrVnVzyCFu10qIyWH7drdrIOKJXbxWiEwEAqYkDMlB3j9xSqnblFItQAB4Qin1GlAPLFdKrQbWAPOAr+S7bkKIwvDUUzqHH15JZ6fGKacM8uyzvVRXu10rISYPaZB7U3TxWukpEJkKBi06OjQGBtyuSWHK+5A8wzAuAi6K89DB+a6LEKLw/OQnfn7ykxI0DX74w37OP3/Q7SoJMels2qSzYIFkFvCaykqoqbFk8r7IWCBgYlk+Wls15syR4ZyZ8swcJiGESMY04ayzynjqqSJKS+Gvf+3liCOkQSeEE2TojjdYcdq1kaFV0sMkMhPtnWxp0ZkzZ8jl2hQeCZiEEJ7X2QmLFlXQ2qoTCJg0N/dQW+t2rYSYvKRB7l2BgMnbb8v7IzITDEbnv8nFkGzIJ04I4WkrV+ocemglra0aH/lImBdekGBJCKfJkC/vCgYtWlr0uL1PQiTS1GSiaZIpL1ty1IQQnvXLX/r5+MfLCYXgssv6eeCBPnT51hLCcdLD5B3j07sHgya9vRrbtklQK9JXXAz19ZZ8trMkQ/KEEJ5jmnDeeaUsXeqjuBjuv7+XY4+V+UpCOGnsApfSGPeq0dTiGrW10s0k0hcIWPLZzpKEmUIIT9mxA446qoKlS/3MnGnx3HPdEiwJkUc1NRbbtun09LhdExHP6OK10oQTmQkGTTlvsiRHTQjhGatX6xx8cCXr1+sceWSYV17poanJ7VoJMbXEZtMS3jP6/khPgchMIGDS1qZhyjXIjMm3oRDCE373Ox+LFpXT0wNf/Wo/Dz/ch08GDQuRd9EeDGmQe1NNDZSXy1wUkZ7YobaBgMXgoEZHh3y2MyXNESGE6y68sJQ//cmHzwe//nUfJ50ka0QI4ZZo+uHI0B35LHqNpkWHVkmjV6Qvet5AJAtmY6PMf8uEBExCCNf09sIJJ5TzxhtF1NSYLFvWw5w5btdKiKlt5kwLv18mh3tZZPK+9DCJzIwmDNFl4fcMyadNCOGKtWs1Fiyo5I03ijjwwCFWrZJgSQgv0HWLpiZpkHtZIGDK+yMyJvMTsydHTAiRdw8+6OOYYyrYtQvOOWeA5uZeSkvdrpUQIkqyaXlbMGixfbtGd7fbNRGFpLIykgVThnNmTr4NhRB5dfnlJVxwQSmWBT//eYgbbuh3u0pCiHFkvRZvk54CkS3pncyOzGESQuTFwACcckoZL73ko6rK4pFHetl/fxlDLYQXBQIm7e0+BgaguNjt2kxdVoJ5+bGpxd/znjxWSBS8QMDknXckYMqUHDEhhOM2bIAFCyp46SUfSg2xalW3BEtCeFgwaGJZGm1t0svkRWMzGQqRvmDQYtMmPWEwLuKTT5oQwlHLlhVx1FGVbNum8clPDrJiRS+VlW7XSgiRTGw2LeE99fWSyVBkJxAw6e3V2L7d7ZoUFhmSJ4RwzLXXFnPrrcVoGvz4x/387/8Oul0lMYUppW4EPgHsBRxgGMYapVQtcB+wDzAAvAWcbxhGZ5ztfwMcD2wdvusvhmFcm4eq55WmjR3y5XXt7RorVhRxxhlht6uSN7qOZDIUWYm9GDJjhoz0SJcETEII24XDcMYZZfz73z7KyiyWLOnl4IPli1m4bglwK7Ai5j4LuMEwjOUASqmfANcDX0hQxvWGYfzMyUp6waxZFppmFcSQrwsuKOXZZ3184APdNDVNnXFGkslQZGN08VqdBQvkdzldEjAJIWzV3g4LF1bQ0aEze7bJ44/3UFPjdq2EAMMwngFQSsXetw1YHvO0lcCX81oxDyoujgz7KoQGeX9/pBds40adpqYhl2uTP4GAxfLl3n9/hLeM9jB5v/fYS+STJoSwzdNP6xx2WCUdHRonnDDIf/8rwZIoHEopnUiw9HCSp31DKfWqUmqJUmq/PFXNFYWSWnz0irn362qnQMCko0NjYMDtmggvs6yxn4sZMyzKy2U4Z6akh0kIYYtbbvFz3XUlaBp8//v9fO1rMl9JFJzbgW4g0ZC7K4DNhmGYSqnPAcuUUnsbhpF2t0ZtbW4ZT+rqqnLaPpno4tGVlaXU1ZUydy4895yz+7TDvHmR/2/fXkZd3cTHvV7/ZMrKIvOV4r2G/fePpB0PhaqYNcve/RbyMXOLV4/ZtGmR/8+YUTHy+Zg9G7ZsKaauzt01A7x6zOKRgEkIkRPThM9+tpTmZh8lJfDHP/bygQ/IuGhRWIYTQuwLnGwYRtwT2DCM1pjbv1NK3QwEgA3p7qerqxvTzG6eTV1dFZ2du7PaNh3d3QBVdHeH6OwcpK6umE2biuno6Eb38MXo4uJioATDGKCzc+xC2E4fM6f19ZVgmj46O3smPDZtWhFQzqpVvVRX2zcUsdCPmRu8fMx27vQBZWzf3kNnZ+SrrbGxjLff1ujs7HWtXm4dM13Xsrpw5eGvQCGE123fDocfXkFzs5/GRosXXuiWYEkUHKXUdcChwKmGYfQned6smNsfBYaA1kTPL3SBgMXgoEZHR2EMdSuE+VZ2KqRMhsJbAgFTzpsMSQ+TECIrzz+v88lPltPXBx/6UJg//7nP01ehhVBK3QacDjQATyiluoBPAZcDa4FnhxNCrDcM47ThbV4BTjQMow34rVKqHjCBXcAphmFM2lzWsXODGhu9n31uqjUACymTofCWYNBi2zadnh6oqHC7NoVBAiYhRMZ+/Ws/3/lOCZYFl1wywOWXy6xj4X2GYVwEXBTnoYQtbcMwDoq5fbwT9fKq2PVajjjC+z3HU20SezST4VR73SJ3o72TOkp5/7PtBRIwCSHSZprwpS+VsmSJD78ffvvbPo4/fuqk8RViKoltVBWCUEijuxsqc8urUVAKJZOh8JbY1OIxqyyIJArjW1AI4brubnjve2HJEj+1tRYrV3ZLsCTEJFZZCTU1VkGl655sw9OsFCMhZfFakY3YxWtFeuRICSFSeu01nQULKnnzTTj88DCrVvUQDLpdKyGE0yKTwwunqTDVelsCAZO2Ng1TRlWJDNTXW/h80juZicL5FhRCuOKPf/TxkY+U090NF18Mf/97H8XuLt0ghHBAtDdDi2lDFVo2ral2xbzQMhkKbygqgqYmmf+WCTlSQoiEvvGNEi66qBRNg7vuCnHzzW7XSAiRT8FgJAtbqqFhXlFIwZ0dYjMZCpEJGc6ZGTlSQogJQiH4yEfKuf/+YqZNs1ixooePf3zSZk8WQiQQCJj09mps3+52TVJrbCys4YN2iM1kKEQmJGFIZuQTJoQY4+234YADKnj11SLmzx9i1aoe5s4tkMvLQghbFVKDPBCYnGsSaUnatIWWyVDkX6Le4UDApL1dY0BWBUmLfMKEECMefriID36wkp07Nf7nfwZ48sleysvdrpUQwi2FlE0rGCys+VZ2KMRMhsIbgkETy9Joa5NzJx3e/wYUQuTFD35QzHnnlWFZcOutIW6+ud/tKgkhXBa7XovXBQImHR06/VPsq6vQMhkKbyiU3uNQCDZscP/7x9tHSQjhuHAYTjqpjF/+soSKCnjiiV7OOkvmKwnhVZZF3tJIz5hhUV5eGNm0og3A1lb3G1f5VGiZDIU3jA7n9Pa5841vlHL44ZX09rpbD+9/AwohHNPaCgsWVPD88z7mzh1i1apu5s+XBT2E8LK5cys54YT8jJXVtGg2LW83qmB0+GAhBHd2KrRMhsIbZs2KnDBeH277wgtFALS2ultPbx8lIYRj/vlPnSOOqGTrVo2Pf3yQZ57ppbra7VoJIVLZvVvj5ZeL8ra/SDYt7zcXCmn4oJ0KKZNhLsLhxAkMROZKSqC+3vvDOb2SOt/bR0kI4YgbbvBz1lnlhMNwzTX93HVXCF2+DYQQcXh9yFe0ET1rlommTc5MeckUylyUXIRC0NRUxQ03yKrpdiqE1OJeOb8n76dLCDGBacIZZ5Rx440llJbC0qW9fOlLg25XSwiRhVAoP/sJBi22bdPp6cnP/rLl90NDQ2H0htmpkDIZZiua+vqnPy1xtyKTTCEsXtvY6I25Vt4+SkII23R2wiGHVPD00z4CAYuXX+7m8MNlvpIQhcqpdMDj1/0ppLV+CuGKeSbSGYI2FYYiJluLSmQvEDBpa9PylkQmG35/5P9uB3be//YTQuTs2Wd1DjmkkrY2jeOPH+SFF3qorXW7VkKIXNjdgEi2wCUURoO8EK6Y262QMhkKbwkELAYGNLZs8f5n2+3vH/l0CTHJ/eIXfk47rZz+frj88n7+8AeZryREISsrc3ZMf6DkP9Q1V1O0ezUQGZIH7l/hTUf0ivnQkNs1yR9Ni7xutyfFi8LjlYQK6XD7goD3v/2EEFkxTTjnnFKuvLIEvx/+8pdeLrkkP/OVNjcvY8WZp9J83FGsOPNUNjcvy8t+hZgKGhujAYwzjRxV+TAAxV1PAlBfb+H3F8ZQt0DAIhzW6Ojwfl3tVCiZDO3g9bl0hcQrCRXS4faFEO8fISFExnbsgCOPrODRR/3MnGnxwgvdHHNMfgYpb25exus3Xk+oox0si1BHO6/feL0ETULYxO93uJFjDZc7PEZP16GpqTAa5FMhAUI8Xs9kaCe7z8PVq3WuvbZ40qYsT/a6CuXz8qH3PM1nP/g72tvdO8e9fYSEEBnZ3LyMe0/4JgeoIt59V+ew/bbwyis9NDTkrw7r7r4Ds39s+i6zP8S6u+/IXyWEmAKcaiBbRMsdvcgSDJps3Oj9JsNkTICQTsKDQslkaAe739vPfKaMW28tmfTrWMU7jyorYfp0b/ceWxY8/b1j+e0F57ga2Hn/208IkZbNzcu47dtruezFn9NvlXFy7V38v9IT6Hwqvz07oS0dGd0vhMhM9Iqxcz0+2vB/Ry9NF0L2uaqNNzKnJjLvqhB6w+xUSJkMc2V3o7mmJnKet7ZO/mMXT6R30oHXbg5Q11xN6Ub7Lpa6+R00Nc8OIXIwMj/n2CNp/vD7aT72SE/M07noG9P4dct30DG5JPA1Pl1/S8qeHSfmGpXOrM/ofiFEdpwa0z/Sw2TFBkwmHR3ayHo43mMx7d2raXr1OGbMmHoJECZjz1qs2GFldr/G6LHz+rA0pzg1nLN61dkAVBnfsq1MNy8ITM2zQ4gsjZmfA0QXL4idp5PvhAfd3fChD5XzdMciKot2cOPckzisevnI44l6dpyaazT3vAvQS0rH3KeXlDL3vAtyKlcIMVY4rDkypt8aaRqMtlKDQRPL0mhttXl/lsm0Fz8+kpEvW7o2PHzQHJhSCRCiCmUuih3sfm8LKW2+E4JBi02bdNvncJVsfdzeAnE3m5/PtT0L4aDNzctYd/cdhLZ0UDqznrnnXUDd2Wdktb2/qhoLi/Du3ZFBwAlWeDP7Qxi338xQf//IHJ5oEALQuHBx7i9snDff1DnppHJ279aYW/UG35v1aXx6eMxzEvXsJJtrlEtdo9uOP/5OvH4hpqrKSovubo1Nm3RmzbK3m6lc3zp8a+yQPIg0VufMsW9/pS33UrztKWas/CCdC3dlVYZlga5Hv5c1AgFTKPTpAAAgAElEQVSTdesmf+AQq77ewudzZthkactvKW37LTuOeNL2srNhd1A4Y4b0MPX0aOzYATU1btcmOTcvhEjAJCadaM/J+KClqrqM3bv6Ujbkx28/uGvn6IMpLsGMee4wO4KQeP7v/3x87WulDA3BOecMcMnCNbx+ow+zfzRgStazM9JLNv5+G+YaNS5cHPe4ShAlhD2CQZM33ihypIF86LRfR26MG5IH9l+F1we22FPOcA+ThkkwaLF8uf1XzKtXfQbf7tVs+2BuvWFOKCpyLpNh1RsXRm5YVnoZKBxm9zkYfUlTtYcp9mJITU1+sulmy833SAIm4UmZNK7HPzfc1xu35+TFH/2IcF8oZe9PvJ6XXNmd8ODb3y7h3nv9FBXBL38Z4v3VS0frretgmpTWNyQ8bsmG3dk512jkvRkXnDnd8ybEZDdrlsUbbzh9xXU04pg1y0LTLM9ehR8ZkkckuOvt1di2TWPmTPv2UbLl4Zy218K7mLFiPrsOvI/BGcfYVKtRwaDp6PtTsfYyetSPHSs/HZWVFu3tkbl0xcX2lj3VhnFGxQ7nPOAAZwImbaATq7gu53JaWnTX4vapeXYIT8tkbk2854Z3xR/WMbBjR9J019G5R4l6XnJhVxAyMACLFpVz773FVFXBU0/18v7qpXHnVe1x5PsTBiPJEkHYNddoc/My1lx/TcLjKanGhcjeUXOe4FD1lqNj+jWzb+R2cXFk2JfdjcqKd35kSzmjQ/K8mwChaPdr6OEdVLx9rSPlO53JsHzjLx0rO13RuXRtbfa/Tq+dL1Fl797CtJdOdaz8fHxe9nh6H1vKqfR3snWrO++T9DAJz3nz9pvSnltjR29QqKOd5mOPzKmMZOxKeLBhAyxaVMH27Tp7VqznB4Ez6PzBNNr7+uIeg5aHHmT6/AVxg6ZkPV7R5+c6D+zN22+CoXDS54yvh1PD9mQ4YOFTSmnAecBZwB6GYSxQSn0IaDAM48/u1i7/rj72BK4+Fj5896Bj+9DCu8f87eXU4kX66LwqzyZA0KILAjtzFT8QMGlv99nb++Kx1VwDgdGe1b32snfuXldXZB2rigpbi81Z5Vvft6Wc4qHNYNaCPrbpX1trUVbm3d7jWLNqWmlpmUddXf7PS+8fHYdFexX+MH++J1JDT3Wbm5cl7CGK17h2ojfIVrpO0+ITWXf3HVlnzdvcvIyffPCHHHl4Kdu3a3xw2hJ+NPtjlGp9hDra486bilpz7ZVx9+mvqo77fH/1tJF9xuvlW790acq6RjMEJnofY8X2vDmVtS9euWuuvZKnTlkkn/fCcjXwBeBXwJ7D97UA33atRh5Qp/3HtrImtI3H3eH0kK+inrVZb7v9V6Oz1b2a9cy3+1UA/Dufy2i78uIeSlvuSRm87Lmn/b0vubwnTnDivbUsqKvewrH7P+XIWkylm36FNtCZcznVqz6X9bbTy7fzgS37UrH28gmPaZpzqcXtVqQPuTZ0ckoHTE410ET2kg3RGt+4fu3HKYY16O6f3oGTT6Vt2aNZn2Nv3HwD371oBz9Z+2NMdM5tuIovz7oiozrE26dF/B/ewV07R3pj4vXyrbrlloT7Gf95Skf/zh00H3cUT52yiDXXXZV0yGS2EvVChnftGnNc8p0OXmTsHOBjhmH8kdHJNeuBvV2rkQfMKn3FsU6A8QkZAgGTtjYtUaLQnE1/7visty2KGZJXUwPl5c6lFtd73854m7J3b6XqzW9kvF2RNsj6H1dT9cbFFHc1J31u7OR9+3grCcCsWc5ktNvyy3qeuuLDdG7amvrJGSjqNqh681KqV/9vzmWVbFmS9bYzKrdFytj6WNzH7U7Fr/XbN2879vtN19xbY839FqWLkqVVFu5INlQsdljburvvwAqnGIri1K96mvTSMraufDbrc2zTP5Zx/i0n80jXFynR+rh6zqf5yIy/ZFUXsz/EmmuvHFloN1nvz5prr0zYc9fbnrhHL5vhkVYoBJYVqU+CVl+oo31MAJNpYJPsnIq+F3LxpCAUAd3Dt6MnS2XMfVNG7Eflk4c/QFeXfQ2IM4/848jtks6xPcqBgMXgoEZHhzMNFm2ox55ytGhvmDP1zGZ9mcq3vpfVvr54wHkjt7Vw8lPdqz1rFW/9INJDlgPLgmDtRk6fdQGzGvsdC4Z3tOfeExSrrOUuAPTBbbaWmwnLGh2yaiVo9tvdw1S62ZlR0lWVYelhckOihpTnh3lNYomSI/irp42Zc1II75EZ6ss6dXd7O3zkC8fyZt8RzPRv5PZ9j2PvstdtqFRuQWR5Q0PCx+zOBDim7OEA5o2bb8g4sEmVcCPU0S4XTwrDP4CblFIlMDKn6YfAI67WygWnHnjfyO33z/uPrQ2dP154VsLHRucGOdUgt6mrzLIcXrw2f/Mnjmj4a8xuk18kdCaTYe7vdfm7N1P1xsU5l/OrL3yJg6t/w+8vOMOxoLBzi73zoso2/Wr4lrtzwSpKIhcjfH3vxH08GLRG5nDZw97jGNVQH3btgkDeAyal1I1KqfVKKUspNT/m/nlKqf8opdYO/39fp+uSaB4HJE+7LJwz97wL0EtKJ9w/uGvn5BompWkJX8vTT+scekg5O8J7cFhVMz/d5wQqfO5fRNdLSmn60IcS9u7YmY48HrM/RMsjSzIObFIm3ND1xIFtAQTmU8glQCOwE5hGpGdpNmnOYcr1t0cpVaSU+rlS6m2l1Dql1HnxnpcPN5w2dnhPviZrOzPkK5Y9jcriLY8MXzF3pp6VxmWOlBuPaRWN/mElb4Q6lcnQK6zh4O2YuY84ds53dtgY2JgDIzd93a/ZV26GKtnAy9cdkvQ50YshTszhitL6c19z7Yh5a6bUkLwlwIeADePuvwP4uWEY84CfA3c6XZFE8zgg+Vwa4ZzGhYtpWnxi3Mcm1TAp02TN9dew/JSPjgQf65cu5eab/ZxxRjmDYY2zZ/6ES4IXe2EqFr7qapoWn8j6hx5K2LuTKNi1VYIeslRZ/3zViS+OYJpJ57tNivOtwA33Ju0BnEEk4cORwD6GYZxmGMbupBuPyvW353+AucC+wFHAlUqpvTJ5HU7J1xXX0SFfHvhSimGNG86rh3cQDFps367R7f61ppxYMQGTliJgAm9nMsyVZY2+rkpzrSOj7ndu60v9pHQ5lA0xU3sWJU/WBE6kFp9YTqbJTuL5+vu/ysn73p5zOdnI+7eeYRjPGIaxKfY+pdRM4BDggeG7HgAOUUrlvspVEuHdiX9n5cqye7aufDbhYyO9CV6IIjIRr75D4UiGO8uid3M7p3+6ih/9qAS/1s93Z5/DSXv8Nv/1TEBDY+vKZxkKJe7daVy4mP0vvczZ9yZB2Yl6t6LznZLN2Sqtb0g6VFEunrjPMAwLeBUwDcPYYhjG84ZhZPQlbcNvz5nAXYZhmIZhdBIJwNLPs++gfAUwlZVQU2M5d4U3y+wVx8+8eszf2lD3SHC3YXx4XGBMYnuYIss0+LuWo4da4z7f6UyGGRnqpa45ycWqDJnW6Ov68H6PsWWL/efhuYfak8IbcGd11TgsK/X5YH8q/njl2NN7997GF0nSfHeMRz5VBIFWwzCGAIb/3zZ8v2N8VVWJHyy0BvkkkipYDXW0EzjZuUXcbKfrSRsCu8LTuXjd47zScwwzfO3cuu/x7FfxYh4rmNrgrp1Jh62tOPNUmo89kjU/utrRZBs1Bx0yoRcr0TpXb9x8Q9IEFrHbJuuBcnJulsjIy8A8m8vM5LdnT8b2Tm1M8Ly8y+cQFSeHumlZZmQ7csa4TsGhfucDpjytTzR2SF4kYJr+0inU/OeouM+3P5Nh9ueW3r/ZrkoAYJpjzzu7zvuG0jUjt9+39zMMOre0mWfV11v4fDb2TsYLFm3qcTv1sCW0v2tvco50TKqFa2trKzN6vp4sKDJN6uqSBFQu8Fp94sm1jqnW+Yka6GjLaT95ZZqUNzbSu3nij8fa3gP50YZfM2CVckDFM3wreH5BxuojQYnDmQl73llHUWnJyDym4unTOfTyy5nzsY+NPKeuror1S5fS8tCDScsqb2zkwIsjE5HNvsTDMMobGmz97E2Fz7FDlgPLlFK/ATYRc7nSMIzcUnDlUaa/U+PFe2/a2/3U1flzKhfAH6eI8fvbZx9Yu9a5cySbcnvHtc0qK0s4aFZk9dENG+DEE+2va90e5RMWAM1o+zRfZ3fMde0qayNVw9vp4R1xy9hvPxgYgKGhKupzmFY6UnZxRfz707F77Lmeyznj94/OYYJI1rdduyqos2Ec0qX7Hzbm7/7+KpqaMi9nwusb8id/PAvZlNFaVgoxa8gnKiMYhM7OEurqSrKt3qiuicPzp1WXwrh9p/N6xi8kXFnaw/wtp1Bdtyb+Bg7xSsC0CZillCoyDGNIKVUENA3fn7aurm5MM/2rPgM7Ey/4WVrfQGenC31+CdTVVXmqPvHYUceXb7o5redtWbkyp/3kla7HDZaWdZ3N/R2XY6Fx2h6/4JMzf+FC5SLpz82QjeO2Y2marVdiB3bsGPv37m527+obOe+i52Cq80jz+zFNi/9cdlmkjgkCPb2klL3PPd+2z95k/xzrupZzQJDEB4isu3TMuPstINuAKZPfno1Ekkw8P/z3+B6ntGT6OxUr+t6MbyO++65FZ2fuk3V27oTp4+4bfy7MnFnC44/72bKl25YRR+NfS9fG1zDL9oz73EQqxh3O7u5+fLW78fsr2bBBs+/zG3O7c0sXFJVntS1MPK6JlMQOpzJupXPPH46UFa+M6dOLgHJeeaUHvz+7C1ix3wFFPb3MiK33ll1pDzXTe7upjd02h/dh586xPUxF+hCvvdZPZ+dAkq3SM/69WbWql8rKzLK8xf3eHAqNPWeyeP1Fu18be/yzKKO3bxBiYrdEZTQ1lbFuHXR25t4eKOvuZ/wvwc5dvQzE7Dvd35qqoYmLJ1ebr+X9d8oT17INw9gCvAJE85meBbw8PE7cMcmyeu1x5Pud3LVIYFIOfxrXGDdNuL39Vu7r+A46Yb4ZPN+dYEnTmH/FlRRPm+bcPsYFS77qagIfP92+8ofCvHn7TRPuTnUeWYODowvsJukV2//Sy8aksxfuMQzjuAT/PpxDmZn89vwF+KJSSh+e43Qq8Nc4z8u7HTucS25Qtv6nY/4OBEx6ezW2b3dmf9NfOCHnMvy7XkTXoanJcmxIXsW6H2b0/MGqBVntZ8yQvDQ4ncmw0vhm2s/Vxn3/57qYqR6zOHFVxaBjyS3sG+Ka+8XC8g12JDhI7/XYmoo/TlBt+WqyKuoj9T/KtTa2cCOt+G1KqRYgADyhlIrmWrwAuFAptRa4cPhvRyULipIlHhDOcTo1tdt6w+V8e8MyVm47nhnV/dx+0BkcVPVvdypjWSnn+NjNV1ZOx1P/tLXMeAkd7DiPSusbJFjykOFAJe6/NLfP+LdHKfWoUio6Xuc+4B3gLWAlcLVhGOttenk5c2qif+W6q8b87XSDXO/PZrj12IZpyZaHgchEdqcCJv/OFzJ6/mDt8bbsd48nZiR9PDp3y6nzoXTT3Rk8e1zAlEaWv2SOVitGbtdMd24BU7vKLX/3FhtKsWOERroBk0l7u2bTHK44AZM/u4DJ7TWsovI+JM8wjIuAi+Lc/ybwvnzWJVlQJFny3DH3vAt4/cbrJ6y1Mxm82zePH264j5BZyUHzOln6ZCnFxfezuXkZ6+6+g9CWDkpn1k/qcy9fry3X8yhRIgnhqjCJfzlTXobP5rfHMIwTY24PAV9Oq6YuaGnR2G+/3Mrw9b2d8jmx2bQWLLB/zmI2jepyX/zurkDA4l//yrVGiWT62u1p9GlWOOnj0UyGjqUWz2kcZvbni2WBro1uP326cwuY2lVuxTve6Bmxxgcv5gDoxROeFwyamKZGW5vG7Nm5na8T9jl8bzZi33c3eWUOkysmc8O00MQGDf6qatA05+bVuGD59tO4e/NVWOicMOM3fK7idrqejgz3iv7r/u/TvHDtdW5XNTM2z1HKhr964pDCaM+QcfvNkdTtGSitb2DueRdI75L3zBn3dyNwGfCIC3XxnEiPQm5X8Mt2px5ZYf96Lc4JBEw2b44kQSie2D7MTcYZv7L7ntSy2M7JTIaZDEya8NOQY5a0sDnaZFUNBps26ViW/dm7PZOWHbAj0LbGvWfaUC9WnIAptvd49uzcvkvi92pl91oqfd6YqjGlAyZ0Pen8hc3Ny6TRlIbxPSSZNjY3Ny8b0xuQaQPX637VdhVP7/gEOkNcOOsS3jetGbM/sr5P48LFbG5ellXDPq98fgjH6afPJFgq8sFQ8qujmdJ8fuqP+wgrzjyV0JYOyhsa2Pvc88cEom/cfEPKjHlRpfUNHP2nJbbWUdjDMIzxg6s2KKU+TyQJw69dqJJnFBfb1aOQuowZMyzKy22c6+AUyyIYNLEsaG3VmDMnt4an3jf29PPvehHC3eBLc/J4lheWynyZ/y4EAibr19v1/ow7J7T0y92+nTFJH3LpYQJo297EHlVdABxR/2d6ev7Ejh1Qk+1Irzj+vekTzg03DbVhlmaWfk8L556wZEJvT4JeytHhnDZ8l9iYVnxO5X9yrIw9kgZMSqm90ynEMIx37KlOnqVIgRxt0IrExgc7oY52Xr/xeoC0j926u++YlEPwQmYpV62/n439+1Gh7+SqOWfSWDKafCvU0T7h+HlWvGApE7qOv6LC1qCwtL6BPY58P23LHh05fr2bN7Pm2it58/abeM+F36Bx4WL2u+RbTJ+/YCSoT9ZwkWF4BaeaiUmuppymJnsCGCuNS/WaFmlY5XPtp2yUtvyaQOB8IHLFfM6c3K6YV70+YTQn+uA2zHQDpqyHI2Ve72DQ4l//sqv3ZXy90z/PiorGbZtjD5MZZwHWlhadmhr7hmxZ/hpaWyPrWNm9xIc+sCXjgEkf2GLHnsf8pZkDcc/GWbPsnJ9oXw+TV6Q6KuuITHCN/v+tBH8XpHhDeWJNyoxtNosX7Jj9IdbdfUfaZUzG47y5fzYXrn2Sjf37MbvkdX4279gxwRIAmjZpg8UxinzMv/z79vagaRqhjnZaHnow7vEL79rFmmuv5InFx470FB/9pyXM/84PEhbpr54mF0g8TCl1n1LqdzH//gq8CNzvdt3yypx4dTgYHLJlGNH4rGaJ2JpNKw69b2POZRR3PTFyxdy54YOZlJu/+RuBgElPj8a4VRiyMz7TnZn+UPnxwwmLQi3ULt+Lop6JaaLTcdDsVRPus3v4XHXlAAMDGp2d9p8z2mBXxttY+sT1jDIuY9x56t+2PO7zSkqgvt605fPS1jbxfSne+ljO5bop6ZlmGIZuGEaRYRg6cB7wR+A9QOnw//8AfMHxWjrESvEF5qvy5OKNnpIo2MkkCJpsmfFW7vwo33r7YXrNao6b/meu2+cMivU4a0VY1tSYRzcUziiATkuaDTsrFGLNtVfyxs03ACSth7rwEluqJhyzDng75t9K4GzDMC50tVb5ZvVPuGvRgcttaeQMlO4T/4Hw2JzlkTkyzvUw6f32fC/OmmWhafY0qrWh3twKiPnOsvRMFgbNZg6TnT0FufTejMte2PE39MFtlG26E20ws7z0xb2vxr3f7p7OBZW/c6RcgOpXz818I21iPpui7tepfukTYE78LojLGvtaSjoSD1EPBCxbPi+7d08so+KdH+PvsjdLbj5lclR+CJxnGMZbhmEMGIbxFnA+cI0zVXNevHTEscy+SX7l3waJgp1MgqDJNAzqvs3f4vbWn2Kh8aWm73Je01WpN/IQzeePzDWymduBYctDD7K5eVnSQH7NdVex4sxT2dy8LI81E+kyDOOqcf9uNAzjCbfrlW+aOXF47GL1Ozo6dPrTbD8lYmnxMyNUvfH1MX8Hgxbbtun09OS2P+K8luEHciw4orgYGhvtCRz8O/8b/wHLRBtMpytnNHgYKktrtgMAupZ5wBSbyTB3uQyjGj8kL/J32aa72GP57IxK0ocmnmxtP2+ia3Pydly2nOhB1TMMEiPGHkM91ErVGxdT0tWMf+eLaZXg18bOg9KSDI0MBu1JGFKRYKTq9JdOi2TpK0CZtIx0YC/gjZj7ZpNGOlfPSpH0wRwszDc1n+Klb56KKZnDpo9rNvyGt/oOplTv5vt7fZbZpdkNO8i74c9BNDscMDLfp6i0lKG+yZGtcN3ddyRP2z7c45fpHDzhHKVUWpdkDcO4x+m6eNlhM+4H7qO1VWPvvW1s4A4r6nt3zN+jQ910lMoluIm/ba7r9QyXAsDs2c70FkR2oVHx1g8o33ArW4/dhOVPNsw/5tjandZtHFszGeaSBXVCwzz7sizNP+G+xumbqRr4L3Bs1uUm2Jt3MuWNO/7Vr5w9cv7ET9090VH+/zeuzMSfr0DA5NFHfTbM4Upct+KtjzEw8+RcCndFJgHTzcCTSql7gU1AEDhn+P7ClCLpg0gt2qjMJUue7cO18mzrQD3fW/8Xdg3V0lj8DlfvdSblvhyHb+STaY4EudH3LZq9b831BduBPEFoSwfzv/ODlEk2onPwJGDyhM+m8RwLmEIBU+JGZ0uLzt572xFsjN/l2DJjG+RK5VJugtdiS8AUMXs2PPusc43fko6/AaCFd6QfMGU09ylz0UyG7vcw2Rcwmb74x7Zra9ZFJlRX2+/ZLJBaeCe+vuE8a1kH3onfh0DAGpnDVV+fy3ufrG7eThiTSNoBk2EYP1FKvQqcARwMbAbONQyjYMev+KqrUw7LE6lF0zfX1VXR2Zl5Cky3h2vl4pXdR3PTptsYws9R1X/nK03fsj2zTj6Y/SGM228eE/iG+3ptTwPuJn9V9dgAP8l5NxkTkRQiwzCOc7sOXvXrl3/KFw4ee+U49x6FRAHM2PvHDvnKJbhJcNHS5oDpr3/VGBqCItvHw2iMNv5SNC5jjqGv+zW7KzJGNJOhHT1MRb3rst94fA9TDu+rlSC43rrV/sb3XnuGaGlJN/thfmljPjNZvvYUQ/Ig0iubW8CUzCQPmACGg6OCDZDGswaTp0r21+6Rp5pMLRMWqS1Qf9nyVZZs/TIaFp+tv47FtX9wu0o5Gdy1cySTXaEGsb7ycsJ9fXGvXA/29IxkzGtcuJjmY49MWM5kS0RSqJRSmmEY1vDthJciDMOYQsMF4jdidD33HoXEi6SObejW11v4fDas/ZSw4WZfQ232bAiHNTo6NJqaHGgARq/ypxy6Zv++9d53MMvjz4eyK5PhtNWfy37jce9vSec/cqzNRN3dGj09UFFhX5lz9uzj1bXeTPqlhW24yJ90SN5owpDDDsv+a1VLMg9RH+jMulw3pf1pUkqVKKWuVUq9o5TaOXzfIqXU15yrnrNSzc0YdKKvd4qLrjsU6mgHy/L2Yq0JmCb86N27WLL1yxRrIb6/12cKPliaLMK9vYkbLjHZ+lIldphqc/A8LPYLIgwMjvsXvW/KGb9mUkODc6m+x08SLyqyZ+2nxAGaHcHF6BwmyD0BwuC09024b9pLp8b8lbzOiV9r9qpXfz7hY05nMkyHNi7Q1gedaVPZfd7PagqzaZOe0/StRIp6cuixI7L216js3t/x70ss+xKGJA6Yqt6YuKZZIcjkiNwMzAf+h9FvhteAL9tdKeGezc3LWHHmqTQfd5QtGcPGl/fm7TcV9LpDO8K1XLTun6zpfT97+Nq4dd/jmVc+cW0I4U3RoXbJ5s3VHHKYzF/yjvfG3J4D7D3uX/S+qSNBK86OBnKiYU++7jUT7gsGzdwbVYl6mGxsqUYDplyPzeC0Qyfc5+sxSHtIngMBk5YkrXQwaNHVZUMmw1zkuFDt2LLiH7+/f/NjlG/6uX37AZqahuju1tjpwPXcGc8eYl9h2c5hStLDVFkJ06fn3ns8RDkAS18+iec1e98ft2QyJO80YK5hGD1qOC2OYRitSqlZzlRN5Fu09yca0GSSMWxz8zL+fc+d9La3jyR+ACaUV8he7zmMGzb+ikGrmIMrn+Ibga8V5HylqNL6hkgAoWmJE6Bomq2NF7dFh9olm6O087U1I0P3hLsMw9gUc3uDm3XxnHFrqwQCJi+8YP+SAFG+HSsJTx8dxhoIWPzrX05dhXYiYMq1ril6kFJ9TzrxPZokPXM0k2Frq868eW6NWLXvNSfroVPhm+i18dr9rKYh7jj3fJpeepreDz9vW7n2s38OE0QvvuT2eTG1yFpj337gx5zbsA+HN341p/K8IJNv14Hxz1dK1QGZL13sFZOsMZirdXffMaH3J52MYZubl/Haj6/FCkdGxoQ62llz7ZVOVtUxms8/8jpiLd16Dg9siUyw/lTdLXy87u58V81WpfUNHP2nJZFMeMneK698PlIsAZBWETHp7pOlF5csed6llDoFOAbYg5jWgmEYOUy0mBz2DIZ5+GG/bckNLK1oTHpvfWDsT30gYNLe7mNgILLeUTbC4dHPdOz+bBm+ZkWCicpKqKmxck8tnnK+VaqAyf6Ro8nmicRmMpw3z+YdWyZoqRvUyeoXKcfKoJck8fEdsjkxZFOjyVkf+RUMgbfz3aZ37HqsRiq0zSN/+3clX78pEDDZsMGeCwympUd6qxpzLM4DMjkifwF+q5SaA6CUagR+BvzRiYrlhVcagx6R6Kp7qoxhxu03xw0yCtF7v30FpfUNI3+bJty08VYe2HIpPm2Q78z+QsEHSxT5mHveBSOBrteV1jcQOPnUuI8FPn56ej+4us7+l142EgTNPe8C9JLShE+XLHneo5T6AXAnkd+tM4hcrPsokM6qoZPO+DVYAgGTcFijvd2eeSu7Ftw37p6x5QaDJpal0daWw/5i2tNbj49Z1NOG3+aSrY+P3LbjinnCIC7NYWdlrb/Naf/xjF8fK5a9i9eOVf52mstNpDw2mbzPSQImmzvQGhsdSM3vIiOc1nJ2I4JBi40bc53DFdnYsjTvrGmVo0xexXeA9cCrwHTgLaANuNqBeuVFbMNYJMkMpkHow6AAACAASURBVGkjc5DeuPmGCXOcCjFxQyLG7TcTGk720R2u5v+9/Q9e7D6e6b4t3DJ3Ee+teM7lGuYu8LFTaFy4mHV331EQgW6oo52tK5+l5pDDRlfS03UCHz+d/S75VsqMdnpJKfMv//6YHqPGhYvZ/9LLEq7MJ1nyPOlcYKFhGJcAA8P/P5nIgupTSPxWzJ6ByBIAOTVOYlpIZvHMsY9p4wO00Wxa2UvU0rW3BWxLAoSE862ijWtvXYC1LZPhsNjzoWL9jeluZcu+IxIfX3PI3mM/Y4aXkm4mfm01z+Ww6kKSYDYQMOnp0dhhw6UoC8315CN2SfubzjCMAcMwLjEMoxKoB6qG/04869DjJBPWWAmvupsmWBahjnZaHnpwJMNdqKN9Ui1sCpHU2oTDrOt9Lxe99U+2DO7J/uX/5da5x1PjnxxZE7eufBYorDlloY52dqxeha+yEjSN0rqZTJ+/gM3Ny5JmuyytbxjTsxSrceFi5l/+/QnnfOzQPeEp0w3DiGYfGFBK+Q3DeI7IEL0pZ3wP06xApOGeW+NktHEWnnZY0mdG58jktL8EjTZtyKZMBcPlBIORjH52XDGfaPg1eGzESjSToV1X98efb2lJOYw6g2OW5PjmOFp7gj2eeW/qJxUQTRseHkfJyH11T0xPkkDGjoshEZWV9mcxdEsmacVHchkahtEZszbGFicqlg+NCxfnsFLy5BO96l5a3xA5LulkNJhEC5tGNW/7FFe++0f6rTJO2eNOrtjrXHy6l6445SbU0c4bN9/gdjUyZoUHIwtNR4P1a69kzbVXTujh9FVXM/+KK1m4fCVH/2lJ0rlI48/5ZAGWcN3bSqloS2YN8GWl1GeB7Um2mYTiN3JmNUYDJpsaJxPmqIwL0GZF6pFbgzwmQIv5Kal+NbMhRIloQ5GLKcGgSW+vxrZtuQR3qYbkeStggsjrtu/qfpxyLHMkKI0vjTlM6UryXMu0GEic/0IMn5utlV8Zc6/e3xb32XYM54wOYW1oMNm8eeK5o4da0i5rQ8/h/GPVYiy9JPWTHZTJ0fCPv0Mp5QdsXzs7n2oOnpgqdCprXLiYo/+0hIVP/cdzV8zy4Ret1/Ob9u+jM8Q3gl/jzJm3uV0lR7Q89KDbVXCMr6w8o4An9pxPFWAJV30XqB2+fRlwEfAT4Buu1chNlkZf4Asjf87Y/Vdqa82ckhskT7YwttySEqivz3Vu0GiD2q65V2NFyoxNgJC9/KcNB+gfKkv6uDaQeOSDXYvXxmUOULH2CuqebISh+L38KZM+2DSHqaKkJ7e5dB62fXvy16X3pZM8dPjiRtXFce8fz87PS0ODhWlOLKd2xf7Qlt6yNRqR+ZJWUXkO9cldyix5SqkVRF55qVLqX+MeDgDPOlGxfOle95bbVfCMzc3LWHf3HYS2dFA6sx5fVVXkiv4UEDLL+P47D9A6sC9VRdu4es6nmVnc6na1RBZCHe2SFnySUEp9zTCMnw3/udYwjHUAw0Px5rpXMxfFXMgaqhhNf1b1+lcJBM7P6/CXSIPcniF5LS06B9tQp1haODIJI/aK+YEHZjtaIH7jsmigI+njudrcPQ+jJcii+Y+hxcm0N+3lT7DjfU/H3TaayXBwEPwTLnlnRg+PbQvU/XOPkdvaUC9WUZzAztakD4lVlPbS0qKz1172J2vo7YXyHNvpK7vO58jaO7PadmiIpN0SlW9+k10H/zlpGRoWQ6aOpY07CRK8P7W1FmVlOQ7nHP6eakiWHW/7SzDzAymL0rAwLd31i/jpHI27gXuIrKj+65h/dxNZtPZ0x2qXB5MpYUEuomswxc5PGurtgyLn1vXwik2huXxt7VO0DuzLPqWr+dm+x3kiWKo55DBJTJKl12+8PudFl4UnxKZxfMm1WnhRnOHkuSc3GNsgCTWckXR/OS9eG9MAyjntdxwlHQ8DNs23wmKoOMn3cYaNOd+O5yjesjTl8zTNmrDmViz/rpcTPhYMmphmjpkMc5KfHiaAthZnxuS1tuZ+AaJzZ/WYv7X+9GeyZJuyf8z+oj0045v81iBYJmXrb6Kk7YHR52s2JUoBGpLlT0r7M5Nm6n6HpWwNG4bxWwCl1ErDMN50vkrCDfHWYLLCg6Bp+KqrJ21P0zM7PsadbddhorOw5vec03id21Uasf2lF9yugvNsWF8pHllLadJ4Ryn1U+A1wK+Uiju5xTCMe/JbLTfFzPsp33fMI4GAxZNP6pktb5OgbADTPyPmoYmf00DA5NFHfZhmelNeJ4qU2fzq8bRoeiT/7rDizmUM1OX2+a1c9wNY8BVqakooL7dheFrSg5pZY67m+eMB6FyY+rc1q4QLjJ28P3t2/lNla6l6mDIKMpM/9z2DPyIyatdeMzdcTFVoB7sPyP4r5p5/X8zJe/9k5O89/jWXrcesxyquTbJVVPLXXbI1nQuDFpalgTa2q6r23wfTX7uQkq5mADqbzhp5LPfhnNEeJnuCHMvS0k7h75RMjsZXlFLvj71DKfV+pdQtNtcpr3zV1amfNMlsbl42ITV4wnVnLItwj7eXbsvWvZu/yy/brgcsvtr0TU8FS1OB5vMTOPlUNF+OY0USkLWUJoUzgWnAWUTm0X42zr/PuFY7lw3ucfyYv4NBk74+ja4um3oUYidZWxMT/AQCFgMDGlu2ZLm/4QbQfc98dsLV7GmvfCrtYnYONiV+8ME6NC3aG+ZA0ofRJ2RfdlrlZl5+tGfNjt67nrnfpb/uYxlulZ+04gBWf+65X+IFpnOtuylt/2tW5XV0B3mzTfHmOzMmPJZsDS0nWGhYccb2RYOl8ezqrS4uhrq6VIs+pyd2Tty2I/M/GyiT8VZnAZeOu+9FYAkwfiZZwWg47vhJPQF+vOjQu2hvUqijnddvvD75fKVJlglvwCzm6nfvY31oPmX6Lq7a62xmla53u1pTir96GurCS2hcuJjp8xdg3H7z6PBYn29syqwsyVpKhc8wjLXAeQBKqX8ahvERl6vkGdbwMK3BqgX4d68GIgFMsa+fvrebYVwwlV6ZYxswlh6Tct+a2EMxOjdIo6Ehm4Ahsr2FRksOQ/t0Uvee2HPFXGPbkf9hxsqjJjxatuluuve7JeOuvbJ3b6Fvr+RNqMh7nfnxjWYytGNem1lcx64Df0/dE9PS3mb8+RTnGelXIEVZ3d25B4VD5XPx9do3r71vsIo1LfuxqSVeUzvd157769KGe5gmDMlLIhi06OrS6emBiops9jq8cC0awWCiDJPp9XpG06JbWtHI0TBLZ2VTqZxk8imy4jy/KMMyPCe6Jk0ik20eRLyhd2Z/CGtwMP4aTJNMx0CAr619ivWh+QRL1vKzecdJsJQnpfUNI+m+j334sZHhco0LF3Psw4+xcPlKFi5fSWntHilKGluev3riD7ispTT5SLAUMT6TnRbTkAwGTW48+1IO2XU6vp0v5r4zfbT3t6RjyYSHc16vZbjupqnndDU7euU5XD4v4XMiV8xzD5iGquKv0VPWei/+HZlf9a586/tJHx95v7OY8D6ayTDbHsCYfVpDGQeDybMuwvTnF2ZSmaSP1petzaCsBCak0s+xOM3ENHV6eicGTPpAl637Sm54SF4GSa2jvZO5z+HSRi6s5MJCY8dhse3x/IcemexxBXCNUkoHGP7/lcP3F6xUw3bW3X1HnmqSH4le71BfH02LT8x2IHpBeH7Xh7l03d/pMadx9LS/cf0+p1Gqh1JvOEUUlSVIX6tpkMOwOV91dUY/tOkOpQt1tLPu7jtQF17C/CuulLWUxJQyOnxotCEZCJjMa4w0HLXBbXG2Sl3qGDFzHko7/m/Cs3NdryXaoDYtPaeGmaalvlIdDFps367R3Z3dPsKDFv0DKb7HhpwZvh55r7Mb8hcI5JLtbLSh219/ahbbJ69ztGc0HamCr6P2Xp7zdFjLV5VbAeMUaUMMmfGDlGmvnEFRT+ogb0BLZ55TCtZwlrkMfodzTS2ujflesni7Y5+49cqkrKGq+Vj6cMpCm4PbdGSyx68DxwOblVLPAW3AQuBCJyqWL6mG7Uy2eRDJXu/Wlc8y//LkV7sK1QMd3+CWltuw0PhCw5VcMMv+yaGFbqivj5pDDpv4pWpZFPmzz5YYu9hsOtnrMhlKFy0T4Og/LeHsNWtkLSUxuU1oZIz+XTf0FNMr7EvQE2/OQ6zKSpg+PYfU4sNzmKqqLPr6su9h0tMImEYz5WXX0Hr++SI6O5P3hKVq1GcjWma2ZUcWr7WhcenyGjjptK2znUvX2nsAAD37fC+r7RPRtSEsLfFnaMazh1HSnnxKyPai96XcT6qse9EseZl0EOa8eO3IG6YRCJgccFm84DiDCNcae4HI8nLAZBhGC3AIcCqRxQJPBQ4dvr9gpRq2M9nmQSR7vaEtHTQuXDypEmGETZ0frv8NS7u+QInWx9VzPs2HZ2Q3gXMq2PnaGuZ/5wfMv+LKMUM0h/riL0yYqWj2umTmnndBRsND0ylTiMlnuAERkzmq5uWP8759Vg4/mnkDOzq87/72x4fvGHuhZMaK/Sdsk8tQt+gclxm1uTV+Ui+Qmntq8b6+SF3feSfPDTUtMpzK9E1P/dw4AgGT1lbNiWSkqdm6bk7qsrJNbmFZOo+8fDJD5XOy2j6RIj1MSUkkYHqqK34wVtS9JlXtUu5n+gsn/H/2zjs+jura47+ZLdrVqlpW33XD9hhjOiFA4oANelGIEzs82iMhDScxJKaHGsAQek8gickzyUsCCQ4JoRqDCM0UQ6jGhbEFLrJVrN5X2t2Z98fs7M7szJ22s0XyfD8fgzQ7c+/d2aJz7jnnd3THMKu0WFPDw+1Op89afN2UkJI3Oq7icJuWFZeSfal8U598lmU5lmXfZln2cZZlN7Ism1uNPxvQ24mebHUQWg6R6BwWzybngU8kusercOGOl/Hp6BdQ492FB+Yswiz/1lwvK6/hxsLYfNtN+PSBexW1bnahF7WtbWjE/MuvSvagiqeJavWkmmyRYAd1GIYpZxjmuwzDXB3/v1J+atJDjjDZN3bcGEnZknaFlfuj6alpCSZEeZqvohHRB7Hw3OqOeVFAcFw0n6uGARjzWzfGeVAYr/yqpWtFJcPOTguvkWGD1h4VtHRJJ5LG8xQ4/wz7FgPARUVAuz0oKuLR3291bcI9vOzRu4ln6AlVUBSfEIkxissF1NWl2bxWmD2R3vfO+G0pjxl/f1iV1rcTzTvBMMw2yc8tDMPsUfuX+WXmjsmY2jNv5aWqO/hTjzsB7136s0nR/+eTweNwafML6I9NxReKX8Bds76OgNti8vqBBsel33fL5VYVYwCMRW1rGxqxcO2TghDEy2+h4dWNWLj2SaLT5CmePFFRB3UYhjkewGcAVgA4DMBPADTHjx9w8LyyhinlBOtja42R0gslFBKMKivTiX16pkxRN0Wo6KCxcSj9vdvqah4ej/Ud80BASGuyakDyUol2E4iRwiHmTkvXS5UMzaN8UUemna84VtCurG8jXW8d/bHCHTYIP9iIm44gynkQCnGgxrvSGuv3L/+Y+BhP6aXMm48wAeluhsjFaABg3c4UNUiDfZXk0fLcNa/V++T/SPLzd6DeA+PczCwte5AMO9LxfECtl5LWcSm1DY2CwEMK+557ZlI4S//q/Alub1mDKNw4p/pOXBy6dDJrWeQnsSh48ArHPF31utnLV6j2bYoMD086RUsHBfcDuIBl2RNYlv0flmW/BOB8AL/O8bqyjLEI06ilALF8rJh/uuKMws/lhnswyGF4mEJfn5X5BIOpsBAoKlJ5Hty4oVEoAxEmmhZ2zK1GITyeuPy55SiGhqGn423yPAXeU25p1rSVDBNriP+gYpxTMVLKtnBRf3RaWnPLF0Dm3KBeapo6omy13bjoKGK8G6EQD1fUiggLYMhB0KiTEscwG2EC0pXiT8qKFxcDpaVqY5mIMMXXPzLjUuEA5bW4LutouqUsy74h+fm1zC8nNzArL8HWO28BF4nIjlcvyk8VW1Ivpb7Nm9C6fp3iOKCMlKnJqfPRiOLYRILjgLtbVuPj4S/DQ43hqmnLMS/wYa6XlVUaXt2IpkXH25w7bo3o4CAWXHMDmtesRnh/B3xV1Zi9fEVaUdvahkZ8+sC9iA6kvFdjUTSvWY3DzjkjzVU75DFzAfw95dg/AKRdwMYwzAwIPQVFygCUsCw7JeW8VQAugCB6BABvsiz703Tnt0Jix5jwWR9ob8eU2gHwbvPRVyqeijde9Q1Eiw6Be2hL4jFvdxNGDroq8buY6rZ3L43ycrNZ+qJRRatLD6s0y1XDiOgDIDavTU/RT3vH3dr3bqD5RgzPWaU5r1XSK95Pzt3VRaGykld1mLxdL2J05iUqlwvXP9X1J3y3ZpHqDO6+dxEtO1Z3JUbuQ3lBGzp1z1LHikOhhygrHgxyCI9aex3F510xReN6nQgThbhKnkmCQQ7t7W5EIsKGgTWoxFjWHabkeSMHXSX7/skmmneZYZibjAzCsuyEllarbWhEuHkbmteulR1vXb8OZQsOy7u0PFIvpb3PPInUyk6xID71OUy2mo+BaBl+8fnj6I7WYYq7DTfPOgOl7vQ7f08kxNo0X1U1wh3tOV6NsI7ahsa0Pz9tTetlThcpXXCyvacdFOwAcDaAv0qOnQEhTS8tWJbdBeAI8XeGYe4H+e/jn1mWTW3inkXkRsbg/AdR/p6yvmVu30rENt6Dni8bl25WyopTGKtaInOYUs+RGuSHHmrSYeLFxrV0IhIim56PGDKpaElKXvdXtqPidfU63GCQx6uvWt8xF2qY7E9X8O37P6LDJMxs3ZhPW8kQwH8+Owa9UwWHSU31zdv3JuHKpENMouST89Cz8BMDq8jkJmBmxhadnWCQw+nMY9YGiTuddUGNU3STxThLDmEoxIPjKLS2Upg+3ew9Un5P7N6dss62F+AqPR2xIkZ/tAw4tGbRu8shyb85AK4CcDKA2QAWx3+fk8kFZovW119XHMtX9S2iYUiQwVE7fzKp/7HDh+PCHf9Gd7QWhwXewK9mn3LAOUuAENFpWnQ8IgP9uV4K4HLbIpgiRlPDHe0JaXISk+k97aDKxQAeZBhmI8MwaxmGeQfAbwFcaOckDMN4AXwbwB/sHNd+BAMiWk4u4XKN7jI5ZtzAlRonqXUGKRGt9Pq1iGNTCRU7+cPGUvJkl3iV3wPu/v/Au38dQsEIOjoojJsfFkmHKf2ajlRorZ5ZKuliIzMuVjmRjHUlQ2Huf733rcT1sYAZUShx7RopY4a1ro0Z7FaTKzIhKkDFa4fEKGw61NVpRZi0X1vKckqedSl+SiIrLoylkpLX+xGmvP0F02PnCs27wLLsD8R/EJ71/7As+yWWZc9hWfbLEHb7JgUjbW2qx/Nx15poGBIKdcTzpfVNdslE55rnu8/FL3c/gghfgP+e+htcOf0nB269Es8DPJ8Xr60nELAlMqsWTVUj3dooh/yHZdm3ABwE4EEA7wN4AMDs+HE7+SaAfSzLfkB4/GyGYTYxDPNiTgQnspJuSxF+BmR9n5pKMHPPWSgstKimxScjEKoRJm7M/JgqBnj5uyej9OOz8d/ML8HzFPbtsyg/DQqxmP0peQBAj7aoHqfAS3rQCERKjzM1tuXi/cTrQyVEI8aqv2Xhejv+KBu7t1Zq6TLRP0uE5wmbAXFc4X16IwAA6uu11qj32hoTffB2Pi/7PR3BED4RXRQIBjkMDlpsgpuhGjOzmOlG+TUIu25SngbwR/uWkxu0isVdPuP9YLLF7OUrsOWOW2Q1R5Tbg/qvf0NWwwQkDcnUuqe8iEKkAccBD+y7F+8O/hfciOCy0E9xWLHddpODVSKD9jTP1Nqw8FXX2FYb5ZD/MAzza5ZlLwTwSMrx+1mWNbflrs0PQY4urQZwC8uyEYZhGgA8xTDMwSzLdhsdvKKiKK3Fide73W5UVhbrnm/kHJH+bj8AwF9YkLyuTf430OOmZWMWdK3D9OkUOju9qKw0V4g9FhHmK/AV4JCZBaj6cQf2/y65ITilzAuUG1+/K2VtqRxULvS8GRoqQmWlqaWiy5M0+j+b9RYO8vwdYO+XnVNa4gdI87u0nYaK1nuA4x5WHG+jKICSP6/SUvlrovcaz50LvPkmMHVqseGATmVlMRATCld4nkJ3tw+VlT6AU7eJ1NbQ3yooA/oD/sSxgdESlPiTfx9cNGXoPdrX69c9BwCGh4sx12RnlDaaAnjtdVRUFOtuxqZeP0rxoGkaRxwRAF5Wv8bX9jf4jr0NCCgFVgBgyOcBxgGGIass0jr3sN/rAj9CYcqUAKDulwMASkc2AJVnJn4vjg/Z2+s3/XkZKiwAhoGSkgAqK4txyCHkc/Ve/w6KAg9j75NMYsZhagbwU8gVic6HDfnjuUYr7S4fduvV4FN2GXmeR9mCw1C24DDVIvsNZy3LWF+dbDMSLcR1u/6O9vGZKHV14aaZZ2KqN/8igWagfX64CryI9E9sR1bErvQ4Uj2Wr7oGC9c+qXKFwyTm+1BPvzsXQrpe2jAMUw/gRBDUX1mWbZf83MQwTAuABQAMiyJ1dw+B46ztmFZWFqO7exAVACKRGDo7BdltLVtGPMcI4YFRlAIYGRlPXFc4HEZAck4kEkNfyry1tVF89hmFzs4Rw3MBANc7hGoA4XAUVSXDGIvIjcL+tmaMRw/SHUdcRyzKobdzkHg/XC5BRGLz5lEceqgxQQmRaDQKVzzSs/Gzw1FxTDfKIHeY+vtHMF6gfr/LozFNg2s0HMGQymvF8xx4nken5Hn19wuvk4jea1xR4cHgoA/NzYMoM9D/trKyWBgzNpqYc8eOCDo7wwDPqd5ftTWMDwrrHB6OAHFbdywWAJB0mGIcjx4D79GRvhEYad27adMogkFzry0fL2fo1HjvbN06hOpq8uc2cc8kFEF4fhSl/fx62z5DtFS9GdloWMgfLS0j26IcT6Fb4x5GxiPgeQo9PcMYmnsrirZfo3reyGgEwynjVFUFwLJRdHaai/aOjgj25uDACDo7B1FSQgOyb5Ikeu9fjucMnWcUmqYsbVyZcZiWA/gXwzBXANgHoB5AFMBppmfNM/Ix7U6L5jWrgVjKF0JcJWzh2idVd9rzQQTADnaOzsMvd/0ZY3wAc/3v49rp34ebntj9k+kCH+ZfdiWKS/x4+6qr8kLhLl3sSo+bvXyFLDIKOOl3BxoMw/ww/qNb8rPILADpNTiR8z0Az5EiRgzD1LMsuy/+8xEAZgBgbZzfMBlp5Kj63aMvYx4Mcvj4YzPmhHw+sc4jxslrXUo/OgOdDfZEqwHA5xPSe6ykD0rrQFpaaOAYtTHS+VtEfj2V9Sfm/kaI6Y4tLTTKysyvUd5/ysz7TpmS53Kl9/dtXfdvsfCEYRSzP1d93GqNmd7nKfz5v4HqxaZHBYjVEhL00zwDhQAsZKgKQyRV8kan/4zoMKkh9lmzMGn8/8kapnTIB9EHw99wLMt+yDDMHADHAagD0AbgbZZlJ7YWNYSmlxMpRY3k4E00x88sr/Z+C2vabgQPGqdO+QO+XXNPrpdkC9xYGJtvWZXrZdhGcOlptqXHiePYKU3uMOEQoz1eyCM/PIAOCE6OXXwfKVEshmHWAbieZdn3ANzKMMzRAGIAxgGcK406ZYds1zDpO0yhEI/ubhrDw0BAfROZQFwlj6cFyWrant4qMf8MVcELmhYa2FpVuhON6r17aUTKv6R43N/yEMarllgam/S62lFfI9ai7N1rVskwOXfinhkWaUhmwvC81GEyMb3KWvqis8D5yfaaVdl4PY4ZWoZOmHPeKcpow1jyOZSBc/TXIajkmXjpEoRCHD7+2MqLltwMAYCpU3n4/Txeab0Ai+p+a2qkTNaYmcHClpAAy7KvMwwTYBjGy7LssJ2Lyja8xotB+4zlzWaCtqb18Z4zyQ+pu6QE7uJiVWlltTSotqb1YB+4L6PrzAa/b70Jr/WdBhoxXBi8GMeWvJTrJSlxu4GouVSAycjBl1xh63h2SJM7TFxYll0EAAzD3Myy7C8yPJei+oFl2VMlP9vpnKWJeeuHivSDDu9GrPgwwhlq0t7yY64xpUCSWNS+bx+NuXONG+SURFacpoGqGg8e/WQVvn3oKsNjAMBApBYlnjYMH/SLxHgkBLUuK8ancB+qquICCirKZK4hrWCjdaPPcjSRi4CK9CIYFGwD88X7SaO3v5/C4GCyrsUIlKEIk1mVPErzmq72YZg1bTNqkKcZGRE/fjwotA3ORG3xTrWz9EYx9B7ytT2G4Xl3yY4FgxzWrXOD44xEytQQ5qUooL6ew+BgepsVucTwyhmGORTAdgD/C0CsTDwReS+9qk90kJwXOf+yK7O4kiRtTeux+fabFY5RdGCA3Iemox0bzlqWELHYdt+d2HzLqgkVPUslzPlw9Wf/xGt9/40A3Y87D1qSn84ScGA4SxSF4NLT4KuuUX2YdNzBIV0y7SxNHKwbd6UffgtTNn5Zd2xeI8JEj+9XXGVdWlx0mJLNLe9rMv8yj3N+PPLGtzFWvUw4oLGVbrV5rZiSFwpZj1DpzaB6VPUwj7GKBt0Ri7degKmvz0ZF+Tj8fqupVZCnIpq7Mv5ficNEW33/St+b5Ne3lt5gbfRMNK41/FnVijAJY1AU8ODW59RP0knjNyorTkeVtmIwyGN8nEJnp7n7k7rRIo5lzWHKjwiTmZX/DkJawjwAYhreawC0vn0nBKQCdcrny9nO9qcP3KusUzJAuKMdm29ZhZcaFmLvU09kYGXZo21sOlZufwV7xuZhhm8zHpy7CLUFGhIvDpmH59G18S1MPe4E0AVytSSntsjBIXtITYhI8ZGGrvH0v2dwdK2UPCXS5rXmkKdshUIc9uxJ33DlPeoF9IDglLW2UqS2hVqjgo/3i7Iis2wdFWOX5xEtOVz3yoIOQRjH3/aINWnxFKNX63r/rl/B05vSwDZxPTnCZDxyIK+JPqqQcgAAIABJREFUIdHZlZ5B3n/kPyxcrw4liezsiZxscZTk866uSaPpssUIjXVpceXmSyjEYcBqhCkPapjMrPwQJOVceQCIp+LlLmfNJmYvXwHKrRLCjcY0JcczRVvTemIUSQEhRspHJnZp2Vv9jbjis6cxwhVjcdla3DLrLHhpS90GHWwm3NGO1vXrUNd4qhBRoij4qmsw//KrnNQ5B4ccEK5P7fghIbXxrCZqTVIvURwr2nKB7Pfqah5ut/lUN16SkgcItVCdndaiP/ID5JqLYJBHJEKho8PkjnnccQkGeezfTyOsIjo7VnsW8fp0Ej1Eo3Os6huJI1Iqm0rg3U+IPgAo3naReuNQHcT7KqbhaTnERTuuQ9l7X1OsXPivDRGmhEAINCOIAwNCLZ3p4eP3eHzqf1lYnMa4cUPf4yskn6RZXJR0mGokDtPgwQ+onEMew6rDkYweW9sMoSTPLRjkMTJqw+c7R5hZ+S4AR0sPMAxzLAS58QlNbUMjPEVKiUE+GtGUHLcbsbGsKQEA89tkec+f267Cb/bdDYDCirqrcV7dTblekkMK3FgYXRvfwsK1T6LhlbeJ6owODg42Y1JF09OjoniuM4Z0V5j3lGFk+kWyx/2tslZYcLmAujrzKV9UwplLpuRZxegOelrNOCVNSFtbldfzLrLiRU+PxaadBo3FgvbHUbTtEtAjkk4vktfZcvNaCEIeBQVWUhGVDhNFAdMv70n+Dg6entd1HXtK4jjovdb79pl8H2aqMSrFJ3wh7XJ4Y69LbV3y53DwexitN1ZSSSGpkmcW69Hj5OwiwSBn2XGbUDVMAK4D8BzDMDcC8DIMczWAxwFMirzycUL/m2wpz4mNZSeL/LcVopwbN3z+KF7oPRc+ehi3zjoNC8ueyfWyDixMyOhMdlVGh/yDYZh5DMNcxzDMbyS/kxQMJjlmU5n0jpGPxwJzdGcJhawY5HKDOhTK/E5yujvmWgYkFSOHNjyeNJ5b3MiM+WcJv3qnKk5xD34C/96HUbLpB8n18MmsjO8ddVNCydDExMI4FFBfbz6CmBR9kKd41tYmf3ON7kLZ+0sQ2K5tSiYV97RrmNZd8XVLjqERQ36kw1x8QFo75PNZNfiT97CmjjRG5iJMRUVAWZkVoRR1NU1rTLAIE8uyzwJohNAj7jUA0wGcxrLsixlaW1Zx+9Xdf5dPvau1GA1qWnS8TGhBC61rmtesnjSNZa3QOV6Ln+14Bc3hI1DnbcZv5p6EkG/CBy8nFL7qGjS88jaCS421VrOrOa2DgxEYhjkDwOsQegCK8uJFAO7N2aJyglnjQe3PPEHCmk/u4pvFSsqXGFUQa5isR5iShr0e4hxW0tPEflHi9Twll0Ev3P0r4vVqWf9G5xUZnn0d+o94XFXSXK/G5+SqWwCYj74II1PxCFX6ESZAiEamUrjnQUNjAVTiheYpN3qPfVlxptVaOj1Cm75gctyks0jTWm9OfdEHgEJ5acz0/OIY6dQAWXvtBXi7IkwTpYaJYRgXwzB/ArCVZdkLWJb9OsuyK1iWfT/D68sa0RH1DuWx0VGFcyOLBvE8wh3t2Hr37ZpOk941B/Ju/QeDX8Flzc9jMFaOE0qewV2zl8JHk7ta5xVuT65XYBvR0RG0Na3HwZdcAU9Jqea5jsCDQw64CUADy7IrIPRBAoCPAehXv09CjBoQXMy8kZU6NjWu7ONb2VQi+z0Y5NDeTsFc+axcJa+2lgdtscbF6P0oKgLKy3lLRew8T6GujgNNC9f3Hmdckc3t1nleGt5ewuikvRiv/CoAEJwm6Pq6piIFkpQ+QV3QbI2a+IPU1ORRV5+G8UslI0yR8oWIlh6Dodk3yk6xFGEysElAw9xnyY5Uv+QIFChe6FzbPjQj/iAv/7/GKOmktFlL51Q68DU1PN5qJrxvJwCGHCaWZWMA/gvptbGesKQ6N2rRIG4srFnvpHfNgbpb//j+n+Gelt+Cgwvfr7kZPw1eleslmcOCkmG+Eh0YSLzPmZWXKFTwRByBB4ccUQVgU/xnXvL//MjXyBrmnq6fvUFlCPUxSD0JvV0v6M4TCnHgOEq1tocExcsjEB6P4DSlDeH5FXStBx3ea3nHnOcpuN2C4bd3L41Y0cGGr7UsdkAwuiMVJ6dEuIyNby76kkyD0xK70L0+JSWvrt7KvdBILU0R+Yj17rAwvv3IIzsaDrGGsyx+RigK4ApqMRCegttfvEd2Dh0jt8ZRrsM8oZBQn2iyfDIxu4jbDbzf/k0LI+THV7yZT859AG5kGGbybKmbQOrckKJBqcelKXik2qRwRzuaFh2P6OgIqEkUrdAjytG4ddcaPNm1Al4qjFUzzkHDlMdyvSzzWPsGyVvE93ltQyPmX36VTAVvwbWrcM6WLY7Ag0OueB/JVDyRswG8m4O15ByjO8YF4S2qV2tjLSUPMJvqpkzZspKWZ8agKnv3lDR3zPV33D09rwHcmMnx1dEydnm3UqxK67WzomQIJFPyAEHsYjR4nqHrKJUoA+BCvYUIk3ws8XrRYZK/52b7n7c4tv2In1Mtp0j78yY+Rwpw+bDytTb85ZVlpldh2GHilGrEwSCHoSEKhFJ/VdT6MIljvbPnlJTl6Ufu8kH0wUxW7UoANQAuZRimE8KrSAHgWZadlonF5RuiQ+SrqlZ3gHgeTYtPADgO7pISxEZGwUcN5CfwvHEZ8UlAX7QCv/j8cfRGq1HpacHNM89EkfvAef75TrijHW1N61Hb0Og4Rg75xIUAXmQY5jwAAYZhXgAwF0L2wwGDdMfZCDyvJrFNMBCJG0D6k4kGtZm0LT6lhkkYJ7Mpea6xVoRCPF57TdgxN3ofKSSNtmCQx3vvqUuXe7r/jbIPvoXR4HkYOvg+Y4PrQDYWJbv3CXU88hMyr2TIJ0YUa7daWmjMOvE++Pb+EZRe0hGfNPaHZt+IouYbwNMFqA/yQJeJZUjWImtcS3i/9vdnRvQBAOiRz8H5QgBtZIM7/ca1qYRCPHp6BPEONXeZtA6jKnlUbAQ8La/NEz+TLS00ysqMbmioRReFsUyX62dKxdAkZhym72RsFfkARelHC3geG85ahqnHnYDW9evURRriMt8HkgNkhi3Dx+LOPasR5b04quhlXBJcSWol5ZBDttxxCz594F5EBwfhKS4BDx7RwUEU1tRg1g9/4jhSDlmHZdlPGYaZB2AJgGcBtAB4lmXZodyuLDcYrmFSNZTsjzDV11uJMMWNL4nXIqrQSXH3vYNo8WGAy562j8Egh5ERCr29wBRyj1s5VHKXPhTi8PTTbqiVh5V98C0AgGuYVVyvMwHhuNZ1KtdoeIBWpcWlESbh9Y3pO0vxK8V1jk47H4W77sMQcxvqKZh3mKSCJKnPMSXCNDAIBD7+AVA4DcNz5PVN2uvUp+LNIxDzBdGzcKvBK/RT8gxFmOJIXwejhRwUBLEF8bZFAwzcqe9PwnyAXBny0EPNOUxUymsVCnEIj6V8R/C87lfOhBF9iPM2gJMBrAGwLv7/UwC8k4F1ZR+DqVWKpp0Ohnmm6we4dffDiPIenFV5Hy6b5jhL+QofjQhOP88jMtCf+HmkrU1X4MTBIVOwLDvCsuzfWZa9i2XZxw5MZ8ncbivHqXzJEv/eWd/JLSgAqqvNGeRqqnzBII/GO5IpVQVtj6P8Pw0o3nYRyJhbtzR9sPT9b6I07uToIRptwSCPaJRCe7udRhz5ORCNRRXniB4jtyYxr2SYXJMoyGFO+EESZXD50L1oD8arl6Giwsp9U6+HEv7rUpxauP+fKNxlPMJnxiB3hfcaOs+o6IN/78PkMcCD4+SfDyAubGG4DYhc9KH/qH9pnpuKbE7TKCNMyq8f7fs0EWuYfgdgMYS0iC/E/38SgN/av6zs4y0rM3yutGmnmb41ByocB9yz5wE8tv8yuKkIrpl+Hr5ZSf6CcMhv9AROHBwyAcMw0xiGeZhhmA8Yhtku/ZfrteUGY397opzxlDz1vjnGCQatpXzJU/I4HDd7Y+L3ks1CvYx7cBO0MLNm6Y65t+dVeLv/rTiHHtkp+11qtEmv7/1CE2FB5ow817C6UIG20a18zlrG/Kxpg+aUDCW9j0RBDmsOV4rqoqbEts5YPIVI2QkI156DwfmCFHlqw+CZVTtTL9YkUwY5TRkTffDvXaM5jvS9baWRrKIOjiInl7mGtyPAXiN7/1ZU8PD7rX22U1GrUSzeeoF+zd8EizAtA7CEZdnnWZbdyrLs8wCWxo9PaNqa1hMb15IQxRomW9G/3QxFS3DpZ+vxwdBilLv341dzGnBI4ICs0Z5UHMgy+A4543EIaeTXAzg/5Z8DAenudBKz6WEG+9SErPdrSY7BIzhFafS7h4ymQOmjt2Pu6X4VFW8ejoI2qRCRsEtPUfLro6XGevPoGeXevrfgGt4Od99GxWMkZ9CsY3v49E9MKxlK5zGf0kdywC28R3jJWLQHgwtWgyucCQAYqz3L/Hipw2dIVCDdcVOVK6ureQviHcZlxUs/PB2Fex4EHd6TOCa85+157UMhZS8mX9taeDu1hDryw842U8PUDqAQQJ/kmB9Am60rygHNa1Zbc3wcZ0mT5pFDccvuP2Kc92N+4UZcOe1HcNMHpDJ9znCXlMDtLySqNFrlQJXBd8gp8wAcz7Ks8yUCuSHCu0vI5/G8iriBTkqeRRsvGOSwbp0bHAeD6dbKddTXc9hmssg71RnhdWqdpkzhUVgYj5bMUj7uHhKUBd39H2Cs9uzEHKKhV18vreexbhDzlAuURCFsylvHAAA6T+lL1OVoO1rm5q6tEvpN7t1LY/p08/25gkEe//mPutiFGpQkKiR/II0Ik9pzNiTAoD12xuMXFrORxPedeLnLJYh3KDYmuChAk0x6eYRJ69nS0YHEWVLMpnOqKyQKtY59ytMV56WSDyp5Ztz8vwBYzzDMjxiG+RrDMD+GUMv0Z4ZhFov/MrPMzOLslttPU8/ZWLXrrxjnfVg69SFcO+O8SeksUW4P4LLYwj0NgktPk0l+k4gODGD28hWKnkrprNtpWuuQI54BcGKuF5F7lAb0WM2ZxLMp8OjpMWtsWE/JGx+n0NlpLYIBAIWFQKE/vWa7g4c8hOGZlyNatED1XHHH3Gw9jjhHIABUVMSvJxjC3r43QYf19pONCD3wtqUjHexfC8CckqFAUuyitZVSFbtQhSdFmKxgfCy/3/xmtu2iAin1eZzPqpi0UhJcaCKcKpxATmmjTKjkJa6Jjch+tyoYkvoe9/kAr1ejp5bqCPkRnDBjMf0k/v9rUo6viP8DhGessl9jDIZhdgEIx/8BwJUsy+p3zEsToky4gyV+s/d2vDWwBC5EcXHoQhxV/Hqul5Qx+Fg065FGl9+PsgWH4eBLrkgc23DWMuJ7uG/zJsy//Co0r1mN8P4O+KqqEw5P85rVwnU0nVB4VMNdUuKo5DnkmgsBvMUwzGcAZLtcLMv+MDdLyh3ymgSyMVTsH8LevRQqKiTfUya/s4waLKEQh6/Mew0DOylUVx9pag4pgUB636mcrxYjs6/HeNVSlL+zUPWcYJDHotpfmRpXaqwb2XEv3rIC/Uc/lbhaCeH6FNUwopOg8bqrURF+DoAZJUNllEEUuyBu08nCmepKaelgZKyiInvmG5p7G4q2X53WGOLndPigaxApOQKlm75rfgwV4YTXX5e/hhQ3pqjlSjyG1DQ4/ftT8smP0Hv8m4nfQyEe3d2CnHlAfRrFqkkU+KwpNeYaww4Ty7IzM7kQCaezLLs5S3MBAGYvX4Etd9wMPhrN5rSTjjDnx3WfP4bW8dkodvXglzPPRKV3wmdsapODtMzY6Ci23n07AKC2oRFtTesRGx0lnr/3qSdQtuAwQaQkBanjQ3K6fNU1WLj2SbQ1rcfnf3gIm2+9Ec1rVmP28hWO4+SQTf4IIAZgGwDyG36yY+E7p6WFxuGHJzdEKEVlhHzsVOPEqLESDPJ47bqTgEGgE/qtNfhETyn5agIBk9kIhBS+aMnhqsc93a8gGGzE5Yt/rjNu8nmnOo3BIIft23UcD15HXYGiCHal/LXSGEB7fMV0FKqqzEQK5O+HZK8tGkcQrvC1PITwNOkeuvr75/XCN/CVkS/LjhW0/xNjNf+tuRYjnHmEOS0yorBGWo5eSnor7cV4tZWSfzXVOg7t7W5wkkgfxY1r3CHOdITJPfSJYk4A2LePxty5Bj6fGtFFn09xaEKQ/VyiPKS2oRHFJX5svOEG8GGzHbUcAKAlPBs37noEo1wxZvs+xnUzvgs37TigmYIbC2PzbTehb/Mmck8wCc1rVus6N7OXr8DWu2+XjSWm37U1rZc9Fu5olzltDg5ZYDGAOpZlB3O9kPzAuDGnNJDNiT6YiTAhrtfg2/t/CAe/b+i61PmKCs2nb5tJqfJ2vYBQ6KtmZ5DNEQzyePlloflt11eaMfX12SbHA4ivIS9//nami4VCZtXOkEgJFJvX7t1LAaXqpxazVygcJrXnWTLjcCBFx6Pkkx+ga8pXwHsrVdZgZ3qfyvC2j0uo3yKerig0BKCicIekcMLICJAI9nDjxKFpxBBTVcs0jlToZO5cI1eQX3tfgdnZ8yMlLx+74DzKMMwmhmF+yzCMca3vNJm5ZAncXtOvogOADX3fwNWfP4FRrghfLf8Lbpx1juMsZQOOw96nntB1lgBjdXq1DY2Yf/lVstqo+ZdfhdqGRjSvWa2Yx5EXd8gymwBU5HoRuce8MMOR7l8CsWRQztulLoVNiDvBqMFSVJT8uXjbhUaXp6AwzZQ8GfOvVB7jOVV5Y23kSmOhEIfRUQrd3RT4girCJcnnoe50kl5ESYRJSwCDEAEhSZTzoOO1KAZNv5Roplzsgox74KPEjMJ/leusqSFI23N6muf2O0yWamQMRnqNO2LG+3CJzsvwsPS4Vg1QqsNk/vlakTMXZ0/FZ6H/dD40rs23CNNClmVbGIYpAHA/gAcBfMfoxRUVRfonEdj57LOIDJiTFncAHm69Hi/3nQkaMVxQfzlOKHUamuYjhTU1qKws1j2v8pwzcNg5ZyiOkxyu8P4OQ+Nmk3xbTyr5vj4gb9f4MoAXGYb5I5Q1TH/IzZKyzJZbUbblDuFnEwbEkum3YpRtTfxe8skP0ElMfTI3tiY8p1Nno264BUxGmDQNXlXDlheMTmLWoPp4UrUy6Y771KmkvlY6z4N4b1JU/4hGt/rxKW8djc4G5ZNzjXcgGOTx/POUCSXD5PyFhcDUqXGxi0PI55e9cyK6TulRbUycWIvJgEeiR1iGDGe1caOBecTzC1ofxVi9hnlqNnWW8FnhVcYRnf3hYR6Ip7e5Bz5EFBS4gjqFWh5FxRCNJY9ZcRBrangU+UfQutfoteTzfD6TNZQmVTMzRV45TCzLtsT/P8YwzG8BPG3m+u7uIXCctRv7/m23WbruQGWc8+LGXX/BrvACFNIDWDXjHNT7zDWLyya0xwsuQg5ZT2boAh9m/fAn6Oy0nslEEkbxVVWnNa7dVFYW59V6Usn39QHprZGmqbQ2rnT4MoB9AP4r5TgP4MBwmD6+1nJaiGvkM/kBFQONsrkm07/rfozOvFT3vFSHoLCQS0o/GYToVJQdpjhEgRN2zLcYHz/VyJTuuB9xhLpjRI/u1hyTB6W6aornJLPZV8MEAMfMehe/Gf8yOjspVFfrvd5qxrq+2AUFHlSkVxKxtE8lz+xYZf/5KmL+6Rhc8HsDY8uJlpAqtQDX2D6dmc06eOrvoQC/Cy46Jgsm1tdLIkxxh0kUkxir/DoGjvibbAyKFyJMqa9azFuNnhN3oLKJ3JZAxOUCBtcEsKVzIYDnDDwfrRomCkgR9XMPfoLxqqXEqKkTYZLAMEwAgJtl2X6GYSgAZwP4SOcy2xjvU1eGd1DSNhbCDTsfwzBXhmkFn+KGmd+Gj87v2q8DyVmi3B64Cv2IDg4mFPHSrTPSqm9ycMgGLMsuyvUaJjZyo9Db1YTxytQ6HnvrRNyDmyxd5y8uNu0wEZlxDqIfr4J7pDlxiB5rF5wFEw6TsoZJTE/TuFeyWiQVo5wUYeJTBTrMRZi0+H7NQvwAPFpa9B0mtV46wSAHljWnsmfu/UQymK29Nz19b8PT97aOwyQft/sr2wE+SlyLGQyvl1c6TNTYfgS5fyks9YICoLqaw9AwrUhSLuh8DuBjAJUM4VGU4DC5bPhYH1K5AZ0GziP1YQIAn59XOEyBnXfB17YWPQuVmm/5IiueTzVM1QBeZRhmE4DNAOYCuCC3S3JI5d2BU3DFZ89imCvFiWX/xG0H/XfeO0uTndSeTIdceS3mrbxUiArt70DzmtVoa0ovVVKsbyqsrVXUNzk4ZIr45pn4M036l8s15gqzhmNq7xwq2qtxdsrYepEnFWMPADwDH2peRjKEXCX12vOpLYG0A01RiBUeJDtUsP8ZgylhEpU8Su64lJUJ8ufa0RZz4hqk67SemxUOCW7G3r00XEPbgNiw7vmpYhdGa6ASEUtbogN2RqvkpL4PuYIacL5gmqOaNfKVnyE6Qv6MBoM8/vcNdZXH4q0/k/1upQ+TfShfLzchVOMK7wEINWwTSlY807As+zkA640bHDLOo+2XYV3PD0CBw3k112PxlCdyvaQDHk9JqawfE4CMKdrVNjTisHPOyPuUModJRT8AMV8kCqUVQsWPpScBNQkYnnUlKG4MhbvuV318bAyQ1lqXbP4xOquWAi7pUXUjb7zyq/D0v6MxO8HxGdVO0xbt6dTENJJp5O59G9Hy4zXHVCNSvhAFXSktHWPyjb7AjhsxPPMywE1OKZU6DhQlNhDNhCFnUPTBIm+tOgGP738CU95uwPiURZJeUSmoOMqi2MXn8zaiEm+j+NNLCLPwSAqUpH+PKJujn6mYT/kydr7R9Xq7XoBrtAWjM4yJpYRCHDZ+pC5X52t9FIOH/E5jUcbeU2UbT0TfMc8pPhORCODx6F1tNZVUveYwHzggd+YczBHlaNy4889Y1/NDFFAj+OXMMx1nKcu4S0pAueXfUJTbA2al8o8VSdFu8203pR1pcnDIMtLS8pkQGqNL/4nHDjhSDbGRg67F8JybiOePjSmNFG/3K+onpxi44+VfVj8vY6gbSOXvqUuBU+A1HYvR6T9THCvafo3s98Jd9yCw805Ta9KLtrjGO+CKpyWaSiuSGLQ8YJ8IRxy3K4rvhU4FAHh7CO+B5OyKhr0A8Fn3AoRD5xGvLP3oHGQiKmRnE1wDs5EfiQ7Cu3+dxrW8zghySjd9D0U7fmF4ZcEgh9ZW8ui+vX/QcIyMvS6ewQ/h7X1DcVxrXsUcpl8vHuB5uAc+lh/Ngxomx2ESyeqHcOLQPV6FC3e8gu2jR6PGuxMPzjkJM/2f5npZkxZPSSnoAnlXN8rtATceAR+Vh6rV1HMADQlxjsPWu293nCaHCYMoBBTnDJZld6f+A6Ah9zaJiKWX+jw2pnJQ0VjV4k4u4btIf3edUGyvoU5GXoLGXCq1QnR4j/I8Tu0mxYeI1zBJTQWpRHffUcrG4ABQ+uGZ8HS/rPpYtFgpSBFfCABBHrysYD9xTVqRAi1jvtA7CjetJ98tmSalhgnQl5eWRiRtiQpltEm8+bELd/8KpR+fDTpVTCUxpHZE7OYnrzU9p5RgkMf4OIWe8u+pPl687WK4B94TF6M+SPzNHCG+D4HSj86Et/N52TFDKZmJKU2+9jyHgvbHUf7OQvh33hcfwYkw5RcZ/TBOTDYNnoBLm19Af6wCxxa/gLtmLUGheyTXy5q0uHw+MCsvkfVC8pSUgud5cOFR5QWxKDbfsgobzlomc4J8VdXEOZzeSQ4TmOsJx41vy05gyt/5SsoRdUMk5pumenxawQbFMW/3y4T6Iwu7wmpQxjIlU43K8epvgh36muq5JR+eDmrcSNm5HsrnWLjnd3ANsYTzleILwSCPvj4KQ0NApGKx6lWusVaUfbCMMCKhKiJuj0x56+j4eeqvx+i0nxDWChR+fjvxMeOop+QBSbGLrq+o93ySX29fDVM2G9fybkJ3XglUTOVvs3QMgiN/3eM3G1sYAfF1GBomm/FULLkBoLWhMHTwrzTn8na9KPvdWBqq1ZQ8Dq5h4TNY1HwD6HCb7mjZwnGY4hTW1uZ6CXnFE/vPxx0tv0cMbnyn+g5cFLrUcM+GSYFWxJGmseDaVXCX6EtxGoamMXPpUtQ2NKK2oREL1z6JhlfehsvvB2LaTYDFGiXRaZq9fIUiSiU730ATWweHfIFhmMUMwywG4GIYZpH4e/zfcgAHRFGdK9yifxKAcP25hsf07/sj/Lt/LTli1Syx2PBWY6NyxLtA9XhB14so3P0bc/Ooov4d7+lVOpaJWVKMTnPNPJVrHBvjMDZVWVeq278pjlqqoYhn0D6RYenzLi0FioqSqYh8AXmDjrKibEf825tB0QdSOidFIVx9msVR1cccYu5AX5my16EeqbdFTI0cHNaQIjCaOaXZK03JYveZKP1A/b5Qkd54k2xrKXkUz0EmthLpMXV9JjmQTGBNDr/44lwvIS/gOOCO3Q/hn10/hYcaw3XTv4OvVfwl18vKPqQ/5C43Flx9PWobGjFvpX5/EcNwHHY+9ZQiXc6ocyONHImKdiQPVysC5eCQhzwc/+eD0G9J/H0NgB8CWJm7pWUPpdGpboiEa//H1Liu4aTcNjmNyKKRSlDPU6Icv7jEnHmS6RoHtbQgQ9LiKsRQAAAYHivBwOGPKB6veH0uqPGuxO+5q99QPmczYheZ6MOUzca1ADC44CFEA4zGleYcvNFp52Ps6Ifh9aYXMxGd9cFBK3o36c19SOmz8Ha/pPrY1Feno/zdk62n0fEReS0hJer65B7HYYozc8kSUD7yrvyBQH+0HBc3N2HT8JdR4W7Dr+ecDCbwsf6FBxDSgtPahkZ4SvRD9kaJhZXpcmacG6lzVdvQiAVXX6+INDm9kxwmGizLzmRZdiaAR8Wf4/9msSwhnszrAAAgAElEQVR7AsuyphqcT1yM7hbrylfZN1cCdYMmndqD4lItQzAXDgSvEmESnp+xCFOSzyuFCFnb8ByQzDBP71uSmXPkMPHqRr9xaXH70ui0+vpkFLoAXIFGBpJOBEXNEaPpZPNZqxQVAWVlPIaH9M9Vfg4zey/dQ0IvJY4z3z+MiqYmDAjnOqIPeQavWhV7YPDp8JG4aMdL6I7W4vDABtw/uwElbqeZbyp8NCJzatRU6qSoqdtpIXV62prWIzaqnR8tg6JkESox0iTt0eT0TnKYqLAs+91cryGfIBuh5gwxuVy49rWR0mMJU6aZyqfyVFwzvml4FGvS2yZThShO0cumqoqH18sbijCVF+xL/Fww93T89P8exL+aryca3Kl1I7kk1ViVil1oQREcLourSGus4k3q4gipo6tjbs6irRei9CPtSK8YnVQuhId/94OgogO6zlgwyGF42Nh7X/37IpNOiMa6NJ5XxRuHpp6cN5psjsMkgSJ105rkPN99Ln65+8+I8AU4vfIBXDF9xYFVr2SScEd7QmihtqERwaXkHOfowAAOufJaoRbJAGJEadt9d2LzLasQGeiXPe4uKUFw6WnqNUoqKnjSeqiFa590nCUHBwIMw+xiGOZThmE+iv9TaFgzDFPIMMxahmGa4+cuydoCM2Q1uIeVIgfk3Vweg/PuURwt+vSyNFehUnBfMp94Nq9yL8xGMaiYga15aS2FSvNPmgbq6ow3chXxF9L4+4fnY1eLHyQzzN/658TPWrvrXSftRrj2bFPzG0fd6JWKXQDAyAytXkw2SYFbqYeS4Ov4V/prUEW5Hv++/5PItauvV0ypS8XT/RKKtl+Dok+vUH1cSjDIYUSz77A93xmu0d0o3vxj9RnGu0FF+lUeUYqkpHLFv/6s+TiQ/JzmQ+NaxyyOs/PZZ8FHjMtsTgY4Dri/5V480nEl3IjiqtByfKvyoVwvKyNQxtq6Gybc0Y4td9yCfzcuwt6nNHpS0TRqGxqx+PlXsODaVZopfC6fkC7X1rSeOKbbX4iDL7mCWKPkqOA5OKTF6SzLHhH/94LK45cDGGBZdjaAbwBYwzAMudOprcgNhoykqJAiRfE0P95djGhx6g4w4G+deHWuVEzT0lSeH3eYUm1/oZ5H+C4erf+h4fFCIV64Lk1ngveUg3enKUCkU2uWaqymil1wnnLilabhwkBUzZm1OY2MG0PJh2fANbhF6OOlebJVxTeyoS+KNihGi0vbU1E1J0ROKMTrRpjocBvcSLmfJiPC3u6X4Gt7TPlAbARTX5uJitfnKB5Sk+FPpbC0WHfu8ncXIz/cJcdhSvDx/erd0Scrw9Ei/Pzz5/Cfwa+i1NWFe2d/FYcWb8z1smwnuPQ0+KprwMdito/NRyPqct9SOA5tTeux4axl2HzrjXD5/Yk1AUg4Pb7qGhx7442obWjUdHjElL3ahkbil56jgufgkDHOAvAQALAsuwPAewDU9a9tx5jJwGmolumjbpRGS47E0JxfYmDB/xpeB2DEqbOrmNv8OJ6BD0ydT4EDxylNJqGeR3ieQ/O0Gt+mXmcsrQ3QNxfTNScrXyojSLWTIkzGxC5obkSjjkWdKW8ehcpX6sgnECzwobm3mprHPfARCrpeQPG2iwCksQHBW7MtSCl5pR+fIw5saIxYjHweDwoVGxhMdX0CQCV1NU1nvfydk4RhOGWPOJ6nNO6pcLyy0vhcTg1THjHS1pbrJWSNnaPzsHLHy2gfn4F5/v/g13MWo8Kr0RxvArP3qScQ7mjP2fyeklJsvft2YQ08j3BHO1rXr8Ps5SvQ8OpGNLz8Fhpe3YiFa5/EzCVCdo+WwyMVgSAJQjgqeA4OlnmUYZhNDMP8lmGYMpXHpwHYLfl9D4BQdpaWCsGAoFzoOX4jokXqstxGUBjgFIXRGReB91aacsi6x6YDAApaH9WUBzZv8Kuk5NlpUKlsRlEUp7rOUIhDRwctNAamvYanCAZ57NtHGdrs139m6T9316hKI1+CAy2KXezZo21CVnY/YD5Vkk9poxEdQkHr36DnQIxO/xl6j1ELCifxdAnKbtR4F8r/0xA/yumOrXV/p2w8HqXvLyU+Tnr+4j0kX2jEYeLxyJvf0T1PGE66jpSxDatZynEPf5r8JTYCyKK2+us36jDlS+PaA7No5wDm5Z7T8Yf268GDxpKKh/E/1ffmekmTmtQaJCCZNifWE7U1rUfzmtUI7++Ar6oa7uJiRAcGVMeTKtzNXr4CW+++HdxYcnfHUcFzcLDMQpZlWxiGKQBwP4AHARizRkxQUWExgy8lBbequAOVlYSUlsovAjM/Af5qzFgVx+nfJ8hdBwIF5LGhTMlLjJFieE317QI8u4Et5wN1pwInPSd7fLRLqO0MFGrNpyRQWICA5PxBCgAo4hhmxi5u/yMQ/BYAoNDvRWH82jYa4DgalZXFKChInj8/XmoVDhcjGDQ2R2VlMQ4+GAiHAZ43sDaK1n4OewrIjxmkfOf1QMPrsjViSHiv0i6XbP6KCsDrBXp6fKis9AGd5Pl5nkJxcfw8EyTme3slsPNPcNcuBwAEAj7yveC1a4XL+p4CDv4W8NnfE8c8biHNkofGPT7oLCBRk6RErFdSu56mXarHDz8c+POd5+K7C9XTWQsK3CgoDyR+Ly8vVDgYhx0GbPg0taF1kvL35GWYU6YUCWOMCCmUrsKa+NrSf/9UvhoCuGRZi9/nBj9CoaxMuW4UCK5HVaUxR0j39ckSjsN0ALF6383Y0L8MNKK4KHgxvlDycq6XdMAiRpHamtbLnJ5wR7ugqudyKxrWBpeeJhNtEH+WOluzl69whB0cHCzAsmxL/P9jDMP8FoCaXPkeANMBiPlL0wCQLSkVuruHwHHmd0wreHlKSDQaRWends9eoxkvYy8txcCCNRgbDKMUwMjIuObYauN27h8A+Ijisb6OFpQBiAzuQ1/qmIMj8AMYHhlTnY+0/tGBDgxJzvfxPABefYzKYnR2Dhq+F+j7RPgHYGR0HMPxMTkuBo6n0dk5KHOYSktdAArx0UcjKC2NGZqns3MQZWXidcNQqIukwHHQfD0Co+MoNDCv9qI2JOYQ7xk9MogKALEYp5i/vj6A7dtj6OwMwz88DtI2AM9TGBwMo7NTWSOuda/E+Ur798ALYHykF4UAhobV3ysA4O4bBamaCgCw80/onP0ACgbHIVZ9RSKReDRH/f0jLOJ/UAmdjUgugs5u4e+49HnFYjHVcQsKgChHNsHHxiIY7h3BlPjvvb0j6OyUp/8FAhRAvPNKenqG0NnJAyiG7+AHMF75VXCdg/D0DkAtnG4KTv76hsPj4HkKfX3Kdfv9R6MIT6GkdjpgILmJ13t9TELTlKWNK8dhOgAIcz6s2vk3tIzNRYDuxS9nnY1q795cLyvvcPn98JSUElP4PCWlcPn9grNjWUZXQEyba16zWhYhAoTaKOlcWo5QbUOj4yA5OKQJwzABAG6WZfsZhqEAnA3gI5VTHwfwEwDvMQwzB8AXAJjrFGuZlGiRjSloBfufhrfnFYynkfria/0LwjWnK44HdvzCwNXmnou/5X8xNOdmwJWMKGSmLFyqkqeUFQek9Tw0AOP1LGLR/969NKAjolpe2G14nZlALd1RWoM1XnkqsOM69WstNz3mZTU2FOLRS433fbTsi+DcJaCj6hkaycEkY/DidBrrNFLrs3stUKSWmqd+rccDFPrJnzf3gNrXj5yKCh5+jTG0CAclMusW67D0IN3T0ekXYrzyVFRQRxlymLTGyiZODZNIvgi928y+8Eys3P4KWsbmYqZvMx6cu9hxlgho9TyiC3xgVl6C2ctX2FIjJKbNkeqVIoMDjhy4g0P2qAbwKsMwmwBsBjAXwAUAEJcYF6vQ7wJQxjBMM4BnAfyYZVl7tj3NksG/WboiA5RSddTT+wYoThlFSIorKA07cd/JyjNRKzTPJBTFqap+1dXxoGkeLS3mnkVSaU7/utOOetTU2JZRKAeSjXFBHVBYeywwB8OzrlQ9T8vQ5V0B4mOFO1MFNAzIilM0huf8kvw4ANfglpQjHGCpj1cKb58LerTFVD1QURH5XNe4fu01RWn0c0pB677xtBAyDdecaWgsY2j1YaIRC8xFfb2xkUq97Tio6jN7lpUGToQpzuwzz0Tz2rW5XoatvNX3Nfyu9XZwcOHkssfwwzrtLxIHqEaX3CUlmLfyUgBQ1AxZgqISDpCvqlp1Tke4wcEhe7As+zmAIwmPHSH5eRjAGdlalxy5waMW7Ugfo4ajWjSFBlIL9iV4BjelMZ8aGkXsGUCtDxMgRApqasz3YiotBYqL49fNtWuV6UHxXMqdJDspwSCP/ftphMOAzweQ9t+1pKV7jnsDFW+qfuxQ+NltGJl1ZXJjIOGIpLdRUPzpJSl9q/jEOtPFNdKMijcOkR3TclQChem/b0ny5GaIln4BA/N/i/HqpfC1/13/AgNQBvowmWl9umj+y1DTccwmToQpzrHXX6/ZgHSi8X9t1+A3rXcBoHB+3VWOs5QG3Ng4APX0OUtI0vlmL1+haELrCDc4ODgoyEYWhNFUY5UIk6/tr5oOk+a0BMNqfMqJ5CVEeoF4zxog8yk7FDjiHEJ6mvn5xbS2nuP/gyHm9nSXmEHUU/IAoLWVIp4DaDsMXOFBGjOmRk7Sa1wr4unbiOJtFyfnsagQpwYd3qc8qPG+DGhEmFIhffxJDXBTObhum+bgY/XfAe+2U1RBvw/TRMNxmCYZUc6N6z//G5p6vw0/PYRbZ52GL5c9m5G5KI8nI+PmG6KqnSl5cpWmsiKJHkwQapDmX36VcIyi4KuuwfzLr3JS8BwcHGRoyXLbja5RquIwAfanyfUf/QzxsYo3D0fZ+9+0dT41vB1PIcBeA0C7CakYYeIp49LiQvNaCrEiBpzHRFOabMBF4B4iG9miLLbYvBaU/eakf2dSxdc/+nH8J7stcF6QrbZjWJPeQaAwfWdNV55cPK8iu6UYPJ8fdUd24jhMEvY+82Sul5AW+8fr8bMdr+Cz8GGo9+7Ag3MXIeRrzth8fESZr26U8qOOSTgJWs5FvhDe32FunRwHT0mpoHYnQS16VNvQiIVrn8Q5mzc79UoODg6qUCmF2bzNKXne/c/BaGqbWg0TANUaJhnceOpIhuYj4el7W5jXwDjdC7damqN007ko3POg5jmhEIfWVgqxGNB73Oua50qRN69N517Yn5IY2HFdookqSfQBQHL9GXCYippXJaKe3ojQJ8pucQ/30BYE3N0ZM+611ruz6Nq0xzdaw5RtCnhjCXS7h7+Q4ZXYR/5bqtmEy883nhHeHzwJlzc/h8FYOb5U+gzunL0MPposYpBrej94P/Gzu6hIkNLOY3xV1abfH5GBflAUBXdJiRM9cnBwsBW7DUd/659B8aLDoz12OLicsChttS3/nt8RHrH+XAo/u1WYWmcMzmewSZIUWcSA7JQEgzyiUQrt7RRiRfMNDx8MchgYoNDfrz2+LjS5x1Gk+EiMVX7d9JCevnc0H7cqdmEWb0r/o0ykeLlp65u/clQaKmu8L8tDM3Daff8kjxZV9nFMxY4apkwQiv4VAd+I7nlPDr2UhdXYg+MwSZkAkQ411u6/EPe2PAgOLny/5iZcUH9VrpdkAF5IceN5RAcGwEcjeaFUSPmUf3jEqJA0lc4ofDSC6NAQFlxzgxM9cnBwsBH7vy9rWq82dN7ItAtUj3t6X9O8jorpG1Cp9OEQzccDn2eu9sfT+6bsd1IUIql4Z86GkKe1WTd8eYpcPR8rno/hg4xIu6cg+3usfN6pYhckx8D+yE32I0FGCcSdd6PU12tvwga2q0u1SzFaw5Sv1AUnjvbcxPQQMkTwG8tyvQRTRDkat+z6A57u+gkKqFGsmnE2GqbYo3CSE9LsbZQuLr8fp6x/FQuuXaVaU6Qm0EC5PUIESQuOw9a7b0db0/oMrt7BweFAIhMpRDQfNjY2ISWv+NPLNS/ztf1NfkCUFdeYbts0bScMAIo8XXDT1gQntEhKogPVPqGeR22tyZ5K+q9JzD9Tcp2Y1kal9fePd5Hb1vIwF/UCoGhCSq7dkopdZGfDMzP9tuzBFd6jclRD9CEAFBeTHR5v35vEx0Sqq42/b/JgT1rBRHL4HIdpgtIbmYqLml/C1pEvotLTgl/PORmzC1P7CxwYJJyWNOuhRBELsaYotQeSmkDDIVdei0VPv6j7TSQKRzg4ODjkP5mxrFyju1SPaxnBdSFjIgol/sy3wyI5kmKkQE9a/LGN56DnhPcSv8ua16bB6LSfajxKARSFSPERGuckcQ03A4954elPrpNUIyYVu8iWw5SxCFOmaph0xp06Nb2NYpf63sWEob6ex7zLNRT88gjHYZKw96kncr0EQ2wZPhYXN7+IvmgVji56Cfce1Igit05n60kG7Yu3Rqdp8NEI3P5CLLjmhrR26aKD+n9wSc6Ukb5JpCa1ANDWtB5PnnIKmhYdjw1nLXOiUQ4ODppkcqdddyc6i1vVheTgiYxcRh4CAaCigtOt54nEPACdrNetrORRUMCnnZIHlw9jVSTFQHFN+uNT0UG4hpQbr6R7KxW70HpPaL1dek54DwOHkGrbVMiR8lq49pyMjOsvKTJ0HmVHc10DjEy/KCvziPh8QHHpxEjLcxymCcbTnefh1t0PI8p7cHbVPbh02kUTtfQqLbhwXNAiLsQQ7mjH1rtvh7vYeh+BdJrFqqXrGR2/rWk9tt59O0ba2gCeTzwXx2lycHAgkznDMXvOh31GYK4ljOXRFmPQtLDDLqS1pXcvtOqYjFKxYR6o2JDh86ViF7xL/W9vYYG2+FQsMBfRkmMMz6n/3rR2H/XeP4MLrGWI6K23y73Y0riZYniuvT07jeyt1Nbmp3BFKgegqT0x4Tjg7j2/wdrOS+ChIrh2+g/wjal/zPWy8gpuLAwKlGqdkZqYg5R0m8XK0vVMjq/WENdJ4XNwcNAit7Uc1ud2DW62cR1JzjnhrxkZV4rWPU+3eW3mEI1RY2sr2XK+cgQDYhfh+u9ZWp158rAQRwM9RywU4vHke0uztJr8pLrGcZgmHLTHeMO5bDIULcGln72AD4dOQrm7A/fPacD8wHv6F05Ikl8u7pISBJeeZurqyOCAap3RKetfxfF33JE47ikpNS333da0HhvOWkZMm0uk6726kSgcoQYpVU8rhc/BwcEhc2gbeby7FGNTGzF48K9Njzxl4wlWF5XXiBEmraxwtbSqUCieypdh0SPerSNOpHWtRsNeIC5aQbsxPuVEaxOYSPHMZ9EHKwSDPIbHArleRk6prZkYwg8TI3EwS9B+H7hIamO93LJ95HDctvthjPM+LCh8C1dO+9EkT8FL/tHgxsZRtuAwdG18S5AgN4Cvqhq1DY2qzsnMJUtQ9EVrX+hi2pwYCRLT5gCozkVaA2nNas8vnRRBBweHyU1ODUeKxsCRfwdNEHEwTI6VUa1Asu1DIQ6joxS6uylUmRgvGOTR1UUjMp6peyGMO3Dowyj74FtwD9kX4VOIXVh+Pc28l/NXVlwdvQgTh+9cfBe+/aXMR0jzlbq6ieEwTWrT2yxGiv6zyYvdZ+OmXY9gnPdh2dTVuHrG5HOWxCiM2hPjxsJgH7jPUH2QSLijPSOiCZlMm1N7fummCDo4OExyMlizY9x4TG8NdLTHlnGyhVZ6lRlpcfl1grHY15eewxQpX6j5OF9QjZEZF1oam/S8jYpd2Ek+SmNrorPeYJBDz/CU7KwlT6mtsb8lQCaYZOZ3euTTjv6De+/Anzp+ARpR/Dx0Ps6oejC7C7DrWyme/qaGr7omoThH2pmKDAidrudffhVxnFQyIZpATJvraE97HrH+qbC21lSKoIODw4FLZiNM2XGYprboN+acKKTbvHZLL/n7PsbpjxkO/hA9J7yvOC6XBLf4emnYA3Kxi8xHDPXe95y30tq4mdqA0Bm3rAzw+Q9sU7ymJpbrJRjiwH6VUph6XO5zq8OcHz9vfhpvDyxBiasH98w+FUcUb8jchKSQlV2pEjyP6kUnEyMoYl2Q1nzNa1ajtqERJz39guGaJrXoz85nn9WsQdJCy5m2wzmrbWjEspdeUsiVOzg4OKjB83nw55uyaw0TLWygRNaEloDaI+J1zfvq0Xlyp+p1nJHXmqIQK5ylPM5Ld+/tv89GxC70JbHtS8kbr1yC/sMeMTFeZtFz8CgKqKvTf/4TLrJmglrHYZp4dG18K6fz7w7PxU+3v4rW8YMwx/8hHpizGJXetsxNSGW+0BQAWtevQ13jqbIIEV3gRd/mTdh69+269UnS6M7Bl1yRFFTQQXpdW9N6vHvDDcJcFqS7tdICHUU7BweHfITzlFu8MvMRJk9vbv/eWkErClFWBgQCQrSl//DHVM9Rcxxqa3nQdFxanC5QvS7GGexOSinP87WtlZ5gbJxUdFIRk2IXuY8wgaIwXk3qSZXGuBmkPjh5vSEjjl5RwHGYJhxGhQUywet9S3Ht5/9AmAugsfxPWDXzO3DTmc3r9FVVZyUNkRsLo/2VlxAbG0sciw4MYO9TTyjqgtRIXaOoRqfnNEmva16zGrGw9RokMW2OhKNo5+DgkE2MGHjh2rMzvAqLJgQfQ9l7kyuKTlFJxbvxqlMNX+fxCE6TViqfXdHEWOFMW8aRIhW7sLoBy5sKn0ws58JIqp8YZbTKaN25aV2fc3hH9GHikSNFhTWtN+Ch1ltAgcfK+stwbu2dGZ9TTIkzI6ighija4C4p0bx/0YEBQ84RaZ1qaDkpqdfZId1d29BIdNLyqf7NwcHhQCD3og+8xZQ819AWS/PlGr11WmleK1ynndZmKCXPANFS4w1ipWj3n7ImdiEn97LimaphMrJe8R6myz/ePUPzcbvS+nibXQeKdyJMEw8uu17uOOfFNZ89jlf6zkQhPYA7D/omjit9IfMT03RCVECv4aoWUtGGeSsvBe0ymDZgYp11jacS63mITork+emda9bRcRTtHBwcJjItdUYEhDKbkuce2mbq/GjgYEvzZAIto9NqE1o9RyvXDqWWMyEVu4gVz8/CYiaGc20G8R5aR7gn6z/+r/QXY2g6u10HJ8I04bDiNFilbSyEn21/BbvH5mN6wTY8MHcxagt2Z2fyFMdQTHFbcO0qVWcguPQ01SiUVCSjec1qcJGI+nwut7WtDY5D6/p1xDojkvOy4OrrFU7W7OUr4PKl7+jIHExH0c7BwSFHWDWiiyr1uwQZHtnilnXJ5h+ZGqbv2JfwdPsDlubKJsEgj74+CkND6o9ThBqfUIhDWxuFKCEL364Ik1X0nERAiDCNhqxuHE7elDwj6003Jc9uhmdervk4xdtcLuKk5E08shUleKe/AVd89iyGuVKcWPYP3HrQ6fDR5tPVtNBz/jbfsgrb7pOn/pGcgYMvuQJ1jcqcbKkzo5Xa5gkENHObteTCteqMzDgvtQ2NOPbGG21xdEQH01G0c3BwyBUk41sP2oC9adwZUzchhpjb0XfUk8YXpTMf7y5GX/EyE+NlBr20LavS4sEgj1iMQlub+vi5VkTUej9IxS5yvZb00B+35/h3TY9qZL2itPyBClcwMUoa3LleQD5R29CIzbfdlNHUvEfaf47ne74HChx+VHsdTir/l/2T0DQWrn0SbU3rsfXu24m1Q3ufegJlCw6TGfximl4qagqC3FgYnz5wr+DQaDhEkcEB+KprVEU1PCWlMjEINcReRyRHyHFYHBwcHHIAIfQwGlphe9pOXR0P9Ns6pO0YkRbXvo7GUSqPG1bJywFSsYtY4UHWB7H/VNuJFc0D5y4BHR2wddyqKn2HKZvPm8qCerIUzhfEm6Xv4Ev9X8zqvGZxIkwStt13Z8acpShHY9XOv+D5nu/DRw/j5pmnZ8ZZAhLPQU/ZDYBhlThSBCk6MKCrLuirqiamz/HgDYlBpNvrKF1ZcQcHB4d8wvpOu5Hr0oswZSJtqq7e9iFtR4wUkCJMNK1uXySvU79vuU7J04usJWqwCE6yvrGf+5S8xsOeM3aiWWfCQM2VXXpjhYV2OTrZj3hNCdVlfU6zOA6ThL3PmEkhME7XeDVW7ngFO0aPQo13Jx6YvQgz/NszMhcgT3HTUnYDjKvEWVWBE+uESOlz0cFBQ+Ok2+soXVlxBwcHBwc5vLvYnnEMGMFTp+Y+bUnPcaiq4uH18sQI01lf/KvqcWmESQ07HaaRGZeYvkZfHdCa2EUS49dmKiVvVuXnGRk3m4IdU8o57Og6MmvzGcFoZGzq1Myuww4ch0lKBqJLHw1+CZc2r8dArALHlazDXbOWoNA9Yvs8UnjwaGtajw1nLUPTouMRGx0lnmvUETItP65SJ6RW+2PGEUun15EdsuIODg4O+YORYiQVJ0NiwXwwa6/6ZUaNPMoFntBs1W5oI8VXWUDLAKRpIXXQrPPg9wNTp5KlxdOpYYoGGNnvw3NutDwWCT2xCz04X9DE2bl+H2TGcbfDKS4v5/HtP75kw2pysDlh0+ZLJnEcJik292H6x/4LcFfLQ+DgxrnVt2Fl8OdZafUUHRjA1rtvT6SfRQb6QalIfptRiVOLEGmJNYhpeHr1RWaENtLpdWSXrLiDg4NDLuDcZbaMI3WG9rZ6CGeZMEp1nDL7yLWhbAyhnsf8H/pQiNy8Nh1jmncFLF+bHCQzYhcJKAqR4iOMLSXH7wOzYiuGe5rZYI6XT+Gxrbk87XFy4jDZLlVuP/m/wiwS/IY9KjwcB9y26/f4V9cF8FJhXD/jO2iseDStMQtra7Hg2lVYcO2qhNNC9L5oWlEXxMdi8JSUpqUSlxohYlZeQow6Ga0RMjp/ur2O7JIVd3BwcMgFfcc8a/uYe/flie6TEQcrl9X+cYw2IbXSxFUrrS296IMN901nCKtiF1J4j9ENgRy/DzImiGBPhGloyIb7k2XRh4lCnnxb5gcHX3IFhlv2oPeD9yyP0R8tx5nco3QAACAASURBVC8+fxw90VpUuPfh5llnomIKD6AUkQFrEj++6hose+kldHYK9T6ik6GmgkcX+IgiCpHBAZz0tLXGuG1N69G8ZjXC+zsU0aPmNatVhR/EGiGpU6Q2DklBDzQN8LzhaJUWtQ2NKC7x48N771N9Dg4ODg75DG9ThEnKnhYPUKEyV66N0glMMMiho8OD0Slfg7/neRPX8WhqyoDogw2Opt77QSZ2EUp7urTWknnMOhNG01vtcJjSHiKO8BxHpl+E0dB5qHjjMLsGntA4DlMK9V9bgv6tW8CFyXU/JLYNH4079vweEb4ARwRew2WhC0DTQCSuQEm5XOBjMVNj0gU+TD3uBDx5yikYaW+XGfoyh0XiBJAcGKvpZ6mOmRg9ApKy3k2LjlfdlZDWCJHGqWs8Fa3r1ykcP7sbws5csgRFXzzRtvEcHBwcsgXnTa2KTt9w3NPiAYxlQpGhqCxl8NhnKHee3I3Kf6t4ijYgpqe1uRoxC8YdplCIQziciT5MdkQcTIhdWHaYjL6Jcu0wmUNPKCQB5da8BUb83inldtXhCwvhvFPB+TLsAU8gHIdJgl7fIi2e6/oe/rr/cgAUzqj8FZZV/q/iHD4Wg8vv1xRhACCcEw7DV1WNqcedIHMmwh3t2HzLKmy+RUjNm718BRauVar7qUWerKafNa9ZrbgnqdEjX1W1rpNGGqdr41uYf/lVxAiWg4ODwwGPq9CmgSj0nPA+Lr9wBC29pP4+2TZK9eezNbJAk2q3dNZgwPgNBgVjs6/XbPNasrFr2OjOEVbFLqxh7F7EvNVwjWdC1CkzuwMUTQHm9tMVlJfbvTYKE81BzSSOwyRBzaDXg+OAX+27H+8NngI3NY7Lgxfg0OKNxPNj4bBm2hwgyIIvfv4VAMCGs5YRz02N9IiQIk9WHRAjCnOzl6/AtntuV0h3Tz3uBNl6Sc/DaUDr4ODgYDO0V/VwLDAHfW4fWloIDpOJNK4h5i4Ub7vQyurMYUNqWc9xb8Dbs8GGxZARHZ/eXgAmhL9ER0uNKJeOqZb5lDxAKXYRKTkSnoEPhRWYWEKk9Ivw9L9DfNzoWL3Hb0T5O1+BK9xifPIMkM0UwkAR4Pfb4DTJsoUch+n/27v3ODmqMuHjv+653yeXycwkPSTBhAOEOwhEBASJooJGXTYL7uLriiu4y/sii6uiCLKLRpcV3lU0aF4WEGVdVhdvGI2KK6isioIieDBcc89kksnMZDKTme5+/6jqmeqequ66dtf0PN/PBzJdXZfTVdXd5+lzznNyJOmDRakJWAsdnGzl75/7Hr8eXkNH7V5uX/G6osESANnSE7VaA5FSaa8z42M89cmbZyRXsEvh7ZebDHO9ay5k+VveMmOdHZsemi5bkSQVQgghwjV65AcZPeLv8heaNc5UKsOOHcErQ4fnnxN4H+4EL2u67QQOLf1b39tns4mSFfbFi7MkEln27fNW3lxXPqfjejV46kPmX+Wp8BYmu5hs8zYfUC6bX7p5WSjlydYvYLLl6OnytBwVyn4jG8MUggTFWyrdy73G4GWPQa6W0EhN1cpDxf35Q8dy9Z8eZs/EERzd9Ev+dcX5zKvbG0oxatumf5ZyNe4ok3GVkc4vuzmY7Lr47fjpT2cWzTo5rNM8VxHMfyWEEG4ppRYopR5SSmml1O+VUt9QSnXZrHe3UmqbUuoJ87+PVKK84O6X62xtGwfVJ2yfS6WyTEw47cNLLSf6MTKzSX099PRkGRz0Vr3q6IC2NvvKeCIRoNUglBqrm66IGXbvTjLUdy2Trcd6Pu7wsXdwcMXHmOg43W8hyyOqDHIhRRapVJZ3ffVhRlb+U4C9ZKfLVE0RT0ASMFm5rLj/aN8lfOyFf2c828RFCzZyw/J3UZsMr9KfHj00Ffy4nTA2LzAJmd0cTHYJGUZ3OXS5M1vJGrt7bJ93Wi6EEGWSBT6ttVZa6+OB54D1Duuu11qfZP53S/mKGK5iLRphOXjk9e5WdFUnmz0Vt1Qqa3TJ87zd9DXZ+5qXQypN+brkATxT/3H2r37M83Gz9QsYXX4d6ablforo4gDFA51Mxl15h064x+Nhy3DfWoKaVCrDtx87h8Ndbwyww/ilFT8w2l7pIkjAZFXbXvqCfGH7Ldy160aSZHh/6u+4tPu20MuRnZzgqU/ezObzVrNl4wYWX/hGmnt7S25XqvteEG66+DX3OAREZiuZ25Yqt3Zu3sQj69ay+bzVPLJubWQtbEKI6qa13qe1/oll0WPA0goVxxX/YyNyXfKMStFYdn6I+843+ooPhbIfw+wJmPr6Mr4Cplx6bsP0314q3eNdb8p7XK4xNLn7aSrxg8+WiYmFF4RVJE/cnqfDi6bP795zXwhtv0EcNq/5RMcr6evLMjCQZPRQeVoW5xIJmCwSRW6OsUwjH3zuv3j0wFpaawb55xUXcVr7T1zt100gNkMmA9ksY7t3sWPTQ5x4zTWs+cljpN7yNsdN/KYND8uJ11xTNCBy21LlRi6j4djuXVPnKcpuiUKIuUEplQSuAr7lsMq1Zre9B5VSx5SxaKHKtWbc8scdM58sexe52TFx7V+s/ndX66VSGQ4M+pu81o6Xlz50wpfpP383YY5DcZcdsHDy2spfr/wTV7zVxE9LULZ+AUPHzcyI7E/x4xe7Bw53XUj/BQdIt62aug67dweo3mctXfLEFMmSZzExPGS7fNvYkdz04lc4lGnnyMbfc8Oyy6lPHna938kh+/26lRkf48nbb+es+89l72M/d1zPb0tNWJZfdBHDQ4eKZucLKxuem1TnQgjhw2eBEeBzNs99BNiptc4opS4HNimljtRau04IvGBBayiFrK+rpavLQxo2U2dnC3S10dUF8+fD3r0NcPxN8PubptZpbWukq8tlGvORma/Ha7laWhpLb5O2z/jn5rh+zpOdhrrDdHW1laxHHnMM/PYp+5WKleXo6RwFLFyYv57n15BoAqC+3t99YlVTU3ofHR1G/Xrfvia6uoCXpn88bWszl4Wgra2Rrq7SwxQAqJ+u4tbWlA4gvJ6nrq42WPhueOo9juu4OXdAyfhy3rwWV+fw+OONfwcOrZjx3IIFre6uw0tG2v3W1iZaA947nZ3NRY/p+pwnwnsf+yUBk4XdXEI/G3wjG3asJ0OSC+bdz7t6gwyk8y83PqhYt7s4BArlSg/uJtW5EEJ4oZS6FVgJXKy1nvFzv9Z6u+Xve5VStwEp4CW3xxgYGCGT8TdGwFrvmJiYpL9/2PN2g4OjTNQY2y1Z0syf/pRleKI9LwP2yPA4/f3uYsDk6AiFU8Bay+WmfjZycLz0a8kcLrkvu310dbXZL3dRLjt795Y+552dTvNb2ZcxZ9686SrZ3r3DWKcqdnutc+oGD9IJHJ7IcKBgW6+vPZ1Ouzp+T08LWqfp7x+jZWyCXMg9PHyI/v5J18crVr7h4TH6+ydc7af98CQN5t/7j7mTeb88z3HdLAnX57iz45VMzDubg+b6xco7OZlxtd8F2eJdvvbvP0h/f+lxh62tCaCV3/0hyesW5z83MDBCR0fpz57WQ+M0ASMjYxzqH57x+tINvdSM7yy5HzA+b5w+Swrfm0Xvy6z394CTZDLh64cr6ZJnseKKK0nUTk9od/fO6/n8jk8D8L7F/+A7WKpr7yiauMF6TCe58UGOKb6rMHFCsTFKdW323RydlgshRDFKqU8ApwJrtdbjDusssfz9eoypJrfbrRu94N1lUqlMXipof8o1QHz2dA8qNqdS8e0sFeLEdNDlK3FAtrxd8mBmavE4mew4tejzXs7x4Ok/4uDKmwKWKBo9PVlqa4Ndh0N97yXd0Mv4orW2z0+2Hud6X9XUq08CpgLZbJbJTC03PH8/m/e/g6bkCJ88ci1ndT5UemMbyYZG1NXvnx67Y3fMyeK/liQbGjnxmmuA8BMnxFWpMUpZhy/p9GH3XSWFEAJAKbUK+DCwGPi5mTL8v8znnlBK5X6rvcccv/Qk8FHgzVpr9z+dx8J0DaavL8vWrckZA9O9DVSPR8C0Y3/pxEjFDJ767UDbW1kDn8kW98PcrIFWtradg0d+GPCbVjzEcSgud5E/ea1UL8HLeymcyKKmxpgLzDqJsFfplpXsO0eTaXR4TyWiubbphqnfo3h2eA3jC1839bicEwA7iVWXPKXUUcA9wAJgALhca/2nch1/y8YN7DnUzQ0vfI2R9DxSDc/y8eWX0Zg85Gt/jd09eWN4etdcyObzVnvL459Mcux1H2L5RRfR3z88ta9i44SqQakxSpPD9k2zmTEjJXu1nQ8hRHS01n/AocaitT7J8ndlUnjldKyCA38IbXepVIbR0QSjoxBV2/zQqi9QP/Awjbv+o8ha5Ul9bbtdsoG9r+0PfHyrlhZoafGetr2rK79uMN79Nlqe/6TPUoTXwuR2H319Gb71rVrSaSKrVAeak6qIsqT/LuLAyQ/QtPVLJd4n7oTTclxMNPved9av6fqxEaQ9/MI63va2P6drc3x6DcXtJ4ANwB1a66OAO4A7y3nwR7ccw3VbvstIupNXdzzIp17xVt/BEomEbfptL5nskg2NHPfhj83Yh5sU37NdqTFKxc5jVPNRCSFERb3pKca7Lgq2j7w5W4zK5+D+wgqQ+wpRwvwB0Gn+nPHF72D4+I1uixRwpfj41e5L+P2usxhdfq3rbZIONbIoK/Pjiy4uuU5Nwl0jaiqVZXIywe7dCWZTF8oopBtyjdLuzsNk5xkl3yduGV0jZ95M4b2FIrq2NS1Tf+4biN/9E5uASSm1CDgFuN9cdD9wit1s61H48Ifhtq2fJUMNf93zca5aUnwC99r29qLjhpwq9AvPfJXrMhWm3J5L8w45jtWyzOnkRBI/CCGqV9Bf2K1d8oxWkP2FAZOnmlXW/H/UFRxj/5m6whQT/qSblgHuAgY/Oro6+LM7/5t04xGR7D8MmdpOhk78Ssn1muvdDbbP3U9Gd7A4VHjtyzBhk3Ex9PvXbGELK9j18pZMpTLs2hXd+Z9sPyGyfef4mccsarEJmIA+YHsuPav57w5zeaSuvrqR9euhqSHNzcvX8dr5DxTfoKaWBAnG9uymtr19RtIGpzFFOzdvYscmd2OhGrt78oKlF77znTk171CpsVq9ay50nN+q0vNRCSFE1LyETRMdZ9guz7Uw7d8fQlUgwM/XriqriST7znyUAycH77IEkG4y5iUeW/xXoeyv0NSv/F664M8QZNvwKszJhLuMidOT1yYCdcnL1ISTet/J7jE1Y1n4rXiVq1739WXJZKILmMYWv4N9qx9j8NTvRHaMwcH8t05UXTG9iNUYpqD8zm+RTsMpp8CPflTLQ2c97bxiIkFdezvp0VEmhg4AxhxLidpa6js7OXzgAM09PZx4zTUsv2hmt4mf3XXnjHE5dmoaGzn52vfn5Zz/2e23247pef6uOznhsktcvtLohZUnv+uyS2hrb+LJ229ndNcu2/P6yo98hF/eeCPpsenzYnfuoihfVOJePoh/GaV8wc2GMs5d3isOY4svpe7A/8xYPn9+lubmLPsHEwU5fT1UtrLmWJ2EcyrtsKTbTqBm5JnIjxOGvj5jfNjQMMzzsf3YGOQ6KFV6wLvbgGnJktzktUk4IUAAXdsG6RHf25dSV5fk/Ft+xIZ753GUPsU4ZlQtTBW4dk4TIIcnQbr1WJL7HonsCBMTCfr7EyyK7AjexSlg2gosUUrVaK3TSqkajIxFW93uwO/8Fp//vFFB+N1Xi7csrXn4Fzyybi0TBw7kLc9OTpJsaGTNw9OtPXb54nNzKdlp7O7JS+LQesa5eftw2nZ0167QctMH5TTfhV+tZ5zLWfefm7fMuv/WM87lmL//0IwEGIXnLqryhS3u5YP4l1HKF1yQMvqd30JEzKGVI5EwKvb79gX5NTxXOSvPL+q5lqG4y7W29O8BPyXevj3BUYE6S7ipC7mrL9Uk3VXAW1pgwYIMW7fGewzTZF03Dz99Ps/3H+SoyI5ivP5KBEy5rpHRMV7TRIlU7YGOkDBSo6+K7AjexSZg0lrvUUo9AVwK3Gf++1utdbjpa4ooliwg1/0ryISpdhPjghEsnf21B4tu29zTw+jOmROFlbv72c7Nm2KVoa9cE+UKIUR1yK/ApVLZGWOYjEqeyx8fcy1MgSqGHratabZd7L9LVTQV2lyldc+eBMwcMlPS1q3JgAFTzszXN9FxOnUHful+Dwn3FfBcV8SspYum996a0VyTydZjqR15ml1dxhj1bduS5GbXDbtLXjbiboXFLF4cdfc181zVNDN42vfo/PUbnNf0eVoTmF1am6YfV1qcxjABXAlcrZR6FrjafFw2xYKenvOMbLKOAUoiUTIZQ5A5lE685pqKz79Uam4kIYQQ5RJOBS+Vsmth8rLvXJe8uFUn3IqmIpbrFvXbrWf62t4uy1lYhk68L7J9T6e0jm8L06JeI4I1WsIMQQKm4WM/l/dY132QbO79UIF05Y2NsGhRhK1M1igo0Bi94qzXJw5i9Qmntf6j1voMrfVR5r+6nMcv1lqzY9ND7Ny8yTboASCTKRlE9K65cHoC20SCxu6eGZnwnCy/6CLf24al2NxIQgghyiBwBSW/EtLXl2XjDy/zv7dsebvkAYz2xX+i9s5OaGnJ8vI2m/qCC1HOo5OdulbhV3anU1r7L//hRW8Kr0BW5nunpTVBe3t+6u0gXefGllye93hoCKbfD5Wp9Pf1RdkiYwk0axoiOUJTUybv+kjSh5hZccWVPH3retvEDLnAINd1LtctjUTCCJZs1rULZoJ0Iat097Mg3RGFEEKEJ6yxEalUhoPjrWzp/AwrBq/1sW8zrXiAFiavv+5na8NMShJNhTY3Pszrr+Qj6lNc8fcnkVkUYQCaayEwA4iJtpOoG34ilF3nkl0cOpSkpfTqtkaO+hSjyz/Agp/OHGHkrYtX4crTk/kaLWGWgCloS9CJt8CTRle/4eEEdId7/bx2bYsi8UOmppVkegTreZ1sPy304wDMn5/hl9uSDB99G21/fD8T6brSG0UsVi1Mlda75kIWX/hGx+dzgYF14linX/uqMYgoNTeSEEKI2SVXsdo/6LPCmM1lUCtndSLMICe6X66dJhAt5tARV/Hi2PlmC1NUKcnzW5jSrccU35OHYCKX7GJoKMA1StaSaXCe59K/6fF2fX3ZcLt8rbp+6s8DQ8mp90WlMhymUln6/s9O9pz9XAR7t7ymiCaUbmprYevWBOPdbzEWVKBrYyEJmCxKzZNkFxjMpSAiyBgsIYQQFZJwruDkuu7kJ35wXznJJo3xIJmGsswxH6mBs55k35k/n3r8i/4rAu2vsBXD/XbZcFs/XMomnbpXuT9+LtnF4FD0aeZLGe9dl78gW6SFKcTAZuhAYqrF7rTUD0Pbrxd9fRm27e1hz4EIEnOXCJIebvx50efd2Jl9beCunWGTgMnCboxOjlNgMJeCiCBjsIQQQvi3bVuCZ56BsFtEuruz1NVlGRz0Vx1Itx7H8NG3MbzqzgCl8Fgpsu3+FzxLXqZ5Oem246YeN7UGG7Vgl4HQjb6+DDt2JEi7m/7IuxkV3uL3lJdgItdiOTRU+erleM/b8x6nW48FIFvbSiqVMbrOmbLZ8Mo7ZNlve8O+0PbrRS5wDbMVLZeprtT98MyO4MnAFy/OMjSUYDg3w4WMYYoXu5TfOU6BQW5ZnFJtR6nS46iEEGIuuvXWep58Eh7/TG5JOGm0k0mjcrJ/fwKW+9h3IsFY37t9lsUvu/KFX6FqbWG6F5cPfufDSaWypNMJ9u5NEE27nREcFEvVfP76n/PjD72q5HqFcskuBg/UwPxgpQzb0KovUNf3N2Qal8xIihBmK97wkGW/FeySB2a2RXMwWXi954rvaOu2usDXfskSo/zbdySnP5YqTAImq2RyRgKH3PJiQULUQcTOzZv42V13MrprV9UHZEIIIWbq6IDnnoOggYFdBa6vL5PXEhLRsIRZp7U1A0Ol13Pid+B9brtduxIUH11URNFsiomCdWau+/iWY30dNpfsYnAwhvdTbSsT888GZl6bMAOmyUn/+xo89SGyE8PM+9260isXMd3ClISjA+1qWu5+KZHgZeu2GjjBXNXnqeg155LasSMRm75wEjBZ2QVLxZaXQW7uo1xXwVzackCCJiGEmCP6+jKMjcHhcQg7ka/RdSzknXrhuVZlt77fSqpzYNHamg0UMPlN7ZzbbteuBLT7P76jRH7Sh2zSnL01UQuMAzA24S8dOhj3U6CkD2Uwo4UpqmyJHtefmP/qUI7b2gqdnVkjeUhYAdOUEi1MW4NHOIunAqYaSAXeXShiErfFQ2O3c1aWSk3OKnMfCSGEyP0ifmjMe8XOeTD/9L7jMOYkbmoC5i1YtChLff10xfzZXcrVdkuWGNd6z+4AlfgiQWhhcHBw5c0cXP4PjC96y9Syw5P+w/JUKsP+wconfShm4cIsTU3T1yaqxBphdRT101LjN+lIUGEcc+HCLA0NWbZvL919tFzkE9KiWKIG/dnbyliSaTL3kRBCiNyYhLGp38/c16DGe/686PN9fZmKjbXwJdQ+XtG97tz4MK+am2Hhwgy7zIAp9GtTcP6ydR2Mrvgo2UQ4QU4qleXgaLwDpkRiOjAFyISY9CEujIApgvu7xPtv+/bgx0wmEyxZMh0wxUF8ShIDxbq4TQwdKGNJps2ltOVCCCHs5cYkjB3ysXGyjomO042/bSo7uWCscoJ3yStX6m2vgiR+2LUrqteU22/hdZ9+bG19SXjMUNbXl4nseoQZK1vv+zDLW29pnKvkfWnMNRVFNb/4axofD29S7R07zBamGGTJk4Ap5uZS2nIhhBD2OjqgvX26hcl/q4NdwBRdBXdWK5o4wZ28SrnH9Ny7g3TJc5P0oQhr64tXqVSGQ4ebfG8fptFl7ydTt8D2OWswG2YrXnt75Sv3YFyHkZEw39dmnvtEedIf9PVl2LZdAqbYqm23H2HptDxqubmPmnt7Ze4jIYSYw5YuhTFzDJP3Pv3O6y9Zko1FhcS98gR3E51nBN6HNRubl2uWSmXZszuEKlrRJpn88ljLF6T1pa8vyxd//Deetinmqa3+5/U5uPLjDLzmBdvnomphqlB1cYawW44TGSMhSKkxkSEdjVQqS39/fLp2SsBUoOe8CzwtL4feNRey9oc/ZM3Dv+Dsrz0owZIQQsxBS5fCIfu51QOpr4f2tkoGTPFs3RrvDZbaGfx3yevryzAWUtemmQrTipP/+Iy78srtNZhetChLoqbOsr2vQnIo9W4mW45iza2P+dtBCdZgNswxTNYWpkqODfR77zk5PP88449EXfEVQ+I3LX9UJGAqsPexn3taLoQQQpTD0qUwdiiaRACd80LdnSfes4rHM8Cyk0plefcXN5qPvHTJizCATRoV3vHeS+yfTyQCHd9vsotCI8fcxv5X/Zr5i6Lp3uc37Xsp+V3yKnevhn0PHTjpq+x71W9Kvv86OsI5blTXxy+Zh6mAZKUTQggRR0uXQro/mkrEvAoGTF6Dv4n2kwPvo1xSqQy/fP50X9sF6yZZZNtEDXvPfYFsbYfjNtZf9/10VwuzdSOqinMqlYFnwt9vFGOY/PxGsGBBfur0wGpaSLesKLlaKC1DiYTv7qxRkRamApKVTgghRBwtXTr9t/dKbPEKx7x58er+UszEgtfOWBaHCpWdxYuzvsqWF3AEGl9jv222fgEknX4zT+QFKUkfgVuYrRtRdc3q6Ym+hamSyVQSifJ0axs4W7P33OlxYmEFy7292Vg1JkvAVECy0gkhhIgja8Dk2dR4FfsaSOe86eVRJ4BINwZ5IbNLfT0sWOh9u44OaA6zdcAV+xYmP8KsqEdV6a+1xIth3vPt7fGp5VsD16iCj0xjrxGA2xwziLq66aA2DklpJGAqkMtK19jdI1nphBBCxEaggCnHodY0f175KiQHV9xQsMR7TS5bMPDcc5e8EFKGu9Xd7f1YiQT09qYjKE0x00F10NaX2dAlLyotzfEpbyUSJ4RzTOP93N1rhCkv9L8ihH0GIwFTgZ2bN7Fl4wbG9uymcVE3K664UoIlIYQQFdfdDTU1UY1hKmMlLxG86jFw7vMMnPUEv9j21hAKFK2eHn8VSD+BVhCTLUcbfzSnqK2FH//Rf91nNnTJA9g7mgJC7tJpid0rPbauEsFmmMfs6mnizf/yTd762e+Htk+/JGCy2Ll5E3/41C2M7d4F2Sxju3fx1C038cxtn6500YQQQsxxiQQ0NhbvWudX1Fny9r3qNzNahYLI1nWQaT6SH/R/0njstY5WxsER3T6HQC9aFG45Sjm07Br2n/Z96DkfgHHsJ3x1I9wuedFV+j+/5SeR7TsOZm8Lk6GvL8O3f/Nm9g6X+c1gQwImC/3Z28hOTsxYvu2b32Dn5k0VKJGIq52bN/HIurVsPm81j6xbK/eHEKIsmhr9Vh6Lb9cY8VyU6ZYVjC+62HxUEKwECF56eo1/PQ+uj3mXPPDfMuVboobJeaunHjY355Z7L781rXjQ2LSrK7pr1dNrTgQd2RgZfy9+sjmcLmiRpqd3cUy/1z7XMleJ8juRgMliYuiA43NbNm4oY0lEnO3cvImnb12f1xL59K3rJWgSQkSuMXDAVLkuQoe73gDAZOuq0Pa5uNcIKuIwKNyJ38Cnp9vYrlLdulpa/J/T+vrwyhFlY2BvbxT3TfACZxoW5+/R5y7DnrzWjYULZ0d3TK8kYHJJ5mGaNtdbV7Zs3EBmfCxvWWZ8TIJqIUTkGhtLr1Nc5QKm8d519J+/i3Tr0aHts8dvhbecXfJ6fG5X5jFMhXItTPHJ+Ra+XEtYmAH3xPxzQthLOOWJKnV6MWG+teKU8EMCJova9nbn59rayliS+JLWFZncWAhROQ1mC5P3rHDFn54MMYgpqqZ55rIAc9X0mq03UXXJm+g802uRZujxGfh0B6jshpHEoNnmqjSbEQAAFkpJREFUUlWbXJfOMJM+TMw7a+rvvaNLQtuvHzU1FT18YNLCFFNHX32tY2icHj00p4ICJ9K6IpMbC1GNlFJHKaV+oZR61vx3pc06NUqpO5RSzymltiilrih3Of2PYSpusvOMSPYbtaaIK/WDJ3+Dfav/J9A+mpr8bTc/lMmE/QejjeYPxYfTgZs1Y6uhIZr2s4kaoz7w3T/6/Igo4xg7NyZbjyvvAc26eFMTLFwYj6BJAiaL3jUXctz1N9oGTdnJiTkVFDiR1hWZ3FiIKrUBuENrfRRwB3CnzTrvAFYAK4HVwE1KqWVlKyHWLnnV3FHKi4grlrWtpJtXRHsMByFkXw9k8rgb+eD96/nvF95e2YKUQRhd8sYXXsjY4ncAkKwx3p+ZrN+LmF+eSsZPe1/zIvtP/1HFjh+XbnkSMBUoNufSXAoKnEjrikxuLES1UUotAk4B7jcX3Q+copTqKlh1HfAlrXVGa90PPAhcUr6SQkN90MpD/AKtis5V42rARdDy+d2+steqoaWVf/ufD5DJzvJ+XcWEOOBm6OT/YHjVFwoPEMq+BwYqdy9k6+ZDjc9m0hDEpVtebaULEEeNi7qNMTo2y+e6FVdcydO3rs/rljcXW1d611woAZIQ1aMP2K61TgNordNKqR3m8n7LekcAL1kev2yuUza5Vofly70GTu7XP/nk8lZQTjwxDfitlAesSJblp3vjGN4TdhjbtbZW7hf2o4/O0NER7PgtLSEVhujSTNfXZzkcwX6POz7ta7vCMVVhZJ7z2zU0iGXL/H6WTL+vTzstzUsvVb59RwImGxIUOMsFCVs2bmBsz24aF3Wz4oorJXgQQggXFixoDbR9fZ0RWPzZ2+uhy0MyolqjAjJvfivML77dOecEq+F2eSkX8NrXtkCNz3EyI0ZZlyyBOofj2pan3jiPnZ3NjudxarvMZPF9lVJvlPGII/B4zYztUqmkt+0ADhu14/r6Wl9lzm3zjW9AMgnz5/ufdPiyy5qoDam2+da3hjxobcyo53W04/0cF8g7z0kjwLnhhiZo8rHfuvwA4fzzg0edq1YF++xxy3oejjuu+DGd7s2uhW1Qa1zrG26Aj30MoLLJ1yRgKrBz86bpxAbJJGQyNHb3SFBgIa0rQogqsxVYopSqMVuXaoDF5nKrl4GlwK/Mx4UtTiUNDIyQyfj7tbirq43DhyepBwaHDjFRN+x6287JDHXA/v2jTKbtt8v1P+zvd79fv9tb+zr27x2B5MxJ491IHhphAcbXtd1xu7rabJd3HE4b5/HAISZq8p+vOfMXkEiQzm2XmQx0bmpGRpkPTKYz7Pewfc3wQWO7SW/bAdQfGKUDGD+cZsjjtoXnLJ2G/v4iGzjtx/x3/35/95NVZ/up1A097vvedJI4PMJCIJPJMBBg34XnbEE6TRLYOzBKtsH7fjsn0lhD1CCvO+j72s9x3BzT7r05td3eYajx1zpXSjKZ8PXDVeXbuGIkL2U2QCYz1bIkAYIQQlQnrfUe4AngUnPRpcBvzXFKVg8A71FKJc3xTWuB/yxfSavH4Mlfr3QRDDZd8tJtq0i3Hju9IPA4l6CTDYvB039M/wUHIthzouDfsJjXzm/mjphlySu/+I21lIDJQlJmCyHEnHUlcLVS6lngavMxSqmHlFKnmet8GXge+BPwGHCz1vqFShR2tptYuCacHc2qiuXsTP4QC4lEJJMNZ+vmM7rs/Qye9t2Q95wbuyPXrlpIlzwLSZkthBBzk9b6j8CMyYi01m+0/J0GripnucIT58Ai7lnywlLOlqY4X+8YSSQ4uPLj4e83G7CFSa5f7EgLk4WkzBZCCFGNElMtMVX2i3dZAx4h3Mq1MEnA5E/83tcSMFnIhKRCCCFmB38Vimy1BRhl6ZIX1jkrZ5e8KrvOs03WDJh8tjAdnv+a8MoiQiEBk4VMSCqEECLeqjGBQAiV+1kRCMb5GogwTSw4D4Bsst7X9qOvuJ6BVz8VZpFEQDKGqYCkzI6XXJp3mfNJCCHETBKEiPgZOm4jNWPbIdngbweJGjJNR4RbqNkkhj+ASMBUZhIAuJdL857LXDi2exdP37oeQM6ZEEJUjfhVjqLh7XUmAnU3lECyomqaSLesqHQpRIikS16BnZs38ci6tWw+bzWPrFvLzs2bQt331DxP2exUABDmMaqJpHkXQggnXoOMKk364Pv1eAkoKnvOAo07i+Ev9ULMRtLCZBF1i0axAGAut5g4tbpJmnchhCgUtOXAuQK9/7TvUzf0eMD9l1v1t6QEa2kSYjaKX6AvAZNF1AGNBAAzFQtSGxd1G61xBSTNuxBC+G1hcjY5bzWT81b7K04QFWkFiV+FTAgRX9Ilz8IxoLGptPsh8zzNVCxIlTTvQgiRb2Le2QBkGpd427Ba52Hybfa02vjrkjd7Xp8Qs4EETBbFApcwxhlJADBTsVY3SfMuhBD5Rpdfx8Crf0+6ZWWlixITZQgMZvU4oNlcdjF3xe++lS55FgvPfBXbvvkN2+fC6JaX216y5E0r1e1O0rwLIYRFIkmmaWmlSxEy6ZJnT1qJhIgLCZgs9j72c8fnwhpnJAFAvhVXXJk3hgmk1U0IISIzq1tL7JQjS16lVds1E2L2kYDJolhQNJfHGUVJWt2EEKIcZlOA4MVseF2zoYxChGf/K39AtqYlwB7i9yOBBEwWTt3DAGnxiJC0ugkhRNTinPQhjDLF8XUJMTdNdp5Z6SKETpI+WNglZQBIveVtUqEXQghRBSSwyFeO8yHnXPgztOpOBk/9TqWLIZAWpjzSPUwIIYSYS8rRXU665Al/xhdfWukiVEYMx1pKwFRAuocJIYQQZRTDypEQQlhJlzwhhBCi6kkrh704d8mTayZEXEjAJIQQQlS7bJyTPviXyAYNKmZDl7wg16y6rreYK+J330rAJIQQQswVVdr9LRvDCtZMs6GMQgg7EjAJIYQQVU+6d9krZxAj10CI2UoCJiGEEEIIIURMxK81VgImIYQQQojIlbESGHhslxDCKhZpxZVSdwMXAHvNRQ9orW+pXImEEEKIalKdSR9mF49BTChBj1xvIcIQi4DJtF5r/blKF0IIIYQQs8VcaEmRoEfMMTFMThOngEmIGXZu3sSWjRsY27ObxkXdrLjiSplYWAghvKr2LloxrGDNNBvKKISwE6eA6Vql1HuB54APa62f8bBtDUAyGezDKOj2UYt7+SDcMu559L956Sv3UFOTpKW3F4CXvnIPdS3NLHr1uRUvXxTiXj6IfxmlfMH5LaNlu5rQClM9Kvs91bwEkmmStfVko7oHW5YCHsrodX0bydp6Yz+Nix33Y7u8qQcOLyVZ2+ju+AHKOlXGJucy2m/XYG7X4/m4ibpm87x0+StzGPdICNd3Nont6yzXdfBxnBnrlqGsfr+nEtky/OqklPoNcITD091AD7BTa51RSl0O/CNwpNY67fIQrwYeCV5SIYQQITgbeLTShYgZ+Z4SQoj48PQ9VZaAySul1ABwitb6JZebNACvBHYCboMsIYQQ4aoBeoFfAeMVLkvcyPeUEEJUnq/vqVh0yVNKLdFabzf/fj3Gl8l2D7sYR37NFEKIOHiu0gWIKfmeEkKIePD8PRWLgAm4RynVDWSAIeDNWuvJCpdJCCGEEEIIMcfFskueEEIIIYQQQsRBstIFEEIIIYQQQoi4koBJCCGEEEIIIRxIwCSEEEIIIYQQDiRgEkIIIYQQQggHEjAJIYQQQgghhIO4pBUvC6XUUcA9wAJgALhca/2ngnVqgH8FLgSywHqt9cYylW8B8GXgFcBh4E/Ae7XW/QXr3Q1cAOw1Fz2gtb6lTGV8ERgz/wP4oNb6+wXrNAP/BpwKTALXaa2/U4ayLQMetCzqBNq11vML1rsJeB+ww1z0M63130ZUpluBtwPLgOO11k+Zy0vei+Z6kd+PdmV0ey+a299NhPdjkXP4IiXuRXO9yO9Hh3O4DBf3o7n9TUR0Txa7lkqpM4E7gSbgReAvtdZ7bPZRkfe0MLj9vKh2fj5P/T5XLfy+//0+Vy2UUg8CyzGmuxkBrtZaPyH3WmlKqRuBm5j+LqyK+2yutTBtAO7QWh8F3IFxIQq9A1gBrARWAzeZFZ9yyAKf1lorrfXxGBNrrXdYd73W+iTzv7IESxZ/Zjn2jAoqcB0wpLVeAVwMbFRKtUZdKK31i5ZynYRRWf2qw+r3WtaNJFgyPQicA7xUsNzNvQjluR/tyujlXoRo70encwil70Uoz/04o4we70eI7p60vZZKqSRwH/C35n34U5yvcUXe02KK28+Laufn89Tvc9XC8/vf73NV5p1a6xO11icDtwJ3mcvlXitCKXUKcCbme7Sa7rM5EzAppRYBpwD3m4vuB05RSnUVrLoO+JLWOmP+mv4gcEk5yqi13qe1/oll0WPA0nIcO2TrMD8MzF9Qfg28oZwFUErVYwQbd5VaN0pa60e11lutyzzci1CG+9GujHG6F+3K51Hk92OpMlbyfixyLU8FxrTWj5rLNwB/7rCbir+n5yqPnxdVzevnqd/non4d5eTz/e/3uaqhtT5gedgBZOReK04p1YARDF5lWVw199mcCZiAPmC71joNYP67w1xudQT5v169bLNO5Mzo+irgWw6rXKuU+r1S6kGl1DFlLBrAV5RSv1NKfV4p1WnzfBzO4ZsxrvdvHJ7/C/M1/EAptbqcBcP9vQgxOJcu7kWo3P1Y6l6EGJxDSt+PUIZ7suBa5p0XrfVeIKmUmtFlsHBdKvS5OEd5+byYi4qdH7/PVSUP73+/z1UVpdRGpdTLwC3AO5F7rZSbgfu01i9allXNfTaXAqbZ5rMY/WY/Z/PcR4AVZvP6N4BN5liXcjhba30i8Eog4VC+OPhrnH/N3wAs11qfAPwz8E2zn7ewV+xehMrdj7PlXoTi9yOU754sdS2FENVL3v8eaK2v0FofAVyP8bksHJg/8p0GfL7SZYnKXAqYtgJLchU589/F5nKrl8nvenSEzTqRMge2rgTWaa0zhc9rrbfnlmut7wVagVQ5ypbrDqG1Hsd4Y5xls1pFz6FSaglwLvAVu+e11ru01hPm35vNsh1XrvLh/l6Eyp/LovciVO5+dHkvQuXPYdH7EcpzT9pcy7zzopRaCGS01vtsNq/45+Ic5uXzYi4qdn78Pld1PL7//T5XlbTWXwbOA7Yh95qTc4FjgBfMhEwp4PsYY7Cr4j6bMwGTNjJrPAFcai66FPitnpn16wHgPUqppNm/dC3wn+Uqp1LqExh9N9eaFUG7dZZY/n49kAa2l6FsLUqpDvPvBPAXGOe00APAe831VmK0AGyKunwW7wS+q7UesHuy4PydhJFxSZenaJ7uRajg/ejmXjTXK/v96OFehJjfjxD9PelwLR8HmpRSrzYfX4lxruxU+hzOWR4/L+acYufH73PlK315+Hj/+32uKiilWpVSfZbHFwP7ALnXHGit12utF2utl2mtl2EEl6/HaJmrivsskc1mK3n8slJKHY2R1nEesB8jraNWSj0EfExr/Wsz8v8c8Dpzs09prb9YpvKtAp4CngUOmYtf0Fq/VSn1BPBGrfUOpdQPgW6MdJdDwAe01o+VoXxHAl8Hasz/ngb+t9Z6Z0H5WoC7gZMxKs//oLX+ZtTls5TzWbNcmyzLrNf4HowvjzRGmtUbtdYPRVSWfwXeBvRgpN0e0FqvcroXbcoa+f1oV0aMwZW296K5TdnuR4fyXYzDvWhTvsjvR6frbD434340l5flnizxufIqjGQOjUynbd1tbheb9/RcV+zzYi7x+Xnq67lqEeD97+u5aqCU6ga+CbRgfN7tw5hK4Tdyr7ljtjJdpI204lVxn82pgEkIIYQQQgghvJgzXfKEEEIIIYQQwisJmIQQQgghhBDCgQRMQgghhBBCCOFAAiYhhBBCCCGEcCABkxBCCCGEEEI4kIBJiCqhlMoqpVZUuhxCCCHiTSm1QSl1g89tv6eUeqfPbW9SSt3nZ1shKqm20gUQYq4y5ym4Qmv9wwoXRQghxCwRxneH1vrKANu+we+2QsxW0sIkhBBCCFEllFLyY7gQIZM3lRAhUEp9CHgPsAjYCnxEa/1f5nPvAa4FUuZzfwm8HzgC+LZSKg3cDPwSuE9rnbLs90XMXxKVUqcD/xc4BmPG9q8D12qtD5fjNQohhKgspdSXmfnd8R/AC8AVwI3Ai8A5SqkHgLOBJuBJ4Cqt9R/M/dwNbNNaf1Qp9RrgPuA24INAGrhea/1vDmX4CcZ31Ual1P8yj/sY8G5gEHif1vp75rrLgbuBU8x1dMG+zgQ+AxwLvAT8H631T5RS84HfmWX+tlKqFXgCuFlrfa/P0yeEb9LCJEQ4nsP4YuoAPg7cp5TqVUpdAtwEXA60A28GBrTWfwW8DFystW7VWn/axTHSGIHWQmA18FrgfWG/ECGEEPFU4rvjXIwf1F5vPv4esBLjh7zfAF8psusejO+vJRiBzx1KqXkui3UGRiC0EPg08P+UUgnzua8Cj5vP/SMwNfZJKbUE+C7wT8B84Drg60qpLq31PuCvgS8ppRZhBHNPSLAkKkVamIQIgdb6AcvDrymlPgycjvHL26e11r8yn9sS4BiPWx6+qJS6E+ML8na/+xRCCFE1btJaH8w90FrflftbKXUTsF8p1aG1PmCz7QRG680k8JBSagRQGK1Cpbyktf6SeZx7gM8D3UqpeuCVwAVa63Hgp0qpb1u2+0vgIa31Q+bjzUqpXwNvBO7RWv/AbCX7EUZAdYKLsggRCQmYhAiBUupyjG53y8xFrRi/qPVhtD6FcYyjMLounAY0Y7x/Hy+6kRBCiLlia+4PpVQNcAtwCdAFZMynFgJ2AdOAGSzljGJ8j7mxK/eH1npUKQXT34H7rUEcRre7PvPvpcAlSqmLLc/XAQ9bHn8R+DvgE1rrAZflESJ00iVPiICUUkuBL2F8qC/QWncCTwEJjC+wVzhsmi14fBAjEMrttwbjiy7nC8AfgZVa63bgevMYQggh5o7C7w675ZcBbwEuwOhqt8xcXs7vjJ3APKVUi2XZEZa/twJf1lp3Wv5r0Vqvh6nvwC8C9wLvk2kzRCVJC5MQwbVgfFH1Ayil3gUcZz63EfiMUupRjD7krwAmtNYvAbuBIy37eRZoVEq9CfgBRkDUYHm+DRgCRpRSRwNX5Y4phBBizij87rDTBowDAxg/xH0i6kIV0lq/ZHax+7hS6nqMbuoXA98yV7kP+JVS6vXADzFal84Etmitt2F8B2YxxjJ9ELhXKXW21jpd5pcihLQwCRGU1vpp4F+AX2B8kR0P/Mx87gGMbhFfBYaBBzH6YgN8EvioUmpQKXWd2a/8fRhB1naMFqdtlkNdh/Gr4TBGi9bXon1lQgghYijvu8NhnXsxur9tB57G3VikKFyGkRRiH0YGv6mkDVrrrRitYNdj/Pi3FfgAkFRKnYrRzf1yM0D6FEbw9KGyll4IUyKbdWrZFUIIIYQQQoi5TVqYhBBCCCGEEMKBBExCCCGEEEII4UACJiGEEEIIIYRwIAGTEEIIIYQQQjiQgEkIIYQQQgghHEjAJIQQQgghhBAOJGASQgghhBBCCAcSMAkhhBBCCCGEAwmYhBBCCCGEEMLB/wf42uremq0mJAAAAABJRU5ErkJggg==\n"},"metadata":{}}]},{"metadata":{"_uuid":"2fc9dc56cc1220a8a07f69ab52dad35af7800572"},"cell_type":"markdown","source":"### Gradient Boosting\n\nThe last model is a gradient boosting decision tree. It's not possible to use GridSearchCV with early stopping (lightgbm), so I am using a custom function for random search."},{"metadata":{"trusted":true,"_uuid":"e7e09e34f28daac2d4e7dca1f58f9882ef4f3e8a","_kg_hide-input":true},"cell_type":"code","source":"fixed_params = {\n    'objective': 'regression_l1',\n    'boosting': 'gbdt',\n    'verbosity': -1,\n    'random_seed': 19,\n    'n_estimators': 80000,\n    'max_depth': -1\n}\n\nparam_grid = {\n    'learning_rate': [0.1, 0.05, 0.01, 0.005],\n    'num_leaves': list(range(2, 60, 2)),\n    'feature_fraction': [0.8, 0.85, 0.9, 0.95, 1],\n    'subsample': [0.8, 0.9, 0.95, 1],\n    'lambda_l1': [0, 0.2, 0.4, 0.6, 0.8, 2, 5],\n    'lambda_l2': [0, 0.2, 0.4, 0.6, 0.8, 2, 5],\n    'min_data_in_leaf': [10, 20, 40, 60, 100],\n    'min_gain_to_split': [0, 0.001, 0.01, 0.1],\n}\n\nbest_score = 999\ndataset = lgb.Dataset(X_train, label=target)  # no need to scale features\n\nfor i in range(600):\n    params = {k: random.choice(v) for k, v in param_grid.items()}\n    params.update(fixed_params)\n    result = lgb.cv(params, dataset, nfold=10, early_stopping_rounds=100,\n                    stratified=False)\n    \n    if result['l1-mean'][-1] < best_score:\n        best_score = result['l1-mean'][-1]\n        best_params = params\n        best_nrounds = len(result['l1-mean'])","execution_count":18,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"08fc682c7ebff2403000a97cf23ea30cda98a65d"},"cell_type":"code","source":"print(\"Best mean score: {:.4f}, num rounds: {}\".format(best_score, best_nrounds))\nprint(best_params)\ngb_pipe = make_pipeline(lgb.LGBMRegressor(**best_params))\ngb_oof = make_predictions(X_train, target, gb_pipe)","execution_count":null,"outputs":[{"output_type":"stream","text":"Best mean score: 2.0061, num rounds: 849\n{'learning_rate': 0.01, 'num_leaves': 6, 'feature_fraction': 0.85, 'subsample': 1, 'lambda_l1': 0.6, 'lambda_l2': 0.2, 'min_data_in_leaf': 20, 'min_gain_to_split': 0, 'objective': 'regression_l1', 'boosting': 'gbdt', 'verbosity': -1, 'random_seed': 19, 'n_estimators': 80000, 'max_depth': -1}\n","name":"stdout"}]},{"metadata":{"_uuid":"1b837edfd77f9eff4b45f452495186d6a2cc11ad"},"cell_type":"markdown","source":"Now let's have a look at the <b>feature importance</b>:"},{"metadata":{"_kg_hide-input":true,"trusted":true,"_uuid":"36c5fa96d7d353285006f7aef366d36eedc41be1"},"cell_type":"code","source":"def plot_feature_importance(x, y, columns):\n    importance_frame = pd.DataFrame()\n    for (train_index, valid_index) in fold_generator(x, y):\n        reg = lgb.LGBMRegressor(**best_params)\n        reg.fit(x.iloc[train_index], y.iloc[train_index],\n                early_stopping_rounds=100, verbose=False,\n                eval_set=[(x.iloc[train_index], y.iloc[train_index]),\n                          (x.iloc[valid_index], y.iloc[valid_index])])\n        fold_importance = pd.DataFrame()\n        fold_importance[\"feature\"] = columns\n        fold_importance[\"gain\"] = reg.booster_.feature_importance(importance_type='gain')\n        #fold_importance[\"split\"] = reg.booster_.feature_importance(importance_type='split')\n        importance_frame = pd.concat([importance_frame, fold_importance], axis=0)\n        \n    mean_importance = importance_frame.groupby('feature').mean().reset_index()\n    mean_importance.sort_values(by='gain', ascending=True, inplace=True)\n    trace = go.Bar(y=mean_importance.feature, x=mean_importance.gain,\n                   orientation='h', marker=dict(color='rgb(49,130,189)'))\n\n    layout = go.Layout(\n        title='Feature importance', height=800, width=600,\n        showlegend=False,\n        xaxis=dict(\n            title='Importance by gain',\n            titlefont=dict(size=14, color='rgb(107, 107, 107)'),\n            domain=[0.15, 1]\n        ),\n    )\n\n    fig = go.Figure(data=[trace], layout=layout)\n    iplot(fig)\n    \nplot_feature_importance(X_train, target, X_train.columns)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"7bf1e7fc30de8ba4e4a609af578f562418f5ed4b"},"cell_type":"markdown","source":"## 7. Submission"},{"metadata":{"trusted":true,"_uuid":"eca824c317d8aa9f14aba2ed04f6d08e783c1827"},"cell_type":"code","source":"X_test = make_test()\nX_test = X_test[X_train.columns]  # feature selection\ngb_oof, gb_sub = make_predictions(X_train, target, gb_pipe,\n                                  test=X_test, plot=False)\nsubmission = pd.read_csv('../input/sample_submission.csv')\nsubmission['time_to_failure'] = gb_sub\nsubmission.to_csv('submission_gb.csv', index=False)\nsubmission.time_to_failure.describe()","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}