projet.ipynb 445 KB
Newer Older
Delaunay Marina's avatar
code    
Delaunay Marina committed
1
{"nbformat":4,"nbformat_minor":0,"metadata":{"accelerator":"GPU","colab":{"name":"Copie de projet.ipynb","provenance":[{"file_id":"1bv09vf91ZdjmAUGEbxEsQg4L6kNbBtC6","timestamp":1621426807839}],"collapsed_sections":[],"toc_visible":true},"kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"markdown","metadata":{"id":"oeRLDwmhmzdk"},"source":["# **Projet ML : Prédiction de box-office**"]},{"cell_type":"markdown","metadata":{"id":"_cPjr7oPn05u"},"source":["https://www.kaggle.com/hoangdang89/ticket-box-prediction"]},{"cell_type":"markdown","metadata":{"id":"q2apPRTaDHu5"},"source":["https://github.com/busyML/Box-Office-Predictions/blob/master/Predicting_Film_Revenue_A_Soft_Intro_to_Neural_Networks.ipynb "]},{"cell_type":"markdown","metadata":{"id":"Qq-eCRRHnGxl"},"source":["**Chargement des bibliothèques**"]},{"cell_type":"code","metadata":{"id":"kszDiG1ehfXj","executionInfo":{"status":"ok","timestamp":1621426590795,"user_tz":-120,"elapsed":731,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["import numpy as np\n","import os, sys, sklearn\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","\n","\n","import pandas as pd\n","pd.options.display.float_format = '{:,.4f}'.format\n","\n","import warnings\n","warnings.filterwarnings(action=\"ignore\", message=\"^internal gelsd\")\n","\n","import csv"],"execution_count":1,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"Qn-Z2DCznVGm"},"source":["**Chargement des données**"]},{"cell_type":"code","metadata":{"id":"1M2LAN6MhsPO","colab":{"base_uri":"https://localhost:8080/","height":643},"executionInfo":{"status":"ok","timestamp":1621426591806,"user_tz":-120,"elapsed":1736,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"2ef5c815-889d-4746-de4d-37303de77257"},"source":["train_set = pd.read_csv('/content/train.csv')\n","test_set = pd.read_csv('/content/test.csv')\n","\n","train_set.head()"],"execution_count":2,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>id</th>\n","      <th>belongs_to_collection</th>\n","      <th>budget</th>\n","      <th>genres</th>\n","      <th>homepage</th>\n","      <th>imdb_id</th>\n","      <th>original_language</th>\n","      <th>original_title</th>\n","      <th>overview</th>\n","      <th>popularity</th>\n","      <th>poster_path</th>\n","      <th>production_companies</th>\n","      <th>production_countries</th>\n","      <th>release_date</th>\n","      <th>runtime</th>\n","      <th>spoken_languages</th>\n","      <th>status</th>\n","      <th>tagline</th>\n","      <th>title</th>\n","      <th>Keywords</th>\n","      <th>cast</th>\n","      <th>crew</th>\n","      <th>revenue</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>1</td>\n","      <td>[{'id': 313576, 'name': 'Hot Tub Time Machine ...</td>\n","      <td>14000000</td>\n","      <td>[{'id': 35, 'name': 'Comedy'}]</td>\n","      <td>NaN</td>\n","      <td>tt2637294</td>\n","      <td>en</td>\n","      <td>Hot Tub Time Machine 2</td>\n","      <td>When Lou, who has become the \"father of the In...</td>\n","      <td>6.5754</td>\n","      <td>/tQtWuwvMf0hCc2QR2tkolwl7c3c.jpg</td>\n","      <td>[{'name': 'Paramount Pictures', 'id': 4}, {'na...</td>\n","      <td>[{'iso_3166_1': 'US', 'name': 'United States o...</td>\n","      <td>2/20/15</td>\n","      <td>93.0000</td>\n","      <td>[{'iso_639_1': 'en', 'name': 'English'}]</td>\n","      <td>Released</td>\n","      <td>The Laws of Space and Time are About to be Vio...</td>\n","      <td>Hot Tub Time Machine 2</td>\n","      <td>[{'id': 4379, 'name': 'time travel'}, {'id': 9...</td>\n","      <td>[{'cast_id': 4, 'character': 'Lou', 'credit_id...</td>\n","      <td>[{'credit_id': '59ac067c92514107af02c8c8', 'de...</td>\n","      <td>12314651</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2</td>\n","      <td>[{'id': 107674, 'name': 'The Princess Diaries ...</td>\n","      <td>40000000</td>\n","      <td>[{'id': 35, 'name': 'Comedy'}, {'id': 18, 'nam...</td>\n","      <td>NaN</td>\n","      <td>tt0368933</td>\n","      <td>en</td>\n","      <td>The Princess Diaries 2: Royal Engagement</td>\n","      <td>Mia Thermopolis is now a college graduate and ...</td>\n","      <td>8.2489</td>\n","      <td>/w9Z7A0GHEhIp7etpj0vyKOeU1Wx.jpg</td>\n","      <td>[{'name': 'Walt Disney Pictures', 'id': 2}]</td>\n","      <td>[{'iso_3166_1': 'US', 'name': 'United States o...</td>\n","      <td>8/6/04</td>\n","      <td>113.0000</td>\n","      <td>[{'iso_639_1': 'en', 'name': 'English'}]</td>\n","      <td>Released</td>\n","      <td>It can take a lifetime to find true love; she'...</td>\n","      <td>The Princess Diaries 2: Royal Engagement</td>\n","      <td>[{'id': 2505, 'name': 'coronation'}, {'id': 42...</td>\n","      <td>[{'cast_id': 1, 'character': 'Mia Thermopolis'...</td>\n","      <td>[{'credit_id': '52fe43fe9251416c7502563d', 'de...</td>\n","      <td>95149435</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>3</td>\n","      <td>NaN</td>\n","      <td>3300000</td>\n","      <td>[{'id': 18, 'name': 'Drama'}]</td>\n","      <td>http://sonyclassics.com/whiplash/</td>\n","      <td>tt2582802</td>\n","      <td>en</td>\n","      <td>Whiplash</td>\n","      <td>Under the direction of a ruthless instructor, ...</td>\n","      <td>64.3000</td>\n","      <td>/lIv1QinFqz4dlp5U4lQ6HaiskOZ.jpg</td>\n","      <td>[{'name': 'Bold Films', 'id': 2266}, {'name': ...</td>\n","      <td>[{'iso_3166_1': 'US', 'name': 'United States o...</td>\n","      <td>10/10/14</td>\n","      <td>105.0000</td>\n","      <td>[{'iso_639_1': 'en', 'name': 'English'}]</td>\n","      <td>Released</td>\n","      <td>The road to greatness can take you to the edge.</td>\n","      <td>Whiplash</td>\n","      <td>[{'id': 1416, 'name': 'jazz'}, {'id': 1523, 'n...</td>\n","      <td>[{'cast_id': 5, 'character': 'Andrew Neimann',...</td>\n","      <td>[{'credit_id': '54d5356ec3a3683ba0000039', 'de...</td>\n","      <td>13092000</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>4</td>\n","      <td>NaN</td>\n","      <td>1200000</td>\n","      <td>[{'id': 53, 'name': 'Thriller'}, {'id': 18, 'n...</td>\n","      <td>http://kahaanithefilm.com/</td>\n","      <td>tt1821480</td>\n","      <td>hi</td>\n","      <td>Kahaani</td>\n","      <td>Vidya Bagchi (Vidya Balan) arrives in Kolkata ...</td>\n","      <td>3.1749</td>\n","      <td>/aTXRaPrWSinhcmCrcfJK17urp3F.jpg</td>\n","      <td>NaN</td>\n","      <td>[{'iso_3166_1': 'IN', 'name': 'India'}]</td>\n","      <td>3/9/12</td>\n","      <td>122.0000</td>\n","      <td>[{'iso_639_1': 'en', 'name': 'English'}, {'iso...</td>\n","      <td>Released</td>\n","      <td>NaN</td>\n","      <td>Kahaani</td>\n","      <td>[{'id': 10092, 'name': 'mystery'}, {'id': 1054...</td>\n","      <td>[{'cast_id': 1, 'character': 'Vidya Bagchi', '...</td>\n","      <td>[{'credit_id': '52fe48779251416c9108d6eb', 'de...</td>\n","      <td>16000000</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>5</td>\n","      <td>NaN</td>\n","      <td>0</td>\n","      <td>[{'id': 28, 'name': 'Action'}, {'id': 53, 'nam...</td>\n","      <td>NaN</td>\n","      <td>tt1380152</td>\n","      <td>ko</td>\n","      <td>마린보이</td>\n","      <td>Marine Boy is the story of a former national s...</td>\n","      <td>1.1481</td>\n","      <td>/m22s7zvkVFDU9ir56PiiqIEWFdT.jpg</td>\n","      <td>NaN</td>\n","      <td>[{'iso_3166_1': 'KR', 'name': 'South Korea'}]</td>\n","      <td>2/5/09</td>\n","      <td>118.0000</td>\n","      <td>[{'iso_639_1': 'ko', 'name': '한국어/조선말'}]</td>\n","      <td>Released</td>\n","      <td>NaN</td>\n","      <td>Marine Boy</td>\n","      <td>NaN</td>\n","      <td>[{'cast_id': 3, 'character': 'Chun-soo', 'cred...</td>\n","      <td>[{'credit_id': '52fe464b9251416c75073b43', 'de...</td>\n","      <td>3923970</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["   id  ...   revenue\n","0   1  ...  12314651\n","1   2  ...  95149435\n","2   3  ...  13092000\n","3   4  ...  16000000\n","4   5  ...   3923970\n","\n","[5 rows x 23 columns]"]},"metadata":{"tags":[]},"execution_count":2}]},{"cell_type":"code","metadata":{"id":"WClsRqWLkz-e","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426591807,"user_tz":-120,"elapsed":1733,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"a5fd5cc4-9e91-4689-c9ba-a9da10146867"},"source":["test_set.info()"],"execution_count":3,"outputs":[{"output_type":"stream","text":["<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 4398 entries, 0 to 4397\n","Data columns (total 22 columns):\n"," #   Column                 Non-Null Count  Dtype  \n","---  ------                 --------------  -----  \n"," 0   id                     4398 non-null   int64  \n"," 1   belongs_to_collection  877 non-null    object \n"," 2   budget                 4398 non-null   int64  \n"," 3   genres                 4382 non-null   object \n"," 4   homepage               1420 non-null   object \n"," 5   imdb_id                4398 non-null   object \n"," 6   original_language      4398 non-null   object \n"," 7   original_title         4398 non-null   object \n"," 8   overview               4384 non-null   object \n"," 9   popularity             4398 non-null   float64\n"," 10  poster_path            4397 non-null   object \n"," 11  production_companies   4140 non-null   object \n"," 12  production_countries   4296 non-null   object \n"," 13  release_date           4397 non-null   object \n"," 14  runtime                4394 non-null   float64\n"," 15  spoken_languages       4356 non-null   object \n"," 16  status                 4396 non-null   object \n"," 17  tagline                3535 non-null   object \n"," 18  title                  4395 non-null   object \n"," 19  Keywords               4005 non-null   object \n"," 20  cast                   4385 non-null   object \n"," 21  crew                   4376 non-null   object \n","dtypes: float64(2), int64(2), object(18)\n","memory usage: 756.0+ KB\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"L1LGsw7woOxI","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426591809,"user_tz":-120,"elapsed":1730,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"dfdf5924-baa7-42bf-aad6-7e287a36725d"},"source":["train_set.info()"],"execution_count":4,"outputs":[{"output_type":"stream","text":["<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 3000 entries, 0 to 2999\n","Data columns (total 23 columns):\n"," #   Column                 Non-Null Count  Dtype  \n","---  ------                 --------------  -----  \n"," 0   id                     3000 non-null   int64  \n"," 1   belongs_to_collection  604 non-null    object \n"," 2   budget                 3000 non-null   int64  \n"," 3   genres                 2993 non-null   object \n"," 4   homepage               946 non-null    object \n"," 5   imdb_id                3000 non-null   object \n"," 6   original_language      3000 non-null   object \n"," 7   original_title         3000 non-null   object \n"," 8   overview               2992 non-null   object \n"," 9   popularity             3000 non-null   float64\n"," 10  poster_path            2999 non-null   object \n"," 11  production_companies   2844 non-null   object \n"," 12  production_countries   2945 non-null   object \n"," 13  release_date           3000 non-null   object \n"," 14  runtime                2998 non-null   float64\n"," 15  spoken_languages       2980 non-null   object \n"," 16  status                 3000 non-null   object \n"," 17  tagline                2403 non-null   object \n"," 18  title                  3000 non-null   object \n"," 19  Keywords               2724 non-null   object \n"," 20  cast                   2987 non-null   object \n"," 21  crew                   2984 non-null   object \n"," 22  revenue                3000 non-null   int64  \n","dtypes: float64(2), int64(3), object(18)\n","memory usage: 539.2+ KB\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"4r0wQmXOyfC6"},"source":["## **Partie 1 : Visualisation des données**"]},{"cell_type":"code","metadata":{"id":"Eslf8K19oTQr","colab":{"base_uri":"https://localhost:8080/","height":566},"executionInfo":{"status":"ok","timestamp":1621426592635,"user_tz":-120,"elapsed":2551,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"a81d820f-5081-45ee-9ef1-cab1d6297154"},"source":["# plot histogram for all columns\n","train_set.hist(bins=50, figsize=(12,9)) # bins is number of groups of values\n","plt.show()"],"execution_count":5,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 864x648 with 6 Axes>"]},"metadata":{"tags":[],"needs_background":"light"}}]},{"cell_type":"code","metadata":{"id":"dM7vGbzHhFWT","executionInfo":{"status":"ok","timestamp":1621426592638,"user_tz":-120,"elapsed":2551,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["most_popular_movies = train_set.sort_values('popularity', ascending=False).head(n=20)\n","most_popular_movies['revenue(million)'] = most_popular_movies['revenue'].apply(lambda x : x//1000000)\n","most_popular_movies['budget(million)'] = most_popular_movies['budget'].apply(lambda x : x//1000000)"],"execution_count":6,"outputs":[]},{"cell_type":"code","metadata":{"id":"hBGt7Gl8hJA7","colab":{"base_uri":"https://localhost:8080/","height":621},"executionInfo":{"status":"ok","timestamp":1621426593037,"user_tz":-120,"elapsed":2947,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"33b53e2a-91f8-46eb-d81e-5ed4d45b9d4d"},"source":["plt.figure(figsize=(12, 10))\n","ax = sns.barplot(y='original_title', x='popularity', data=most_popular_movies, order=most_popular_movies.sort_values('popularity', ascending=False).original_title, orient='h')\n","for p in ax.patches:\n","        ax.annotate('{}'.format(int(p.get_width())), (p.get_width(), p.get_y()+0.5), fontsize=12)\n","plt.title('Top 20 Most Popular Movies', fontsize=12)\n","plt.ylabel('')\n","plt.show()"],"execution_count":7,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 864x720 with 1 Axes>"]},"metadata":{"tags":[],"needs_background":"light"}}]},{"cell_type":"code","metadata":{"id":"LV6Iujt-M1gc","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426593038,"user_tz":-120,"elapsed":2945,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"1346dc9a-2435-4127-973f-9192daaa95e3"},"source":["label_column = \"popularity\"\n","\n","info = train_set[label_column].copy()\n","info.describe()"],"execution_count":8,"outputs":[{"output_type":"execute_result","data":{"text/plain":["count   3,000.0000\n","mean        8.4633\n","std        12.1040\n","min         0.0000\n","25%         4.0181\n","50%         7.3749\n","75%        10.8910\n","max       294.3370\n","Name: popularity, dtype: float64"]},"metadata":{"tags":[]},"execution_count":8}]},{"cell_type":"code","metadata":{"id":"AOpKw-YYabcp","executionInfo":{"status":"ok","timestamp":1621426593039,"user_tz":-120,"elapsed":2942,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["train_set['popularity'] = train_set['popularity'].mask(train_set['popularity'] >= 40, 40)"],"execution_count":9,"outputs":[]},{"cell_type":"code","metadata":{"id":"pT6ZBbSXybjB","executionInfo":{"status":"ok","timestamp":1621426593040,"user_tz":-120,"elapsed":2941,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["most_popular_movies = train_set.sort_values('popularity', ascending=False).head(n=20)\n","most_popular_movies['revenue(million)'] = most_popular_movies['revenue'].apply(lambda x : x//1000000)\n","most_popular_movies['budget(million)'] = most_popular_movies['budget'].apply(lambda x : x//1000000)"],"execution_count":10,"outputs":[]},{"cell_type":"code","metadata":{"id":"-2xhm6vqyozt","colab":{"base_uri":"https://localhost:8080/","height":621},"executionInfo":{"status":"ok","timestamp":1621426593334,"user_tz":-120,"elapsed":3232,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"9707ef09-27fa-478a-e44a-da17725ea4fd"},"source":["plt.figure(figsize=(12, 10))\n","ax = sns.barplot(y='original_title', x='popularity', data=most_popular_movies, order=most_popular_movies.sort_values('popularity', ascending=False).original_title, orient='h')\n","for p in ax.patches:\n","        ax.annotate('{}'.format(int(p.get_width())), (p.get_width(), p.get_y()+0.5), fontsize=12)\n","plt.title('Top 20 Most Popular Movies', fontsize=12)\n","plt.ylabel('')\n","plt.show()"],"execution_count":11,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 864x720 with 1 Axes>"]},"metadata":{"tags":[],"needs_background":"light"}}]},{"cell_type":"code","metadata":{"id":"7LymUlYpZH_e","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426593335,"user_tz":-120,"elapsed":3229,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"939d8a6f-34e0-4340-85b4-405fcdd61ccb"},"source":["label_column = \"popularity\"\n","\n","info = train_set[label_column].copy()\n","info.describe()"],"execution_count":12,"outputs":[{"output_type":"execute_result","data":{"text/plain":["count   3,000.0000\n","mean        7.9653\n","std         5.8379\n","min         0.0000\n","25%         4.0181\n","50%         7.3749\n","75%        10.8910\n","max        40.0000\n","Name: popularity, dtype: float64"]},"metadata":{"tags":[]},"execution_count":12}]},{"cell_type":"markdown","metadata":{"id":"7-Ng-JGIoqBN"},"source":["**Création du jeu de données**"]},{"cell_type":"code","metadata":{"id":"VLQ3tppNhl_1","executionInfo":{"status":"ok","timestamp":1621426593336,"user_tz":-120,"elapsed":3227,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["SPLITTER = \" \"\n","COLUMN_EXCLUDE_PATTERN = \"id|popularity|belongs_to_collection|homepage|tagline|revenue\"\n","\n","TEXT_FIELDS =[  (\"genres\", \"name\"), \n","                (\"production_companies\", \"id\"),                \n","                (\"production_countries\", \"iso_3166_1\"),\n","                (\"spoken_languages\", \"name\"),\n","                (\"Keywords\", \"name\"),\n","                (\"cast\", \"name\"),                         \n","                (\"crew\", \"name\"),              \n","             ]\n","\n","TEXT_FIELDS2 =[ (\"production_companies\", \"name\"),\n","                (\"cast\", \"character\"),  \n","                (\"cast\", \"job\"),  \n","                (\"cast\", \"profile_path\"),\n","                (\"crew\", \"job\"),\n","                (\"crew\", \"department\"),\n","             ]"],"execution_count":13,"outputs":[]},{"cell_type":"code","metadata":{"id":"TczD8c7Dho9O","executionInfo":{"status":"ok","timestamp":1621426593336,"user_tz":-120,"elapsed":3225,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["label_column = \"popularity\"\n","X_train = train_set.copy().drop(train_set.filter(regex=COLUMN_EXCLUDE_PATTERN), axis=1)\n","y_train = train_set[label_column].copy()"],"execution_count":14,"outputs":[]},{"cell_type":"code","metadata":{"id":"R4-oCu_uozfn","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426593337,"user_tz":-120,"elapsed":3224,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"47d4b714-4ae7-46b0-e5bc-cc6679ac0cba"},"source":["X_train.shape"],"execution_count":15,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(3000, 16)"]},"metadata":{"tags":[]},"execution_count":15}]},{"cell_type":"markdown","metadata":{"id":"YM1OzSRHnfcY"},"source":["## **Partie 2 : Pré-Traitement des données**"]},{"cell_type":"markdown","metadata":{"id":"yPjVlz-RpFEy"},"source":["Visualisation des variables : budget, le titre original, la popularité et la langue original."]},{"cell_type":"code","metadata":{"id":"6EEyhUOjh00B","colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1621426593338,"user_tz":-120,"elapsed":3221,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"66e56065-0a6a-48f5-88de-417b80c6a2ed"},"source":["from sklearn.base import BaseEstimator, TransformerMixin\n","\n","class LimitedColumnsFilter(BaseEstimator, TransformerMixin):\n","    def __init__(self, filters):\n","        self.filters = filters   \n","\n","    def fit(self, X, y=None):\n","        return self\n","\n","    def transform(self, X, y=None):        \n","        return X.copy().filter(items=self.filters)\n","\n","filters = ('budget', 'original_title','revenue', 'original_language')\n","result = LimitedColumnsFilter(filters).transform(X_train)\n","result.head()"],"execution_count":16,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>budget</th>\n","      <th>original_title</th>\n","      <th>original_language</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>14000000</td>\n","      <td>Hot Tub Time Machine 2</td>\n","      <td>en</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>40000000</td>\n","      <td>The Princess Diaries 2: Royal Engagement</td>\n","      <td>en</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>3300000</td>\n","      <td>Whiplash</td>\n","      <td>en</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1200000</td>\n","      <td>Kahaani</td>\n","      <td>hi</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>0</td>\n","      <td>마린보이</td>\n","      <td>ko</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["     budget                            original_title original_language\n","0  14000000                    Hot Tub Time Machine 2                en\n","1  40000000  The Princess Diaries 2: Royal Engagement                en\n","2   3300000                                  Whiplash                en\n","3   1200000                                   Kahaani                hi\n","4         0                                      마린보이                ko"]},"metadata":{"tags":[]},"execution_count":16}]},{"cell_type":"markdown","metadata":{"id":"tPqzDsCEpapu"},"source":["On convertie les dates de type \"2/20/15\" en date \"2015-02-20\" pour être plus facilement utilisable dans le modèle."]},{"cell_type":"code","metadata":{"id":"JGgNpKEMh5K9","colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1621426593676,"user_tz":-120,"elapsed":3555,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"3aac02de-5f8b-48e2-d8a7-db4c84098725"},"source":["class DateTimeImputer(BaseEstimator, TransformerMixin):\n","    def __init__(self, replace=True):\n","        self.replace = replace\n","        pass\n","\n","    def fit(self, X, y=None):\n","        return self\n","        \n","    def transform(self, X, y=None):\n","        clone_X = X.copy()                            \n","        for feature in X.select_dtypes(include=[np.object]).columns:\n","            try:\n","                clone_X[feature] = pd.to_datetime(X[feature], infer_datetime_format=True)\n","            except:\n","                pass\n","        return clone_X\n","\n","result = DateTimeImputer().transform(X_train)\n","filters = list(X_train.filter(like=\"date\").columns)\n","result[filters].head()"],"execution_count":17,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>release_date</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>2015-02-20</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2004-08-06</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2014-10-10</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>2012-03-09</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2009-02-05</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["  release_date\n","0   2015-02-20\n","1   2004-08-06\n","2   2014-10-10\n","3   2012-03-09\n","4   2009-02-05"]},"metadata":{"tags":[]},"execution_count":17}]},{"cell_type":"markdown","metadata":{"id":"vmnLFVhhqGRP"},"source":["On transforme la date de \"2015-02-20\" en trois valeurs : y=2015, m=02, d=20"]},{"cell_type":"code","metadata":{"id":"mGbFYLRWh6uM","colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1621426593942,"user_tz":-120,"elapsed":3817,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"cc46bc18-830a-45bc-9169-e69f2365e9dc"},"source":["class DateDissolver(BaseEstimator, TransformerMixin):\n","    def __init__(self, replace=False): # no *args or **kargs\n","        self.replace = replace\n","        pass\n","\n","    def fit(self, X, y=None):\n","        return self  \n","\n","    def transform(self, X, y=None):\n","        clone_X = X.copy()\n","        for feature in X.select_dtypes(include=[np.datetime64]).columns:\n","            if self.replace:\n","                clone_X = clone_X.drop([feature], axis=1)                  \n","            try:        \n","                clone_X['{0}_Y'.format(feature)] = X[feature].dt.year\n","                clone_X['{0}_M'.format(feature)] = X[feature].dt.month\n","                clone_X['{0}_D'.format(feature)] = X[feature].dt.day\n","            except:\n","                pass\n","        return clone_X\n","\n","result = DateTimeImputer().transform(X_train)\n","result = DateDissolver(replace=True).transform(result)\n","filters = list(result.filter(like=\"date\").columns)\n","result[filters].head()"],"execution_count":18,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>release_date_Y</th>\n","      <th>release_date_M</th>\n","      <th>release_date_D</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>2015</td>\n","      <td>2</td>\n","      <td>20</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2004</td>\n","      <td>8</td>\n","      <td>6</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2014</td>\n","      <td>10</td>\n","      <td>10</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>2012</td>\n","      <td>3</td>\n","      <td>9</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2009</td>\n","      <td>2</td>\n","      <td>5</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["   release_date_Y  release_date_M  release_date_D\n","0            2015               2              20\n","1            2004               8               6\n","2            2014              10              10\n","3            2012               3               9\n","4            2009               2               5"]},"metadata":{"tags":[]},"execution_count":18}]},{"cell_type":"markdown","metadata":{"id":"Kdoat_lCqgpR"},"source":["Visualisation des variables : budget, popularité et durée"]},{"cell_type":"code","metadata":{"id":"jUB5NZFPh8wR","colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1621426593943,"user_tz":-120,"elapsed":3814,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"b29d002b-b91c-48ab-ea50-fdbb6f020cc1"},"source":["class NumberFilter(BaseEstimator, TransformerMixin):\n","    def __init__(self): # no *args or **kargs\n","        pass\n","    def fit(self, X, y=None):\n","        return self  # nothing else to do\n","    def transform(self, X, y=None):      \n","        return X.copy().select_dtypes(include=[np.int64, np.float64])        \n","\n","result = NumberFilter().transform(X_train)\n","result.head()"],"execution_count":19,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>budget</th>\n","      <th>runtime</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>14000000</td>\n","      <td>93.0000</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>40000000</td>\n","      <td>113.0000</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>3300000</td>\n","      <td>105.0000</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1200000</td>\n","      <td>122.0000</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>0</td>\n","      <td>118.0000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["     budget  runtime\n","0  14000000  93.0000\n","1  40000000 113.0000\n","2   3300000 105.0000\n","3   1200000 122.0000\n","4         0 118.0000"]},"metadata":{"tags":[]},"execution_count":19}]},{"cell_type":"markdown","metadata":{"id":"GZvyDEKwuM55"},"source":["Liste des variables : result.columns"]},{"cell_type":"code","metadata":{"id":"S-hO07Mkh-ed","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426593945,"user_tz":-120,"elapsed":3813,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"34f1a386-e8c5-45ea-c75a-296891fec4fb"},"source":["class CategoryFilter(BaseEstimator, TransformerMixin):\n","    def __init__(self):\n","        pass\n","\n","    def fit(self, X, y=None):\n","        return self\n","\n","    def transform(self, X, y=None):        \n","        return X.copy().select_dtypes(include=[np.object])        \n","\n","result = CategoryFilter().transform(X_train)\n","result.columns"],"execution_count":20,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Index(['genres', 'original_language', 'original_title', 'overview',\n","       'poster_path', 'production_companies', 'production_countries',\n","       'release_date', 'spoken_languages', 'status', 'title', 'Keywords',\n","       'cast', 'crew'],\n","      dtype='object')"]},"metadata":{"tags":[]},"execution_count":20}]},{"cell_type":"markdown","metadata":{"id":"u1NkEtfyuuA6"},"source":["Traitement des autres variables: belongs_to_collection, genres, production_companies, production_countries, spoken_languages, cast, crew, keywords"]},{"cell_type":"code","metadata":{"id":"-DUAi16XiAAF","colab":{"base_uri":"https://localhost:8080/","height":452},"executionInfo":{"status":"ok","timestamp":1621426598588,"user_tz":-120,"elapsed":8451,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"67641d2c-d4ae-4b42-dfaa-608a47c189c7"},"source":["import ast\n","\n","class InfoExtractor(BaseEstimator, TransformerMixin):\n","    def __init__(self, field, replace=False):\n","        self.field = field\n","        self.replace = replace\n","\n","    def fit(self, X, y=None):\n","        return self\n","\n","    def transform(self, X, y=None):\n","        clone_X = X.copy()   \n","        for feature, field_name in self.field:\n","            if self.replace:\n","                clone_X[feature] = X[feature].apply(lambda x: self.extract_field(x, field_name))\n","            else:\n","                clone_X[\"{0}_{1}\".format(feature, field_name)] = X[feature].apply(lambda x: self.extract_field(x, field_name))\n","        return clone_X\n","        \n","    def extract_field(self, data, field_name):\n","        if(data is not np.nan):\n","            info = ast.literal_eval(data)            \n","            result = SPLITTER.join(\"{}\".format(x[field_name]).replace(SPLITTER, \"_\") for x in info)\n","            return result\n","        return np.nan\n","    \n","infoExtractor = InfoExtractor(field=TEXT_FIELDS, replace=True)\n","result = infoExtractor.transform(X_train)\n","filters = list(result.filter(regex=\"cast|production|genres|languages|Keywords|crew\").columns)\n","result[filters].head()\n"],"execution_count":21,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>genres</th>\n","      <th>production_companies</th>\n","      <th>production_countries</th>\n","      <th>spoken_languages</th>\n","      <th>Keywords</th>\n","      <th>cast</th>\n","      <th>crew</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Comedy</td>\n","      <td>4 60 8411</td>\n","      <td>US</td>\n","      <td>English</td>\n","      <td>time_travel sequel hot_tub duringcreditsstinger</td>\n","      <td>Rob_Corddry Craig_Robinson Clark_Duke Adam_Sco...</td>\n","      <td>Kelly_Cantley Steve_Pink Josh_Heald Josh_Heald...</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Comedy Drama Family Romance</td>\n","      <td>2</td>\n","      <td>US</td>\n","      <td>English</td>\n","      <td>coronation duty marriage falling_in_love</td>\n","      <td>Anne_Hathaway Julie_Andrews H√©ctor_Elizondo J...</td>\n","      <td>Garry_Marshall Charles_Minsky John_Debney Whit...</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Drama</td>\n","      <td>2266 3172 32157</td>\n","      <td>US</td>\n","      <td>English</td>\n","      <td>jazz obsession conservatory music_teacher new_...</td>\n","      <td>Miles_Teller J.K._Simmons Melissa_Benoist Aust...</td>\n","      <td>Terri_Taylor Richard_Henderson Jeffrey_Stott H...</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Thriller Drama</td>\n","      <td>NaN</td>\n","      <td>IN</td>\n","      <td>English हिन्दी</td>\n","      <td>mystery bollywood police_corruption crime indi...</td>\n","      <td>Vidya_Balan Nawazuddin_Siddiqui Parambrata_Cha...</td>\n","      <td>Sujoy_Ghosh Sujoy_Ghosh Sujoy_Ghosh</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Action Thriller</td>\n","      <td>NaN</td>\n","      <td>KR</td>\n","      <td>한국어/조선말</td>\n","      <td>NaN</td>\n","      <td>Kim_Kang-woo Jo_Jae-hyeon Park_Si-yeon Kim_Joo...</td>\n","      <td>Jong-seok_Yoon Jong-seok_Yoon</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                        genres  ...                                               crew\n","0                       Comedy  ...  Kelly_Cantley Steve_Pink Josh_Heald Josh_Heald...\n","1  Comedy Drama Family Romance  ...  Garry_Marshall Charles_Minsky John_Debney Whit...\n","2                        Drama  ...  Terri_Taylor Richard_Henderson Jeffrey_Stott H...\n","3               Thriller Drama  ...                Sujoy_Ghosh Sujoy_Ghosh Sujoy_Ghosh\n","4              Action Thriller  ...                      Jong-seok_Yoon Jong-seok_Yoon\n","\n","[5 rows x 7 columns]"]},"metadata":{"tags":[]},"execution_count":21}]},{"cell_type":"markdown","metadata":{"id":"OHKl84QBD6Eu"},"source":["Remplace les valeurs nulles"]},{"cell_type":"code","metadata":{"id":"GiRrNo2iiCAw","colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1621426602830,"user_tz":-120,"elapsed":12689,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"c3e2cb75-d0fe-429f-d1e3-fbbb5030fee8"},"source":["from keras.preprocessing.text import Tokenizer\n","\n","class TextEncoder(BaseEstimator, TransformerMixin):\n","    def __init__(self, field=None, replace=False):\n","        self.field = field\n","        self.replace = replace\n","\n","    def fit(self, X, y=None):\n","        return self\n","\n","    def transform(self, X, y=None):\n","        clone_X = X.copy()\n","        if self.field is None:\n","            self.field = X.copy().select_dtypes(include=[np.object], exclude=[np.datetime64]).columns               \n","        for feature in self.field:               \n","            if self.replace:               \n","                clone_X[feature] = pd.Series(data=self.encode_textBySum(X[feature]), index=clone_X.index)\n","            else:                \n","                clone_X[\"{0}_{1}\".format(feature, 'count')] = pd.Series(data=self.encode_textBySum(X[feature]), index=clone_X.index)\n","        return clone_X\n","        \n","    def encode_textBySum(self, df_feature):\n","        tokenizer = Tokenizer()\n","        clone_feature = df_feature.copy().fillna('')        \n","        tokenizer.fit_on_texts(clone_feature)\n","       \n","        encoded_docs = tokenizer.texts_to_matrix(clone_feature, mode='tfidf')\n","        encoded_nums = np.sum(encoded_docs,axis=1)\n","        return encoded_nums\n","    \n","    def encode_textForOneHot(self, df_feature):\n","        tokenizer = Tokenizer()\n","        clone_feature = df_feature.copy().fillna('')        \n","        tokenizer.fit_on_texts(clone_feature)\n","        encoded_docs = tokenizer.texts_to_matrix(clone_feature, mode='binary')        \n","        encoded_onehot = pd.DataFrame(data=encoded_docs).applymap(\"{:1.0f}\".format).apply(\"\".join, axis=1)                      \n","        return encoded_onehot\n","\n","\n","infoExtractor = InfoExtractor(field=[(\"cast\", \"name\")], replace=True)                \n","textEncoder = TextEncoder(field=[\"cast\"], replace=True)\n","result = infoExtractor.transform(X_train)\n","result = textEncoder.transform(result)\n","filters = list(result.filter(regex=\"date|cast\").columns)\n","result[filters].head()"],"execution_count":22,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>release_date</th>\n","      <th>cast</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>2/20/15</td>\n","      <td>233.3424</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>8/6/04</td>\n","      <td>192.8608</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>10/10/14</td>\n","      <td>565.7832</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>3/9/12</td>\n","      <td>92.8054</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2/5/09</td>\n","      <td>61.5024</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["  release_date     cast\n","0      2/20/15 233.3424\n","1       8/6/04 192.8608\n","2     10/10/14 565.7832\n","3       3/9/12  92.8054\n","4       2/5/09  61.5024"]},"metadata":{"tags":[]},"execution_count":22}]},{"cell_type":"markdown","metadata":{"id":"vq6axZV1vT7M"},"source":["**Transformation Pipelines**\n","\n","Données utilisées dans le modèle"]},{"cell_type":"code","metadata":{"id":"ZBn1KEFmiFTq","colab":{"base_uri":"https://localhost:8080/","height":226},"executionInfo":{"status":"ok","timestamp":1621426611402,"user_tz":-120,"elapsed":21258,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"6bc5ba3c-376a-402c-d245-d43a4cdd9709"},"source":["from sklearn.pipeline import Pipeline\n","from sklearn.preprocessing import StandardScaler\n","from sklearn.preprocessing import MinMaxScaler\n","from sklearn.impute import SimpleImputer\n","\n","\n","text_pipeline = Pipeline([        \n","        ('inf_extor'   , InfoExtractor(field=TEXT_FIELDS, replace=True)),\n","        ('txt_encoder' , TextEncoder(replace=True)),\n","    ])\n","\n","result = DateTimeImputer().transform(X_train)\n","result = DateDissolver(replace=True).transform(result)\n","result = text_pipeline.fit_transform(result)\n","result.head()"],"execution_count":23,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>budget</th>\n","      <th>genres</th>\n","      <th>original_language</th>\n","      <th>original_title</th>\n","      <th>overview</th>\n","      <th>poster_path</th>\n","      <th>production_companies</th>\n","      <th>production_countries</th>\n","      <th>runtime</th>\n","      <th>spoken_languages</th>\n","      <th>status</th>\n","      <th>title</th>\n","      <th>Keywords</th>\n","      <th>cast</th>\n","      <th>crew</th>\n","      <th>release_date_Y</th>\n","      <th>release_date_M</th>\n","      <th>release_date_D</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>14000000</td>\n","      <td>1.3649</td>\n","      <td>0.7722</td>\n","      <td>29.7698</td>\n","      <td>86.2111</td>\n","      <td>8.0070</td>\n","      <td>10.7776</td>\n","      <td>0.8390</td>\n","      <td>93.0000</td>\n","      <td>0.7634</td>\n","      <td>0.6936</td>\n","      <td>29.5467</td>\n","      <td>27.5602</td>\n","      <td>233.3424</td>\n","      <td>724.6678</td>\n","      <td>2015</td>\n","      <td>2</td>\n","      <td>20</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>40000000</td>\n","      <td>6.8065</td>\n","      <td>0.7722</td>\n","      <td>32.2026</td>\n","      <td>232.2292</td>\n","      <td>8.0070</td>\n","      <td>3.8840</td>\n","      <td>0.8390</td>\n","      <td>113.0000</td>\n","      <td>0.7634</td>\n","      <td>0.6936</td>\n","      <td>31.5978</td>\n","      <td>29.8190</td>\n","      <td>192.8608</td>\n","      <td>86.3571</td>\n","      <td>2004</td>\n","      <td>8</td>\n","      <td>6</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>3300000</td>\n","      <td>1.0846</td>\n","      <td>0.7722</td>\n","      <td>7.3139</td>\n","      <td>75.4125</td>\n","      <td>8.0070</td>\n","      <td>17.8654</td>\n","      <td>0.8390</td>\n","      <td>105.0000</td>\n","      <td>0.7634</td>\n","      <td>0.6936</td>\n","      <td>7.3139</td>\n","      <td>90.1565</td>\n","      <td>565.7832</td>\n","      <td>612.9589</td>\n","      <td>2014</td>\n","      <td>10</td>\n","      <td>10</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1200000</td>\n","      <td>2.6527</td>\n","      <td>4.2594</td>\n","      <td>7.3139</td>\n","      <td>318.5386</td>\n","      <td>8.0070</td>\n","      <td>0.0000</td>\n","      <td>3.6266</td>\n","      <td>122.0000</td>\n","      <td>4.7455</td>\n","      <td>0.6936</td>\n","      <td>7.3139</td>\n","      <td>50.7610</td>\n","      <td>92.8054</td>\n","      <td>29.8478</td>\n","      <td>2012</td>\n","      <td>3</td>\n","      <td>9</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>0</td>\n","      <td>3.1861</td>\n","      <td>4.9688</td>\n","      <td>7.3139</td>\n","      <td>99.7664</td>\n","      <td>8.0070</td>\n","      <td>0.0000</td>\n","      <td>4.8785</td>\n","      <td>118.0000</td>\n","      <td>8.7627</td>\n","      <td>0.6936</td>\n","      <td>12.3545</td>\n","      <td>0.0000</td>\n","      <td>61.5024</td>\n","      <td>30.5540</td>\n","      <td>2009</td>\n","      <td>2</td>\n","      <td>5</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["     budget  genres  ...  release_date_M  release_date_D\n","0  14000000  1.3649  ...               2              20\n","1  40000000  6.8065  ...               8               6\n","2   3300000  1.0846  ...              10              10\n","3   1200000  2.6527  ...               3               9\n","4         0  3.1861  ...               2               5\n","\n","[5 rows x 18 columns]"]},"metadata":{"tags":[]},"execution_count":23}]},{"cell_type":"code","metadata":{"id":"Ef15Dh4niRlk","colab":{"base_uri":"https://localhost:8080/","height":300},"executionInfo":{"status":"ok","timestamp":1621426611403,"user_tz":-120,"elapsed":21258,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"0567d6ad-90ca-4db5-8357-0065f303d453"},"source":["from sklearn.pipeline import Pipeline\n","from sklearn.preprocessing import StandardScaler\n","from sklearn.preprocessing import MinMaxScaler\n","from sklearn.impute import SimpleImputer\n","\n","num_pipeline = Pipeline([        \n","        ('num_filter', NumberFilter()),\n","        ('imputer'   , SimpleImputer(strategy=\"median\")),     # fill nan/empty cells        \n","        ('mm_scaler' , MinMaxScaler(feature_range=(-1, 1))),  # feature scaling\n","    ])\n","\n","result = num_pipeline.fit_transform(X_train)\n","pd.DataFrame(data=result).describe()"],"execution_count":24,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>0</th>\n","      <th>1</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>3,000.0000</td>\n","      <td>3,000.0000</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>-0.8814</td>\n","      <td>-0.3618</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>0.1949</td>\n","      <td>0.1306</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>-1.0000</td>\n","      <td>-1.0000</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>-1.0000</td>\n","      <td>-0.4438</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>-0.9579</td>\n","      <td>-0.3846</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>-0.8474</td>\n","      <td>-0.3018</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>1.0000</td>\n","      <td>1.0000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["               0          1\n","count 3,000.0000 3,000.0000\n","mean     -0.8814    -0.3618\n","std       0.1949     0.1306\n","min      -1.0000    -1.0000\n","25%      -1.0000    -0.4438\n","50%      -0.9579    -0.3846\n","75%      -0.8474    -0.3018\n","max       1.0000     1.0000"]},"metadata":{"tags":[]},"execution_count":24}]},{"cell_type":"code","metadata":{"id":"wOYgLmUYiV5I","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426611403,"user_tz":-120,"elapsed":21256,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"12d0c358-1919-424b-9df3-126130020c3f"},"source":["from sklearn.pipeline import Pipeline\n","from sklearn.preprocessing import OneHotEncoder\n","from sklearn.impute import SimpleImputer\n","\n","cat_pipeline = Pipeline([\n","        ('cat_filter', CategoryFilter()),\n","        ('imputer'   , SimpleImputer(strategy='constant', fill_value='Missing')),  # fill nan/empty cells\n","    ])\n","\n","filters = ('budget', 'original_language')\n","result = LimitedColumnsFilter(filters).transform(X_train)\n","result = cat_pipeline.fit_transform(result)\n","result[0]"],"execution_count":25,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array(['en'], dtype=object)"]},"metadata":{"tags":[]},"execution_count":25}]},{"cell_type":"code","metadata":{"id":"RBWhaRNoiXxU","executionInfo":{"status":"ok","timestamp":1621426611403,"user_tz":-120,"elapsed":21255,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["from sklearn.compose import ColumnTransformer\n","from sklearn.compose import make_column_transformer\n","from sklearn.compose import make_column_selector\n","\n","\n","full_pipeline = make_column_transformer(              \n","    (num_pipeline , make_column_selector(dtype_include=[np.int64, np.float64])),            \n",")\n","\n","\n","filters = list(X_train.filter(regex=\"date|budget|original_language\").columns)\n","result = LimitedColumnsFilter(filters).transform(X_train)\n","result = full_pipeline.fit_transform(result)"],"execution_count":26,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"mB1R4IWZvmVb"},"source":["## **Partie 3 : Création du modèle**"]},{"cell_type":"markdown","metadata":{"id":"2HYIlQ4-g3UK"},"source":["#### Modèle 1"]},{"cell_type":"code","metadata":{"id":"9xLvkTL4iNwH","executionInfo":{"status":"ok","timestamp":1621426611404,"user_tz":-120,"elapsed":21254,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["import tensorflow as tf\n","from tensorflow import keras\n","from keras.layers import Dense\n","from tensorflow.keras import Model\n","from keras.layers import Dropout"],"execution_count":27,"outputs":[]},{"cell_type":"code","metadata":{"id":"hsZc0pSTiZhw","executionInfo":{"status":"ok","timestamp":1621426619711,"user_tz":-120,"elapsed":29559,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["X_train_pp_df = DateTimeImputer().transform(X_train)\n","X_train_pp_df = DateDissolver(replace=True).transform(X_train_pp_df)\n","X_train_pp_df = text_pipeline.fit_transform(X_train_pp_df)\n","\n","X_train_pp = full_pipeline.fit_transform(X_train_pp_df)"],"execution_count":28,"outputs":[]},{"cell_type":"code","metadata":{"id":"z5gBn-3aigDg","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426639237,"user_tz":-120,"elapsed":49083,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"39a9938a-c303-4976-f1de-03539675ab1d"},"source":["model_1 = tf.keras.Sequential()\n","\n","model_1.add(Dense(16, input_dim=18, activation='relu'))\n","model_1.add(Dropout(0.6))\n","model_1.add(Dense(8, activation='relu'))\n","\n","# No need to modify the code under this\n","model_1.add(Dense(1, activation = 'linear'))\n","model_1.summary()\n","\n","# Compile the model\n","model_1.compile(loss='mse', optimizer='adam', metrics=['accuracy'])\n","\n","\n","print(X_train_pp.shape)\n","print(y_train.shape)\n","history = model_1.fit(X_train_pp, y_train, epochs=100, batch_size=32, validation_split=0.2)"],"execution_count":29,"outputs":[{"output_type":"stream","text":["Model: \"sequential\"\n","_________________________________________________________________\n","Layer (type)                 Output Shape              Param #   \n","=================================================================\n","dense (Dense)                (None, 16)                304       \n","_________________________________________________________________\n","dropout (Dropout)            (None, 16)                0         \n","_________________________________________________________________\n","dense_1 (Dense)              (None, 8)                 136       \n","_________________________________________________________________\n","dense_2 (Dense)              (None, 1)                 9         \n","=================================================================\n","Total params: 449\n","Trainable params: 449\n","Non-trainable params: 0\n","_________________________________________________________________\n","(3000, 18)\n","(3000,)\n","Epoch 1/100\n","75/75 [==============================] - 1s 7ms/step - loss: 84.4125 - accuracy: 0.0000e+00 - val_loss: 47.8255 - val_accuracy: 0.0000e+00\n","Epoch 2/100\n","75/75 [==============================] - 0s 2ms/step - loss: 54.9187 - accuracy: 0.0000e+00 - val_loss: 35.6070 - val_accuracy: 0.0000e+00\n","Epoch 3/100\n","75/75 [==============================] - 0s 2ms/step - loss: 50.5575 - accuracy: 0.0000e+00 - val_loss: 35.2577 - val_accuracy: 0.0000e+00\n","Epoch 4/100\n","75/75 [==============================] - 0s 2ms/step - loss: 48.7238 - accuracy: 0.0000e+00 - val_loss: 33.9941 - val_accuracy: 0.0000e+00\n","Epoch 5/100\n","75/75 [==============================] - 0s 2ms/step - loss: 40.0666 - accuracy: 0.0000e+00 - val_loss: 33.2654 - val_accuracy: 0.0000e+00\n","Epoch 6/100\n","75/75 [==============================] - 0s 2ms/step - loss: 43.3679 - accuracy: 0.0000e+00 - val_loss: 32.4224 - val_accuracy: 0.0000e+00\n","Epoch 7/100\n","75/75 [==============================] - 0s 2ms/step - loss: 43.8945 - accuracy: 0.0000e+00 - val_loss: 32.3221 - val_accuracy: 0.0000e+00\n","Epoch 8/100\n","75/75 [==============================] - 0s 2ms/step - loss: 38.0782 - accuracy: 0.0000e+00 - val_loss: 31.9151 - val_accuracy: 0.0000e+00\n","Epoch 9/100\n","75/75 [==============================] - 0s 2ms/step - loss: 36.8868 - accuracy: 0.0000e+00 - val_loss: 31.4708 - val_accuracy: 0.0000e+00\n","Epoch 10/100\n","75/75 [==============================] - 0s 2ms/step - loss: 42.4069 - accuracy: 0.0000e+00 - val_loss: 31.6785 - val_accuracy: 0.0000e+00\n","Epoch 11/100\n","75/75 [==============================] - 0s 2ms/step - loss: 33.8426 - accuracy: 0.0000e+00 - val_loss: 30.9307 - val_accuracy: 0.0000e+00\n","Epoch 12/100\n","75/75 [==============================] - 0s 2ms/step - loss: 38.3300 - accuracy: 0.0000e+00 - val_loss: 30.4796 - val_accuracy: 0.0000e+00\n","Epoch 13/100\n","75/75 [==============================] - 0s 2ms/step - loss: 33.8000 - accuracy: 0.0000e+00 - val_loss: 29.2869 - val_accuracy: 0.0000e+00\n","Epoch 14/100\n","75/75 [==============================] - 0s 3ms/step - loss: 33.5738 - accuracy: 0.0000e+00 - val_loss: 28.4907 - val_accuracy: 0.0000e+00\n","Epoch 15/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.1124 - accuracy: 0.0000e+00 - val_loss: 28.9575 - val_accuracy: 0.0000e+00\n","Epoch 16/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.5832 - accuracy: 0.0000e+00 - val_loss: 28.0248 - val_accuracy: 0.0000e+00\n","Epoch 17/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.0700 - accuracy: 0.0000e+00 - val_loss: 27.5151 - val_accuracy: 0.0000e+00\n","Epoch 18/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.7749 - accuracy: 0.0000e+00 - val_loss: 26.0225 - val_accuracy: 0.0000e+00\n","Epoch 19/100\n","75/75 [==============================] - 0s 2ms/step - loss: 31.7589 - accuracy: 0.0000e+00 - val_loss: 25.1884 - val_accuracy: 0.0000e+00\n","Epoch 20/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.7500 - accuracy: 0.0000e+00 - val_loss: 25.2243 - val_accuracy: 0.0000e+00\n","Epoch 21/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.2112 - accuracy: 0.0000e+00 - val_loss: 24.7873 - val_accuracy: 0.0000e+00\n","Epoch 22/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.7356 - accuracy: 0.0000e+00 - val_loss: 24.1183 - val_accuracy: 0.0000e+00\n","Epoch 23/100\n","75/75 [==============================] - 0s 2ms/step - loss: 32.7682 - accuracy: 0.0000e+00 - val_loss: 23.6397 - val_accuracy: 0.0000e+00\n","Epoch 24/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.4834 - accuracy: 0.0000e+00 - val_loss: 23.0549 - val_accuracy: 0.0000e+00\n","Epoch 25/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.6674 - accuracy: 0.0000e+00 - val_loss: 23.2016 - val_accuracy: 0.0000e+00\n","Epoch 26/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.3177 - accuracy: 0.0000e+00 - val_loss: 22.6292 - val_accuracy: 0.0000e+00\n","Epoch 27/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.5256 - accuracy: 0.0000e+00 - val_loss: 22.8510 - val_accuracy: 0.0000e+00\n","Epoch 28/100\n","75/75 [==============================] - 0s 3ms/step - loss: 28.8240 - accuracy: 0.0000e+00 - val_loss: 22.9707 - val_accuracy: 0.0000e+00\n","Epoch 29/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.8295 - accuracy: 0.0000e+00 - val_loss: 22.2404 - val_accuracy: 0.0000e+00\n","Epoch 30/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.4772 - accuracy: 0.0000e+00 - val_loss: 22.2522 - val_accuracy: 0.0000e+00\n","Epoch 31/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.7370 - accuracy: 0.0000e+00 - val_loss: 22.2915 - val_accuracy: 0.0000e+00\n","Epoch 32/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.5615 - accuracy: 0.0000e+00 - val_loss: 22.5216 - val_accuracy: 0.0000e+00\n","Epoch 33/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.8073 - accuracy: 0.0000e+00 - val_loss: 22.7872 - val_accuracy: 0.0000e+00\n","Epoch 34/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.7629 - accuracy: 0.0000e+00 - val_loss: 22.3500 - val_accuracy: 0.0000e+00\n","Epoch 35/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.5689 - accuracy: 0.0000e+00 - val_loss: 22.3029 - val_accuracy: 0.0000e+00\n","Epoch 36/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.8265 - accuracy: 0.0000e+00 - val_loss: 22.2981 - val_accuracy: 0.0000e+00\n","Epoch 37/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.9910 - accuracy: 0.0000e+00 - val_loss: 22.3253 - val_accuracy: 0.0000e+00\n","Epoch 38/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.9437 - accuracy: 0.0000e+00 - val_loss: 22.3001 - val_accuracy: 0.0000e+00\n","Epoch 39/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.4239 - accuracy: 0.0000e+00 - val_loss: 22.4117 - val_accuracy: 0.0000e+00\n","Epoch 40/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.0069 - accuracy: 0.0000e+00 - val_loss: 22.3648 - val_accuracy: 0.0000e+00\n","Epoch 41/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.3425 - accuracy: 0.0000e+00 - val_loss: 22.3174 - val_accuracy: 0.0000e+00\n","Epoch 42/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9133 - accuracy: 0.0000e+00 - val_loss: 21.9997 - val_accuracy: 0.0000e+00\n","Epoch 43/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.8871 - accuracy: 0.0000e+00 - val_loss: 22.3071 - val_accuracy: 0.0000e+00\n","Epoch 44/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.7325 - accuracy: 0.0000e+00 - val_loss: 22.6331 - val_accuracy: 0.0000e+00\n","Epoch 45/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.7189 - accuracy: 0.0000e+00 - val_loss: 22.1971 - val_accuracy: 0.0000e+00\n","Epoch 46/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.1330 - accuracy: 0.0000e+00 - val_loss: 22.2516 - val_accuracy: 0.0000e+00\n","Epoch 47/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.6433 - accuracy: 0.0000e+00 - val_loss: 22.1719 - val_accuracy: 0.0000e+00\n","Epoch 48/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.8994 - accuracy: 0.0000e+00 - val_loss: 21.9938 - val_accuracy: 0.0000e+00\n","Epoch 49/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.6498 - accuracy: 0.0000e+00 - val_loss: 22.1617 - val_accuracy: 0.0000e+00\n","Epoch 50/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.8678 - accuracy: 0.0000e+00 - val_loss: 22.0428 - val_accuracy: 0.0000e+00\n","Epoch 51/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.6351 - accuracy: 0.0000e+00 - val_loss: 22.2621 - val_accuracy: 0.0000e+00\n","Epoch 52/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.2925 - accuracy: 0.0000e+00 - val_loss: 22.0862 - val_accuracy: 0.0000e+00\n","Epoch 53/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.7943 - accuracy: 0.0000e+00 - val_loss: 22.1124 - val_accuracy: 0.0000e+00\n","Epoch 54/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.2518 - accuracy: 0.0000e+00 - val_loss: 22.1939 - val_accuracy: 0.0000e+00\n","Epoch 55/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.3013 - accuracy: 0.0000e+00 - val_loss: 22.1619 - val_accuracy: 0.0000e+00\n","Epoch 56/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.1624 - accuracy: 0.0000e+00 - val_loss: 21.8906 - val_accuracy: 0.0000e+00\n","Epoch 57/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.2153 - accuracy: 0.0000e+00 - val_loss: 21.9989 - val_accuracy: 0.0000e+00\n","Epoch 58/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.2761 - accuracy: 0.0000e+00 - val_loss: 22.0653 - val_accuracy: 0.0000e+00\n","Epoch 59/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.5078 - accuracy: 0.0000e+00 - val_loss: 22.0094 - val_accuracy: 0.0000e+00\n","Epoch 60/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.7965 - accuracy: 0.0000e+00 - val_loss: 21.9712 - val_accuracy: 0.0000e+00\n","Epoch 61/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.4243 - accuracy: 0.0000e+00 - val_loss: 22.0753 - val_accuracy: 0.0000e+00\n","Epoch 62/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.7483 - accuracy: 0.0000e+00 - val_loss: 22.1186 - val_accuracy: 0.0000e+00\n","Epoch 63/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.3053 - accuracy: 0.0000e+00 - val_loss: 21.8552 - val_accuracy: 0.0000e+00\n","Epoch 64/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.3140 - accuracy: 0.0000e+00 - val_loss: 22.0565 - val_accuracy: 0.0000e+00\n","Epoch 65/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.6077 - accuracy: 0.0000e+00 - val_loss: 21.9815 - val_accuracy: 0.0000e+00\n","Epoch 66/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.4499 - accuracy: 0.0000e+00 - val_loss: 21.5939 - val_accuracy: 0.0000e+00\n","Epoch 67/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.4167 - accuracy: 0.0000e+00 - val_loss: 22.2193 - val_accuracy: 0.0000e+00\n","Epoch 68/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.0007 - accuracy: 0.0000e+00 - val_loss: 22.4591 - val_accuracy: 0.0000e+00\n","Epoch 69/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.6767 - accuracy: 0.0000e+00 - val_loss: 21.8518 - val_accuracy: 0.0000e+00\n","Epoch 70/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.5578 - accuracy: 0.0000e+00 - val_loss: 21.9746 - val_accuracy: 0.0000e+00\n","Epoch 71/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.8746 - accuracy: 0.0000e+00 - val_loss: 21.9271 - val_accuracy: 0.0000e+00\n","Epoch 72/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.3325 - accuracy: 0.0000e+00 - val_loss: 21.7955 - val_accuracy: 0.0000e+00\n","Epoch 73/100\n","75/75 [==============================] - 0s 3ms/step - loss: 23.2851 - accuracy: 0.0000e+00 - val_loss: 22.4780 - val_accuracy: 0.0000e+00\n","Epoch 74/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.3965 - accuracy: 0.0000e+00 - val_loss: 22.2110 - val_accuracy: 0.0000e+00\n","Epoch 75/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.1233 - accuracy: 0.0000e+00 - val_loss: 21.8598 - val_accuracy: 0.0000e+00\n","Epoch 76/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.4416 - accuracy: 0.0000e+00 - val_loss: 21.8230 - val_accuracy: 0.0000e+00\n","Epoch 77/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.1349 - accuracy: 0.0000e+00 - val_loss: 22.4025 - val_accuracy: 0.0000e+00\n","Epoch 78/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.1901 - accuracy: 0.0000e+00 - val_loss: 21.7866 - val_accuracy: 0.0000e+00\n","Epoch 79/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.4093 - accuracy: 0.0000e+00 - val_loss: 21.7628 - val_accuracy: 0.0000e+00\n","Epoch 80/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.2030 - accuracy: 0.0000e+00 - val_loss: 21.6532 - val_accuracy: 0.0000e+00\n","Epoch 81/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.2585 - accuracy: 0.0000e+00 - val_loss: 21.6662 - val_accuracy: 0.0000e+00\n","Epoch 82/100\n","75/75 [==============================] - 0s 2ms/step - loss: 21.5560 - accuracy: 0.0000e+00 - val_loss: 21.8027 - val_accuracy: 0.0000e+00\n","Epoch 83/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.3003 - accuracy: 0.0000e+00 - val_loss: 21.8776 - val_accuracy: 0.0000e+00\n","Epoch 84/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.8438 - accuracy: 0.0000e+00 - val_loss: 21.7846 - val_accuracy: 0.0000e+00\n","Epoch 85/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.2745 - accuracy: 0.0000e+00 - val_loss: 21.8963 - val_accuracy: 0.0000e+00\n","Epoch 86/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.3511 - accuracy: 0.0000e+00 - val_loss: 22.0034 - val_accuracy: 0.0000e+00\n","Epoch 87/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.3021 - accuracy: 0.0000e+00 - val_loss: 21.9989 - val_accuracy: 0.0000e+00\n","Epoch 88/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.9292 - accuracy: 0.0000e+00 - val_loss: 21.9540 - val_accuracy: 0.0000e+00\n","Epoch 89/100\n","75/75 [==============================] - 0s 2ms/step - loss: 21.0494 - accuracy: 0.0000e+00 - val_loss: 21.4949 - val_accuracy: 0.0000e+00\n","Epoch 90/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9125 - accuracy: 0.0000e+00 - val_loss: 21.6270 - val_accuracy: 0.0000e+00\n","Epoch 91/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.6411 - accuracy: 0.0000e+00 - val_loss: 21.7500 - val_accuracy: 0.0000e+00\n","Epoch 92/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.4328 - accuracy: 0.0000e+00 - val_loss: 21.3521 - val_accuracy: 0.0000e+00\n","Epoch 93/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.0127 - accuracy: 0.0000e+00 - val_loss: 21.6318 - val_accuracy: 0.0000e+00\n","Epoch 94/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.5246 - accuracy: 0.0000e+00 - val_loss: 22.2244 - val_accuracy: 0.0000e+00\n","Epoch 95/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.5129 - accuracy: 0.0000e+00 - val_loss: 21.7589 - val_accuracy: 0.0000e+00\n","Epoch 96/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.6450 - accuracy: 0.0000e+00 - val_loss: 21.5657 - val_accuracy: 0.0000e+00\n","Epoch 97/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.0172 - accuracy: 0.0000e+00 - val_loss: 21.7809 - val_accuracy: 0.0000e+00\n","Epoch 98/100\n","75/75 [==============================] - 0s 2ms/step - loss: 21.0013 - accuracy: 0.0000e+00 - val_loss: 22.2921 - val_accuracy: 0.0000e+00\n","Epoch 99/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.0373 - accuracy: 0.0000e+00 - val_loss: 22.1305 - val_accuracy: 0.0000e+00\n","Epoch 100/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.2428 - accuracy: 0.0000e+00 - val_loss: 21.7624 - val_accuracy: 0.0000e+00\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":533},"id":"Ftz5aTwLh8vd","executionInfo":{"status":"ok","timestamp":1621426639812,"user_tz":-120,"elapsed":49657,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"573fe758-0d45-4651-e0f9-4c6af9ba297c"},"source":["keras.utils.plot_model(model_1, show_shapes=True)"],"execution_count":30,"outputs":[{"output_type":"execute_result","data":{"image/png":"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\n","text/plain":["<IPython.core.display.Image object>"]},"metadata":{"tags":[]},"execution_count":30}]},{"cell_type":"markdown","metadata":{"id":"TAN4NxnShUMM"},"source":["#### Modèle 2 "]},{"cell_type":"code","metadata":{"id":"TGkE4B7yhYVx","executionInfo":{"status":"ok","timestamp":1621426648346,"user_tz":-120,"elapsed":58189,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["X_train_pp_df = DateTimeImputer().transform(X_train)\n","X_train_pp_df = DateDissolver(replace=True).transform(X_train_pp_df)\n","X_train_pp_df = text_pipeline.fit_transform(X_train_pp_df)\n","\n","X_train_pp = full_pipeline.fit_transform(X_train_pp_df)"],"execution_count":31,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"EKFDdqlDhZTd","executionInfo":{"status":"ok","timestamp":1621426667231,"user_tz":-120,"elapsed":77072,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"16f698d4-aa6a-4d9f-ccc8-01612389d676"},"source":["model_2 = keras.models.Sequential([\n","      keras.layers.Dense(72, input_shape=(18,), activation=\"relu\"),\n","      keras.layers.Dropout(0.6),\n","      keras.layers.Dense(128, activation=\"relu\"),\n","      keras.layers.Dropout(0.6),\n","      keras.layers.Dense(256, activation=\"relu\"),\n","      keras.layers.Dropout(0.6),\n","      keras.layers.Dense(128, activation=\"relu\"),\n","      keras.layers.Dropout(0.6),\n","      keras.layers.Dense(72, activation=\"relu\"),\n","      keras.layers.Dropout(0.6),\n","      keras.layers.Dense(6)\n","    ])\n","\n","model_2.summary()\n","\n","# Compile the model\n","model_2.compile(loss='mse', optimizer='adam')\n","\n","print(X_train_pp.shape)\n","print(y_train.shape)\n","history = model_2.fit(X_train_pp, y_train, epochs=100, batch_size=32, validation_split=0.2)"],"execution_count":32,"outputs":[{"output_type":"stream","text":["Model: \"sequential_1\"\n","_________________________________________________________________\n","Layer (type)                 Output Shape              Param #   \n","=================================================================\n","dense_3 (Dense)              (None, 72)                1368      \n","_________________________________________________________________\n","dropout_1 (Dropout)          (None, 72)                0         \n","_________________________________________________________________\n","dense_4 (Dense)              (None, 128)               9344      \n","_________________________________________________________________\n","dropout_2 (Dropout)          (None, 128)               0         \n","_________________________________________________________________\n","dense_5 (Dense)              (None, 256)               33024     \n","_________________________________________________________________\n","dropout_3 (Dropout)          (None, 256)               0         \n","_________________________________________________________________\n","dense_6 (Dense)              (None, 128)               32896     \n","_________________________________________________________________\n","dropout_4 (Dropout)          (None, 128)               0         \n","_________________________________________________________________\n","dense_7 (Dense)              (None, 72)                9288      \n","_________________________________________________________________\n","dropout_5 (Dropout)          (None, 72)                0         \n","_________________________________________________________________\n","dense_8 (Dense)              (None, 6)                 438       \n","=================================================================\n","Total params: 86,358\n","Trainable params: 86,358\n","Non-trainable params: 0\n","_________________________________________________________________\n","(3000, 18)\n","(3000,)\n","Epoch 1/100\n","75/75 [==============================] - 1s 6ms/step - loss: 76.2553 - val_loss: 57.5431\n","Epoch 2/100\n","75/75 [==============================] - 0s 2ms/step - loss: 48.7325 - val_loss: 47.4983\n","Epoch 3/100\n","75/75 [==============================] - 0s 3ms/step - loss: 43.9904 - val_loss: 43.9457\n","Epoch 4/100\n","75/75 [==============================] - 0s 2ms/step - loss: 40.1893 - val_loss: 38.7467\n","Epoch 5/100\n","75/75 [==============================] - 0s 2ms/step - loss: 32.2035 - val_loss: 35.7323\n","Epoch 6/100\n","75/75 [==============================] - 0s 3ms/step - loss: 33.5930 - val_loss: 31.7209\n","Epoch 7/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.9358 - val_loss: 32.6950\n","Epoch 8/100\n","75/75 [==============================] - 0s 2ms/step - loss: 31.8284 - val_loss: 25.0589\n","Epoch 9/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.5812 - val_loss: 27.1045\n","Epoch 10/100\n","75/75 [==============================] - 0s 2ms/step - loss: 33.2563 - val_loss: 26.8468\n","Epoch 11/100\n","75/75 [==============================] - 0s 3ms/step - loss: 32.3253 - val_loss: 27.7596\n","Epoch 12/100\n","75/75 [==============================] - 0s 3ms/step - loss: 30.6932 - val_loss: 25.9193\n","Epoch 13/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.5612 - val_loss: 24.7736\n","Epoch 14/100\n","75/75 [==============================] - 0s 2ms/step - loss: 32.1753 - val_loss: 25.4560\n","Epoch 15/100\n","75/75 [==============================] - 0s 2ms/step - loss: 32.2846 - val_loss: 24.1441\n","Epoch 16/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.4033 - val_loss: 22.2135\n","Epoch 17/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.2154 - val_loss: 24.1173\n","Epoch 18/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.3555 - val_loss: 22.8526\n","Epoch 19/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.9653 - val_loss: 22.2340\n","Epoch 20/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.3373 - val_loss: 23.2557\n","Epoch 21/100\n","75/75 [==============================] - 0s 2ms/step - loss: 31.1755 - val_loss: 22.8815\n","Epoch 22/100\n","75/75 [==============================] - 0s 2ms/step - loss: 31.0475 - val_loss: 21.9307\n","Epoch 23/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.4724 - val_loss: 22.3846\n","Epoch 24/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.3731 - val_loss: 24.4609\n","Epoch 25/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.6902 - val_loss: 23.3821\n","Epoch 26/100\n","75/75 [==============================] - 0s 2ms/step - loss: 29.0715 - val_loss: 21.8402\n","Epoch 27/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.4086 - val_loss: 22.0120\n","Epoch 28/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.8702 - val_loss: 22.5013\n","Epoch 29/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.2043 - val_loss: 22.6356\n","Epoch 30/100\n","75/75 [==============================] - 0s 2ms/step - loss: 30.0470 - val_loss: 23.2710\n","Epoch 31/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.4841 - val_loss: 20.7977\n","Epoch 32/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.7128 - val_loss: 21.1288\n","Epoch 33/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.6414 - val_loss: 22.0939\n","Epoch 34/100\n","75/75 [==============================] - 0s 3ms/step - loss: 28.9319 - val_loss: 24.3793\n","Epoch 35/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.8285 - val_loss: 22.2643\n","Epoch 36/100\n","75/75 [==============================] - 0s 3ms/step - loss: 27.9684 - val_loss: 20.7502\n","Epoch 37/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.9640 - val_loss: 22.2909\n","Epoch 38/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.0390 - val_loss: 21.4142\n","Epoch 39/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.5874 - val_loss: 21.2774\n","Epoch 40/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.3661 - val_loss: 22.1915\n","Epoch 41/100\n","75/75 [==============================] - 0s 3ms/step - loss: 25.7780 - val_loss: 21.3495\n","Epoch 42/100\n","75/75 [==============================] - 0s 3ms/step - loss: 25.1385 - val_loss: 21.9155\n","Epoch 43/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9556 - val_loss: 20.2459\n","Epoch 44/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.7965 - val_loss: 21.8323\n","Epoch 45/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.5015 - val_loss: 23.0072\n","Epoch 46/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.2294 - val_loss: 21.2709\n","Epoch 47/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.2824 - val_loss: 20.7341\n","Epoch 48/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.2927 - val_loss: 21.1740\n","Epoch 49/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.3628 - val_loss: 21.5633\n","Epoch 50/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.2770 - val_loss: 21.2092\n","Epoch 51/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.4530 - val_loss: 20.2386\n","Epoch 52/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.5793 - val_loss: 21.0687\n","Epoch 53/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.3414 - val_loss: 21.7791\n","Epoch 54/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.9527 - val_loss: 20.6619\n","Epoch 55/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.6506 - val_loss: 20.4156\n","Epoch 56/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.5382 - val_loss: 21.8521\n","Epoch 57/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.8670 - val_loss: 21.5419\n","Epoch 58/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.3154 - val_loss: 20.5356\n","Epoch 59/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.8741 - val_loss: 22.3273\n","Epoch 60/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.5678 - val_loss: 21.0604\n","Epoch 61/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.0707 - val_loss: 21.8840\n","Epoch 62/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.8905 - val_loss: 21.6098\n","Epoch 63/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.5359 - val_loss: 22.1888\n","Epoch 64/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.6294 - val_loss: 21.4380\n","Epoch 65/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.4197 - val_loss: 21.4841\n","Epoch 66/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.1388 - val_loss: 20.9881\n","Epoch 67/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.4511 - val_loss: 21.3594\n","Epoch 68/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.0554 - val_loss: 20.8115\n","Epoch 69/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.7750 - val_loss: 21.3549\n","Epoch 70/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.1947 - val_loss: 21.2668\n","Epoch 71/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9143 - val_loss: 21.3695\n","Epoch 72/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.9926 - val_loss: 21.7009\n","Epoch 73/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.9523 - val_loss: 21.1704\n","Epoch 74/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.8589 - val_loss: 20.8749\n","Epoch 75/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.2160 - val_loss: 22.1444\n","Epoch 76/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9401 - val_loss: 22.0519\n","Epoch 77/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.6728 - val_loss: 21.1593\n","Epoch 78/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.8483 - val_loss: 20.7673\n","Epoch 79/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.6277 - val_loss: 20.8428\n","Epoch 80/100\n","75/75 [==============================] - 0s 3ms/step - loss: 24.7760 - val_loss: 21.1206\n","Epoch 81/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.5907 - val_loss: 20.5003\n","Epoch 82/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.8642 - val_loss: 20.7560\n","Epoch 83/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.9080 - val_loss: 20.9750\n","Epoch 84/100\n","75/75 [==============================] - 0s 2ms/step - loss: 26.0959 - val_loss: 20.5888\n","Epoch 85/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.0471 - val_loss: 20.8891\n","Epoch 86/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.8839 - val_loss: 21.3166\n","Epoch 87/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.6707 - val_loss: 20.6604\n","Epoch 88/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.2142 - val_loss: 20.8855\n","Epoch 89/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.6100 - val_loss: 21.0493\n","Epoch 90/100\n","75/75 [==============================] - 0s 3ms/step - loss: 25.9725 - val_loss: 20.7624\n","Epoch 91/100\n","75/75 [==============================] - 0s 3ms/step - loss: 23.8718 - val_loss: 20.9937\n","Epoch 92/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.5818 - val_loss: 20.8534\n","Epoch 93/100\n","75/75 [==============================] - 0s 2ms/step - loss: 28.0264 - val_loss: 21.0619\n","Epoch 94/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.0233 - val_loss: 20.7947\n","Epoch 95/100\n","75/75 [==============================] - 0s 2ms/step - loss: 24.2687 - val_loss: 20.5390\n","Epoch 96/100\n","75/75 [==============================] - 0s 3ms/step - loss: 26.4951 - val_loss: 20.4195\n","Epoch 97/100\n","75/75 [==============================] - 0s 2ms/step - loss: 22.1806 - val_loss: 20.3187\n","Epoch 98/100\n","75/75 [==============================] - 0s 2ms/step - loss: 27.2957 - val_loss: 21.7693\n","Epoch 99/100\n","75/75 [==============================] - 0s 2ms/step - loss: 25.9404 - val_loss: 20.8691\n","Epoch 100/100\n","75/75 [==============================] - 0s 2ms/step - loss: 23.9441 - val_loss: 20.8266\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"_uEtde0Ni_wT","executionInfo":{"status":"ok","timestamp":1621426667843,"user_tz":-120,"elapsed":77682,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"ef85b70e-ab22-4cc4-d1a2-8db13c1149e3"},"source":["keras.utils.plot_model(model_2, show_shapes=True)"],"execution_count":33,"outputs":[{"output_type":"execute_result","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAbUAAAULCAYAAACqJTzLAAAABmJLR0QA/wD/AP+gvaeTAAAgAElEQVR4nOzdeVgUZ7Y/8G+zNg3dLCKIKMriEgQ1RiaCOsY4QxIZQQSURJJRJxk0CyJqFLcYRBRxkIuR67iEO1cSFdSgMWJyjYOON+ovGSUaTAwuKIiIKPsiCOf3h7c7No1IQzfdXZ7P8/Tz6FtvVZ2qXg5d/dZ7REREYIwxxgTASNcBMMYYY5rCSY0xxphgcFJjjDEmGJzUGGOMCYZJ24bTp08jOTlZF7Ewxhhjnebr64uYmBilNpVvakVFRdi3b1+PBcUY65wzZ87gzJkzug7DoBQXF/PnmUCdOXMGp0+fVmlX+aYml5WVpdWAGGPqCQsLA8DvTXVkZmZixowZfM4ESP5+aIt/U2OMMSYYnNQYY4wJBic1xhhjgsFJjTHGmGBwUmOMMSYYnNQYe8YcOXIE1tbW+PLLL3Udil6aO3cuRCKR4hEREaHS59ixY4iNjcX+/fvh5uam6Pvmm2+q9PX394dUKoWxsTGGDRuGc+fO9cRhdFtrays2bdoEPz+/J/b5/PPP4ePjA6lUigEDBmD27NkoLS1VLD906BASExPR0tKitF52drbSOba3t9dY3JzUGHvGcGGOp7Ozs0NOTg4uX76MnTt3Ki376KOPkJqaimXLliEkJATXrl2Du7s7evXqhYyMDHz11VdK/b/55htkZWVhypQpyM/Px6hRo3ryULqkoKAAv//97xETE4P6+vp2++zduxczZ85EWFgYiouLcfDgQZw8eRKvvfYaHj58CAAIDAyEWCzGpEmTUFlZqVg3KCgIxcXFOHnyJCZPnqzR2DmpMfaMCQgIQFVVFaZMmaLrUNDQ0NDhNwFdsbCwwKuvvorBgwfD3Nxc0b5+/Xrs2bMHmZmZkEqlSuukpqbCyMgIkZGRqKqq6umQNebHH3/E0qVLMW/ePIwcOfKJ/f7+97+jb9++WLx4MaytrTFy5EjExMQgLy8PZ8+eVfSbP38+RowYgcmTJyuSnUgkgrOzM8aPH49BgwZpNH5Oaowxndm5cyfKysp0HUanXLlyBStXrsTHH38MsVisstzPzw/R0dG4desWFi1apIMINWPEiBHYv38/Zs6cqZTQ2yoqKoKTkxNEIpGirX///gCAGzduKPVdvXo18vLykJKSop2gH8NJjbFnyKlTp+Di4gKRSIRPPvkEAJCWlgZLS0tIJBIcPHgQr732GmQyGfr164fdu3cr1k1NTYVYLIaDgwPmzp0LJycniMVi+Pn5Kf1lHhUVBTMzM/Tp00fR9t5778HS0hIikQjl5eUAgOjoaCxcuBBXr16FSCSCh4cHAODo0aOQyWRYu3ZtT5ySTktNTQURITAw8Il94uPjMXjwYOzYsQPHjh3rcHtEhOTkZDz33HMwNzeHra0tpk6dil9++UXRp7PPDQC0tLRg1apVcHFxgYWFBYYPH469e/d276A74ObmpvIHifz3NDc3N6V2W1tbTJgwASkpKVq//M1JjbFnyLhx4/Ddd98ptb377rtYsGABGhoaIJVKsXfvXly9ehVubm5455130NzcDOBRspo1axbq6+sxf/58FBYW4ty5c3j48CH++Mc/oqioCMCjD//p06cr7WPLli34+OOPldpSUlIwZcoUuLu7g4hw5coVAFAMKmhtbdXKOeiqr776CkOGDIFEInliHwsLC/zXf/0XjIyM8M4776Curu6JfVevXo3Y2FgsX74cZWVlOHnyJIqKijB+/HjcuXMHQOefGwBYunQpNmzYgE2bNuH27duYMmUK3njjDfzwww+aOwmPWbZsGUpLS7F582bU1NQgPz8fKSkpeOWVVzBmzBiV/s8//zxu3bqFH3/8USvxyHFSY4wp+Pn5QSaToXfv3ggPD0ddXR1u3ryp1MfExETx7cLT0xNpaWmoqalBenq6RmIICAhAdXU1Vq5cqZHtaUJdXR2uX78Od3f3p/b19fXFggULUFhYiKVLl7bbp6GhAcnJyZg2bRoiIiJgbW0Nb29vbN26FeXl5di2bZvKOh09N42NjUhLS0NwcDBCQkJgY2ODFStWwNTUVGPPS1sTJkzAkiVLEBUVBZlMBi8vL9TU1GDHjh3t9pf/dnbx4kWtxCPHSY0x1i4zMzMAUPo20J7Ro0dDIpEoXTYTmrKyMhBRh9/SHhcfH48hQ4Zgy5YtOHXqlMry/Px81NbWYvTo0UrtPj4+MDMzU7qc2562z83ly5dRX18PLy8vRR8LCwv06dNHa8/L8uXLsW3bNnz77beora3FtWvX4OfnB19fX8W39sfJz538W6i2cFJjjHWbubk57t69q+swtKaxsREAOhw48TixWIz09HSIRCLMmTMHDQ0NSsvlw9utrKxU1rWxsUFNTY1a8ckvc65YsULp/q8bN248cUh+d9y+fRuJiYn461//ipdffhmWlpZwdXXF9u3bUVJSgqSkJJV1LCwsAPx2LrWFkxpjrFuam5tRWVmJfv366ToUrZF/ILe9ibgj8gKWBQUFWLNmjdIyGxsbAGg3eXXlXPbu3RsAsGnTJhCR0qO9mmPdVVBQgJaWFvTt21epXSaTwc7ODvn5+SrrNDU1AfjtXGoLJzXGWLfk5uaCiJQGB5iYmDz1sqUhcXBwgEgkUvv+szVr1mDo0KE4f/68UruXlxesrKxUBnGcPXsWTU1NeOGFF9TaT//+/SEWi5GXl6fWel0lT7q3b99Waq+pqcH9+/cVQ/sfJz93jo6OWo2NkxpjTC2tra2oqKjAw4cPceHCBURHR8PFxQWzZs1S9PHw8MD9+/eRnZ2N5uZm3L17V+XeJeDRzB0lJSUoLCxETU0NmpubkZOTo3dD+iUSCdzc3FBcXKzWevLLkMbGxirtCxcuxIEDB5CRkYHq6mpcvHgR8+bNg5OTEyIjI9Xez+zZs7F7926kpaWhuroaLS0tKC4uViSe8PBwODo6amSaLldXV0ycOBHbt2/HyZMn0dDQgKKiIkXcf/nLX1TWkZ87b2/vbu+/Q9TG3r17qZ1mxpiOhYaGUmhoaLe2sXnzZurTpw8BIIlEQoGBgbRlyxaSSCQEgAYNGkRXr16lbdu2kUwmIwA0YMAA+vXXX4mIKDIykkxNTcnZ2ZlMTExIJpPR1KlT6erVq0r7uXfvHk2cOJHEYjG5urrSBx98QIsXLyYA5OHhQTdv3iQionPnztGAAQPIwsKCxo0bR6WlpXTkyBGSSqUUHx/frWMl6trnWWRkJDk7O6u0R0VFkampKdXX1yvaDhw4QO7u7gSA7O3t6f333293m4sXL6agoCClttbWVkpKSqJBgwaRqakp2draUnBwMF2+fFnRR53n5sGDB7RkyRJycXEhExMT6t27N4WEhFB+fj4REQUHBxMAWrVqVYfHf/r0aRo7diw5OTkRAAJAffr0IT8/Pzpx4oSiX3l5OUVHR5OHhweZm5uTlZUVjR07lr744ot2txsQEEDOzs7U2tqq1D5//nzq1atXhzG150nvB05qjBkITSS17oqMjCQ7OzudxqAOTSa1goICMjExoV27dmkqvB7V0tJC48ePp507d/b4vsvLy0ksFtPGjRtVlmk6qfHlR8aYWtQZLGGoGhoa8PXXX6OgoEAxwMHDwwNxcXGIi4tDbW2tjiNUT0tLC7Kzs1FTU4Pw8PAe3//q1asxcuRIREVFAXg0m0pJSQlOnTqluOleUzipMcZYG/fv31dMaDxnzhxFe2xsLMLCwhAeHm5Qkxbn5uZi//79yMnJ6fS9dpqSnJyMvLw8HDlyBKampgCAgwcPKiY0blvVoLu0ktTefvttSKVSiESiHhuNo2mJiYkYOnQoLCwsYGlpiaFDh2LlypWorq7u0vaEUMPqzJkzeO6552BkZASRSARHR0fEx8frOiwlbetb9enTp916WEx9y5YtQ3p6OqqqquDq6op9+/bpOiSt2Lp1q9KQ+IyMDKXla9euRVRUFNatW6ejCNU3adIkfPbZZ0rzcfaEgwcP4sGDB8jNzYWtra2iferUqUrnWD4fqCaYaGxLj9mxYwf+8Ic/4PXXX9fG5nvEv/71L7zzzjt46623YGFhgZycHMycORNnz57FN998o/b2SAA1rMaMGYOff/4Zr776Kr7++mtcvnxZcb+NvggJCUFISAg8PDxQXl6uVLCQdU9CQgISEhJ0HYZe8Pf3h7+/v67D0HtBQUEICgrq0X3y5ccnMDMzw3vvvYfevXvDysoKYWFhmDp1Kv7nf/5H5d6MzuAaVtohpGNhjHWfVr6pAVCqsWOIDhw4oNLm7OwMAAb3I3FbhlTD6mmEdCyMse7TyDc1IkJSUhKGDBkCc3NzWFtbY/HixSr9Oqr3o07doBMnTuB3v/sdJBIJZDIZvL29Fb91abOmUEFBAWxsbDBgwAC11hN6DSt9OxZ1/etf/4Knpyesra0hFovh7e2Nr7/+GsCj34flv8+5u7srZoaYPXs2JBIJrK2tcejQIQAdv/Y2bNgAiUQCqVSKsrIyLFy4EM7Ozrh8+XKXYmaMPUHbMf5dua9j+fLlJBKJ6G9/+xtVVFRQfX09bdmyhQDQ+fPnFf0WLVpE5ubmtG/fPqqoqKBly5aRkZERff/994rtAKBvv/2WqqqqqKysjMaPH0+WlpbU1NRERES1tbUkk8koMTGRGhoaqLS0lKZNm0Z3797t1D7U1dTURMXFxbR582YyNzfv8j0qRUVFBIA2b96sdN6edrxEj+6bsbS0pEuXLlFjYyPl5+eTj48PSaVSxU2sREQzZ84kR0dHpf0mJSURAMX5ISIKCQkhd3d3pX6HDx8mqVRKcXFxTz2WV155hQBQRUWFXh4LEZG7uztZW1s/9ViIiLKysmj16tV0//59unfvHo0ZM0bpvpmQkBAyNjamW7duKa33xhtv0KFDhxT/7+zre/78+bR582aaNm0a/fzzz52KkUg/7lMzNHzfrXBp7T61hoYGbNq0CX/4wx8QExMDGxsbWFhYwM7OTqmfOvV+OqobVFhYiOrqagwbNgxisRiOjo7Yv38/7O3ttVJTqH///ujXrx9Wr16NDRs2YMaMGV07UR0QUg0rfTgWdYWGhuKjjz6Cra0t7OzsEBgYiHv37ilmnZ83bx5aWlqU4quursb333+PyZMnA1Dv9b1+/Xq8//772L9/P4YOHdpzB8rYM6DbSe3KlSuor6/HpEmTOuzX1Xo/besGubm5wcHBAREREVi9ejUKCwu7vY+OFBUVoaysDJ9//jn+8Y9/4Pnnn9fqbzhCqmFlqMciv5dGfpPxyy+/jMGDB+PTTz9VjGLds2cPwsPDFXP69VQ9q3379imVFuFHxw/5H6G6joMfmn886ZaSbg8UkU9SKS998CSP1/tZsWKF0jInJ6dO78/CwgLHjx/H0qVLsXbtWsTFxWH69OlIT0/X2D4eZ2pqit69e8Pf3x+urq4YPHgwEhISkJKS0qXtaZKQaljp8li++uorJCUlIT8/H9XV1SpJWCQSYe7cuYiJicG3336LP/zhD/jv//5vfPbZZ4o+2njttWfMmDFYsGCBxrYndKdPn0ZKSorGfldn+mPTpk3ttnc7qYnFYgDAgwcPOuz3eL2f6Ojobu1z2LBh+PLLL3H37l0kJydj/fr1GDZsmGL6F03soz0eHh4wNjZut1ZQTxNSDauePpaTJ0/i3//+NxYsWICbN28iODgY06ZNw6effoq+ffti8+bN+PDDD5XWmTVrFpYtW4YdO3agf//+kMlkSgOGNPn67ki/fv0wffp0rW1fiFJSUvicCVBWVla77d2+/Ojl5QUjIyOcOHGiw36aqvdTUlKCS5cuAXj0QbJu3TqMGjUKly5d0tg+7t27hzfeeEOlXV4Yr71aQT1NSDWsevpY/v3vf8PS0hIAcPHiRTQ3N+Pdd9+Fm5sbxGIxRCLV21FsbW0xY8YMZGdnY+PGjXjnnXeUlvd0PSvGWPu6ndR69+6NkJAQ7Nu3Dzt37kR1dTUuXLiAbdu2KfXrTL2fzigpKcHcuXPxyy+/oKmpCefPn8eNGzcwZswYje3D0tIS33zzDY4fP664HHX+/Hn8+c9/hqWlJWJiYjq9LU0RUg0rbR/LkzQ3N+POnTvIzc1VJDUXFxcAwLFjx9DY2IiCggKl2wseN2/ePDx48ACHDx9WuYleU689xlg3tR0O2ZUhsDU1NfT2229Tr169yMrKisaNG0erVq0iANSvXz/68ccfiajjej+drRtUWFhIfn5+ZGtrS8bGxtS3b19avnw5PXz48Kn7UEdgYCC5urqSlZUVmZubk7u7O4WHh9PFixfV2g6RcGpYnTlzhoYNG0ZGRkaKGktr167Vq2P5z//8T0V9q44eBw4cUOxryZIlZGdnRzY2NhQWFkaffPIJASB3d3el2wyIiJ5//nmKjY1t9/x09NpLTEwkCwsLAkD9+/fv0q0hPKRffTykX7ie9H4QESlPSpiZmYkZM2YIYq5CoZg7dy6ysrJw7949XYfSbYZ+LAEBAfjkk0/g6ura4/sOCwsD8OTfEpgq/jwTrie9H3juRwMhpBpWhnQsj1/OvHDhAsRisU4SGmOsc56ZpPbLL7906t6HzhbQ0/T2mH5asmQJCgoK8Ouvv2L27NlYs2aNrkNiWjZ37lyl93B7pYuOHTuG2NhYlVJHb775pkpff39/SKVSGBsbY9iwYTh37lxPHEa3tba2YtOmTR1OGP7555/Dx8cHUqkUAwYMwOzZs5UqYxw6dAiJiYkqf8hmZ2crnWN7e3vNBd72eiRfg9YvsbGxZGZmRgBo4MCBlJWVpeuQuswQj2X58uVkZGRE/fv3V5oSSxf4NzX1deXzLDIykuzs7CgnJ4cuX75MjY2NSstXrVpFU6ZMoerqakWbu7s79erViwDQ4cOHVbaZk5NDQUFBXTsIHfj1119p7NixBIBGjBjRbp89e/YQAEpMTKTKyko6f/48ubm50ciRI6m5uVnRLyUlhSZMmKA0rV5raysVFxfTyZMnafLkyUrT0nXWk94PnNQYMxD6kNTq6+vJ19fXYPbR1aTm7Ozc7rJ169bR4MGDqaGhQand3d2dPvvsMzIyMiJnZ2eqrKxUWm5ISS0vL4+mTZtGGRkZNHLkyCcmtYkTJ1Lfvn2ptbVV0SYfZHXq1CmlvlFRUeTr66uU7OTmz5+v0aT2zFx+ZIx1X0+U+tHXckJXrlzBypUr8fHHHysmnXicn58foqOjcevWLSxatEgHEWrGiBEjsH//fsycORPm5uZP7FdUVAQnJyel+zrl9/C2vf1m9erVyMvL65GZmDipMSZgRITk5GTFBNK2traYOnWq0nyU3Sn1YwilkTQlNTUVRITAwMAn9omPj8fgwYOxY8cOHDt2rMPtdea5UacklzbLbrXHzc1N5Y8P+e9pbm5uSu22traYMGECUlJStD8Ste1XN778yJh+6srlx1WrVpGZmRnt2rWLKisr6cKFCzRq1Ciyt7en0tJSRb/ulPrRt9JIj9Pk5Uc3Nzfy9PRsdx13d3e6fv06ERF99913ZGRkRAMHDqTa2loiav/yY2efm86WddJ02S0iohdffPGJlx9zc3PJ1NSUUlNTqbq6mn766Sd67rnn6JVXXmm3f2xsrEo5MiK+/MgY66SGhgYkJydj2rRpiIiIgLW1Nby9vbF161aUl5erzPrTHYZSGqmr6urqcP36dbi7uz+1r6+vLxYsWIDCwkIsXbq03T5deW46KuukjbJbTzNhwgQsWbIEUVFRkMlk8PLyQk1NDXbs2NFu/0GDBgF4NDWdNnFSY0yg8vPzUVtbi9GjRyu1+/j4wMzM7InTgWmCvpUT6q6ysjIQESQSSaf6x8fHY8iQIdiyZQtOnTqlsry7z03bsk49VfroccuXL8e2bdvw7bffora2FteuXYOfnx98fX1RVFSk0l9+7u7cuaOVeOQ4qTEmUJWVlQAAKysrlWU2NjaoqanR6v6FVBqpsbERADocOPE4sViM9PR0iEQizJkzBw0NDUrLNf3cPF766PH7v27cuIH6+nq1ttUZt2/fRmJiIv7617/i5ZdfhqWlJVxdXbF9+3aUlJQgKSlJZR0LCwsAv51LbeGkxphA2djYAEC7H5DaLvUjpNJIwG8fyOrMhuPr64uYmBgUFBSo3LSv6efm8dJH9OhWLcXj9OnTam2rM+QVS/r27avULpPJYGdn1255rqamJgC/nUtt4aTGmEB5eXnBysoKP/zwg1L72bNn0dTUhBdeeEHRpulSP0IqjQQADg4OEIlEqKqqUmu9NWvWYOjQoTh//rxSuzrPTWf0dOkjedJtW4GipqYG9+/fb7c8l/zcOTo6ajU2TmqMCZRYLMbChQtx4MABZGRkoLq6GhcvXsS8efPg5OSEyMhIRd/ulvoRUmmk9kgkEri5uaG4uFit9eSXIY2NjVXaO/vcdHY/Tyt9FB4eDkdHR41M0+Xq6oqJEydi+/btOHnyJBoaGlBUVKSI+y9/+YvKOvJz5+3t3e39d6jtcEge0s+YfurKkP7W1lZKSkqiQYMGkampKdna2lJwcDBdvnxZqV93yhbpS2mk9mhySH9UVBSZmppSfX29ou3AgQOKUkf29vb0/vvvt7vNxYsXqwzp78xzo05Zp6eV3QoODiYAtGrVqg6P//Tp0zR27FhycnJSlGrq06cP+fn50YkTJxT9ysvLKTo6mjw8PMjc3JysrKxo7Nix9MUXX7S73YCAAHJ2dlaagYRI80P6OakxZiD0YZqs9sjnStRHmkxqBQUFZGJi0qVaePqgpaWFxo8fTzt37uzxfZeXl5NYLKaNGzeqLOP71BhjeseQygl1RkNDA77++msUFBQoBjh4eHggLi4OcXFxqK2t1XGE6mlpaUF2djZqamp0Ujlk9erVGDlyJKKiogA8mk2lpKQEp06dwpUrVzS6L05qjDHWxv379/Hqq69i8ODBmDNnjqI9NjYWYWFhCA8PV3vQiC7l5uZi//79yMnJ6fS9dpqSnJyMvLw8HDlyBKampgCAgwcPwtnZGePHj8dXX32l0f1xUmOMddmyZcuQnp6OqqoquLq6Yt++fboOqdu2bt2qNCQ+IyNDafnatWsRFRWFdevW6ShC9U2aNAmfffaZ0tybPeHgwYN48OABcnNzYWtrq2ifOnWq0jmWz/2pCSYa2xJj7JmTkJCAhIQEXYfR4/z9/eHv76/rMPReUFAQgoKCenSf/E2NMcaYYHBSY4wxJhic1BhjjAkGJzXGGGOC8cSBIpmZmT0ZB2PsKeTTDPF7s/Pkk/nyOROe4uLi9id+bns3tvwOfH7wgx/84Ac/9PnR3owiIiIiMMa0SiQSYe/evZg+fbquQ2FM0Pg3NcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4IhIiLSdRCMCUlkZCQuX76s1Hbu3Dm4urrC1tZW0WZsbIx//OMf6NevX0+HyJhgmeg6AMaExtHREdu2bVNpv3DhgtL/3dzcOKExpmF8+ZExDXvjjTee2sfMzAyzZs3SfjCMPWP48iNjWuDl5YVLly6ho7fX5cuXMXjw4B6MijHh429qjGnBW2+9BWNj43aXiUQijBgxghMaY1rASY0xLXj99dfR0tLS7jJjY2P8+c9/7uGIGHs28OVHxrTEz88PZ8+eRWtrq1K7SCRCUVERnJ2ddRQZY8LF39QY05I333wTIpFIqc3IyAjjxo3jhMaYlnBSY0xLwsLCVNpEIhHeeustHUTD2LOBkxpjWmJvb49JkyYpDRgRiUQIDg7WYVSMCRsnNca0KCIiQjGs39jYGK+88gp69eql46gYEy5Oaoxp0bRp02BmZgYAICJEREToOCLGhI2TGmNaZGlpiT/96U8AHs0iMmXKFB1HxJiwcVJjTMtmzpwJAAgODoalpaWOo2FM2AR7n1pmZiZmzJih6zAYY0zvhIaGIisrS9dhaIXgZ+nfu3evrkNgBmbTpk0AgAULFmhsmxkZGQgPD4eJiTDfcqdPn0ZKSgq/3wyA/PUtVMJ8hz1m+vTpug6BGRj5X7CafO0EBgZCLBZrbHv6KCUlhd9vBkCo39Dk+Dc1xnqA0BMaY/qCkxpjjDHB4KTGGGNMMDipMcYYEwxOaowxxgSDkxpjWnLkyBFYW1vjyy+/1HUoeu/YsWOIjY3F/v374ebmBpFIBJFIhDfffFOlr7+/P6RSKYyNjTFs2DCcO3dOBxGrr7W1FZs2bYKfn98T+3z++efw8fGBVCrFgAEDMHv2bJSWliqWHzp0CImJiU8sQMs4qTGmNQKd10DjPvroI6SmpmLZsmUICQnBtWvX4O7ujl69eiEjIwNfffWVUv9vvvkGWVlZmDJlCvLz8zFq1CgdRd55BQUF+P3vf4+YmBjU19e322fv3r2YOXMmwsLCUFxcjIMHD+LkyZN47bXX8PDhQwC/3RoyadIkVFZW9uQhGAxOaoxpSUBAAKqqqvRivseGhoYOvyHoyvr167Fnzx5kZmZCKpUqLUtNTYWRkREiIyNRVVWlowi778cff8TSpUsxb948jBw58on9/v73v6Nv375YvHgxrK2tMXLkSMTExCAvLw9nz55V9Js/fz5GjBiByZMnK5Id+w0nNcaeATt37kRZWZmuw1By5coVrFy5Eh9//HG79/H5+fkhOjoat27dwqJFi3QQoWaMGDEC+/fvx8yZM2Fubv7EfkVFRXByclKqlt6/f38AwI0bN5T6rl69Gnl5eUhJSdFO0AaMk8RdVZYAACAASURBVBpjWnDq1Cm4uLhAJBLhk08+AQCkpaXB0tISEokEBw8exGuvvQaZTIZ+/fph9+7dinVTU1MhFovh4OCAuXPnwsnJCWKxGH5+fkp/sUdFRcHMzAx9+vRRtL333nuwtLSESCRCeXk5ACA6OhoLFy7E1atXIRKJ4OHhAQA4evQoZDIZ1q5d2xOnREVqaiqICIGBgU/sEx8fj8GDB2PHjh04duxYh9sjIiQnJ+O5556Dubk5bG1tMXXqVPzyyy+KPp19DgCgpaUFq1atgouLCywsLDB8+HCtTgPm5uam8oeH/Pc0Nzc3pXZbW1tMmDABKSkpfJm7LRKovXv3koAPj2lRaGgohYaGdns7RUVFBIA2b96saFu+fDkBoG+//ZaqqqqorKyMxo8fT5aWltTU1KToFxkZSZaWlnTp0iVqbGyk/Px88vHxIalUSjdv3lT0mzlzJjk6OirtNykpiQDQ3bt3FW0hISHk7u6u1O/w4cMklUopLi6u28falfebm5sbeXp6trvM3d2drl+/TkRE3333HRkZGdHAgQOptraWiIhycnIoKChIaZ1Vq1aRmZkZ7dq1iyorK+nChQs0atQosre3p9LSUkW/zj4HixYtInNzc9q3bx9VVFTQsmXLyMjIiL7//nu1jvNxL774Io0YMaLdZbm5uWRqakqpqalUXV1NP/30Ez333HP0yiuvtNs/NjaWAND58+fVikFTr299xd/UGNMBPz8/yGQy9O7dG+Hh4airq8PNmzeV+piYmCi+dXh6eiItLQ01NTVIT0/XSAwBAQGorq7GypUrNbI9ddTV1eH69etwd3d/al9fX18sWLAAhYWFWLp0abt9GhoakJycjGnTpiEiIgLW1tbw9vbG1q1bUV5ejm3btqms09Fz0NjYiLS0NAQHByMkJAQ2NjZYsWIFTE1NNXb+25owYQKWLFmCqKgoyGQyeHl5oaamBjt27Gi3/6BBgwAAFy9e1Eo8hoqTGmM6Jq+M3dzc3GG/0aNHQyKRKF1OM1RlZWUgIkgkkk71j4+Px5AhQ7BlyxacOnVKZXl+fj5qa2sxevRopXYfHx+YmZkpXbZtT9vn4PLly6ivr4eXl5eij4WFBfr06aO18798+XJs27YN3377LWpra3Ht2jX4+fnB19cXRUVFKv3l5+7OnTtaicdQcVJjzICYm5vj7t27ug6j2xobGwGgw4ETjxOLxUhPT4dIJMKcOXPQ0NCgtFw+vN3KykplXRsbG9TU1KgVX11dHQBgxYoVinvmRCIRbty48cQh+d1x+/ZtJCYm4q9//StefvllWFpawtXVFdu3b0dJSQmSkpJU1rGwsADw27lkj3BSY8xANDc3o7KyEv369dN1KN0m/0BW5yZiX19fxMTEoKCgAGvWrFFaZmNjAwDtJq+unLPevXsDeFR7jIiUHqdPn1ZrW51RUFCAlpYW9O3bV6ldJpPBzs4O+fn5Kus0NTUB+O1cskc4qTFmIHJzc0FEGDNmjKLNxMTkqZct9ZGDgwNEIpHa95+tWbMGQ4cOxfnz55Xavby8YGVlhR9++EGp/ezZs2hqasILL7yg1n769+8PsViMvLw8tdbrKnnSvX37tlJ7TU0N7t+/rxja/zj5uXN0dNR+gAaEkxpjeqq1tRUVFRV4+PAhLly4gOjoaLi4uGDWrFmKPh4eHrh//z6ys7PR3NyMu3fvqtzTBAB2dnYoKSlBYWEhampq0NzcjJycHJ0N6ZdIJHBzc0NxcbFa68kvQxobG6u0L1y4EAcOHEBGRgaqq6tx8eJFzJs3D05OToiMjFR7P7Nnz8bu3buRlpaG6upqtLS0oLi4WJF4wsPD4ejoqJFpulxdXTFx4kRs374dJ0+eRENDA4qKihRx/+Uvf1FZR37uvL29u71/QdHl0Ett4iH9rKs0MeR58+bN1KdPHwJAEomEAgMDacuWLSSRSAgADRo0iK5evUrbtm0jmUxGAGjAgAH066+/EtGjIf2mpqbk7OxMJiYmJJPJaOrUqXT16lWl/dy7d48mTpxIYrGYXF1d6YMPPqDFixcTAPLw8FAM/z937hwNGDCALCwsaNy4cVRaWkpHjhwhqVRK8fHx3TpWoq6936KiosjU1JTq6+sVbQcOHCB3d3cCQPb29vT++++3u+7ixYtVhvS3trZSUlISDRo0iExNTcnW1paCg4Pp8uXLij7qPAcPHjygJUuWkIuLC5mYmFDv3r0pJCSE8vPziYgoODiYANCqVas6PM7Tp0/T2LFjycnJiQAQAOrTpw/5+fnRiRMnFP3Ky8spOjqaPDw8yNzcnKysrGjs2LH0xRdftLvdgIAAcnZ2ptbW1g7335bQh/QL9lOfkxrrKn1400dGRpKdnZ1OY1BHV95vBQUFZGJiQrt27dJSVNrV0tJC48ePp507d/b4vsvLy0ksFtPGjRvVXlcfXt/axJcfGdNTQp+J3cPDA3FxcYiLi0Ntba2uw1FLS0sLsrOzUVNTg/Dw8B7f/+rVqzFy5EhERUX1+L71HSe1Drz99tuQSqUQiUQ99oOxpiUmJmLo0KGwsLCApaUlhg4dipUrV6K6ulrtbbUtCyJ/mJmZwcHBAS+99BKSkpJQUVGhhSNhQhQbG4uwsDCEh4cb1KTFubm52L9/P3Jycjp9r52mJCcnIy8vD0eOHIGpqWmP7tsQcFLrwI4dO7B9+3Zdh9Et//rXv/DOO+/g5s2buHPnDtasWYPExESEhoaqva3Hy4JYW1uDiNDa2oqysjJkZmbC1dUVS5YswbBhw1RGobHOW7ZsGdLT01FVVQVXV1fs27dP1yFp1dq1axEVFYV169bpOpROmzRpEj777DOleTd7wsGDB/HgwQPk5ubC1ta2R/dtKEx0HQDTLjMzM7z33nuKWdDDwsKQlZWFrKws3L59G05OTt3avkgkgo2NDV566SW89NJLCAgIwIwZMxAQEIBff/0V1tbWmjiMZ0pCQgISEhJ0HUaP8vf3h7+/v67D0HtBQUEICgrSdRh6jb+pPcXjZSAM0YEDB1TKejg7OwOAVn7HCA0NxaxZs1BWVoatW7dqfPuMMdYRTmqPISIkJSVhyJAhMDc3h7W1NRYvXqzSr6OSFOqUtjhx4gR+97vfQSKRQCaTwdvbW/FblzbLXhQUFMDGxgYDBgxQtGmyDIn8PqqcnBxFm6GfM8aYgdD18Ett6coQ4+XLl5NIJKK//e1vVFFRQfX19bRlyxaV8g5PK0nRmdIWtbW1JJPJKDExkRoaGqi0tJSmTZumKBei6bIXTU1NVFxcTJs3byZzc3OVYdTqlCFxd3cna2vrJy6vrq4mANS/f39FmyGdM6EPedYGvoXGcAj99S3YV6G6b7L6+nqSSCT0xz/+Ual99+7dSkmtoaGBJBIJhYeHK61rbm5O7777LhH99gHd0NCg6CNPjleuXCEiop9++okA0OHDh1Vi6cw+1OXo6EgAqFevXvQf//EfSnWj1PW0pEZEJBKJyMbGhogM75wJ/U2vDZzUDIfQX988UOT/XLlyBfX19Zg0aVKH/bpakqJtaQs3Nzc4ODggIiIC8+fPx6xZszBw4MBu7aMjRUVFqKysxPnz5xEbG4tt27bh+PHjcHBw6NL2OlJXVwcigkwmA2CY56y4uBiZmZlqr/eskk/yy+dM/xUXFwtiUuwn0nVW1RZ1/3I8cuQIAVCZHaDtN7X//d//VUx10/YxZswYImr/W8f27dsJAP3888+Ktp9++on+9Kc/kYmJCYlEIpoxYwbV19d3ah/d8euvvxIAmj9/fpfWf9o3tXPnzhEA8vf3JyLDO2ehoaFP3BY/+CGEh5C/qfFAkf8jHyH44MGDDvtpsiTFsGHD8OWXX6KkpARLlizB3r17sXHjRq2XvfDw8ICxsXG75Sw04ejRowCA1157DYBhnrPQ0FCV7fDjyQ/5gBxdx8GPpz+6co+qIeGk9n+8vLxgZGSEEydOdNhPUyUpSkpKcOnSJQCPPvTXrVuHUaNG4dKlSxrbx7179/DGG2+otMtrN7VXzqK7SktLsWnTJvTr1w9z5swBYFjnjDFm2Dip/Z/evXsjJCQE+/btw86dO1FdXY0LFy5g27ZtSv06U5KiM0pKSjB37lz88ssvaGpqwvnz53Hjxg2MGTNGY/uwtLTEN998g+PHj6O6uhrNzc04f/48/vznP8PS0hIxMTGKvuqWISEi1NbWorW1FUSEu3fvYu/evRg7diyMjY2RnZ2t+E3NkM4ZY8zAkUB1ZTRWTU0Nvf3229SrVy+ysrKicePG0apVqwgA9evXj3788Uci6rgkRWdLWxQWFpKfnx/Z2tqSsbEx9e3bl5YvX04PHz586j7UERgYSK6urmRlZUXm5ubk7u5O4eHhdPHiRaV+nSlDcujQIRo+fDhJJBIyMzMjIyMjAqAY6fi73/2O4uLi6N69eyrrGtI5E/roMG3g0Y+GQ+ivbxERke5SqvZkZmZixowZEOjhMS0KCwsDAGRlZek4EsPB7zfDIfTXN19+ZIwxJhic1AzML7/8olL6pb2HLmo8McaYrnFSMzBDhw7t1LDdPXv26DpUxjrt2LFjiI2NVanZ9+abb6r09ff3h1QqhbGxMYYNG4Zz587pIOLOe+mll574x6eVlZWiX1xcHDw9PSGTyWBubg4PDw98+OGHShOPHzp0CImJiYIvINsdnNQYYzr10UcfITU1FcuWLVOq2derVy9kZGTgq6++Uur/zTffICsrC1OmTEF+fj5GjRqlo8i7b9y4cYp/Hz9+HO+//z4KCwtRXl6OhIQEpKSkKH4DA4DAwECIxWJMmjQJlZWVughZ73FSY0wPNTQ0wM/Pz+D38TTr16/Hnj17kJmZCalUqrQsNTUVRkZGiIyMNKiq2G2JxWJUV1erXE2JjIzEhx9+qOhnZWWFyMhI2NnZQSqVYvr06QgODsbRo0dRVFSk6Dd//nyMGDECkydPxsOHD3VxSHqNkxpjemjnzp0oKysz+H105MqVK1i5ciU+/vhjlZp/AODn54fo6GjcunULixYt0kGEmnH06FGVhF1UVISffvoJL7/8sqLt8OHDMDY2Vupnb28PAKivr1dqX716NfLy8pCSkqKlqA0XJzXGNICIkJycjOeeew7m5uawtbXF1KlTlSZTjoqKgpmZGfr06aNoe++992BpaQmRSITy8nIAQHR0NBYuXIirV69CJBLBw8MDqampEIvFcHBwwNy5c+Hk5ASxWAw/Pz+cPXtWI/sANFtX72lSU1NBRAgMDHxin/j4eAwePBg7duzAsWPHOtxeZ54DdWr3abM+3/r16zF//vyn9rt16xYsLCzg6uqq1G5ra4sJEyYgJSWFb6Noq0fviutBfDMo66qu3Jy6atUqMjMzo127dlFlZSVduHCBRo0aRfb29lRaWqroN3PmTHJ0dFRaNykpiQAo6sIREYWEhJC7u7tSv8jISLK0tKRLly5RY2Mj5efnk4+PD0mlUrp586ZG9qFOXb3HdeX95ubmRp6enu0uc3d3p+vXrxMR0XfffUdGRkY0cOBAqq2tJSKinJwcCgoKUlqns89BZ2r3EWm+pqFccXExeXp6UktLS4f96urqSCqVUlRUVLvLY2NjCVCu9dgZQr/5mr+pMdZNDQ0NSE5OxrRp0xAREQFra2t4e3tj69atKC8vV5lqrTtMTEwU30Q8PT2RlpaGmpoapKena2T7AQEBqK6uxsqVKzWyvSepq6vD9evX4e7u/tS+vr6+WLBgAQoLC7F06dJ2+3TlOfDz84NMJkPv3r0RHh6Ouro63Lx5EwDQ2NiItLQ0BAcHIyQkBDY2NlixYgVMTU27fa7Xr1+PDz74AEZGHX/8JiQkwMnJCfHx8e0uHzRoEADg4sWL3YpHaDipMdZN+fn5qK2txejRo5XafXx8YGZmpnR5UNNGjx4NiUTS5Tp7ulJWVgYigkQi6VT/+Ph4DBkyBFu2bMGpU6dUlnf3OWhbu08bNQ2BR/OXHjp0CLNmzeqw34EDB5CZmYmvv/5a5fc4Ofm5u3PnTpfjESJOaox1k3xo9eP3HMnZ2NigpqZGq/s3NzfH3bt3tboPTWtsbATwKPbOEIvFSE9Ph0gkwpw5c9DQ0KC0XNPPQV1dHQBgxYoVSveV3bhxQ2XQhjoSExPxzjvvtDswRm7Pnj1Yv349cnNzFUVw22NhYQHgt3PJHuGkxlg32djYAEC7H5yVlZVarTLc3Nys9X1og/wDWZ2biH19fRETE4OCggKsWbNGaZmmnwNt1DQsLS3F559/jnffffeJfTZv3oyMjAwcP34cffv27XB7TU1NAH47l+wRTmqMdZOXlxesrKzwww8/KLWfPXsWTU1NeOGFFxRtJiYmiktcmpCbmwsiwpgxY7S2D21wcHCASCRS+/6zNWvWYOjQoTh//rxSuzrPQWdooz5fYmIiIiIiYGdnp7KMiLBkyRJcvHgR2dnZ7X7jbEt+7hwdHTUWoxBwUmOsm8RiMRYuXIgDBw4gIyMD1dXVuHjxIubNmwcnJydERkYq+np4eOD+/fvIzs5Gc3Mz7t69ixs3bqhs087ODiUlJSgsLERNTY0iSbW2tqKiogIPHz7EhQsXEB0dDRcXF6XfaLqzD3Xr6nWVRCKBm5sbiouL1VpPfhmy7f1c6jwHnd3P0+rzhYeHw9HRsVPTdN25cweffvopFixY0O7yS5cuYcOGDdi+fTtMTU1VptPauHGjyjryc+ft7a3WsQkdJzXGNOCjjz5CQkIC4uLiYG9vjwkTJmDgwIHIzc2FpaWlot+7776LiRMn4vXXX8eQIUOwZs0axeUjX19fxcwR8+bNg4ODAzw9PTF58mTcv38fwKPfT7y9vWFhYYHx48dj8ODB+Oc//6n021R399FTAgICkJ+fr/T72BdffAEPDw9cvXoVPj4++OCDD1TWGzNmjFKBW7nOPAdpaWnYtGkTAGD48OG4du0atm/fjoULFwIAXn31VRQUFAAAUlJSsGDBAiQmJqJXr15wcnJCdHQ0KioqADy6/FdWVoaDBw8+9Vg3bNiAwMBAuLi4tLucunCv2ffffw9nZ2cMHz5c7XUFTXd3E2gX36fGukpf7+OJjIwkOzs7XYfRrq683woKCsjExIR27dqlpai0q6WlhcaPH087d+7s8X2Xl5eTWCymjRs3qr2uvr6+NYW/qTFmQIQ0O7uHhwfi4uIQFxenNBO9IWhpaUF2djZqamp0UuZp9erVGDlyJKKionp83/qOkxpjTGdiY2MRFhaG8PBwg5q0ODc3F/v370dOTk6n77XTlOTkZOTl5eHIkSMwNTXt0X0bAk5qjBmAZcuWIT09HVVVVXB1dcW+fft0HZLGrF27FlFRUVi3bp2uQ+m0SZMm4bPPPlOaY7MnHDx4EA8ePEBubi5sbW17dN+GwkTXATDGni4hIQEJCQm6DkNr/P394e/vr+sw9F5QUBCCgoJ0HYZe429qjDHGBIOTGmOMMcHgpMYYY0wwOKkxxhgTDMEPFAkLC9N1CMzAnDlzBgC/dtQhn7KJz5n+O3PmjNJcoUIjIhJmLfDTp08jOTlZ12EwBgDIycnB888/3+NDwBlrj7zigRAJNqkxpk9EIhH27t2L6dOn6zoUxgSNf1NjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgcFJjjDEmGJzUGGOMCQYnNcYYY4LBSY0xxphgmOg6AMaEprKyEkSk0l5XV4eKigqlNisrK5iamvZUaIwJnojae/cxxrrs5Zdfxj//+c+n9jM2NsatW7fg6OjYA1Ex9mzgy4+Madjrr78OkUjUYR8jIyP8/ve/54TGmIZxUmNMw0JDQ2Fi0vGVfZFIhLfeequHImLs2cFJjTENs7W1hb+/P4yNjZ/Yx8jICMHBwT0YFWPPBk5qjGlBREQEWltb211mYmKCgIAAWFtb93BUjAkfJzXGtCAwMBDm5ubtLmtpaUFEREQPR8TYs4GTGmNaIJFIEBwc3O5wfQsLC0yePFkHUTEmfJzUGNOSN954A83NzUptpqamCA0NhYWFhY6iYkzYOKkxpiWvvPKKyu9mzc3NeOONN3QUEWPCx0mNMS0xNTVFeHg4zMzMFG02NjaYNGmSDqNiTNg4qTGmRa+//jqampoAPEpyERERT72HjTHWdTxNFmNa1Nrair59++LOnTsAgFOnTmHs2LE6joox4eJvaoxpkZGREd58800AgJOTE/z8/HQcEWPCpnfXQYqLi/Hdd9/pOgzGNMbe3h4A8OKLLyIrK0vH0TCmOf3794evr6+uw1Cid5cfMzMzMWPGDF2HwRhj7ClCQ0P17g81vfumJqdnuZYxtYWFhQEAsrKysG/fPoSGhuo4Iv0n/6OW3//6T/761jf8mxpjPYATGmM9g5MaY4wxweCkxhhjTDA4qTHGGBMMTmqMMcYEg5MaY4wxweCkxpieO3LkCKytrfHll1/qOhS9d+zYMcTGxmL//v1wc3ODSCSCSCRSzOryOH9/f0ilUhgbG2PYsGE4d+6cDiLuvJdeeklxPG0fVlZWin5xcXHw9PSETCaDubk5PDw88OGHH6K2tlbR59ChQ0hMTERLS4suDkWrOKkxpuf4nq3O+eijj5Camoply5YhJCQE165dg7u7O3r16oWMjAx89dVXSv2/+eYbZGVlYcqUKcjPz8eoUaN0FHn3jRs3TvHv48eP4/3330dhYSHKy8uRkJCAlJQUpfvKAgMDIRaLMWnSJFRWVuoiZK3hpMaYngsICEBVVRWmTJmi61DQ0NCgl/NXrl+/Hnv27EFmZiakUqnSstTUVBgZGSEyMhJVVVU6irD7xGIxqqurQURKj8jISHz44YeKflZWVoiMjISdnR2kUimmT5+O4OBgHD16FEVFRYp+8+fPx4gRIzB58mQ8fPhQF4ekFZzUGGOdtnPnTpSVlek6DCVXrlzBypUr8fHHH0MsFqss9/PzQ3R0NG7duoVFixbpIELNOHr0qErCLioqwk8//YSXX35Z0Xb48GEYGxsr9ZPPP1pfX6/Uvnr1auTl5SElJUVLUfc8TmqM6bFTp07BxcUFIpEIn3zyCQAgLS0NlpaWkEgkOHjwIF577TXIZDL069cPu3fvVqybmpoKsVgMBwcHzJ07F05OThCLxfDz88PZs2cV/aKiomBmZoY+ffoo2t577z1YWlpCJBKhvLwcABAdHY2FCxfi6tWrEIlE8PDwAPDow1Ymk2Ht2rU9cUpUpKamgogQGBj4xD7x8fEYPHgwduzYgWPHjnW4PSJCcnIynnvuOZibm8PW1hZTp07FL7/8oujT2ecAAFpaWrBq1Sq4uLjAwsICw4cPx969e7t30P9n/fr1mD9//lP73bp1CxYWFnB1dVVqt7W1xYQJE5CSkiKcy9ykZ/bu3Ut6GBZjagsNDaXQ0NBub6eoqIgA0ObNmxVty5cvJwD07bffUlVVFZWVldH48ePJ0tKSmpqaFP0iIyPJ0tKSLl26RI2NjZSfn08+Pj4klUrp5s2bin4zZ84kR0dHpf0mJSURALp7966iLSQkhNzd3ZX6HT58mKRSKcXFxXX7WLvy/ndzcyNPT892l7m7u9P169eJiOi7774jIyMjGjhwINXW1hIRUU5ODgUFBSmts2rVKjIzM6Ndu3ZRZWUlXbhwgUaNGkX29vZUWlqq6NfZ52DRokVkbm5O+/bto4qKClq2bBkZGRnR999/r9ZxtlVcXEyenp7U0tLSYb+6ujqSSqUUFRXV7vLY2FgCQOfPn1dr/5p6fWsaf1NjzID5+flBJpOhd+/eCA8PR11dHW7evKnUx8TERPGtw9PTE2lpaaipqUF6erpGYggICEB1dTVWrlypke2po66uDtevX4e7u/tT+/r6+mLBggUoLCzE0qVL2+3T0NCA5ORkTJs2DREREbC2toa3tze2bt2K8vJybNu2TWWdjp6DxsZGpKWlITg4GCEhIbCxscGKFStgamra7fO/fv16fPDBBzAy6vhjPCEhAU5OToiPj293+aBBgwAAFy9e7FY8+oKTGmMCYWZmBgBobm7usN/o0aMhkUiULqcZqrKyMhARJBJJp/rHx8djyJAh2LJlC06dOqWyPD8/H7W1tRg9erRSu4+PD8zMzJQu27an7XNw+fJl1NfXw8vLS9HHwsICffr06db5LykpwaFDhzBr1qwO+x04cACZmZn4+uuvVX6Pk5OfO3l1dkPHSY2xZ5C5uTnu3r2r6zC6rbGxEcCj4+kMsViM9PR0iEQizJkzBw0NDUrL5cPbH7/vS87GxgY1NTVqxVdXVwcAWLFihdJ9ZTdu3FAZtKGOxMREvPPOO+0OjJHbs2cP1q9fj9zcXAwcOPCJ/SwsLAD8di4NHSc1xp4xzc3NqKysRL9+/XQdSrfJP5DVuYnY19cXMTExKCgowJo1a5SW2djYAEC7yasr56x3794AgE2bNqkMxT99+rRa25IrLS3F559/jnffffeJfTZv3oyMjAwcP34cffv27XB7TU1NAH47l4aOkxpjz5jc3FwQEcaMGaNoMzExeeplS33k4OAAkUik9v1na9aswdChQ3H+/Hmldi8vL1hZWeGHH35Qaj979iyamprwwgsvqLWf/v37QywWIy8vT631OpKYmIiIiAjY2dmpLCMiLFmyBBcvXkR2dna73zjbkp87R0dHjcWoS5zUGBO41tZWVFRU4OHDh7hw4QKio6Ph4uKi9HuMh4cH7t+/j+zsbDQ3N+Pu3bu4ceOGyrbsSEn4mAAAIABJREFU7OxQUlKCwsJC1NTUoLm5GTk5OTob0i+RSODm5obi4mK11pNfhmx7P5dYLMbChQtx4MABZGRkoLq6GhcvXsS8efPg5OSEyMhItfcze/Zs7N69G2lpaaiurkZLSwuKi4tx+/ZtAEB4eDgcHR07NU3XnTt38Omnn2LBggXtLr906RI2bNiA7du3w9TUVGU6rY0bN6qsIz933t7eah2bvuKkxpge++STT+Dj4wMAWLJkCYKCgpCWloZNmzYBAIYPH45r165h+/btWLhwIQDg1VdfRUFBgWIbjY2N8Pb2hoWFBcaPH4/Bgwfjn//8p9LvUO+++y4mTpyI119/HUOGDMGaNWsUl6N8fX0VM1HMmzcPDg4O8PT0xOTJk3H//v0eOQ8dCQgIQH5+vtLvY1988QU8PDxw9epV+Pj44IMPPlBZb8yYMYiJiVFp/+ijj5CQkIC4uDjY29tjwoQJGDhwIHJzc2FpaQkAaj0HKSkpWLBgARITE9GrVy84OTkhOjoaFRUVAB5d/isrK8PBgwefeqwbNmxAYGAgXFxc2l1OXbjX7Pvvv4ezszOGDx+u9rp6SXd3E7SP71NjQqEP9/FERkaSnZ2dTmNQR1fe/wUFBWRiYkK7du3SUlTa1dLSQuPHj6edO3f2+L7Ly8tJLBbTxo0b1V5XH17f7eFvaowJnBBnYn+ch4cH4uLiEBcXpzQTvSFoaWlBdnY2ampqEB4e3uP7X716NUaOHImoqKge37e2CDKpvf3225BKpRCJRBr9gdaQtba2YtOmTd2ajLZtOQ/5w8zMDA4ODnjppZeQlJSkuKzCWE+JjY1FWFgYwsPDDWrS4tzcXOzfvx85OTmdvtdOU5KTk5GXl4cjR47A1NS0R/etTYJMajt27MD27dt1HYbeKCgowO9//3vExMR0696Yx8t5WFtbg4jQ2tqKsrIyZGZmwtXVFUuWLMGwYcNURo+xnrds2TKkp6ejqqoKrq6u2Ldvn65D0qq1a9ciKioK69at03UonTZp0iR89tlnSvNu9oSDBw/iwYMHyM3Nha2tbY/uW9sEmdSEpjvlPn788UcsXboU8+bNw8iRIzUcGSASiWBjY4OXXnoJ6enpyMzMxJ07dxTlUgydvpZa6YyEhAQ8ePAARITr168jNDRU1yFpnb+/P9avX6/rMPReUFAQYmNjVUZ/CoFgk5pIJNJ1CBrTnXIfI0aMwP79+zFz5sxOz7rQHaGhoZg1axbKysqwdetWre9P2/Sx1Apj7MkEkdSICElJSRgyZAjMzc1hbW2NxYsXK/XZsGEDJBIJpFIpysrKsHDhQjg7O+Py5cudKjXR2TIe8nietr3ulvvQNE2WD5Hf/5STkwOAzz1jrAfpcORlu7oypHf58uUkEonob3/7G1VUVFB9fT1t2bJFpZyCvFTE/PnzafPmzTRt2jT6+eefO11qorNlPDq7ve6U++iKF198kUaMGNHuMnXKh7i7u5O1tfUTl1dXVxMA6t+/v6LtWTz3+jrkWZ/xLT2GQ19f33r36lH3RV1fX08SiYT++Mc/KrXv3r37iUmtoaFBaX0rKysKDw9XWv///b//RwCUPuQjIyNVPsy///57AkAff/yx2tvTp6SmjqclNSIikUhENjY2iv8/i+deX9/0+oyTmuHQ19e3SY9+LdSCK1euoL6+HpMmTerS+t0tNdG2jEd3tycEdXV1ICLIZLIO+z0L5/7MmTMICwvr8f0aKvmUTXzO9N+ZM2eU5g/VFwb/m5r8TSCfDVtdmig18XgZD02XrjBEv/76KwBg6NChHfbjc88Y0zSD/6Ymryf04MGDLq3f3VITbct4aLp0hSE6evQoAOC1117rsN+zcO7HjBmDrKysHt+vocrMzMSMGTP4nBkAff02bfDf1Ly8vGBkZIQTJ050ef3ulJpoW8ZDne0ZarmPjpSWlmLTpk3o168f5syZ02FfPveMMU0z+KTWu3dvhISEYN++fdi5cyeqq6tx4cIFbNu2rVPrq1tq4mllPNTZXnfKfWiauuVDiAi1tbVobW0FEeHu3bvYu3cvxo4dC2NjY2RnZz/1NzU+94wxjdPpMJV2dGX0U01NDb399tvUq1cvsrKyonHjxtGqVasIAPXr149+/PFHSkxMJAsLC8VQ88dn9G5tbaWkpCQaNGgQmZqakq2tLQUHB9Ply5eV9hMZGUmmpqbk7OxMJiYmJJPJaOrUqXT16lWlfp3d3r1792jixIkkFovJ1dWVPvjgA1q8eDEB+P/s3XtQFGfWP/BvAwPDwAwXUcQLysUbXuB1NQmjLkmsNau8apCIZMWsZrMhJgYRNYoXYhAJBBcpXCnjZamtmAh4iRojJqUuSVlarvsqEfEXg0YUNIgoyF0Qzu+PLBOHGXAGZuihPZ+qqUq6n+4+3W3PoXuefg75+vpquqpfuHCBhgwZQvb29jR58mStrulPc/bsWZo0aRJ5eHgQAAJA/fv3J7VaTd99952m3bFjx0ipVFJCQkKH6zpy5AiNGzeOFAoF2drakpWVFQHQ9HR87rnnKD4+nu7fv6+13LN67C21d5gl496PvYel/vsWiLpQgMeM2p6pW1hYAIB33nkH+/btw/3798UO5ZnTG499228O/PuQ4Sz5+mfaLPXfd69//NjTpF7Gw5LxsWeMPQ0ntV7mxx9/1Cn9ou8jRm0mxsR24sQJxMbG6pRJWrBggU7badOmQalUwtraGqNHj8aFCxdEiNhwL774YofX+5OvscTHx8PPzw8qlQp2dnbw9fXFBx98oFVr7siRI0hOTpbkH4qc1AxkKWU8Ro4cCfp1JJhOP1lZWaLEZw6WcuyZZfvwww+Rnp6ONWvWaJVJ6tOnD/bs2YOvv/5aq/23336Lffv2YebMmSgsLMT48eNFirz7Jk+erPnvU6dOYcmSJSguLkZFRQUSExORlpam1QV/1qxZkMvlmDp1qub9TqngpGagZ7GMh6XgY991PVE6xxLK8yQlJSErKws5OTlQKpVa89LT02FlZYXIyMheXQ5JLpejurpa5w/YyMhIfPDBB5p2jo6OiIyMhKurK5RKJcLCwhASEoLjx4+jpKRE027p0qXw9/fHjBkz8PjxYzF2ySw4qTEmYT1ROkfs8jzXrl3D+vXr8dFHH2kGY3iSWq1GdHQ0bt++jRUrVogQoWkcP35cJ2GXlJTg8uXLePnllzXTjh49qlMnzc3NDQB0igRv2LAB+fn5SEtLM1PUPY+TGmMWhMxcOsfQMj7dLc9jylJGT5Oeng4iwqxZszpsk5CQgOHDh2PXrl04ceJEp+sz5BxkZGTAwcEBCoUChw8fxvTp06FSqTBo0CDs3btXa30tLS2Ii4uDp6cn7O3tMW7cOGRnZ3dvp/8rKSkJS5cufWq727dvw97eHl5eXlrTXVxcEBQUhLS0NOn0OO3B1wcMwu+pMKnoyns8PVE6x9AyPt3ZhjGljJ7Ulevf29ub/Pz89M7z8fGhGzduEBHRmTNnyMrKioYOHUq1tbVERJSbm0uzZ8/WWsbQc9BWeeLkyZP08OFDKi8vpylTppCDgwM1NTVp2q1YsYLs7Oxo//79VFlZSWvWrCErKys6f/68UfvZXmlpKfn5+VFLS0un7erq6kipVFJUVJTe+bGxsToVTQxhqe+p8Z0aYxaioaEBqampmDNnDiIiIuDk5ISxY8di+/btqKioMHiUHEPY2Nho7kT8/PyQkZGBmpoaZGZmmmT9wcHBqK6uxvr1602yvo7U1dXhxo0b8PHxeWrbwMBALFu2DMXFxVi9erXeNl05B2q1GiqVCn379kV4eDjq6upw69YtAEBjYyMyMjIQEhKC0NBQODs7Y926dZDJZN0+1klJSXj//fdhZdX513hiYiI8PDyQkJCgd/6wYcMAAAUFBd2Kx1JwUmPMQohZOqd9GZ/eory8HEQEhUJhUPuEhASMGDEC27Ztw+nTp3Xmd/cc2NraAoBmOLWrV6+ivr4eY8aM0bSxt7dH//79u3Ws79y5gyNHjmiGiOvIwYMHkZOTg2+++Ubn97g2bcfu7t27XY7HknBSY8xCiF0658kyPr1FY2MjgF9jN4RcLkdmZiYEQcCbb76JhoYGrfmmPgd1dXUAgHXr1mm9V3bz5k2dThvGSE5Oxl//+le9HWPaZGVlISkpCXl5eRg6dGiH7ezt7QH8dix7O05qjFkIMUvntC/j01u0fSEb8xJxYGAgYmJiUFRUhI0bN2rNM/U5aKvzuGXLFp2u+GfPnjVqXW3KysrwxRdf4N133+2wzdatW7Fnzx6cOnUKAwYM6HR9TU1NAH47lr0dJzXGLISYpXPal/ExxzbMoV+/fhAEwej3zzZu3IiRI0fi4sWLWtO7Ww6pvcGDB0MulyM/P9+o5TqTnJyMiIgIuLq66swjIqxatQoFBQU4dOiQ3jvO9tqOnbu7u8liFBMnNcYsRE+WznlaGZ/ubsPYUkZdpVAo4O3tjdLSUqOWa3sM2f59LmPLIRmynUWLFmHv3r3IyMhAdXU1WlpaUFpail9++QUAEB4eDnd3d4OG6bp79y7+8Y9/YNmyZXrnX7lyBZ988gl27twJmUymM5zW5s2bdZZpO3Zjx441at8sFSc1xizIhx9+iMTERMTHx8PNzQ1BQUEYOnQo8vLy4ODgoGn37rvv4qWXXsLrr7+OESNGYOPGjZrHR4GBgZqRIxYvXox+/frBz88PM2bMwIMHDwD8+vvJ2LFjYW9vjylTpmD48OH417/+pfXbVHe30VOCg4NRWFio9fvYl19+CV9fX1y/fh0TJ07E+++/r7PcCy+8gJiYGJ3phpyDjIwMbNmyBQAwbtw4/Pzzz9i5cyeWL18OAPjjH/+IoqIiAEBaWhqWLVuG5ORk9OnTBx4eHoiOjkZlZSWAXx//lZeX4/Dhw0/d108++QSzZs2Cp6en3vnUhXfNzp8/j4EDB2LcuHFGL2uRxHubQD9+T41JhaW+xxMZGUmurq5ih6FXV67/oqIisrGx0arT15u0tLTQlClTaPfu3T2+7YqKCpLL5bR582ajl7XUf998p8bYM0hKo7P7+voiPj4e8fHxWiPR9wYtLS04dOgQampqRKmssWHDBgQEBCAqKqrHt20unNQYY71ebGws5s6di/Dw8F41aHFeXh4OHDiA3Nxcg9+1M5XU1FTk5+fj2LFjkMlkPbptc+KkxtgzRMplfDZt2oSoqCh8/PHHYodisKlTp+Lzzz/XGmOzJxw+fBiPHj1CXl4eXFxcenTb5mYjdgCMsZ6TmJiIxMREscMwm2nTpmHatGlih2HxZs+ejdmzZ4sdhlnwnRpjjDHJ4KTGGGNMMjipMcYYkwxOaowxxiSDkxpjjDHJsNjej4IgiB0CYybB/5aNx8esd3jttdfEDkGHQNSFwcLMqLS0FGfOnBE7DMZMat68eYiOjkZgYKDYoTBmMoMHD7a4f9MWl9QYkyJBEJCdnY2wsDCxQ2FM0vg3NcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSQYnNcYYY5LBSY0xxphkcFJjjDEmGZzUGGOMSYaN2AEwJjV79+5FTU2NzvQTJ06gqqpKa1pISAj69u3bU6ExJnkCEZHYQTAmJQsXLsQ///lPyGQyzbS2y0wQBABAS0sLHB0dUV5eDjs7O1HiZEyK+PEjYyb2+uuvAwCam5s1n8ePH+Px48ea/7e2tsbcuXM5oTFmYnynxpiJPX78GO7u7njw4EGn7U6ePImXX365h6Ji7NnAd2qMmZiNjQ1ef/11rceP7bm5uSEoKKgHo2Ls2cBJjTEzeP3119Hc3Kx3nkwmw4IFC2Btbd3DUTEmffz4kTEzICJ4enqitLRU7/x///vfmDhxYg9HxZj08Z0aY2YgCAIiIiL0PoIcPHgwJkyYIEJUjEkfJzXGzETfI0iZTIaFCxdquvYzxkyLHz8yZkYjR47E1atXtaZdvnwZo0ePFikixqSN79QYM6MFCxZoPYL08/PjhMaYGXFSY8yMIiIi8PjxYwC/Pnr885//LHJEjEkbP35kzMwmTJiA//u//4MgCCguLoanp6fYITEmWXynxpiZvfHGGwCA559/nhMaY2YmmVH6586dK3YIjOnV2NgIQRDw6NEj/nfKLFZMTAwCAwPFDqPbJHOntn///g5fdGWsI6Wlpdi/f79ZtyGXy+Hu7o5BgwaZdTs9ia83adm/fz9KSkrEDsMkJHOnBgDLli1DWFiY2GGwXiQnJwfz5s3Dvn37zLqda9euwdfX16zb6EmCIPD1JiFSem9SMndqjFkyKSU0xiwZJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGTODYsWNwcnLCV199JXYovdKJEycQGxuLAwcOwNvbG4IgQBAELFiwQKfttGnToFQqYW1tjdGjR+PChQsiRGy4F198UbM/7T+Ojo6advHx8fDz84NKpYKdnR18fX3xwQcfoLa2VtPmyJEjSE5ORktLixi70itwUmPMBHi0ua778MMPkZ6ejjVr1iA0NBQ///wzfHx80KdPH+zZswdff/21Vvtvv/0W+/btw8yZM1FYWIjx48eLFHn3TZ48WfPfp06dwpIlS1BcXIyKigokJiYiLS1N64X9WbNmQS6XY+rUqaiqqhIjZIvHSY0xEwgODsbDhw8xc+ZMsUNBQ0MD1Gq12GEYJCkpCVlZWcjJyYFSqdSal56eDisrK0RGRuLhw4ciRdh9crkc1dXVICKtT2RkJD744ANNO0dHR0RGRsLV1RVKpRJhYWEICQnB8ePHtV6MXrp0Kfz9/TFjxgzNYNnsN5zUGJOY3bt3o7y8XOwwnuratWtYv349PvroI8jlcp35arUa0dHRuH37NlasWCFChKZx/PhxnYRdUlKCy5cv4+WXX9ZMO3r0KKytrbXaubm5AQDq6+u1pm/YsAH5+flIS0szU9S9Fyc1xrrp9OnT8PT0hCAI+Pvf/w4AyMjIgIODAxQKBQ4fPozp06dDpVJh0KBB2Lt3r2bZ9PR0yOVy9OvXD++88w48PDwgl8uhVqtx7tw5TbuoqCjY2tqif//+mmnvvfceHBwcIAgCKioqAADR0dFYvnw5rl+/DkEQNC99Hz9+HCqVCps2beqJQ2KQ9PR0EBFmzZrVYZuEhAQMHz4cu3btwokTJzpdHxEhNTUVo0aNgp2dHVxcXPDqq6/ixx9/1LQx9LwAQEtLC+Li4uDp6Ql7e3uMGzcO2dnZ3dvp/0pKSsLSpUuf2u727duwt7eHl5eX1nQXFxcEBQUhLS2NH323RxIBgLKzs8UOg/Uy2dnZZIrLoKSkhADQ1q1bNdPWrl1LAOjkyZP08OFDKi8vpylTppCDgwM1NTVp2kVGRpKDgwNduXKFGhsbqbCwkCZOnEhKpZJu3bqlaTd//nxyd3fX2m5KSgoBoHv37mmmhYaGko+Pj1a7o0ePklKppPj4+G7vK5Fprjdvb2/y8/PTO8/Hx4du3LhBRERnzpwhKysrGjp0KNXW1hIRUW5uLs2ePVtrmbi4OLK1taXPPvuMqqqq6NKlSzR+/Hhyc3OjsrIyTTtDz8uKFSvIzs6O9u/fT5WVlbRmzRqysrKi8+fPd2u/S0tLyc/Pj1paWjptV1dXR0qlkqKiovTOj42NJQB08eLFbsVDJK3vT75TY8zM1Go1VCoV+vbti/DwcNTV1eHWrVtabWxsbDR3GH5+fsjIyEBNTQ0yMzNNEkNwcDCqq6uxfv16k6yvu+rq6nDjxg34+Pg8tW1gYCCWLVuG4uJirF69Wm+bhoYGpKamYs6cOYiIiICTkxPGjh2L7du3o6KiAjt27NBZprPz0tjYiIyMDISEhCA0NBTOzs5Yt24dZDJZt89JUlIS3n//fVhZdf71m5iYCA8PDyQkJOidP2zYMABAQUFBt+KRGk5qjPUgW1tbAEBzc3On7SZMmACFQqH16ExKysvLQURQKBQGtU9ISMCIESOwbds2nD59Wmd+YWEhamtrMWHCBK3pEydOhK2trdajXH3an5erV6+ivr4eY8aM0bSxt7dH//79u3VO7ty5gyNHjmDhwoWdtjt48CBycnLwzTff6Pwe16bt2N29e7fL8UgRJzXGLJSdnR3u3bsndhhm0djYCODXfTSEXC5HZmYmBEHAm2++iYaGBq35bd3bn3zvq42zszNqamqMiq+urg4AsG7dOq33ym7evKnTacMYycnJ+Otf/6q3Y0ybrKwsJCUlIS8vD0OHDu2wnb29PYDfjiX7FSc1xixQc3MzqqqqJFWD7UltX8jGvEQcGBiImJgYFBUVYePGjVrznJ2dAUBv8urKcezbty8AYMuWLTpd8c+ePWvUutqUlZXhiy++wLvvvtthm61bt2LPnj04deoUBgwY0On6mpqaAPx2LNmvOKkxZoHy8vJARHjhhRc002xsbJ762LK36NevHwRBMPr9s40bN2LkyJG4ePGi1vQxY8bA0dER//nPf7Smnzt3Dk1NTfjd735n1HYGDx4MuVyO/Px8o5brTHJyMiIiIuDq6qozj4iwatUqFBQU4NChQ3rvONtrO3bu7u4mi1EKOKkxZgFaW1tRWVmJx48f49KlS4iOjoanp6fWby++vr548OABDh06hObmZty7dw83b97UWZerqyvu3LmD4uJi1NTUoLm5Gbm5uRbVpV+hUMDb29vo6tltjyHbv88ll8uxfPlyHDx4EHv27EF1dTUKCgqwePFieHh4IDIy0ujtLFq0CHv37kVGRgaqq6vR0tKC0tJS/PLLLwCA8PBwuLu7GzRM1927d/GPf/wDy5Yt0zv/ypUr+OSTT7Bz507IZDKd4bQ2b96ss0zbsRs7dqxR+yZ5Iva8NClIqEsq6zmm6NK/detW6t+/PwEghUJBs2bNom3btpFCoSAANGzYMLp+/Trt2LGDVCoVAaAhQ4bQTz/9RES/dumXyWQ0cOBAsrGxIZVKRa+++ipdv35dazv379+nl156ieRyOXl5edH7779PK1euJADk6+ur6f5/4cIFGjJkCNnb29PkyZOprKyMjh07RkqlkhISErq1r21Mcb1FRUWRTCaj+vp6zbSDBw+Sj48PASA3NzdasmSJ3mVXrlyp06W/tbWVUlJSaNiwYSSTycjFxYVCQkLo6tWrmjbGnJdHjx7RqlWryNPTk2xsbKhv374UGhpKhYWFREQUEhJCACguLu6p+xoTE0MREREdzi8oKCAAHX5SUlJ0lgkODqaBAwdSa2vrU7f/NFL6/uSkxp5ppnpPrTsiIyPJ1dVV1BiMZYrrraioiGxsbOizzz4zUVQ9q6WlhaZMmUK7d+/u8W1XVFSQXC6nzZs3m2R9Uvr+5MePjFmAZ3HUdV9fX8THxyM+Pl5rJPreoKWlBYcOHUJNTQ3Cw8N7fPsbNmxAQEAAoqKienzblo6TGmNMNLGxsZg7dy7Cw8N71aDFeXl5OHDgAHJzcw1+185UUlNTkZ+fj2PHjkEmk/XotnsDTmr/9dZbb0GpVEIQBJP2eBJTY2MjRo4ciXXr1hm9bPu6Vm0fW1tb9OvXDy+++CJSUlJQWVlphsifHWvWrEFmZiYePnwILy8v7N+/X+yQetymTZsQFRWFjz/+WOxQDDZ16lR8/vnnWmNx9oTDhw/j0aNHyMvLg4uLS49uu7fgpPZfu3btws6dO8UOw6TWrl2Lq1evdmnZJ+taOTk5gYjQ2tqK8vJy5OTkwMvLC6tWrcLo0aN1ulEzwyUmJuLRo0cgIty4cQOvvfaa2CGJYtq0aUhKShI7DIs3e/ZsxMbG6vT+ZL/hpCZRZ86cweXLl026TkEQ4OzsjBdffBGZmZnIycnB3bt3NbXEGGNMbJzUniAIgtghmERDQwNWrlxp9lpLr732GhYuXIjy8nJs377drNtijDFDPLNJjYiQkpKCESNGwM7ODk5OTli5cqVOu85qKhlTm+m7777Dc889B4VCAZVKhbFjx6K6uvqp2+iKtWvX4r333tMM9dOeKWtrtb0cnJubq5nWG48ZY0wiRH6lwGRg5HsWa9euJUEQ6G9/+xtVVlZSfX09bdu2Tac+0dNqKhlSm6m2tpZUKhUlJydTQ0MDlZWV0Zw5czQ1sExZt+n06dM0a9YsIiK6d+8eAaC1a9dqtTGmtpaPjw85OTl1OL+6upoA0ODBgzXTetMxs4T31HojY683ZtmkdD4lczUbc1Lq6+tJoVDQH/7wB63pe/fu1UpqDQ0NpFAoKDw8XGtZOzs7evfdd4noty/ohoYGTZu25Hjt2jUiIrp8+TIBoKNHj+rEYsg2DFVfX08TJkyg0tJSIuo4qRnjaUmNiEgQBHJ2diai3nfMOKl1jZS+BJm0zucz+fjx2rVrqK+vx9SpUztt19WaSu1rM3l7e6Nfv36IiIjAhg0bUFxc3O1t6LNmzRq8/fbbGDhwoFHLdUddXR2ICCqVCkDvO2Zt2r+6wJ/OPwAwb9480ePgj+nOp1TYiB2AGNoGAu3oN6c2T9ZUav+ul4eHh8Hbs7e3x6lTp7B69Wps2rQJ8fHxCAsLQ2Zmpsm2cfr0aRQUFCA1NdXgZUzhp59+AgCMHDkSQO86Zk/i3+OMM2/ePERHRyMwMFDsUJgJzJs3T+wQTOaZTGptBfoePXrUabsnaypFR0d3a5ujR4/GV199hXv37iE1NRVJSUkYPXq0Zoid7m5j9+7dOHnypN4S8Zs2bcKmTZtw/vx5ncrA3XX8+HEAwPTp0wH0rmP2pLCwMJOs51kVW0QbAAAgAElEQVQxb948BAYG8nGTCCkltWfy8eOYMWNgZWWF7777rtN2pqqpdOfOHVy5cgXAr1/6H3/8McaPH48rV66YbBuZmZk6xQzbqiavXbsWRGTyhFZWVoYtW7Zg0KBBePPNNwH0rmPGGJOeZzKp9e3bF6Ghodi/fz92796N6upqXLp0CTt27NBqZ0hNJUPcuXMH77zzDn788Uc0NTXh4sWLuHnzJl544QWTbcMYxtbWIiLU1taitbVVkyyzs7MxadIkWFtb49ChQ5rf1KR6zBhjvYR4fVRMC0b23qmpqaG33nqL+vTpQ46OjjR58mSKi4sjADRo0CD64YcfiKjzmkqG1mYqLi4mtVpNLi4uZG1tTQMGDKC1a9fS48ePn7qN7uio96MhtbWOHDlC48aNI4VCQba2tmRlZUUAND0dn3vuOYqPj6f79+/rLNubjhn3fuwaY683ZtmkdD4FIiKR8qlJCYKA7OxsfsbPjJKTk4N58+ZBIpdBj+HrTVqkdD6fycePjDHGpImTmgX78ccfDXrHRIwihYx11YkTJxAbG6tT3mjBggU6badNmwalUglra2uMHj0aFy5cECFi47W2tmLLli1Qq9V658fHx8PPzw8qlQp2dnbw9fXFBx98oLdY6hdffIGJEydCqVRiyJAhWLRoEcrKyjTzjxw5guTk5Gey0KxeIj/+NBlI6Jkw6zn8m1rXdPV6i4uLo5kzZ1J1dbVmmo+PD/Xp06fDEWRyc3Np9uzZ3Yq3J/300080adIkAkD+/v562wQFBdG2bdvo/v37VF1dTdnZ2SSTyeiPf/yjVrusrCwCQMnJyVRVVUUXL14kb29vCggIoObmZk27tLQ0CgoKosrKyi7FLKXvT75TY0xkDQ0NHf5F35u28TRJSUnIyspCTk4OlEql1rz09HRYWVkhMjKyV5cx+uGHH7B69WosXrwYAQEBHbZzdHREZGQkXF1doVQqERYWhpCQEBw/fhwlJSWadp9++ikGDBiAlStXwsnJCQEBAYiJiUF+fj7OnTunabd06VL4+/tjxowZePz4sVn30dJxUmNMZLt370Z5eXmv30Znrl27hvXr1+Ojjz7SDH7wJLVajejoaNy+fRsrVqwQIULT8Pf3x4EDBzB//nzY2dl12O7o0aM6hT7d3NwAAPX19ZppJSUl8PDw0BrKavDgwQCAmzdvai2/YcMG5Ofnm73klKXjpMaYkYgIqampGDVqFOzs7ODi4oJXX31Va9zJqKgo2Nraon///ppp7733HhwcHCAIAioqKgAA0dHRWL58Oa5fvw5BEODr64v09HTI5XL069cP77zzDjw8PCCXy6FWq7X+Ou/ONgDTliB6mvT0dBARZs2a1WGbhIQEDB8+HLt27cKJEyc6XZ8h58CYMkeWUMro9u3bsLe3h5eXl2aat7e3zh8jbb+neXt7a013cXFBUFAQ0tLSnu3evOI+/TQdSOiZMOs5XflNLS4ujmxtbemzzz6jqqoqunTpEo0fP57c3NyorKxM027+/Pnk7u6utWxKSgoB0JTQISIKDQ0lHx8frXaRkZHk4OBAV65cocbGRiosLKSJEyeSUqmkW7dumWQbxpQgas/Y683b25v8/Pz0zvPx8aEbN24QEdGZM2fIysqKhg4dSrW1tUSk/zc1Q8+BIWWOiExb/qnN888/3+Fvau3V1dWRUqmkqKgorel5eXkkk8koPT2dqqur6fLlyzRq1Ch65ZVX9K4nNjZWp3yWIaT0/cl3aowZoaGhAampqZgzZw4iIiLg5OSEsWPHYvv27aioqNAZlaY7bGxsNHcifn5+yMjIQE1NDTIzM02y/uDgYFRXV2P9+vUmWV9H6urqcOPGDfj4+Dy1bWBgIJYtW4bi4mKsXr1ab5uunAO1Wg2VSoW+ffsiPDwcdXV1uHXrFgCgsbERGRkZCAkJQWhoKJydnbFu3TrIZDKTHeunSUxMhIeHBxISErSmBwUFYdWqVYiKioJKpcKYMWNQU1ODXbt26V3PsGHDAAAFBQVmj9lScVJjzAiFhYWora3VGUdz4sSJsLW11Xo8aGoTJkyAQqHocnkdsZSXl4OIoFAoDGqfkJCAESNGYNu2bTh9+rTO/O6eg/ZljsxRysgYBw8eRE5ODr755hudDjRr167Fjh07cPLkSdTW1uLnn3+GWq1GYGCgVoeSNm3H+O7du2aP21JxUmPMCFVVVQB+7b3WnrOzM2pqasy6fTs7O81A1b1FY2MjAHTaceJJcrkcmZmZEAQBb775JhoaGrTmm/ocPFnK6Mn3P2/evKnVacMcsrKykJSUhLy8PAwdOlRr3i+//ILk5GS8/fbbePnll+Hg4AAvLy/s3LkTd+7cQUpKis767O3tAfx2zJ9FnNQYM4KzszMA6P3irKqqwqBBg8y27ebmZrNvwxzavmiNeTk4MDAQMTExKCoqwsaNG7XmmfocPFkuidpVujh79qxR6zLG1q1bsWfPHpw6dQoDBgzQmV9UVISWlhadeSqVCq6urigsLNRZpqmpCcBvx/xZxEmNMSOMGTMGjo6O+M9//qM1/dy5c2hqasLvfvc7zTQbGxvNIy5TyMvLAxHhhRdeMNs2zKFfv34QBMHo9882btyIkSNH4uLFi1rTjTkHhujpUkZEhFWrVqGgoACHDh3Se8cJQJOc21eeqKmpwYMHDzRd+5/Udozd3d1NHHXvwUmNMSPI5XIsX74cBw8exJ49e1BdXY2CggIsXrwYHh4eiIyM1LT19fXFgwcPcOjQITQ3N+PevXs67xYBgKurK+7cuYPi4mLU1NRoklRraysqKyvx+PFjXLp0CdHR0fD09MTChQtNsg1jSxB1lUKhgLe3t6bivKHaHkO2f5/LmHNg6HaeVsooPDwc7u7uJhmm68qVK/jkk0+wc+dOyGQynWHvNm/eDADw8vLCSy+9hJ07d+L7779HQ0MDSkpKNPv3l7/8RWfdbcd47Nix3Y6z1xKx56VJQUJdUlnP6UqX/tbWVkpJSaFhw4aRTCYjFxcXCgkJoatXr2q1u3//Pr300kskl8vJy8uL3n//fVq5ciUBIF9fX03X/AsXLtCQIUPI3t6eJk+eTGVlZRQZGUkymYwGDhxINjY2pFKp6NVXX6Xr16+bbBuGlCDqiLHXW1RUFMlkMqqvr9dMO3jwIPn4+BAAcnNzoyVLluhdduXKlTpd+g05B4aWOSJ6eimjkJAQAkBxcXGd7ufZs2dp0qRJ5OHhQQAIAPXv35/UajV99913RERUUFCgmafvk5KSollfRUUFRUdHk6+vL9nZ2ZGjoyNNmjSJvvzyS73bDw4OpoEDB1Jra2uncbYnpe9PTmrsmWapYz9GRkaSq6ur2GF0yNjrraioiGxsbOizzz4zY1Tm09LSQlOmTKHdu3eLHUqHKioqSC6X0+bNm41eVkrfn/z4kTELJaVR1319fREfH4/4+Hi9I9FbspaWFhw6dAg1NTUWXRFjw4YNCAgIQFRUlNihiIqTGmOsR8TGxmLu3LkIDw/vVYMW5+Xl4cCBA8jNzTX4Xbuelpqaivz8fBw7dgwymUzscETFSY0xC7NmzRpkZmbi4cOH8PLywv79+8UOyWQ2bdqEqKgofPzxx2KHYrCpU6fi888/1xpj05IcPnwYjx49Ql5eHlxcXMQOR3Q2YgfAGNOWmJiIxMREscMwm2nTpmHatGlihyEZs2fPxuzZs8UOw2LwnRpjjDHJ4KTGGGNMMjipMcYYkwxOaowxxiRDUh1FzDn4KJOmtn8zOTk5IkfS+/D1xiyRQCSNut+CIIgdAmOM9VrZ2dkICwsTO4xuk8ydmkRyM5MoQRAk86XBmCXj39QYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJhkBEJHYQjElJZGQkrl69qjXtwoUL8PLygouLi2aatbU1/vnPf2LQoEE9HSJjkmUjdgCMSY27uzt27NihM/3SpUta/+/t7c0JjTET48ePjJnYn/70p6e2sbW1xcKFC80fDGPPGH78yJgZjBkzBleuXEFnl9fVq1cxfPjwHoyKMenjOzXGzOCNN96AtbW13nmCIMDf358TGmNmwEmNMTN4/fXX0dLSoneetbU1/vznP/dwRIw9G/jxI2Nmolarce7cObS2tmpNFwQBJSUlGDhwoEiRMSZdfKfGmJksWLAAgiBoTbOyssLkyZM5oTFmJpzUGDOTuXPn6kwTBAFvvPGGCNEw9mzgpMaYmbi5uWHq1KlaHUYEQUBISIiIUTEmbZzUGDOjiIgITbd+a2trvPLKK+jTp4/IUTEmXZzUGDOjOXPmwNbWFgBARIiIiBA5IsakjZMaY2bk4OCA//3f/wXw6ygiM2fOFDkixqSNkxpjZjZ//nwAQEhICBwcHESOhjFpE/09tfZdnhljjPVe2dnZCAsLE237FjFKf3R0NAIDA8UOg7EuO3v2LNLS0pCdna13/p49exAeHg4bG4u45CzGvHnz+PqXkHnz5okdgmXcqYmd2RnrrpycHMybN6/DAYwbGxshl8t7OCrLx9e/tFjC+eTf1BjrAZzQGOsZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYsyDHjh2Dk5MTvvrqK7FDsXgnTpxAbGwsDhw4AG9vbwiCAEEQsGDBAp2206ZNg1KphLW1NUaPHo0LFy6IELHxWltbsWXLFqjVar3z4+Pj4efnB5VKBTs7O/j6+uKDDz5AbW2tTtsvvvgCEydOhFKpxJAhQ7Bo0SKUlZVp5h85cgTJyckdFrftLTipMWZBuGavYT788EOkp6djzZo1CA0Nxc8//wwfHx/06dMHe/bswddff63V/ttvv8W+ffswc+ZMFBYWYvz48SJFbriioiL8/ve/R0xMDOrr6/W2OXXqFJYsWYLi4mJUVFQgMTERaWlpOmWPsrOzMX/+fMydOxelpaU4fPgwvv/+e0yfPh2PHz8GAMyaNQtyuRxTp05FVVWV2ffPXDipMWZBgoOD8fDhQ4sYI7KhoaHDOwQxJSUlISsrCzk5OVAqlVrz0tPTYWVlhcjISDx8+FCkCLvvhx9+wOrVq7F48WIEBAR02M7R0RGRkZFwdXWFUqlEWFgYQkJCcPz4cZSUlGjaffrppxgwYABWrlwJJycnBAQEICYmBvn5+Th37pym3dKlS+Hv748ZM2Zokl1vw0mNMabX7t27UV5eLnYYWq5du4b169fjo48+0vvun1qtRnR0NG7fvo0VK1aIEKFp+Pv748CBA5g/fz7s7Ow6bHf06FGten3Ar3X8AGjd3ZWUlMDDw0NrWMLBgwcDAG7evKm1/IYNG5Cfn4+0tLRu74cYOKkxZiFOnz4NT09PCIKAv//97wCAjIwMODg4QKFQ4PDhw5g+fTpUKhUGDRqEvXv3apZNT0+HXC5Hv3798M4778DDwwNyuRxqtVrrL/GoqCjY2tqif//+mmnvvfceHBwcIAgCKioqAPw6dN3y5ctx/fp1CIIAX19fAMDx48ehUqmwadOmnjgkOtLT00FEmDVrVodtEhISMHz4cOzatQsnTpzodH1EhNTUVIwaNQp2dnZwcXHBq6++ih9//FHTxtBzAAAtLS2Ii4uDp6cn7O3tMW7cuA6HTjOX27dvw97eHl5eXppp3t7eOn+gtP2e5u3trTXdxcUFQUFBSEtL652Pw0lkACg7O1vsMBjrluzsbDLF5VRSUkIAaOvWrZppa9euJQB08uRJevjwIZWXl9OUKVPIwcGBmpqaNO0iIyPJwcGBrly5Qo2NjVRYWEgTJ04kpVJJt27d0rSbP38+ubu7a203JSWFANC9e/c000JDQ8nHx0er3dGjR0mpVFJ8fHy395XI+Ovf29ub/Pz89M7z8fGhGzduEBHRmTNnyMrKioYOHUq1tbVERJSbm0uzZ8/WWiYuLo5sbW3ps88+o6qqKrp06RKNHz+e3NzcqKysTNPO0HOwYsUKsrOzo/3791NlZSWtWbOGrKys6Pz58wbvY3vPP/88+fv7G9S2rq6OlEolRUVFaU3Py8sjmUxG6enpVF1dTZcvX6ZRo0bRK6+8onc9sbGxBIAuXrxoVKyW8H3Od2qM9RJqtRoqlQp9+/ZFeHg46urqcOvWLa02NjY2mrsOPz8/ZGRkoKamBpmZmSaJITg4GNXV1Vi/fr1J1meMuro63LhxAz4+Pk9tGxgYiGXLlqG4uBirV6/W26ahoQGpqamYM2cOIiIi4OTkhLFjx2L79u2oqKjAjh07dJbp7Bw0NjYiIyMDISEhCA0NhbOzM9atWweZTGay4/80iYmJ8PDwQEJCgtb0oKAgrFq1ClFRUVCpVBgzZgxqamqwa9cuvesZNmwYAKCgoMDsMZsaJzXGeqG2atrNzc2dtpswYQIUCoXW47Teqry8HEQEhUJhUPuEhASMGDEC27Ztw+nTp3XmFxYWora2FhMmTNCaPnHiRNja2mo9ttWn/Tm4evUq6uvrMWbMGE0be3t79O/fv0eO/8GDB5GTk4NvvvlGpwPN2rVrsWPHDpw8eRK1tbX4+eefoVarERgYqNWhpE3bMb57967Z4zY1TmqMSZydnR3u3bsndhjd1tjYCACddpx4klwuR2ZmJgRBwJtvvomGhgat+W3d1h0dHXWWdXZ2Rk1NjVHx1dXVAQDWrVuneWdOEATcvHmzwy75ppKVlYWkpCTk5eVh6NChWvN++eUXJCcn4+2338bLL78MBwcHeHl5YefOnbhz5w5SUlJ01mdvbw/gt2Pem3BSY0zCmpubUVVVhUGDBokdSre1fdEa83JwYGAgYmJiUFRUhI0bN2rNc3Z2BgC9yasrx6xv374AgC1btoCItD5nz541al3G2Lp1K/bs2YNTp05hwIABOvOLiorQ0tKiM0+lUsHV1RWFhYU6yzQ1NQH47Zj3JpzUGJOwvLw8EBFeeOEFzTQbG5unPra0RP369YMgCEa/f7Zx40aMHDkSFy9e1Jo+ZswYODo64j//+Y/W9HPnzqGpqQm/+93vjNrO4MGDIZfLkZ+fb9RyXUVEWLVqFQoKCnDo0CG9d5wANMn5l19+0ZpeU1ODBw8eaLr2P6ntGLu7u5s4avPjpMaYhLS2tqKyshKPHz/GpUuXEB0dDU9PTyxcuFDTxtfXFw8ePMChQ4fQ3NyMe/fu6byrBACurq64c+cOiouLUVNTg+bmZuTm5orWpV+hUMDb2xulpaVGLdf2GLL9+1xyuRzLly/HwYMHsWfPHlRXV6OgoACLFy+Gh4cHIiMjjd7OokWLsHfvXmRkZKC6uhotLS0oLS3VJJTw8HC4u7ubZJiuK1eu4JNPPsHOnTshk8m0HnkKgoDNmzcDALy8vPDSSy9h586d+P7779HQ0ICSkhLN/v3lL3/RWXfbMR47dmy34+xxIva8JCLL6ALKWHeZokv/1q1bqX///gSAFAoFzZo1i7Zt20YKhYIA0LBhw+j69eu0Y8cOUqlUBICGDBlCP/30ExH92qVfJpPRwIEDycbGhlQqFb366qt0/fp1re3cv3+fXnrpJZLL5eTl5UXvv/8+rVy5kgCQr6+vpvv/hQsXaMiQIWRvb0+TJ0+msrIyOnbsGCmVSkpISOjWvrYx9vqPiooimUxG9fX1mmkHDx4kHx8fAkBubm60ZMkSvcuuXLlSp0t/a2srpaSk0LBhw0gmk5GLiwuFhITQ1atXNW2MOQePHj2iVatWkaenJ9nY2FDfvn0pNDSUCgsLiYgoJCSEAFBcXFyn+3n27FmaNGkSeXh4EAACQP379ye1Wk3fffcdEREVFBRo5un7pKSkaNZXUVFB0dHR5OvrS3Z2duTo6EiTJk2iL7/8Uu/2g4ODaeDAgdTa2tppnO1Zwvc5JzXGTMBU76l1R2RkJLm6uooag7GMvf6LiorIxsaGPvvsMzNGZT4tLS00ZcoU2r17t9ihdKiiooLkcjlt3rzZ6GUt4fucHz8yJiG9fYT1p/H19UV8fDzi4+P1jkRvyVpaWnDo0CHU1NQgPDxc7HA6tGHDBgQEBCAqKkrsULqk1ye1t956C0qlEoIg9NgPtJbKmDIUT9O+nEfbx9bWFv369cOLL76IlJQUVFZWmmFPGOtYbGws5s6di/Dw8F41aHFeXh4OHDiA3Nxcg9+162mpqanIz8/HsWPHIJPJxA6nS3p9Utu1axd27twpdhgWwdAyFIZ4spyHk5MTiAitra0oLy9HTk4OvLy8sGrVKowePVqn9xjreWvWrEFmZiYePnwILy8v7N+/X+yQzGrTpk2IiorCxx9/LHYoBps6dSo+//xzrXE3Lcnhw4fx6NEj5OXlwcXFRexwuk7Uh59kmmewe/fu7dI4ZZaovr6eAgMDu7RscHAwPX78WGtaWFgYAdAa+88YPj4+5OTkpHfevn37yMrKivr160dVVVVdWr8l6c6xt4Tf1HojU1z/zHJYwvns9XdqALTKKfR23Sn3YWgZClN57bXXsHDhQpSXl2P79u0mX39Ps8RSK4wx4/S6pEZESElJwYgRI2BnZwcnJyesXLlSq80nn3wChUIBpVKJ8vJyLF++HAMHDsTVq1cNKjVhaBmPtnietr7ulvvoDn1lKExZPqTt/afc3FwAfOwZYyIT9T6RjL9dXbt2LQmCQH/729+osrKS6uvradu2bTqPH9tKRSxdupS2bt1Kc+bMof/3//6fwaUmDC3jYej6ulPuo6s6KkNhTPmQzh4/EhFVV1cTABo8eLBm2rN47PnxY9cYe/0zy2YJ51P0q9CYg1BfX08KhYL+8Ic/aE3X95ta2xdrQ0OD1vKOjo4UHh6utfy///1vAqD1JR8ZGanzZX7+/HkCQB999JHR6xMjqa1du5aGDx9O1dXVXV7H05IaEZEgCOTs7Ky13Wft2HNS6xpL+BJkpmMJ59Omx24JTeDatWuor6/H1KlTu7R8d0tNtC/j0d31mVNbGYpvv/1WpwyFKdXV1YGIoFKpOm33rBz7nJwcUbbbm5lzsF/27OlVSa1tPLK20bCNZYpSE0+W8TB16QpTycrKQmpqKvLy8vSO2m1KP/30EwBg5MiRnbZ7Vo79vHnzRNlub5aWloa0tDSxw2AS0auSmlwuBwA8evSoS8t3t9RE+zIepi5dYQpbt27FN998g1OnTnU4arcpHT9+HAAwffr0Tts9C8ce+LXzCjOcIAjIzs5GWFiY2KEwE7CEnui9qvfjmDFjYGVlhe+++67Ly3en1ET7Mh7GrM/c5T7IwDIUplRWVoYtW7Zg0KBBePPNNzttK+VjzxizHL0qqfXt2xehoaHYv38/du/ejerqaly6dAk7duwwaHljS008rYyHMevrTrkPQxhahgKA0eVDiAi1tbVobW0FEeHevXvIzs7GpEmTYG1tjUOHDj31NzUpH3vGmAURsZMKERnfW6ampobeeust6tOnDzk6OtLkyZMpLi6OANCgQYPohx9+oOTkZLK3t9d0NX9yRG9DSk0QGV7Gw9D1dafchyGMKUNhSPmQI0eO0Lhx40ihUJCtrS1ZWVkRAE1Px+eee47i4+Pp/v37Wss9i8eeiHs/dpWx1z+zbJZwPoX/BiIaS32m/s4772Dfvn24f/++2KE8c3rjsc/JycG8efP4NzUjWer1z7rGEs5nr3r82NOkXsbDkvGxZ4x1BSc1C/bjjz/q/Dam72PJtZkYY6wncVLTw1LKeIwcORL066gvnX6ysrJEic8cLOXYM8t34sQJxMbG6tT+W7BggU7badOmQalUwtraGqNHj8aFCxdEiNh4ra2t2LJlC9Rqtd75xtRQ/OKLLzBx4kQolUoMGTIEixYtQllZmWb+kSNHkJyc3PufkojzU95vYAE/LDLWXdxRpGu6ev3HxcXRzJkztYaA8/HxoT59+hAAOnr0qM4yubm5NHv27G7F25N++uknmjRpEgEgf39/vW2CgoJo27ZtdP/+faqurqbs7GySyWT0xz/+UatdVlYWAaDk5GSqqqqiixcvkre3NwUEBFBzc7OmXVpaGgUFBVFlZWWXYraE73O+U2NMIhoaGjr8i743beNpkpKSkJWVhZycHJ0h4NLT02FlZYXIyMheVRW7vR9++AGrV6/G4sWLERAQ0GE7R0dHREZGwtXVFUqlEmFhYQgJCcHx48dRUlKiaffpp59iwIABWLlyJZycnBAQEICYmBjk5+drDSm3dOlS+Pv7Y8aMGXj8+LFZ99FcOKkxJhE9UQ9O7Jpz165dw/r16/HRRx9pRhh6klqtRnR0NG7fvo0VK1aIEKFp+Pv748CBA5g/fz7s7Ow6bGdoDcWSkhJ4eHhojfgxePBgANB5Z3PDhg3Iz8/vtUOXcVJjTCRk5npwhtam627NOVPW53ua9PR0EBFmzZrVYZuEhAQMHz4cu3btwokTJzpdnyHnICMjAw4ODlAoFDh8+DCmT58OlUqFQYMGYe/evVrra2lpQVxcHDw9PWFvb49x48YhOzu7ezttJH01FL29vXX+GGn7Pc3b21truouLC4KCgpCWltY7X1ER9eEnWcYzWMa6qyu/qfVEPThDa9N1ZxvG1Odrz9jr39vbm/z8/PTO8/HxoRs3bhAR0ZkzZ8jKyoqGDh1KtbW1RKT/NzVDz0FbOaWTJ0/Sw4cPqby8nKZMmUIODg7U1NSkabdixQqys7Oj/fv3U2VlJa1Zs4asrKzo/PnzBu9je88//3yHv6m111ENxby8PJLJZJ3CDw4AACAASURBVJSenk7V1dV0+fJlGjVqFL3yyit61xMbG6tTzssQlvB9zndqjImgoaEBqampmDNnDiIiIuDk5ISxY8di+/btqKioMHjoN0PY2Nho7kT8/PyQkZGBmpoaZGZmmmT9wcHBqK6uxvr1602yvo7U1dXhxo0b8PHxeWrbwMBALFu2DMXFxVi9erXeNl05B2q1GiqVCn379kV4eDjq6upw69YtAEBjYyMyMjIQEhKC0NBQODs7Y926dZDJZCY71k+TmJgIDw8PJCQkaE0PCgrCqlWrEBUVBZVKhTFjxqCmpga7du3Su55hw4YBAAoKCswes6lxUmNMBGLWg2tfm663KC8vBxFBoVAY1D4hIQEjRozAtm3bcPr0aZ353T0Htra2AKAZI/Tq1auor6/HmDFjNG3s7e3Rv3//HjnWbTUUv/nmG50ONGvXrsWOHTtw8uRJ1NbW4ueff4ZarUZgYKBWh5I2bcf47t27Zo/b1DipMSYCsevBPVmbrrdobGwEgE47TjxJLpcjMzMTgiDgzTffRENDg9Z8U5+Duro6AMC6deu0Bke4efOmVqcNc8jKykJSUhLy8vIwdOhQrXm//PILkpOT8fbbb+Pll1+Gg4MDvLy8sHPnTty5cwcpKSk667O3twfw2zHvTTipMSYCMevBta9N11u0fdEa83JwYGAgYmJiUFRUhI0bN2rNM/U5aCtevGXLFp0BEsxZ3Xvr1q3Ys2cPTp06pbcocFFREVpaWnTmqVQquLq6orCwUGeZpqYmAL8d896EkxpjIhCzHlz72nTm2IY59OvXD4IgGP3+2caNGzFy5EhcvHhRa3p3a/y1N3jwYMjlcuTn5xu1XFeRgTUU25LzL7/8ojW9pqYGDx480HTtf1LbMXZ3dzdx1ObHSY0xEfRkPbin1abr7jaMrc/XVQqFAt7e3igtLTVqubbHkO3f5zK2xp8h21m0aBH27t2LjIwMVFdXo6WlBaWlpZqEEh4eDnd3d5MM02VoDUUvLy+89NJL2LlzJ77//ns0NDSgpKREs39/+ctfdNbddozHjh3b7Th7nIg9L4nIMrqAMtZdXenS3xP14AytTdedbRhSn68jxl7/UVFRJJPJqL6+XjPt4MGD5OPjQwDIzc2NlixZonfZlStX6nTpN+QcbNu2jRQKBQGgYcOG0fXr12nHjh2kUqkIAA0ZMoR++uknIiJ69OgRrVq1ijw9PcnGxob69u1LoaGhVFhYSEREISEhBIDi4uI63c+zZ8/SpEmTyMPDQ1MTsX///qRWq+m7774jIuNqKFZUVFB0dDT5+vqSnZ0dOTo60qRJk+jLL7/Uu/3g4GAaOHAgtba2dhpne5bwfc5JjTETsNSxHyMjI8nV1VXsMDpk7PVfVFRENjY2WsVne5OWlhaaMmUK7d69W+xQOlRRUUFyuZw2b95s9LKW8H3Ojx8Zk7heP+r6E3x9fREfH4/4+Hi9I9FbspaWFhw6dAg1NTUWXS5qw4YNCAgIQFRUlNihdAknNcZYrxIbG4u5c+ciPDy8Vw1anJeXhwMHDiA3N9fgd+16WmpqKvLz83Hs2DHIZDKxw+kSTmqMSZSUa9Nt2rQJUVFR+Pjjj8UOxWBTp07F559/rjXGpiU5fPgwHj16hLy8PLi4uIgdTpfZiB0AY8w8EhMTkZiYKHYYZjNt2jRMmzZN7DAkY/bs2Zg9e7bYYXQb36kxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyLKKjyJYtW7Bv3z6xw2Csy9qGFZo7d67IkfQ+fP0zUxL++xa4aPhLgD0LcnNz8T//8z8W252bMVOJiYlBYGCgaNsXPakx9iwQBAHZ2dkICwsTOxTGJI1/U2OMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGTYiB0AY1JTVVUFItKZXldXh8rKSq1pjo6OkMlkPRUaY5InkL6rjzHWZS+//DL+9a9/PbWdtbU1bt++DXd39x6IirFnAz9+ZMzEXn/9dQiC0GkbKysr/P73v+eExpiJcVJjzMRee+012Nh0/mRfEAS88cYbPRQRY88OTmqMmZiLiwumTZsGa2vrDttYWVkhJCSkB6Ni7NnASY0xM4iIiEBra6veeTY2NggODoaTk1MPR8WY9HFSY8wMZs2aBTs7O73zWlpaEBER0cMRMfZs4KTGmBkoFAqEhITo7a5vb2+PGTNmiBAVY9LHSY0xM/nTn/6E5uZmrWkymQyvvfYa7O3tRYqKMWnjpMaYmbzyyis6v5s1NzfjT3/6k0gRMSZ9nNQYMxOZTIbw8HDY2tpqpjk7O2Pq1KkiRsWYtHFSY8yMXn/9dTQ1NQH4NclFREQ89R02xljX8TBZjJlRa2srBgwYgLt37wIATp8+jUmTJokcFWPSxXdqjJmRlZUVFixYAADw8PCAWq0WOSLGpE0yz0FycnLEDoExvdzc3AAAzz//PPbt2ydyNIzpp1arMWjQILHD6DbJPH582gCyjDHGOpadnY2wsDCxw+g2ST1+zM7OBhHxhz8Gf7KzswHA7NvZt2+f6Ptqyg9fb9L6SImkkhpjluq1114TOwTGngmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGTODYsWNwcnLCV199JXYoFu/EiROIjY3FgQMH4O3tDUEQIAiCZuSVJ02bNg1KpRLW1tYYPXo0Lly4IELExmttbcWWLVs6HEEmPj4efn5+UKlUsLOzg6+vLz744APU1tbqtP3iiy8wceJEKJVKDBkyBIsWLUJZWZlm/pEjR5CcnIyWlhaz7U9vwkmNMROQ2rs+5vLhhx8iPT0da9asQWhoKH7++Wf4+PigT58+2LNnD77++mut9t9++y327duHmTNnorCwEOPHjxcpcsMVFRXh97//PWJiYlBfX6+3zalTp7BkyRIUFxejoqICiYmJSEtLw9y5c7XaZWdnY/78+Zg7dy5KS0tx+PBhfP/995g+fToeP34M4Ncq63K5HFOnTkVVVZXZ98/ScVJjzASCg4Px8OFDzJw5U+xQ0NDQYJFjTCYlJSErKws5OTlQKpVa89LT02FlZYXIyEg8fPhQpAi774cffsDq1auxePFiBAQEdNjO0dERkZGRcHV1hVKpRFhYGEJCQnD8+HGUlJRo2n366acYMGAAVq5cCScnJwQEBCAmJgb5+fk4d+6cpt3SpUvh7++PGTNmaJLds4qTGmMSs3v3bpSXl4sdhpZr165h/fr1+OijjyCXy3Xmq9VqREdH4/bt21ixYoUIEZqGv78/Dhw4gPnz58POzq7DdkePHoW1tbXWtLYxQp+8uyspKYGHh4fWMICDBw8GANy8eVNr+Q0bNiA/Px9paWnd3o/ejJMaY910+vRpeHp6QhAE/P3vfwcAZGRkwMHBAQqFAocPH8b06dOhUqkwaNAg7N27V7Nseno65HI5+vXrh3feeQceHh6Qy+VQq9Vaf4lHRUXB1tYW/fv310x777334ODgAEEQUFFRAQCIjo7G8uXLcf36dQiCAF9fXwDA8ePHoVKpsGnTpp44JDrS09NBRJg1a1aHbRISEjB8+HDs2rULJ06c6HR9RITU1FSMGjUKdnZ2cHFxwauvvooff/xR08bQcwAALS0tiIuLg6enJ+zt7TFu3DjNEGo95fbt27C3t4eXl5dmmre3t84fKG2/p3l7e2tNd3FxQVBQENLS0p7tx+EkEQAoOztb7DBYL5OdnU2muAxKSkoIAG3dulUzbe3atQSATp48SQ8fPqTy8nKaMmUKOTg4UFNTk6ZdZGQkOTg40JUrV6ixsZEKCwtp4sSJpFQq6datW5p28+fPJ3d3d63tpqSkEAC6d++eZlpoaCj5+PhotTt69CgplUqKj4/v9r4SGX+9eXt7k5+fn955Pj4+dOPGDSIiOnPmDFlZWdHQoUOptraWiIhyc3Np9uzZWsvExcWRra0tffbZZ1RVVUWXLl2i8ePHk5ubG5WVlWnaGXoOVqxYQXZ2drR//36qrKykNWvWkJWVFZ0/f97gfWzv+eefJ39/f4Pa1tXVkVKppKioKK3peXl5JJPJKD09naqrq+ny5cs0atQoeuWVV/SuJzY2lgDQxYsXjYpVSt+ffKfGmJmp1WqoVCr07dsX4eHhqKurw61bt7Ta2NjYaO46/Pz8kJGRgZqaGmRmZpokhuDgYFRXV2P9+vUmWZ8x6urqcOPGDfj4+Dy1bWBgIJYtW4bi4mKsXr1ab5uGhgakpqZizpw5iIiIgJOTE8aOHYvt27ejoqICO3bs0Fmms3PQ2NiIjIwM/H/27jwqijPdH/i3oDcaaBZFRBGEBiWCCTHqCNGjGU/MJE5EJApOdEK85ke8MYS4hODCKCBqcJSLgeS6HM6MGhaXi8YRk2u85J5MjDFXjIiJCxFBjSIEZFW25/dHho5to3ZDQzXF8zmHP/LWW1VPV6X6sarfep/Q0FCEhYXB0dERq1atglwuN9vxf5zk5GS4ubkhKSlJr33y5MmIjY1FdHQ0NBoNAgICUFdXhx07dnS6HV9fXwBAUVFRj8dsqTipMdaLFAoFAKClpeWR/caOHQu1Wq33OK2vqqioABFBrVYb1T8pKQkjR45Eeno6vvrqK4PlxcXFqK+vx9ixY/Xax40bB4VCoffYtjMPnoMLFy6gsbERAQEBuj42NjYYPHhwrxz/AwcOIDc3F5999pnBAJqVK1di27Zt+OKLL1BfX4+ffvoJwcHBCAoK0htQ0qHjGHdUWu+POKkxZqGUSiVu374tdhjddvfuXQB45MCJ+6lUKmRmZkIQBCxYsABNTU16yzuGrdvZ2Rms6+joiLq6OpPia2hoAACsWrVK986cIAi4evXqQ4fkm0t2djY2bNiAgoICDB8+XG/Zzz//jI0bN+L//b//h9///vewtbWFl5cXtm/fjhs3biAlJcVgezY2NgB+O+b9ESc1xixQS0sLampqJFGJuOOL1pSXg4OCgrBkyRJcunQJiYmJesscHR0BoNPk1ZVj5uLiAgDYsmWLQZ2xEydOmLQtU2zduhW7d+/G8ePHMWTIEIPlly5dQltbm8EyjUYDZ2dnFBcXG6zT3NwM4Ldj3h9xUmPMAhUUFICIMGHCBF2bTCZ77GNLSzRo0CAIgmDy+2eJiYnw8/NDYWGhXntAQADs7Ozw3Xff6bWfPHkSzc3NeOaZZ0zaz7Bhw6BSqXDmzBmT1usqIkJsbCyKioqQl5fX6R0nAF1y/vnnn/Xa6+rq8Msvv+iG9t+v4xi7urqaOeq+g5MaYxagvb0d1dXVaG1txdmzZxETEwMPDw9ERkbq+vj4+OCXX35BXl4eWlpacPv2bYN3lQDA2dkZN27cQGlpKerq6tDS0oL8/HzRhvSr1Wp4e3vj2rVrJq3X8Rjywfe5VCoVli5digMHDmD37t2ora1FUVERFi1aBDc3N0RFRZm8n9dffx1ZWVnIyMhAbW0t2tracO3aNV1CiYiIgKurq1mm6Tp//jw++OADbN++HXK5XO+RpyAI2LRpEwDAy8sLzz33HLZv347//d//RVNTE8rLy3Wf79/+7d8Mtt1xjEePHt3tOPssEUdemhUkNCSV9R5zDOnfunUrDR48mACQWq2mGTNmUHp6OqnVagJAvr6+VFJSQtu2bSONRkMAyNPTky5evEhEvw7pl8vlNHToUJLJZKTRaGjmzJlUUlKit5+qqip67rnnSKVSkZeXF7399tu0fPlyAkA+Pj664f+nT58mT09PsrGxoYkTJ9LNmzfpyJEjZG9vT0lJSd36rB1Mvd6io6NJLpdTY2Ojru3AgQOk1WoJAA0cOJAWL17c6brLly83GNLf3t5OKSkp5OvrS3K5nJycnCg0NJQuXLig62PKObh37x7FxsaSh4cHyWQycnFxobCwMCouLiYiotDQUAJA8fHxj/ycJ06coGeffZbc3NwIAAGgwYMHU3BwMH355ZdERFRUVKRb1tlfSkqKbnuVlZUUExNDPj4+pFQqyc7Ojp599ln6r//6r073P336dBo6dCi1t7c/Ms4HSen7k5Ma69fM9Z5ad0RFRZGzs7OoMZjK1Ovt0qVLJJPJaNeuXT0YVc9pa2ujSZMm0c6dO8UO5aEqKytJpVLRpk2bTF5XSt+f/PiRMQsg9RnWfXx8kJCQgISEhE5nordkbW1tyMvLQ11dHSIiIsQO56HWrFmDwMBAREdHix2KqDip/cvChQthb28PQRB67Qdjc0tKSjJ4Pi8Igt77N8Z6sCxIx59CocCgQYMwZcoUpKSkoLq6ugc+CZOiuLg4zJ49GxEREX1q0uKCggLs378f+fn5Rr9r19s2b96MM2fO4MiRI5DL5WKHIypOav+yY8cObN++XewwLMb9ZUEcHBxARGhvb0dFRQVyc3Ph5eWF2NhY+Pv7G4xCY8ZbsWIFMjMzcefOHXh5eWHfvn1ih9Sj1q1bh+joaKxfv17sUIw2depU7NmzR2/eTUty8OBB3Lt3DwUFBXBychI7HNFxUpOYXbt2Gbxrc+7cObNsWxAEODo6YsqUKcjMzERubi5u3bqlK7vCTJecnIx79+6BiHDlyhW88sorYofU46ZNm4YNGzaIHYZkhISEIC4uzmCUaH/FSe0+95d3YI/3yiuvIDIyEhUVFfj444/FDocxxvpvUiMipKSkYOTIkVAqlXBwcMDy5csN+j2qJIUppS2+/PJLjB8/Hmq1GhqNBqNHj0Ztbe1j99ETzFmGpOM9qvz8fF2bFI8ZY6yPEHHkpVnBxCGpK1euJEEQ6K9//StVV1dTY2MjpaenG5RteFxJCmNKW9TX15NGo6GNGzdSU1MT3bx5k2bNmqUrF2KusheJiYnk7u5Ojo6OJJfLafjw4RQSEkLffvutXj9TypBotVpycHB46PLa2loCQMOGDeuTx8wShvT3RaZeb8yySel8SuZqNuWkNDY2klqtpueff16vPSsrSy+pNTU1kVqtpoiICL11lUol/fu//zsR/fYF3dTUpOvTkRwvX75MRETnzp0jAHT48GGDWIzZh7HKysro9OnTVFdXR/fu3aMTJ07Q008/TTY2NnTu3DmTttXhcUmNiEgQBHJ0dCSivnfMOKl1jZS+BJm0zqesl28MLcLly5fR2NiIqVOnPrJfV0tSPFjawtvbG4MGDcK8efPwzjvvIDIyUjcjtznLXgwbNkxvPrgJEyYgMzMTgYGBSE9PR0ZGhknbM0ZDQwOICBqNBkDfO2YdZs+e3aX1+rMtW7Zg7969YofBmJ5++Ztax/xoHbNzP4y5SlLY2Njg+PHjmDhxItatWwdvb29ERESgqampx8tejB49GtbW1rh48WK3t9WZju36+fkBkMYxY4z1Xf3yTk2lUgEA7t2798h+95ekiImJ6dY+/f398emnn+L27dvYvHkzNmzYAH9/f90MBebYR2fa29vR3t5udC0rUx09ehQA8OKLLwLou8eM7zhMIwgC3n33XcyZM0fsUJgZSGnkd7+8UwsICICVlRW+/PLLR/YzV0mKGzdu4Pz58wB+/dJfv349xowZg/Pnz5u17MULL7xg0Hbq1CkQEYKCgrq9/QfdvHkTW7Zsgbu7OxYsWACg7x0zxpi09Muk5uLigrCwMOzbtw87d+5EbW0tzp49i23btun1M6YkhTFu3LiBN998Ez/++COam5tRWFiIq1evYsKECWbbBwBcv34d2dnZqKmpQUtLC06cOIGFCxfCw8MDixYt0vUztQwJEaG+vh7t7e0gIty+fRs5OTl49tlnYW1tjby8PN1van3tmDHGJEbccSrmAxNH79TV1dHChQtpwIABZGdnRxMnTqT4+HgCQO7u7vT9998T0aNLUhhb2qK0tJSCg4PJycmJrK2taciQIbRy5UpqbW197D5MsXTpUtJqtWRra0symYzc3d3pjTfeoBs3buj1M6YMyaFDh+jJJ58ktVpNCoWCrKysCIBupOP48eMpISGBqqqqDNbtS8eMRz92janXG7NsUjqfAhGRWAnVnARBQE5ODj/jZybJzc1FeHg4JHIZ9Bq+3qRFSuezXz5+ZIwxJk2c1CzYjz/+2GkpmQf/LLnGE2MPOnbsGOLi4gzKG82fP9+g77Rp02Bvbw9ra2v4+/vj9OnTIkRsvISEBIwaNQoajQZKpRI+Pj547733DGrImVImqqWlBcnJyfDx8YFCoYCjoyMCAgJQWloKADh06BA2btwo+Zp8xuKkZsH8/PwMZtzv7C87O1vsUBkzyl/+8hekpaVhxYoVeuWNBgwYgN27d+Mf//iHXv/PP/8ce/fuxcsvv4zi4mKMGTNGpMiNc/z4cSxevBilpaWorKxEcnIyUlNTu/Vyf3h4OP7+979jz549aGxsxA8//ACtVqtLlDNmzIBKpcLUqVNRU1Njro/SZ3FSY0xkTU1NCA4O7vP7eJwNGzYgOzsbubm5sLe311uWlpYGKysrREVF9ekyRnZ2doiKioKzszPs7e0xZ84chIaG4ujRoygvL9fra0yZqOzsbOTl5WHv3r343e9+B5lMBjc3Nxw8eFDvru6dd97BU089hZdeegmtra298lktFSc1xkS2c+dOVFRU9Pl9PMrly5exevVqrF27Vjf5wf2Cg4MRExOD69evY9myZSJEaB6HDx82qGs2cOBAAOjSbDcfffQRxowZg9GjRz+275o1a3DmzBmkpqaavB8p4aTGmImICJs3b8YTTzwBpVIJJycnzJw5U2/eyejoaCgUCr1qyW+99RZsbW0hCAIqKysBADExMVi6dClKSkogCAJ8fHyQlpYGlUqFQYMG4c0334SbmxtUKhWCg4Nx8uRJs+wDMG8JosdJS0sDEWHGjBkP7ZOUlIQRI0Zgx44dOHbs2CO3Z8w5MKXMUU+WMrp+/TpsbGzg5eVl0nrNzc345ptvEBgYaFR/JycnTJ48Gampqf17NG/vvkHQcyCh9yxY7+nKe2rx8fGkUCho165dVFNTQ2fPnqUxY8bQwIED6ebNm7p+r776Krm6uuqtm5KSQgB0JXSIiMLCwkir1er1i4qKIltbWzp//jzdvXuXiouLady4cWRvb09lZWVm2YcpJYgeZOr15u3tTaNGjep0mVarpStXrhAR0ddff01WVlY0fPhwqq+vJyKi/Px8CgkJ0VvH2HNgTJkjIvOVMnpQQ0MD2dvbU3R0tF67MWWirly5QgAoMDCQpkyZQoMHDyalUkl+fn704YcfUnt7u8H+4uLiDMpnGUNK3598p8aYCZqamrB582bMmjUL8+bNg4ODA0aPHo2PP/4YlZWVBrPSdIdMJtPdiYwaNQoZGRmoq6tDZmamWbY/ffp01NbWYvXq1WbZ3sM0NDTgypUr0Gq1j+0bFBSEd999F6WlpXj//fc77dOVcxAcHAyNRgMXFxdERESgoaEBZWVlAIC7d+8iIyMDoaGhCAsLg6OjI1atWgW5XN7tY52cnAw3NzckJSXptb/22ms4dOgQysvLUV9fj6ysLJSVlWHy5MkoLi4GAN1AEBcXF6xbtw7FxcW4desWZs6cicWLF+OTTz4x2J+vry8AoKioqFtx92Wc1BgzQXFxMerr6zF27Fi99nHjxkGhUOg9HjS3sWPHQq1Wd7m8jlgqKipARFCr1Ub1T0pKwsiRI5Geno6vvvrKYHl3z8GDZY56opQRABw4cAC5ubn47LPPDAbGDBs2DE8//TTs7OygUCh0ZaKampqQnp4OALpJyP39/REcHAxnZ2c4ODhg7dq1cHBw6DR5dxzjW7dudTnuvo6TGmMm6BgybWdnZ7DM0dERdXV1Pbp/pVKJ27dv9+g+zO3u3bsAYHSlCJVKhczMTAiCgAULFqCpqUlvubnPQU+UMsrOzsaGDRtQUFCgqwP4OA+WiXJzcwMA3W+jHRQKBTw9PVFSUmKwDRsbGwC/HfP+iJMaYyZwdHQEgE6/OGtqauDu7t5j+25paenxffSEji9aU14ODgoKwpIlS3Dp0iUkJibqLTP3Obi/XBI9MMT+xIkTJm0LALZu3Yrdu3fj+PHjGDJkiNHrPVgmys7ODr6+vrpqFfdrbW2Fg4ODQXtzczOA3455f8RJjTETBAQEwM7ODt99951e+8mTJ9Hc3IxnnnlG1yaTyXSPuMyhoKAARIQJEyb02D56wqBBgyAIgsnvnyUmJsLPzw+FhYV67aacA2OYq5QRESE2NhZFRUXIy8vr9E6yg7FlosLDw1FYWIiffvpJ19bY2IirV692Osy/4xi7urp256P0aZzUGDOBSqXC0qVLceDAAezevRu1tbUoKirCokWL4ObmhqioKF1fHx8f/PLLL8jLy0NLSwtu376Nq1evGmzT2dkZN27cQGlpKerq6nRJqr29HdXV1WhtbcXZs2cRExMDDw8PREZGmmUfppYg6iq1Wg1vb29dxXljdTyGfPC9L1POgbH7eVwpo4iICLi6uj5ymq7z58/jgw8+wPbt2yGXyw2mwNq0aZOur7FlopYsWQJPT09ERkairKwMVVVViI2NRVNTU6cDaTqOsTHvtUmWeAMvzQsSGpLKek9XhvS3t7dTSkoK+fr6klwuJycnJwoNDaULFy7o9auqqqLnnnuOVCoVeXl50dtvv03Lly8nAOTj46Mbmn/69Gny9PQkGxsbmjhxIt28eZOioqJILpfT0KFDSSaTkUajoZkzZ1JJSYnZ9mFMCaKHMfV6i46OJrlcTo2Njbq2AwcObZqaBgAAIABJREFUkFarJQA0cOBAWrx4cafrLl++3GBIvzHnwNgyR0SPL2UUGhpKACg+Pv6hn7GoqIgAPPQvJSVF19fYMlFEROXl5TR37lxycnIipVJJ48ePp/z8/E5jmD59Og0dOrTT4f6PIqXvT05qrF+z1HpqUVFR5OzsLHYYD2Xq9Xbp0iWSyWS0a9euHoyq57S1tdGkSZNo586dYofyUJWVlaRSqWjTpk0mryul709+/MiYhZLSrOs+Pj5ISEhAQkKCwYz1lq6trQ15eXmoq6uz6IoYa9asQWBgIKKjo8UORVSc1BhjvSIuLg6zZ89GREREn5q0uKCgAPv370d+fr7R79r1ts2bN+PMmTM4cuQI5HK52OGIipMaYxZmxYoVyMzMxJ07d+Dl5YV9+/aJHZLZrFu3DtHR0Vi/fr3YoRht6tSp2LNnj94cm5bk4MGDuHfvHgoKCuDk5CR2OKKTiR0AY0xfcnIykpOTxQ6jx0ybNg3Tpk0TOwzJCAkJQUhIiNhhWAy+U2OMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZApE06n4LgiB2CIwx1mfl5ORgzpw5YofRbZIZ0p+TkyN2CIw9VHh4OGJiYvRmYGfMkgQHB4sdgllI5k6NMUsmCIJk/iXMmCXj39QYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xhhjksFJjTHGmGRwUmOMMSYZMrEDYExqsrKyUFdXZ9B+7Ngx1NTU6LWFhobCxcWlt0JjTPIEIiKxg2BMSiIjI/G3v/0Ncrlc19ZxmQmCAABoa2uDnZ0dKioqoFQqRYmTMSnix4+MmdncuXMBAC0tLbq/1tZWtLa26v7b2toas2fP5oTGmJnxnRpjZtba2gpXV1f88ssvj+z3xRdf4Pe//30vRcVY/8B3aoyZmUwmw9y5c/UePz5o4MCBmDx5ci9GxVj/wEmNsR4wd+5ctLS0dLpMLpdj/vz5sLa27uWoGJM+fvzIWA8gInh4eODatWudLv/2228xbty4Xo6KMenjOzXGeoAgCJg3b16njyCHDRuGsWPHihAVY9LHSY2xHtLZI0i5XI7IyEjd0H7GmHnx40fGepCfnx8uXLig13bu3Dn4+/uLFBFj0sZ3aoz1oPnz5+s9ghw1ahQnNMZ6ECc1xnrQvHnz0NraCuDXR4+vvfaayBExJm38+JGxHjZ27Fj83//9HwRBQGlpKTw8PMQOiTHJ4js1xnrYn//8ZwDA7373O05ojPUw0Wfpnz17ttghMNaj7t69C0EQcO/ePf7/nUnekiVLEBQUJNr+Rb9T27dv30NfUGWsr7h27Rr27dvX6TKVSgVXV1e4u7v3clSWj69/adm3bx/Ky8tFjUH0OzUAePfddzFnzhyxw2Csy3JzcxEeHo69e/d2uvzy5cvw8fHp5agsnyAIfP1LiCW8fyn6nRpj/QEnNMZ6Byc1xhhjksFJjTHGmGRwUmOMMSYZnNQYY4xJBic1xizIkSNH4ODggE8//VTsUCzesWPHEBcXh/3798Pb2xuCIEAQBMyfP9+g77Rp02Bvbw9ra2v4+/vj9OnTIkRsvISEBIwaNQoajQZKpRI+Pj547733UF9fr9cvKSlJ97nv/wsICDDYZktLC5KTk+Hj4wOFQgFHR0cEBASgtLQUAHDo0CFs3LgRbW1tvfERewwnNcYsCM9aZ5y//OUvSEtLw4oVKxAWFoaffvoJWq0WAwYMwO7du/GPf/xDr//nn3+OvXv34uWXX0ZxcTHGjBkjUuTGOX78OBYvXozS0lJUVlYiOTkZqamp3Xp5Pzw8HH//+9+xZ88eNDY24ocffoBWq9UlyhkzZkClUmHq1Kmoqakx10fpdZzUGLMg06dPx507d/Dyyy+LHQqampoQHBwsdhgGNmzYgOzsbOTm5sLe3l5vWVpaGqysrBAVFYU7d+6IFGH32dnZISoqCs7OzrC3t8ecOXMQGhqKo0ePGrzcvGvXLhCR3t+5c+f0+mRnZyMvLw979+7F7373O8hkMri5ueHgwYN6d3XvvPMOnnrqKbz00ku6ibj7Gk5qjLFO7dy5ExUVFWKHoefy5ctYvXo11q5dC5VKZbA8ODgYMTExuH79OpYtWyZChOZx+PBhWFtb67UNHDgQANDY2Gjy9j766COMGTMGo0ePfmzfNWvW4MyZM0hNTTV5P5aAkxpjFuKrr76Ch4cHBEHAhx9+CADIyMiAra0t1Go1Dh48iBdffBEajQbu7u7IysrSrZuWlgaVSoVBgwbhzTffhJubG1QqFYKDg3Hy5Eldv+joaCgUCgwePFjX9tZbb8HW1haCIKCyshIAEBMTg6VLl6KkpASCIOheHj969Cg0Gg3WrVvXG4fEQFpaGogIM2bMeGifpKQkjBgxAjt27MCxY8ceuT0iwubNm/HEE09AqVTCyckJM2fOxI8//qjrY+w5AIC2tjbEx8fDw8MDNjY2ePLJJ5GTk9O9D/0v169fh42NDby8vExar7m5Gd988w0CAwON6u/k5ITJkycjNTW1bz4OJ5EBoJycHLHDYKxbcnJyyByXU3l5OQGgrVu36tpWrlxJAOiLL76gO3fuUEVFBU2aNIlsbW2publZ1y8qKopsbW3p/PnzdPfuXSouLqZx48aRvb09lZWV6fq9+uqr5OrqqrfflJQUAkC3b9/WtYWFhZFWq9Xrd/jwYbK3t6eEhIRuf1Yi069/b29vGjVqVKfLtFotXblyhYiIvv76a7KysqLhw4dTfX09ERHl5+dTSEiI3jrx8fGkUCho165dVFNTQ2fPnqUxY8bQwIED6ebNm7p+xp6DZcuWkVKppH379lF1dTWtWLGCrKys6NSpU0Z/xs40NDSQvb09RUdH67UnJiaSu7s7OTo6klwup+HDh1NISAh9++23uj5XrlwhABQYGEhTpkyhwYMHk1KpJD8/P/rwww+pvb3dYH9xcXEEgAoLC02K0xK+z/lOjbE+Ijg4GBqNBi4uLoiIiEBDQwPKysr0+shkMt1dx6hRo5CRkYG6ujpkZmaaJYbp06ejtrYWq1evNsv2TNHQ0IArV65Aq9U+tm9QUBDeffddlJaW4v333++0T1NTEzZv3oxZs2Zh3rx5cHBwwOjRo/Hxxx+jsrIS27ZtM1jnUefg7t27yMjIQGhoKMLCwuDo6IhVq1ZBLpd3+/gnJyfDzc0NSUlJeu2vvfYaDh06hPLyctTX1yMrKwtlZWWYPHkyiouLAUA3EMTFxQXr1q1DcXExbt26hZkzZ2Lx4sX45JNPDPbn6+sLACgqKupW3GLgpMZYH6RQKAD8Okz7UcaOHQu1Wq33OK2vqqioABFBrVYb1T8pKQkjR45Eeno6vvrqK4PlxcXFqK+vx9ixY/Xax40bB4VCoffYtjMPnoMLFy6gsbFRb+CFjY0NBg8e3K3jf+DAAeTm5uKzzz4zGBgzbNgwPP3007Czs4NCocCECROQmZmJpqYmpKenAwCUSiUAwN/fH8HBwXB2doaDgwPWrl0LBweHTpN3xzG+detWl+MWCyc1xiROqVTi9u3bYofRbXfv3gXw25f046hUKmRmZkIQBCxYsABNTU16yzuGrdvZ2Rms6+joiLq6OpPia2hoAACsWrVK752xq1evdmlwB/DrqMUNGzagoKAAw4cPN2qd0aNHw9raGhcvXgQAuLm5AYDu99IOCoUCnp6eKCkpMdiGjY0NgN+OeV/CSY0xCWtpaUFNTY0karl1fNGa8nJwUFAQlixZgkuXLiExMVFvmaOjIwB0mry6csxcXFwAAFu2bDEYYn/ixAmTtgUAW7duxe7du3H8+HEMGTLE6PXa29vR3t6uS/52dnbw9fXF+fPnDfq2trbCwcHBoL25uRnAb8e8L+GkxpiEFRQUgIgwYcIEXZtMJnvsY0tLNGjQIAiCYPL7Z4mJifDz80NhYaFee0BAAOzs7PDdd9/ptZ88eRLNzc145plnTNrPsGHDoFKpcObMGZPWexARITY2FkVFRcjLy+v0TrLDCy+8YNB26tQpEJFe9enw8HAUFhbip59+0rU1Njbi6tWrnQ7z7zjGrq6u3fkoouCkxpiEtLe3o7q6Gq2trTh79ixiYmLg4eGByMhIXR8fHx/88ssvyMvLQ0tLC27fvo2rV68abMvZ2Rk3btxAaWkp6urq0NLSgvz8fNGG9KvVanh7e5tcKbvjMeSD732pVCosXboUBw4cwO7du1FbW4uioiIsWrQIbm5uiIqKMnk/r7/+OrKyspCRkYHa2lq0tbXh2rVr+PnnnwEAERERcHV1feQ0XefPn8cHH3yA7du3Qy6XG0yBtWnTJl3f69evIzs7GzU1NWhpacGJEyewcOFCeHh4YNGiRbp+S5YsgaenJyIjI1FWVoaqqirExsaiqamp04E0HcfYmPfaLI54Ay9/BQsYAspYd5ljSP/WrVtp8ODBBIDUajXNmDGD0tPTSa1WEwDy9fWlkpIS2rZtG2k0GgJAnp6edPHiRSL6dUi/XC6noUOHkkwmI41GQzNnzqSSkhK9/VRVVdFzzz1HKpWKvLy86O2336bly5cTAPLx8dEN/z99+jR5enqSjY0NTZw4kW7evElHjhwhe3t7SkpK6tZn7WDq9R8dHU1yuZwaGxt1bQcOHCCtVksAaODAgbR48eJO112+fLnBkP729nZKSUkhX19fksvl5OTkRKGhoXThwgVdH1POwb179yg2NpY8PDxIJpORi4sLhYWFUXFxMRERhYaGEgCKj49/6GcsKioiAA/9S0lJ0fVdunQpabVasrW1JZlMRu7u7vTGG2/QjRs3DLZbXl5Oc+fOJScnJ1IqlTR+/HjKz8/vNIbp06fT0KFDOx3u/yiW8H3OSY0xMzDXe2rdERUVRc7OzqLGYCpTr/9Lly6RTCajXbt29WBUPaetrY0mTZpEO3fuFDuUh6qsrCSVSkWbNm0yeV1L+D7nx4+MSUhfn2H9cXx8fJCQkICEhASDGestXVtbG/Ly8lBXV4eIiAixw3moNWvWIDAwENHR0WKH0iWc1BhjfUpcXBxmz56NiIiIPjVpcUFBAfbv34/8/Hyj37XrbZs3b8aZM2dw5MgRyOVyscPpkj6f1BYuXAh7e3sIgtDtUUd93caNG+Hn5wcbGxvY2trCz88Pq1evRm1trcnberBGVcefQqHAoEGDMGXKFKSkpKC6uroHPgkz1YoVK5CZmYk7d+7Ay8sL+/btEzukHrVu3TpER0dj/fr1YoditKlTp2LPnj16825akoMHD+LevXsoKCiAk5OT2OF0nagPP8k8z2CzsrK6NE+Z1EyfPp02bdpEFRUVVFdXR7m5uSSXy+n555/v8ja1Wi05ODgQ0a8/qldXV9P//M//UGRkJAmCQG5ubt2e104KLOE3tb7IHNc/sxyWcD77/J2a1HSnhpVCocBbb70FFxcX2NnZYfbs2Zg5cyb++7//WzekuDsEQYCjoyOmTJmCzMxM5Obm4tatW7oaYH2dpdYPY4wZTxJJTRAEsUMwm+7UsDpw4IBBjamhQ4cCQI/8qP7KK68gMjISFRUV+Pjjj82+/d5mifXDGGOm6XNJjYiQkpKCkSNHQqlUwsHBAcuXL9fr88EHH0CtVsPe3h4VFRVYunQphg4digsXLhhVP8nY2lQd8Txue92tYdUdly5dgqOjIzw9PXVt5qyJ1fFSb35+PgA+9owxkYn68JNMfwa7cuVKEgSB/vrXv1J1dTU1NjZSenq6wW9qHfWP3nnnHdq6dSvNmjWLfvjhB6PrJxlbm8rY7XWnhpWpmpub6dq1a7R161ZSKpUG7/SYUhPr/t/UOlNbW0sAaNiwYbq2/njs+Te1rjH1+meWzRLOp+hXoSkHobGxkdRqtcHAh84GinR8sTY1Nemtb2dnRxEREXrrf/vttwRA70s+KirK4Mv81KlTBIDWrl1r8vZ6M6m5uroSABowYAD9x3/8h14RQ1M9LqkREQmCQI6Ojrr/7o/HnpNa11jClyAzH0s4n7LeuiM0h8uXL6OxsRFTp07t0vrdrZ/0YG2q7m6vp5SXl6OmpgaFhYWIi4vDtm3bcPz4cQwaNMjs+2poaAARQaPRPLJffzn2Uvp9t7eEh4cjPDxc7DCYRPSppNYxyWZHiQdTmaN+0v21qcxdj8lc5HI5XFxcMG3aNHh5eWHEiBFITk5Gamqq2ffVUbPJz8/vkf36y7HPyckRZb99VXh4OGJiYvRmlGd9lyX846RPJbWOkX337t3r0vrdrZ/0YG0qc9dj6gk+Pj6wtrbWlXY3t6NHjwIAXnzxxUf26y/Hfs6cOaLst68KDw9HUFAQHzeJsISk1qdGPwYEBMDKygpffvlll9fvTv2kB2tTmbK9nq5hVVVVhT/96U8G7ZcuXUJbWxuGDRtm9n3evHkTW7Zsgbu7OxYsWPDIvlI+9owxy9GnkpqLiwvCwsKwb98+7Ny5E7W1tTh79iy2bdtm1Pqm1k96XG0qU7bXnRpWxrC1tcXnn3+O48ePo7a2Fi0tLSgsLMRrr70GW1tbLFmyRNfX1JpYRIT6+nq0t7eDiHD79m3k5OTg2WefhbW1NfLy8h77m5qUjz1jzIKIOkyFTB8tU1dXRwsXLqQBAwaQnZ0dTZw4keLj4wkAubu70/fff08bN24kGxsb3VDz+4e0G1M/icj42lTGbq87NayMNWPGDPLy8iI7OztSKpWk1WopIiKCioqK9PoZUxPr0KFD9OSTT5JarSaFQkFWVlYEQDfScfz48ZSQkEBVVVV66/XXY8+jH7vG1OufWTZLOJ/CvwIRjSAIyMnJsbhn6m+++Sb27t2LqqoqsUPpd/risc/NzUV4eDhEvpz6HEu9/lnXWML57FOPH3ub1GtTWTI+9oyxruCkZsF+/PFHg9Ivnf1ZcsFBxnrKsWPHEBcXZ1Amaf78+QZ9p02bBnt7e1hbW8Pf3x+nT58WIWLjJSQkYNSoUdBoNFAqlfDx8cF7771nMIdrUlJSp98JAQEBBttsaWlBcnIyfHx8oFAo4OjoiICAAJSWlgIADh06hI0bN/b9f1CK+vCTLOMZ7IPi4uJIoVAQABo+fDjt3btX7JD6jb567Pk3ta7p6vUfHx9PL7/8MtXW1uratFotDRgwgADQ4cOHDdbJz8+nkJCQbsXbWyZPnkzp6elUVVVFtbW1lJOTQ3K5nP7whz/o9UtMTCQABn/+/v4G2wwNDaWRI0fSN998Qy0tLXTjxg2aMWOG3m/uqampNHnyZKquru5S3JbwfS76VWgJB4Gx7rKEpNbY2EhBQUF9ah9duf7Xr19PI0aM0JuGjejXpLZnzx6ysrKioUOHUk1Njd7yvpTUpk+fTq2trXptc+bMIQB6858mJiYazO3amaysLBIEgc6ePfvYvtHR0RQUFEQtLS0mx20J3+f8+JExieiN0jlil+e5fPkyVq9ejbVr1xqUWQKA4OBgxMTE4Pr161i2bJkIEZrH4cOHYW1trdc2cOBAAEBjY6PJ2/voo48wZswYjB49+rF916xZgzNnzvTIDES9gZMaYyKhHi6dY2wZn+6W5zFnKaPHSUtLAxFhxowZD+2TlJSEESNGYMeOHTh27Ngjt2fMOcjIyICtrS3UajUOHjyIF198ERqNBu7u7sjKytLbXltbG+Lj4+Hh4QEbGxs8+eSTZps67fr167CxsYGXl5dJ6zU3N+Obb75BYGCgUf2dnJwwefJkpKam9s3RvKLeJ5Jl3K4y1l1defzYG6VzjC3j0519mFLK6EGmXv/e3t40atSoTpdptVq6cuUKERF9/fXXZGVlRcOHD6f6+noi6vzxo7HnoKPyxBdffEF37tyhiooKmjRpEtna2upVwVi2bBkplUrat28fVVdX04oVK8jKyopOnTpl9GfsTENDA9nb21N0dLRee2JiIrm7u5OjoyPJ5XIaPnw4hYSE0Lfffqvrc+XKFQJAgYGBNGXKFBo8eDAplUry8/OjDz/8kNrb2w32FxcXZ1D5xBiW8H3Od2qMiaCpqQmbN2/GrFmzMG/ePDg4OGD06NH4+OOPUVlZafQsOcaQyWS6O5FRo0YhIyMDdXV1yMzMNMv2p0+fjtraWqxevdos23uYhoYGXLlyBVqt9rF9g4KC8O6776K0tBTvv/9+p326cg6Cg4Oh0Wjg4uKCiIgINDQ0oKysDABw9+5dZGRkIDQ0FGFhYXB0dMSqVasgl8u7fayTk5Ph5uaGpKQkvfbXXnsNhw4dQnl5Oerr65GVlYWysjJMnjxZN99rx4hJFxcXrFu3DsXFxbh16xZmzpyJxYsX45NPPjHYn6+vLwCgqKioW3GLgZMaYyIQs3TOg2V8+oqKigoQEdRqtVH9k5KSMHLkSKSnp+Orr74yWN7dc6BQKABAN53ahQsX0NjYqDec3sbGBoMHD+7WsT5w4AByc3Px2Wefwd7eXm/ZsGHD8PTTT8POzg4KhQITJkxAZmYmmpqakJ6eDuDX6hYA4O/vj+DgYDg7O8PBwQFr166Fg4NDp8m74xjfunWry3GLhZMaYyIQu3TO/WV8+oq7d+8C+O1L+nFUKhUyMzMhCAIWLFiApqYmveXmPgcNDQ0AgFWrVum9M3b16tUuDe4AgOzsbGzYsAEFBQUYPny4UeuMHj0a1tbWurJQbm5uAKD7bbSDQqGAp6cnSkpKDLZhY2MD4Ldj3pdwUmNMBGKWznmwjE9f0fFFa8rLwUFBQViyZAkuXbqExMREvWXmPgcddR63bNkC+vV1Kd3fiRMnTNoWAGzduhW7d+/G8ePHMWTIEKPXa29vR3t7uy7529nZwdfXF+fPnzfo29raCgcHB4P25uZmAL8d876EkxpjIhCzdM6DZXx6Yh89YdCgQRAEAXfu3DFpvcTERPj5+aGwsFCvvbvlkB40bNgwqFQqnDlzxqT1HkREiI2NRVFREfLy8jq9k+zwwgsvGLSdOnUKRKRXeDU8PByFhYX46aefdG2NjY24evVqp8P8O46xq6trdz6KKDipMSaC3iyd87gyPt3dh6mljLpKrVbD29sb165dM2m9jseQD773ZWo5JGP28/rrryMrKwsZGRmora1FW1sbrl27hp9//hkAEBERAVdX10dO03X+/Hl88MEH2L59O+RyucEUWJs2bdL1vX79OrKzs1FTU4OWlhacOHECCxcuhIeHBxYtWqTrt2TJEnh6eiIyMhJlZWWoqqpCbGwsmpqaOh1I03GMjXmvzeKIN/DyV7CAIaCMdVdXhvT3RukcY8v4dGcfxpQyehhTr//o6GiSy+XU2Nioaztw4ABptVoCQAMHDqTFixd3uu7y5csNhvQbcw7S09NJrVYTAPL19aWSkhLatm0baTQaAkCenp508eJFIiK6d+8excbGkoeHB8lkMnJxcaGwsDAqLi4mol+nqgJA8fHxD/2MRUVFnU591fGXkpKi67t06VLSarVka2tLMpmM3N3d6Y033qAbN24YbLe8vJzmzp1LTk5OpFQqafz48ZSfn99pDNOnT6ehQ4d2Otz/USzh+5yTGmNmYAnTZHUmKiqKnJ2dxQ7joUy9/i9dukQymcyoqaEsUVtbG02aNIl27twpdigPVVlZSSqVijZt2mTyupbwfc6PHxmTuD4/6/p9fHx8kJCQgISEBIMZ6y1dW1sb8vLyUFdXZ9GVNdasWYPAwEBER0eLHUqXcFJjjPUpcXFxmD17NiIiIkweNCKmgoIC7N+/H/n5+Ua/a9fbNm/ejDNnzuDIkSOQy+Vih9MlnNQYk6gVK1YgMzMTd+7cgZeXF/bt2yd2SGazbt06REdHY/369WKHYrSpU6diz549enNsWpKDBw/i3r17KCgogJOTk9jhdJlM7AAYYz0jOTkZycnJYofRY6ZNm4Zp06aJHYZkhISEICQkROwwuo3v1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhkWMVCkK5N9MmZJOv4fzs3NFTmSvoevf2ZOwr/eAhcvAEEQc/eMMcbMKCcnB3PmzBFt/6LfqYmcUxnrFYIgiH6xM9Yf8G9qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJEMgIhI7CMakJCoqChcuXNBrO336NLy8vODk5KRrs7a2xt/+9je4u7v3doiMSZZM7AAYkxpXV1ds27bNoP3s2bN6/+3t7c0JjTEz48ePjJnZn/70p8f2USgUiIyM7PlgGOtn+PEjYz0gICAA58+fx6MurwsXLmDEiBG9GBVj0sd3aoz1gD//+c+wtrbudJkgCHjqqac4oTHWAzipMdYD5s6di7a2tk6XWVtb47XXXuvliBjrH/jxI2M9JDg4GCdPnkR7e7teuyAIKC8vx9ChQ0WKjDHp4js1xnrI/PnzIQiCXpuVlRUmTpzICY2xHsJJjbEeMnv2bIM2QRDw5z//WYRoGOsfOKkx1kMGDhyIqVOn6g0YEQQBoaGhIkbFmLRxUmOsB82bN083rN/a2hovvPCcKnRkAAAgAElEQVQCBgwYIHJUjEkXJzXGetCsWbOgUCgAAESEefPmiRwRY9LGSY2xHmRra4s//vGPAH6dReTll18WOSLGpI2TGmM97NVXXwUAhIaGwtbWVuRoGJM2ybyn9uDQacYYY8bLycnBnDlzxA6j2yQ1S39MTAyCgoLEDoP1ISdOnEBqaipycnJ6dD+7d+9GREQEZDJpXHLh4eF8vUlIeHi42CGYjaTu1KTyLw3We3JzcxEeHv7IiYfN4e7du1CpVD26j97E15u0SOl88m9qjPUCKSU0xiwZJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGzODIkSNwcHDAp59+KnYoFu/YsWOIi4vD/v374e3tDUEQIAgC5s+fb9B32rRpsLe3h7W1Nfz9/XH69GkRIjZeQkICRo0aBY1GA6VSCR8fH7z33nuor6/X65eUlKT73Pf/BQQEGGyzpaUFycnJ8PHxgUKhgKOjIwICAlBaWgoAOHToEDZu3PjQorT9DSc1xsxAIm/G9Li//OUvSEtLw4oVKxAWFoaffvoJWq0WAwYMwO7du/GPf/xDr//nn3+OvXv34uWXX0ZxcTHGjBkjUuTGOX78OBYvXozS0lJUVlYiOTkZqampnZYhMlZ4eDj+/ve/Y8+ePWhsbMQPP/wArVarS5QzZsyASqXC1KlTUVNTY66P0mdxUmPMDKZPn447d+5YxNyOTU1NCA4OFjsMAxs2bEB2djZyc3Nhb2+vtywtLQ1WVlaIiorCnTt3RIqw++zs7BAVFQVnZ2fY29tjzpw5CA0NxdGjR1FeXq7Xd9euXSAivb9z587p9cnOzkZeXh727t2L3/3ud5DJZHBzc8PBgwf17ureeecdPPXUU3jppZfQ2traK5/VUnFSY0xidu7ciYqKCrHD0HP58mWsXr0aa9eu7fSdveDgYMTExOD69etYtmyZCBGax+HDh/Xq5wG/1tUDgMbGRpO399FHH2HMmDEYPXr0Y/uuWbMGZ86cQWpqqsn7kRJOaox101dffQUPDw8IgoAPP/wQAJCRkQFbW1uo1WocPHgQL774IjQaDdzd3ZGVlaVbNy0tDSqVCoMGDcKbb74JNzc3qFQqBAcH4+TJk7p+0dHRUCgUGDx4sK7trbfegq2tLQRBQGVlJYBfp4pbunQpSkpKIAgCfHx8AABHjx6FRqPBunXreuOQGEhLSwMRYcaMGQ/tk5SUhBEjRmDHjh04duzYI7dHRNi8eTOeeOIJKJVKODk5YebMmfjxxx91fYw9BwDQ1taG+Ph4eHh4wMbGBk8++aTZpk67fv06bGxs4OXlZdJ6zc3N+OabbxAYGGhUfycnJ0yePBmpqan9+3E4SQQAysnJETsM1sfk5OSQOS6D8vJyAkBbt27Vta1cuZIA0BdffEF37tyhiooKmjRpEtna2lJzc7OuX1RUFNna2tL58+fp7t27VFxcTOPGjSN7e3sqKyvT9Xv11VfJ1dVVb78pKSkEgG7fvq1rCwsLI61Wq9fv8OHDZG9vTwkJCd3+rESmX2/e3t40atSoTpdptVq6cuUKERF9/fXXZGVlRcOHD6f6+noiIsrPz6eQkBC9deLj40mhUNCuXbuopqaGzp49S2PGjKGBAwfSzZs3df2MPQfLli0jpVJJ+/bto+rqalqxYgVZWVnRqVOnjP6MnWloaCB7e3uKjo7Wa09MTCR3d3dydHQkuVxOw4cPp5CQEPr22291fa5cuUIAKDAwkKZMmUKDBw8mpVJJfn5+9OGHH1J7e7vB/uLi4ggAFRYWmhSnlL4/+U6NsR4WHBwMjUYDFxcXREREoKGhAWVlZXp9ZDKZ7q5j1KhRyMjIQF1dHTIzM80Sw/Tp01FbW4vVq1ebZXumaGhowJUrV6DVah/bNygoCO+++y5KS0vx/vvvd9qnqakJmzdvxqxZszBv3jw4ODhg9OjR+Pjjj1FZWYlt27YZrPOoc3D37l1kZGQgNDQUYWFhcHR0xKpVqyCXy7t9/JOTk+Hm5oakpCS99tdeew2HDh1CeXk56uvrkZWVhbKyMkyePBnFxcUAoBsI4uLignXr1qG4uBi3bt3CzJkzsXjxYnzyyScG+/P19QUAFBUVdSvuvoyTGmO9qKMKdktLyyP7jR07Fmq1Wu9xWl9VUVEBIoJarTaqf1JSEkaOHIn09HR89dVXBsuLi4tRX1+PsWPH6rWPGzcOCoVC77FtZx48BxcuXEBjY6PewAsbGxsMHjy4W8f/wIEDyM3NxWeffWYwMGbYsGF4+umnYWdnB4VCgQkTJiAzMxNNTU1IT08HACiVSgCAv78/goOD4ezsDAcHB6xduxYODg6dJu+OY3zr1q0ux93XcVJjzEIplUrcvn1b7DC67e7duwB++5J+HJVKhczMTAiCgAULFqCpqUlvecewdTs7O4N1HR0dUVdXZ1J8DQ0NAIBVq1bpvTN29erVLg3uAH4dtbhhwwYUFBRg+PDhRq0zevRoWFtb4+LFiwAANzc3AND9XtpBoVDA09MTJSUlBtuwsbEB8Nsx7484qTFmgVpaWlBTUwN3d3exQ+m2ji9aU14ODgoKwpIlS3Dp0iUkJibqLXN0dASATpNXV46Zi4sLAGDLli0GQ+xPnDhh0rYAYOvWrdi9ezeOHz+OIUOGGL1ee3s72tvbdcnfzs4Ovr6+OH/+vEHf1tZWODg4GLQ3NzcD+O2Y90ec1BizQAUFBSAiTJgwQdcmk8ke+9jSEg0aNAiCIJj8/lliYiL8/PxQWFio1x4QEAA7Ozt89913eu0nT55Ec3MznnnmGZP2M2zYMKhUKpw5c8ak9R5ERIiNjUVRURHy8vI6vZPs8MILLxi0nTp1CkSkV3g1PDwchYWF+Omnn3RtjY2NuHr1aqfD/DuOsaura3c+Sp/GSY0xC9De3o7q6mq0trbi7NmziImJgYeHByIjI3V9fHx88MsvvyAvLw8tLS24ffs2rl69arAtZ2dn3LhxA6Wlpairq0NLSwvy8/NFG9KvVqvh7e2Na9eumbRex2PIB9/7UqlUWLp0KQ4cOIDdu3ejtrYWRUVFWLRoEdzc3BAVFWXyfl5//XVkZWUhIyMDtbW1aGtrw7Vr1/Dzzz8DACIiIuDq6vrIabrOnz+PDz74ANu3b4dcLjeYAmvTpk26vtevX0d2djZqamrQ0tKCEydOYOHChfDw8MCiRYt0/ZYsWQJPT09ERkairKwMVVVViI2NRVNTU6cDaTqOsTHvtUmWeAMvzQsSGpLKeo85hvRv3bqVBg8eTABIrVbTjBkzKD09ndRqNQEgX19fKikpoW3btpFGoyEA5OnpSRcvXiSiX4f0y+VyGjp0KMlkMtJoNDRz5kwqKSnR209VVRU999xzpFKpyMvLi95++21avnw5ASAfHx/d8P/Tp0+Tp6cn2djY0MSJE+nmzZt05MgRsre3p6SkpG591g6mXm/R0dEkl8upsbFR13bgwAHSarUEgAYOHEiLFy/udN3ly5cbDOlvb2+nlJQU8vX1JblcTk5OThQaGkoXLlzQ9THlHNy7d49iY2PJw8ODZDIZubi4UFhYGBUXFxMRUWhoKAGg+Pj4h37GoqIiAvDQv5SUFF3fpUuXklarJVtbW5LJZOTu7k5vvPEG3bhxw2C75eXlNHfuXHJyciKlUknjx4+n/Pz8TmOYPn06DR06tNPh/o8ipe9PTmqsXzPXe2rdERUVRc7OzqLGYCpTr7dLly6RTCajXbt29WBUPaetrY0mTZpEO3fuFDuUh6qsrCSVSkWbNm0yeV0pfX/y40fGLIDUZ1j38fFBQkICEhISDGast3RtbW3Iy8tDXV0dIiIixA7nodasWYPAwEBER0eLHYqoOKn9y8KFC2Fvbw9BELr9g7GYHlemwlgPlgXp+FMoFBg0aBCmTJmClJQUVFdX98wHYZITFxeH2bNnIyIiok9NWlxQUID9+/cjPz/f6HftetvmzZtx5swZHDlyBHK5XOxwRMVJ7V927NiB7du3ix1Gtz2uTIWx7i8L4uDgACJCe3s7KioqkJubCy8vL8TGxsLf399gFBoz3ooVK5CZmYk7d+7Ay8sL+/btEzukHrVu3TpER0dj/fr1YoditKlTp2LPnj16825akoMHD+LevXsoKCiAk5OT2OGITiZ2AMx8OspUfP/997rRTx1lKsxBEAQ4OjpiypQpmDJlCqZPn47w8HBMnz4dFy9e7PS9GfZoycnJSE5OFjuMXjVt2jRMmzZN7DAkIyQkBCEhIWKHYTH4Tu0+giCIHUK3mFKmwhxeeeUVREZGoqKiAh9//HGv7JMxxh6l3yY1IkJKSgpGjhwJpVIJBwcHLF++3KDfo0pSmFLa4ssvv8T48eOhVquh0WgwevRo1NbWPnYfxjKlTIU5y5B0vEeVn5+va+srx4wxJkFiD780F5g4JHXlypUkCAL99a9/perqampsbKT09HSDsg2PK0lhTGmL+vp60mg0tHHjRmpqaqKbN2/SrFmzdOVCzFH2wpQyFaaUIdFqteTg4PDQ5bW1tQSAhg0b1ueOGZFlDOnvi0y93phlk9L5lMzVbMpJaWxsJLVaTc8//7xee1ZWll5Sa2pqIrVaTREREXrrKpVK+vd//3ci+u0LuqmpSdenIzlevnyZiIjOnTtHAOjw4cMGsRizD2N0vPj5/PPP0z//+U+qqqqimpoaev/99wkA7d692+ht3e9xSY2ISBAEcnR0NPrzWMoxI+Kk1lVS+hJk0jqf/XKgyOXLl9HY2IipU6c+sl9XS1I8WNrC29sbgwYNwrx58/DOO+8gMjJSN3O3ucpePFimosPatWvx0UcfYdu2bXj11VeN3p6xGhoaQETQaDQA+tYxu19ubm6X1uvPujLZL2M9Tuysai4w4V8aR44cIQAGswM8eKf2z3/+86FT3kyYMIGIOr/r2L59OwGgH374Qdd27tw5+uMf/0gymYwEQaDw8HBqbGw0ah/GqKurIwA0b948g2VPPvkkDR061Oht3e9xd2qnT58mADRt2jQi6lvHjOi3OzX+47/+/ieVO7V+OVBEpVIBAO7du/fIfuYsSeHv749PP/0UN27cQGxsLHJycrBp0yaz7aMrZSrM4ejRowCAF198EUDfOmb3e3A7/PfoPwDIyckRPQ7+M9/5lIp+mdQCAgJgZWWFL7/88pH9zFWS4saNG7pk4+LigvXr12PMmDE4f/682fYBmF6mortu3ryJLVu2wN3dHQsWLADQ944ZY0xa+mVSc3FxQVhYGPbt24edO3eitrYWZ8+eNSiPbkxJCmPcuHEDb775Jn788Uc0NzejsLAQV69exYQJE8y2D8D4MhWmliEhItTX16O9vR1EhNu3byMnJwfPPvssrK2tkZeXp/tNra8dM8aYxJBEwMRnwnV1dbRw4UIaMGAA2dnZ0cSJEyk+Pp4AkLu7O33//fdE9OiSFMaWtigtLaXg4GBycnIia2trGjJkCK1cuZJaW1sfuw9TGVOmwpgyJIcOHaInn3yS1Go1KRQKsrKyIgC6kY7jx4+nhIQEqqqqMli3Lx0zHv3YNaZeb8yySel8CkTSeKAqCAJycnIwZ84csUNhfUhubi7Cw8Ml97tCT+PrTVqkdD775eNHxhhj0sRJzYL9+OOPBqVfOvuz5BpPjDHWmzipWTA/Pz+jhuNmZ2eLHSpjRjt27Bji4uIMavbNnz/foO+0adNgb28Pa2tr+Pv74/Tp0yJEbLr29nZs2bJFbyKE+yUkJGDUqFHQaDRQKpXw8fHBe++912mJqE8++QTjxo2Dvb09PD098frrr+PmzZu65YcOHcLGjRslX2jWaOL8lGd+kNAPnaz38ECRrunq9RYfH08vv/wy1dbW6tq0Wi0NGDCAgM6nRcvPz6eQkJBuxdubLl68SM8++ywBoKeeeqrTPpMnT6b09HSqqqqi2tpaysnJIblcTn/4wx/0+mVnZxMA2rhxI9XU1FBhYSF5e3tTYGAgtbS06PqlpqbS5MmTqbq6uksxS+n7k+/UGBNZU1PTQ/9F35f28TgbNmxAdnY2cnNzYW9vr7csLS0NVlZWiIqK6lNVsR/0/fff4/3338eiRYseWTHDzs4OUVFRcHZ2hr29PebMmYPQ0FAcPXoU5eXlun7/+Z//iSFDhmD58uVwcHBAYGAglixZgjNnzuDkyZO6fu+88w6eeuopvPTSS2htbe3Rz2jpOKkxJrKdO3eioqKiz+/jUS5fvozVq1dj7dq1uhl97hccHIyYmBhcv34dy5YtEyFC83jqqaewf/9+vPrqq7r5WDtz+PBhWFtb67UNHDgQwK8TJnQoLy+Hm5ubXq3HYcOGAQCuXr2qt/6aNWtw5swZpKamdvtz9GWc1BgzERFh8+bNeOKJJ6BUKuHk5ISZM2fqTaYcHR0NhUKBwYMH69reeust2NraQhAEVFZWAgBiYmKwdOlSlJSUQBAE+Pj4IC0tDSqVCoMGDcKbb74JNzc3qFQqBAcH6/3rvDv7AMxbV+9x0tLSQESYMWPGQ/skJSVhxIgR2LFjB44dO/bI7RlzDkyp3WcJ9fmuX78OGxsbeHl56dq8vb0N/jHS8Xuat7e3XruTkxMmT56M1NTU/v2KirhPP/8/e/ceFMWZ7g/828DAMDDDJSjiBeViVMToMboJo65Ra80xnKiIIlkxa0w2JCdmgrcoXohBRAguUhg5ruihtiRREAkaV0xKPSRlxXKTo0TEXwwaUNQgoALDTRCe3x85TBwHdAZm6KF9PlVTlbz9dvcz3XY/dM/b/ZgPJHRPmPWe7vymFhMTQ/b29rRv3z6qqamhCxcu0Pjx48nDw4MqKip0/RYtWkSenp568yYlJREAXV04IqLQ0FDy8/PT6xcZGUlOTk506dIlam5upuLiYpo4cSIplUq6fv26WdZhSl29R5l6vPn6+lJAQECn0/z8/Ki0tJSIiL777juysbGhYcOGUX19PRF1/puasfvAmNp9ROarz/ewF154ocvf1B7V0NBASqWSNBqNXntBQQHJZDJKTU2luro6unjxIo0aNYpefvnlTpcTHR1NgH5NSGNI6fzJV2qMmaCpqQnJycmYN28eIiIi4OLigjFjxmDXrl2orq42eNVaT9jZ2emuRAICApCWlgatVouMjAyzLD84OBh1dXXYuHGjWZbXlYaGBpSWlsLPz++JfYOCgrB8+XKUlZXpvdrtYd3ZB2q1GiqVCv369UN4eDgaGhpw/fp1AEBzczPS0tIQEhKC0NBQuLq6YsOGDZDJZGbb1k8SHx8PLy8vxMXF6bVPnToVa9asgUajgUqlQmBgILRaLfbs2dPpcoYPHw4AKCoqsnjM1oqTGmMmKC4uRn19PSZMmKDXPnHiRNjb2+vdHjS3CRMmQKFQdLtmnFgqKytBRFAoFEb1j4uLw4gRI7Bz506cPn3aYHpP98GjtfssUZ/PFLm5ucjOzsZXX31lMIBm/fr12L17N06ePIn6+nr88ssvUKvVCAoK0htQ0qFjG9++fdvicVsrTmqMmaCmpgbAb6PXHuXq6gqtVmvR9Ts4OKCqqsqi6zC35uZmAHjswImHyeVyZGRkQBAELF26FE1NTXrTzb0PGhoaAAAbNmzQe6nBtWvX9AZtWMKBAweQkJCAgoICXRHcDr/++isSExPx9ttvY/r06XBycoKPjw/S09Nx69YtJCUlGSzP0dERwO/b/GnESY0xE7i6ugJApyfOmpoaDB482GLrbm1ttfg6LKHjRGvKw8FBQUFYsWIFSkpKsHnzZr1p5t4HlqjPZ4wdO3YgMzMTp06dwsCBAw2ml5SUoK2tzWCaSqWCu7s7iouLDeZpaWkB8Ps2fxpxUmPMBIGBgXB2dsYPP/yg13727Fm0tLTg+eef17XZ2dnpbnGZQ0FBAYgIL774osXWYQn9+/eHIAgmP3+2efNmjBw5EufPn9drN2UfGKO36/MREdasWYOioiLk5eV1esUJQJecHy2npNVqcffuXd3Q/od1bGNPT08zR913cFJjzARyuRwrV65Ebm4uMjMzUVdXh6KiIrz77rvw8vJCZGSkrq+/vz/u3r2LvLw8tLa2oqqqyuDZIgBwd3fHrVu3UFZWBq1Wq0tS7e3tuHfvHh48eIALFy4gKioK3t7eWLJkiVnWYWpdve5SKBTw9fXFjRs3TJqv4zbko89zmbIPjF3Pk+rzhYeHw9PT0yyv6bp06RI++eQTpKenQyaTGbzLddu2bQAAHx8fTJs2Denp6fj222/R1NSE8vJy3fd78803DZbdsY0tURS4zxBx5KVZQUJDUlnv6c6Q/vb2dkpKSqLhw4eTTCYjNzc3CgkJocuXL+v1u3PnDk2bNo3kcjn5+PjQ+++/T6tXryYA5O/vrxuaf+7cORo6dCg5OjrS5MmTqaKigiIjI0kmk9GgQYPIzs6OVCoVzZ07l65evWq2dRhTV68rph5vGo2GZDIZNTY26tpyc3PJz8+PAJCHhwctW7as03lXr15tMKTfmH1gbO0+oifX5wsJCSEAFBMT89jveebMGZo0aRJ5eXkRAAJAAwYMILVaTd988w0RERUVFemmdfZJSkrSLa+6upqioqLI39+fHBwcyNnZmSZNmkRffPFFp+sPDg6mQYMGUXt7+2PjfJSUzp+c1NhTzVrf/RgZGUnu7u5ih9ElU4+3kpISsrOzo3379lkwKstpa2ujKVOm0N69e8UOpUvV1dUkl8tp27ZtJs8rpfMn335kzEpJ6a3r/v7+iI2NRWxsbKdvordmbW1tyMvLg1arteoyT5s2bcK4ceOg0WjEDkVUnNQYY70iOjoaCxYsQHh4eJ96aXFBQQEOHTqE/Px8o5+1623JyckoLCzEsWPHIJPJxA5HVJzUGLMy69atQ0ZGBmpra+Hj44OcnByxQzKbLVu2QKPRYOvWrWKHYrQZM2bgs88+03vHpjU5fPgw7t+/j4KCAri5uYkdjujsxA6AMaYvPj4e8fHxYodhMTNnzsTMmTPFDkMy5syZgzlz5ogdhtXgKzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEmGpAaKbN++HQcPHhQ7DNaHdLxWaMGCBSJH0vfw8caskUAkjbrffFJi1iw/Px//9m//ZrXDwhlbsWIFgoKCxA6jxyST1BizZoIgICsrC2FhYWKHwpik8W9qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmOMMcngpMYYY0wyOKkxxhiTDDuxA2BMampqakBEBu0NDQ24d++eXpuzszNkMllvhcaY5AnU2dHHGOu26dOn43/+53+e2M/W1hY3b96Ep6dnL0TF2NOBbz8yZmavvfYaBEF4bB8bGxv88Y9/5ITGmJlxUmPMzObPnw87u8ff2RcEAa+//novRcTY04OTGmNm5ubmhpkzZ8LW1rbLPjY2NggJCenFqBh7OnBSY8wCIiIi0N7e3uk0Ozs7BAcHw8XFpZejYkz6OKkxZgGzZ8+Gg4NDp9Pa2toQERHRyxEx9nTgpMaYBSgUCoSEhHQ6XN/R0RGvvPKKCFExJn2c1BizkD//+c9obW3Va5PJZJg/fz4cHR1FiooxaeOkxpiFvPzyywa/m7W2tuLPf/6zSBExJn2c1BizEJlMhvDwcNjb2+vaXF1dMWPGDBGjYkzaOKkxZkGvvfYaWlpaAPyW5CIiIp74DBtjrPv4NVmMWVB7ezsGDhyI27dvAwBOnz6NSZMmiRwVY9LFV2qMWZCNjQ0WL14MAPDy8oJarRY5IsakTfT7INnZ2WKHwJhFeXh4AABeeOEFHDx4UORoGLMstVqNwYMHi7Z+0W8/PunFr4wxxvqOrKwshIWFibZ+0a/UAPE3AmM9lZ2djYULF3ZaRw0AcnJyMH/+/F6OyvoJgsDHv4RYw0UK/6bGWC/ghMZY7+CkxhhjTDI4qTHGGJMMTmqMMcYkg5MaY4wxyeCkxhhjTDI4qTFmRY4dOwYXFxd8+eWXYodi9U6cOIHo6GgcOnQIvr6+EAQBgiDo3uDysJkzZ0KpVMLW1hajR4/GuXPnRIjYdO3t7di+fXuXb6KJjY1FQEAAVCoVHBwc4O/vjw8//BD19fUGfT///HNMnDgRSqUSQ4cOxRtvvIGKigrd9CNHjiAxMRFtbW0W+z69gZMaY1aEX8VqnI8++gipqalYt24dQkND8csvv8DPzw/PPPMMMjMz8c9//lOv/9dff42DBw/i1VdfRXFxMcaPHy9S5MYrKSnBH//4R6xYsQKNjY2d9jl16hSWLVuGsrIyVFdXIz4+HikpKViwYIFev6ysLCxatAgLFizAjRs3cPjwYXz77beYNWsWHjx4AOC3au1yuRwzZsxATU2Nxb+fpXBSY8yKBAcHo7a2Fq+++qrYoaCpqckq31WZkJCAAwcOIDs7G0qlUm9aamoqbGxsEBkZidraWpEi7Lkff/wRa9euxbvvvotx48Z12c/Z2RmRkZFwd3eHUqlEWFgYQkJCcPz4cZSXl+v6/f3vf8fAgQOxevVquLi4YNy4cVixYgUKCwtx9uxZXb8PPvgAY8eOxSuvvKJLdn0NJzXGWKf27t2LyspKscPQc+XKFWzcuBEff/wx5HK5wXS1Wo2oqCjcvHkTq1atEiFC8xg7diwOHTqERYsWwcHBoct+R48eha2trV5bx7tGH766Ky8vh5eXl94bP4YMGQIAuHbtmt78mzZtQmFhIVJSUnr8PcTASY0xK3H69Gl4e3tDEAR8+umnAL9mSeEAACAASURBVIC0tDQ4OTlBoVDg8OHDmDVrFlQqFQYPHoz9+/fr5k1NTYVcLkf//v3xzjvvwMvLC3K5HGq1Wu8vcY1GA3t7ewwYMEDX9t5778HJyQmCIKC6uhoAEBUVhZUrV+Lq1asQBAH+/v4AgOPHj0OlUmHLli29sUkMpKamgogwe/bsLvvExcXh2WefxZ49e3DixInHLo+IkJycjFGjRsHBwQFubm6YO3cufvrpJ10fY/cBALS1tSEmJgbe3t5wdHTEc889h6ysrJ59aRPdvHkTjo6O8PHx0bX5+voa/IHS8Xuar6+vXrubmxumTp2KlJSUvnk7nEQGgLKyssQOg7EeycrKInMcTuXl5QSAduzYoWtbv349AaCTJ09SbW0tVVZW0pQpU8jJyYlaWlp0/SIjI8nJyYkuXbpEzc3NVFxcTBMnTiSlUknXr1/X9Vu0aBF5enrqrTcpKYkAUFVVla4tNDSU/Pz89PodPXqUlEolxcbG9vi7Epl+/Pv6+lJAQECn0/z8/Ki0tJSIiL777juysbGhYcOGUX19PRER5efn05w5c/TmiYmJIXt7e9q3bx/V1NTQhQsXaPz48eTh4UEVFRW6fsbug1WrVpGDgwPl5OTQvXv3aN26dWRjY0Pff/+90d/xUS+88AKNHTvWqL4NDQ2kVCpJo9HotRcUFJBMJqPU1FSqq6ujixcv0qhRo+jll1/udDnR0dEEgM6fP29SrNZwPucrNcb6CLVaDZVKhX79+iE8PBwNDQ24fv26Xh87OzvdVUdAQADS0tKg1WqRkZFhlhiCg4NRV1eHjRs3mmV5pmhoaEBpaSn8/Pye2DcoKAjLly9HWVkZ1q5d22mfpqYmJCcnY968eYiIiICLiwvGjBmDXbt2obq6Grt37zaY53H7oLm5GWlpaQgJCUFoaChcXV2xYcMGyGQys23/J4mPj4eXlxfi4uL02qdOnYo1a9ZAo9FApVIhMDAQWq0We/bs6XQ5w4cPBwAUFRVZPGZz46TGWB9kb28PAGhtbX1svwkTJkChUOjdTuurKisrQURQKBRG9Y+Li8OIESOwc+dOnD592mB6cXEx6uvrMWHCBL32iRMnwt7eXu+2bWce3QeXL19GY2MjAgMDdX0cHR0xYMCAXtn+ubm5yM7OxldffWUwgGb9+vXYvXs3Tp48ifr6evzyyy9Qq9UICgrSG1DSoWMbd1Rs70s4qTEmcQ4ODqiqqhI7jB5rbm4GgMcOnHiYXC5HRkYGBEHA0qVL0dTUpDe9Y9i6s7Ozwbyurq7QarUmxdfQ0AAA2LBhg+6ZOUEQcO3atS6H5JvLgQMHkJCQgIKCAgwbNkxv2q+//orExES8/fbbmD59OpycnODj44P09HTcunULSUlJBstzdHQE8Ps270s4qTEmYa2traipqRG1ErG5dJxoTXk4OCgoCCtWrEBJSQk2b96sN83V1RUAOk1e3dlm/fr1AwBs374dRKT3OXPmjEnLMsWOHTuQmZmJU6dOYeDAgQbTS0pK0NbWZjBNpVLB3d0dxcXFBvO0tLQA+H2b9yWc1BiTsIKCAhARXnzxRV2bnZ3dE29bWqP+/ftDEASTnz/bvHkzRo4cifPnz+u1BwYGwtnZGT/88INe+9mzZ9HS0oLnn3/epPUMGTIEcrkchYWFJs3XXUSENWvWoKioCHl5eZ1ecQLQJedff/1Vr12r1eLu3bu6of0P69jGnp6eZo7a8jipMSYh7e3tuHfvHh48eIALFy4gKioK3t7eWLJkia6Pv78/7t69i7y8PLS2tqKqqsrgWSUAcHd3x61bt1BWVgatVovW1lbk5+eLNqRfoVDA19cXN27cMGm+jtuQjz7PJZfLsXLlSuTm5iIzMxN1dXUoKirCu+++Cy8vL0RGRpq8njfeeAP79+9HWloa6urq0NbWhhs3bugSSnh4ODw9Pc3ymq5Lly7hk08+QXp6OmQymd4tT0EQsG3bNgCAj48Ppk2bhvT0dHz77bdoampCeXm57vu9+eabBsvu2MZjxozpcZy9TsSRl0RkHUNAGespcwzp37FjBw0YMIAAkEKhoNmzZ9POnTtJoVAQABo+fDhdvXqVdu/eTSqVigDQ0KFD6eeffyai34b0y2QyGjRoENnZ2ZFKpaK5c+fS1atX9dZz584dmjZtGsnlcvLx8aH333+fVq9eTQDI399fN/z/3LlzNHToUHJ0dKTJkydTRUUFHTt2jJRKJcXFxfXou3Yw9fjXaDQkk8mosbFR15abm0t+fn4EgDw8PGjZsmWdzrt69WqDIf3t7e2UlJREw4cPJ5lMRm5ubhQSEkKXL1/W9TFlH9y/f5/WrFlD3t7eZGdnR/369aPQ0FAqLi4mIqKQkBACQDExMY/9nmfOnKFJkyaRl5cXASAANGDAAFKr1fTNN98QEVFRUZFuWmefpKQk3fKqq6spKiqK/P39ycHBgZydnWnSpEn0xRdfdLr+4OBgGjRoELW3tz82zkdZw/mckxpjZmCu59R6IjIyktzd3UWNwVSmHv8lJSVkZ2dH+/bts2BUltPW1kZTpkyhvXv3ih1Kl6qrq0kul9O2bdtMntcazud8+5ExCenrb1h/En9/f8TGxiI2NrbTN9Fbs7a2NuTl5UGr1SI8PFzscLq0adMmjBs3DhqNRuxQuqXPJ7W33noLSqUSgiD02g+0fUVzczNGjhyJDRs2mDzvo+U8Oj729vbo378/XnrpJSQlJeHevXsWiJyxrkVHR2PBggUIDw/vUy8tLigowKFDh5Cfn2/0s3a9LTk5GYWFhTh27BhkMpnY4XRLn09qe/bsQXp6uthhWKX169fj8uXL3Zr34XIeLi4uICK0t7ejsrIS2dnZ8PHxwZo1azB69GiD0WOs961btw4ZGRmora2Fj48PcnJyxA7JorZs2QKNRoOtW7eKHYrRZsyYgc8++0zvvZvW5PDhw7h//z4KCgrg5uYmdjjd1ueTmtSYq9zHd999h4sXL5ohot8JggBXV1e89NJLyMjIQHZ2Nm7fvq0rl9LXWWupFWPEx8fj/v37ICKUlpZi/vz5YodkcTNnzkRCQoLYYUjGnDlzEB0dbTBKtK+RRFJ7uJxCX2eOch9NTU1YvXq1xUtHzJ8/H0uWLEFlZSV27dpl0XX1BmsstcIYM02fS2pEhKSkJIwYMQIODg5wcXHB6tWr9fp88sknUCgUUCqVqKysxMqVKzFo0CBcvnzZqFITxpbx6IjnScvrabkPU61fvx7vvfee7g0HjzJn+ZCO55/y8/MB8LZnjIlMxJGXRGT6END169eTIAj0t7/9je7du0eNjY20c+dOgzIJHaUiPvjgA9qxYwfNmzeP/t//+39Gl5owtoyHscvrSbkPU5w+fZpmz55NRERVVVUEgNavX6/Xx5TyIX5+fuTi4tLl9Lq6OgJAQ4YM0bU9jdveGob090WmHv/MulnD/hT9KDRlIzQ2NpJCoaA//elPeu379+/vMqk1NTXpze/s7Ezh4eF68//rX/8iAHon+cjISIOT+ffff08A6OOPPzZ5eb1xYm1sbKQJEybQjRs3iKjrpGaKJyU1IiJBEMjV1VX3/0/jtuek1j3WcBJk5mMN+9Ouly4IzeLKlStobGzEjBkzujV/T0tNPFrGo6fLM7d169bh7bffxqBBg3ptnQ0NDSAiqFSqx/aT+rbvsGDBAlHW25dt374dBw8eFDsMJhF96je1jveRdfVb0ZOYo9TEw2U8zF26oidOnz6NoqIivPXWW722TgD4+eefAQAjR458bD8pb3vGmPXoU1dqcrkcAHD//v1uzd/TUhOPlvEwd+mKnti7dy9OnjwJGxvDv1O2bNmCLVu24Pvvvze4sump48ePAwBmzZr12H5S3vYP4ysO0wiCgOXLlyMsLEzsUJgZWMNI9D51pRYYGAgbGxt888033Z6/J6UmHi3jYcryLF3uIyMjw6CGU8dVzfr160FEZk9oFRUV2L59OwYPHoylS5c+tq+Utz1jzHr0qaTWr18/hIaGIicnB3v37kVdXR0uXLiA3bt3GzW/qaUmnlTGw5Tl9aTch7mZWj6EiFBfX4/29nZdsszKysKkSZNga2uLvLy8J/6mxtueMdYrxBqh0gEmjpbRarX01ltv0TPPPEPOzs40efJkiomJIQA0ePBg+vHHHykxMZEcHR11Q80ffqO3MaUmiIwv42Hs8npS7qO7uhr9aEz5kCNHjtBzzz1HCoWC7O3tycbGhgDoRjr+4Q9/oNjYWLpz547efE/rtufRj91j6vHPrJs17E/h/wIRjSAIyMrKsrp76u+88w4OHjyIO3fuiB3KU6cvbvvs7GwsXLgQIh9OfY61Hv+se6xhf/ap24+9TeplPKwZb3vGWHdwUrNiP/30k0Hpl84+1lybiTFLOXHiBKKjow3KJC1evNig78yZM6FUKmFra4vRo0fj3LlzIkRsuvb2dmzfvr3LF23HxsYiICAAKpUKDg4O8Pf3x4cffthprbnPP/8cEydOhFKpxNChQ/HGG2+goqJCN/3IkSNITEzs+39Qinrzk6zjHuyjoqOjyd7engDQsGHD6ODBg2KH9NToq9uef1Prnu4e/zExMfTqq69SXV2drs3Pz4+eeeYZAkBHjx41mCc/P5/mzJnTo3h7088//0yTJk0iADR27NhO+0ydOpV27txJd+7cobq6OsrKyiKZTEb//u//rtfvwIEDBIASExOppqaGzp8/T76+vjRu3DhqbW3V9UtJSaGpU6fSvXv3uhWzNZzPRT8KrWEjMNZT1pDUGhsbKSgoqE+tozvH/9atW+nZZ5/Vew0b0W9J7bPPPiMbGxsaNGgQ1dTU6E3vS0mtsLCQ5s2bR5mZmTRu3Lguk1pwcDA9ePBAry0sLIwA6L0nddq0aTRw4EBqb2/XtX366acEgE6fPq03v0ajoaCgIL1kZyxrOJ/z7UfGJKI3SueIXZ7nypUr2LhxIz7++GPdyxgeplarERUVhZs3b2LVqlUiRGgeY8eOxaFDh7Bo0SI4ODh02e/o0aMG9c88PDwAAI2Njbq28vJyeHl56T0cPWTIEAAweLxl06ZNKCwstHjpKkvhpMaYSMjCpXOMLePT0/I85ixl9CSpqakgIsyePbvLPnFxcXj22WexZ88enDhx4rHLM2YfpKWlwcnJCQqFAocPH8asWbOgUqkwePBg7N+/X295bW1tiImJgbe3NxwdHfHcc88hKyurZ1/aRDdv3oSjoyN8fHx0bb6+vgZ/jHT8nubr66vX7ubmhqlTpyIlJaVvjuYV9TqRrONylbGe6s7tx94onWNsGZ+erMOUUkaPMvX49/X1pYCAgE6n+fn5UWlpKRERfffdd2RjY0PDhg2j+vp6Iur89qOx+6Cj8sTJkyeptraWKisracqUKeTk5EQtLS26fqtWrSIHBwfKycmhe/fu0bp168jGxoa+//57o7/jo1544YUubz8+qqGhgZRKJWk0Gr32goICkslklJqaSnV1dXTx4kUaNWoUvfzyy50uJzo62qDyiTGs4XzOV2qMiaCpqQnJycmYN28eIiIi4OLigjFjxmDXrl2orq42+i05xrCzs9NdiQQEBCAtLQ1arRYZGRlmWX5wcDDq6uqwceNGsyyvKw0NDSgtLYWfn98T+wYFBWH58uUoKyvD2rVrO+3TnX2gVquhUqnQr18/hIeHo6GhAdevXwcANDc3Iy0tDSEhIQgNDYWrqys2bNgAmUxmtm39JPHx8fDy8kJcXJxe+9SpU7FmzRpoNBqoVCoEBgZCq9Viz549nS5n+PDhAICioiKLx2xunNQYE4GYpXMeLePTV1RWVoKIoFAojOofFxeHESNGYOfOnTh9+rTB9J7uA3t7ewDQvU7t8uXLaGxsRGBgoK6Po6MjBgwY0CvbOjc3F9nZ2fjqq6+gVCr1pq1fvx67d+/GyZMnUV9fj19++QVqtRpBQUEoLy83WFbHNr59+7bF4zY3TmqMiUDs0jkPl/HpK5qbmwHgsQMnHiaXy5GRkQFBELB06VI0NTXpTTf3PmhoaAAAbNiwQe850mvXrukN2rCEAwcOICEhAQUFBRg2bJjetF9//RWJiYl4++23MX36dDg5OcHHxwfp6em4desWkpKSDJbn6OgI4Pdt3pdwUmNMBGKWznm0jE9f0XGiNeXh4KCgIKxYsQIlJSXYvHmz3jRz74OOOo/bt283qJhx5swZk5Zlih07diAzMxOnTp3CwIEDDaaXlJSgra3NYJpKpYK7uzuKi4sN5mlpaQHw+zbvSzipMSYCMUvnPFrGxxLrsIT+/ftDEATU1taaNN/mzZsxcuRInD9/Xq+9p+WQHjVkyBDI5XIUFhaaNF93ERHWrFmDoqIi5OXldXrFCUCXnH/99Ve9dq1Wi7t37+qG9j+sYxt7enqaOWrL46TGmAh6s3TOk8r49HQdppYy6i6FQgFfX1/cuHHDpPk6bkM++jyXqeWQjFnPG2+8gf379yMtLQ11dXVoa2vDjRs3dAklPDwcnp6eZnlN16VLl/DJJ58gPT0dMpnM4PV527ZtAwD4+Phg2rRpSE9Px7fffoumpiaUl5frvt+bb75psOyObTxmzJgex9nrRBx5SUTWMQSUsZ7qzpD+3iidY2wZn56sw5hSRl0x9fjXaDQkk8mosbFR15abm0t+fn4EgDw8PGjZsmWdzrt69WqDIf3G7IOdO3eSQqEgADR8+HC6evUq7d69m1QqFQGgoUOH0s8//0xERPfv36c1a9aQt7c32dnZUb9+/Sg0NJSKi4uJiCgkJIQAUExMzGO/55kzZ2jSpEnk5eVFAAgADRgwgNRqNX3zzTdERFRUVKSb1tknKSlJt7zq6mqKiooif39/cnBwIGdnZ5o0aRJ98cUXna4/ODiYBg0apPcGEmNYw/mckxpjZmANr8nqTGRkJLm7u4sdRpdMPf5LSkrIzs5Or05fX9LW1kZTpkyhvXv3ih1Kl6qrq0kul9O2bdtMntcazud8+5Exievzb11/iL+/P2JjYxEbG9vpm+itWVtbG/Ly8qDVaq26ssamTZswbtw4aDQasUPpFk5qjLE+JTo6GgsWLEB4eLjJg0bEVFBQgEOHDiE/P9/oZ+16W3JyMgoLC3Hs2DHIZDKxw+kWTmqMSdS6deuQkZGB2tpa+Pj4ICcnR+yQzGbLli3QaDTYunWr2KEYbcaMGfjss8/03rFpTQ4fPoz79++joKAAbm5uYofTbXZiB8AYs4z4+HjEx8eLHYbFzJw5EzNnzhQ7DMmYM2cO5syZI3YYPcZXaowxxiSDkxpjjDHJ4KTGGGNMMjipMcYYkwxOaowxxiRD+L+nwMULQBDEXD1jjDEzysrKQlhYmGjrF31If1ZWltghMGZxCxcuRFRUFIKCgsQOhTGLUqvVoq5f9Cs1xp4GgiCI/hcsY08D/k2NMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSYSd2AIxJzf79+6HVag3aT5w4gZqaGr22kJAQ9OvXr7dCY0zyBCIisYNgTEqWLFmCf/zjH5DJZLq2jsNMEAQAQFtbG5ydnVFZWQkHBwdR4mRMivj2I2Nm9tprrwEAWltbdZ8HDx7gwYMHuv+3tbXFggULOKExZmZ8pcaYmT148ACenp64e/fuY/udPHkS06dP76WoGHs68JUaY2ZmZ2eH1157Te/246M8PDwwderUXoyKsacDJzXGLOC1115Da2trp9NkMhkWL14MW1vbXo6KMenj24+MWQARwdvbGzdu3Oh0+r/+9S9MnDixl6NiTPr4So0xCxAEAREREZ3eghwyZAgmTJggQlSMSR8nNcYspLNbkDKZDEuWLNEN7WeMmRfffmTMgkaOHInLly/rtV28eBGjR48WKSLGpI2v1BizoMWLF+vdggwICOCExpgFcVJjzIIiIiLw4MEDAL/devzLX/4ickSMSRvffmTMwiZMmID//d//hSAIKCsrg7e3t9ghMSZZfKXGmIW9/vrrAIAXXniBExpjFiaZt/QvWLBA7BAY61RzczMEQcD9+/f53ymzWitWrEBQUJDYYfSYZK7UcnJyunzQlbGu3LhxAzk5ORZdh1wuh6enJwYPHmzR9fQmPt6kJScnB+Xl5WKHYRaSuVIDgOXLlyMsLEzsMFgfkp2djYULF+LgwYMWXc+VK1fg7+9v0XX0JkEQ+HiTECk9NymZKzXGrJmUEhpj1oyTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmNmcOzYMbi4uODLL78UOxSrd+LECURHR+PQoUPw9fWFIAgQBAGLFy826Dtz5kwolUrY2tpi9OjROHfunAgRm669vR3bt2+HWq3udHpsbCwCAgKgUqng4OAAf39/fPjhh6ivrzfo+/nnn2PixIlQKpUYOnQo3njjDVRUVOimHzlyBImJiWhra7PY9+lLOKkxZgb8tjnjfPTRR0hNTcW6desQGhqKX375BX5+fnjmmWeQmZmJf/7zn3r9v/76axw8eBCvvvoqiouLMX78eJEiN15JSQn++Mc/YsWKFWhsbOy0z6lTp7Bs2TKUlZWhuroa8fHxSElJMXg4PysrC4sWLcKCBQtw48YNHD58GN9++y1mzZqle6fo7NmzIZfLMWPGDNTU1Fj8+1k7TmqMmUFwcDBqa2vx6quvih0KmpqaurxCEFNCQgIOHDiA7OxsKJVKvWmpqamwsbFBZGQkamtrRYqw53788UesXbsW7777LsaNG9dlP2dnZ0RGRsLd3R1KpRJhYWEICQnB8ePH9R6C/vvf/46BAwdi9erVcHFxwbhx47BixQoUFhbi7Nmzun4ffPABxo4di1deeUWX7J5WnNQYk5i9e/eisrJS7DD0XLlyBRs3bsTHH38MuVxuMF2tViMqKgo3b97EqlWrRIjQPMaOHYtDhw5h0aJFcHBw6LLf0aNHYWtrq9fm4eEBAHpXd+Xl5fDy8tJ7OHrIkCEAgGvXrunNv2nTJhQWFiIlJaXH36Mv46TGWA+dPn0a3t7eEAQBn376KQAgLS0NTk5OUCgUOHz4MGbNmgWVSoXBgwdj//79unlTU1Mhl8vRv39/vPPOO/Dy8oJcLodardb7S1yj0cDe3h4DBgzQtb333ntwcnKCIAiorq4GAERFRWHlypW4evUqBEHQPfR9/PhxqFQqbNmypTc2iYHU1FQQEWbPnt1ln7i4ODz77LPYs2cPTpw48djlERGSk5MxatQoODg4wM3NDXPnzsVPP/2k62PsPgCAtrY2xMTEwNvbG46OjnjuueeQlZXVsy9tops3b8LR0RE+Pj66Nl9fX4M/UDp+T/P19dVrd3Nzw9SpU5GSkvJ03w4niQBAWVlZYofB+pisrCwyx2FQXl5OAGjHjh26tvXr1xMAOnnyJNXW1lJlZSVNmTKFnJycqKWlRdcvMjKSnJyc6NKlS9Tc3EzFxcU0ceJEUiqVdP36dV2/RYsWkaenp956k5KSCABVVVXp2kJDQ8nPz0+v39GjR0mpVFJsbGyPvyuR6cebr68vBQQEdDrNz8+PSktLiYjou+++IxsbGxo2bBjV19cTEVF+fj7NmTNHb56YmBiyt7enffv2UU1NDV24cIHGjx9PHh4eVFFRoetn7D5YtWoVOTg4UE5ODt27d4/WrVtHNjY29P333xv9HR/1wgsv0NixY43q29DQQEqlkjQajV57QUEByWQySk1Npbq6Orp48SKNGjWKXn755U6XEx0dTQDo/PnzJsUqpfMnX6kxZmFqtRoqlQr9+vVDeHg4GhoacP36db0+dnZ2uquOgIAApKWlQavVIiMjwywxBAcHo66uDhs3bjTL8kzR0NCA0tJS+Pn5PbFvUFAQli9fjrKyMqxdu7bTPk1NTUhOTsa8efMQEREBFxcXjBkzBrt27UJ1dTV2795tMM/j9kFzczPS0tIQEhKC0NBQuLq6YsOGDZDJZGbb/k8SHx8PLy8vxMXF6bVPnToVa9asgUajgUqlQmBgILRaLfbs2dPpcoYPHw4AKCoqsnjM1oqTGmO9yN7eHgDQ2tr62H4TJkyAQqHQu53WV1VWVoKIoFAojOofFxeHESNGYOfOnTh9+rTB9OLiYtTX12PChAl67RMnToS9vb3ebdvOPLoPLl++jMbGRgQGBur6ODo6YsCAAb2y/XNzc5GdnY2vvvrKYADN+vXrsXv3bpw8eRL19fX45ZdfoFarERQU1Olb9Tu28e3bty0et7XipMaYlXJwcEBVVZXYYfRYc3MzADx24MTD5HI5MjIyIAgCli5diqamJr3pHcPWnZ2dDeZ1dXWFVqs1Kb6GhgYAwIYNG3TPzAmCgGvXrnU5JN9cDhw4gISEBBQUFGDYsGF603799VckJibi7bffxvTp0+Hk5AQfHx+kp6fj1q1bSEpKMlieo6MjgN+3+dOIkxpjVqi1tRU1NTWSqMHWcaI15eHgoKAgrFixAiUlJdi8ebPeNFdXVwDoNHl1Z5v169cPALB9+3YQkd7nzJkzJi3LFDt27EBmZiZOnTqFgQMHGkwvKSlBW1ubwTSVSgV3d3cUFxcbzNPS0gLg923+NOKkxpgVKigoABHhxRdf1LXZ2dk98balNerfvz8EQTD5+bPNmzdj5MiROH/+vF57YGAgnJ2d8cMPP+i1nz17Fi0tLXj++edNWs+QIUMgl8tRWFho0nzdRURYs2YNioqKkJeX1+kVJwBdvxaVHQAAIABJREFUcv7111/12rVaLe7evasb2v+wjm3s6elp5qj7Dk5qjFmB9vZ23Lt3Dw8ePMCFCxcQFRUFb29vLFmyRNfH398fd+/eRV5eHlpbW1FVVWXwrBIAuLu749atWygrK4NWq0Vrayvy8/NFG9KvUCjg6+trcqXsjtuQjz7PJZfLsXLlSuTm5iIzMxN1dXUoKirCu+++Cy8vL0RGRpq8njfeeAP79+9HWloa6urq0NbWhhs3bugSSnh4ODw9Pc3ymq5Lly7hk08+QXp6OmQymd4tT0EQsG3bNgCAj48Ppk2bhvT0dHz77bdoampCeXm57vu9+eabBsvu2MZjxozpcZx9logjL80KEhqSynqPOYb079ixgwYMGEAASKFQ0OzZs2nnzp2kUCgIAA0fPpyuXr1Ku3fvJpVKRQBo6NCh9PPPPxPRb0P6ZTIZDRo0iOzs7EilUtHcuXPp6tWreuu5c+cOTZs2jeRyOfn4+ND7779Pq1evJgDk7++vG/5/7tw5Gjp0KDk6OtLkyZOpoqKCjh07RkqlkuLi4nr0XTuYerxpNBqSyWTU2Nioa8vNzSU/Pz8CQB4eHrRs2bJO5129erXBkP729nZKSkqi4cOHk0wmIzc3NwoJCaHLly/r+piyD+7fv09r1qwhb29vsrOzo379+lFoaCgVFxcTEVFISAgBoJiYmMd+zzNnztCkSZPIy8uLABAAGjBgAKnVavrmm2+IiKioqEg3rbNPUlKSbnnV1dUUFRVF/v7+5ODgQM7OzjRp0iT64osvOl1/cHAwDRo0iNrb2x8b56OkdP7kpMaeauZ6Tq0nIiMjyd3dXdQYTGXq8VZSUkJ2dna0b98+C0ZlOW1tbTRlyhTau3ev2KF0qbq6muRyOW3bts3keaV0/uTbj4xZAam/Yd3f3x+xsbGIjY3t9E301qytrQ15eXnQarUIDw8XO5wubdq0CePGjYNGoxE7FFFxUmOM9Yro6GgsWLAA4eHhfeqlxQUFBTh06BDy8/ONftautyUnJ6OwsBDHjh2DTCYTOxxRcVL7P2+99RaUSiUEQei1UVDm9tJLLxn86Nzx6WqEVVcerXXV8bG3t0f//v3x0ksvISkpCffu3bPQt3k6rFu3DhkZGaitrYWPjw9ycnLEDsmitmzZAo1Gg61bt4oditFmzJiBzz77TO+9m9bk8OHDuH//PgoKCuDm5iZ2OKLjpPZ/9uzZg/T0dLHDsJjJkyeb1P/hWlcuLi4gIrS3t6OyshLZ2dnw8fHBmjVrMHr0aIOh1cx48fHxuH//PogIpaWlmD9/vtghWdzMmTORkJAgdhiSMWfOHERHRxuMEn1acVKTELlcjrq6OoMHSCMjI/Hhhx/2ePmCIMDV1RUvvfQSMjIykJ2djdu3b+tqiTHGmNg4qT3k4ZpFfdHx48cN3h1XXl6OixcvYvr06WZf3/z587FkyRJUVlZi165dZl8+Y4yZ6qlNakSEpKQkjBgxAg4ODnBxccHq1asN+j2uzpIp9Zq++eYb/OEPf4BCoYBKpcKYMWNQV1f3xHX0VEJCAj744AO9NnPW1up4ODg/P1/X1te3GWOsDxPxcQKzgonPWaxfv54EQaC//e1vdO/ePWpsbKSdO3ca1CJ6Up0lY+o11dfXk0qlosTERGpqaqKKigqaN2+ergaWJWo5ERHduHGDAgICqK2tTa/dlNpafn5+5OLi0uX0uro6AkBDhgzRtfWlbWYNz6n1RaYeb8y6SWl/SuZoNmWnNDY2kkKhoD/96U967fv379dLak1NTaRQKCg8PFxvXgcHB/rP//xPIvr9BN3U1KTr05Ecr1y5QkREFy9eJAB09OhRg1iMWUd3LVu2jP7rv/6rR8t4UlIjIhIEgVxdXYmo720zTmrdI6WTIJPW/nwqbz9euXIFjY2NmDFjxmP7dbfO0qP1mnx9fdG/f39ERERg06ZNKCsr6/E6nuTWrVs4cuSI3rsDLaGhoQFEBJVKBaDvbrOuHoXgT+cfAFi4cKHocfDHfPtTKuzEDkAMHS/97Cg50ZWH6yxt2LBBb5qXl5fR63N0dMSpU6ewdu1abNmyBbGxsQgLC0NGRobZ1vGoxMRE/PWvf4VcLu/2Mozx888/AwBGjhwJoO9uM/49zjQLFy5EVFQUgoKCxA6FmcHChQvFDsFsnsqk1nGiv3///mP7PVxnKSoqqkfrHD16NL788ktUVVUhOTkZCQkJGD16tO61O+ZYR4eKigp8/vnnuHz5slmW9zjHjx8HAMyaNQtA391mYWFhZlnO02LhwoUICgri7SYRUkpqT+Xtx8DAQNjY2OCbb755bD9z1Vm6desWLl26BOC3k/7WrVsxfvx4XLp0ySK1nBITExEREQF3d3ezLbMzFRUV2L59OwYPHoylS5cC6LvbjDEmDU9lUuvXrx9CQ0ORk5ODvXv3oq6uDhcuXMDu3bv1+hlTZ8kYt27dwjvvvIOffvoJLS0tOH/+PK5du4YXX3zRbOvocPv2bfz3f/83li9f3mUfU2trERHq6+vR3t4OIkJVVRWysrIwadIk2NraIi8vT/ebWl/cZowxCRF5oIrZwMTRO1qtlt566y165plnyNnZmSZPnkwxMTEEgAYPHkw//vgjET2+zpKx9ZrKyspIrVaTm5sb2dra0sCBA2n9+vX04MGDJ67DVCtWrKCIiIjH9jGmttaRI0foueeeI4VCQfb29mRjY0MAdCMd//CHP1BsbCzduXPHYN6+tM149GP3mHq8Mesmpf0pEBGJlVDNSRAEZGVl8T1+ZpLs7GwsXLgQEjkMeg0fb9Iipf35VN5+ZIwxJk2c1KzYTz/9ZNQzJtZcuJAxY5w4cQLR0dEGJY8WL15s0HfmzJlQKpWwtbXF6NGjce7cOREiNp6xJaFiY2MREBAAlUoFBwcH+Pv748MPP9QrqnrkyBEkJiZKvqhsTzyVQ/r7ipEjR/JtMSZ5H330Ec6fP4/PPvsMSqUSoaGh8Pf3R01NDTIzMxEeHo7g4GBd/6+//hrHjx/Hrl27kJeXJ2LkPfdwSahTp05h2bJlCA8Ph0wmQ35+PiIiIlBUVKR7t+rs2bNRWlqKGTNmIC8vD66urmKFbrX4So0xkTU1NUGtVvf5dXRHQkICDhw4gOzsbIMKE6mpqbCxsUFkZGSfLm1kbEkoZ2dnREZGwt3dHUqlEmFhYQgJCcHx48dRXl6u6/fBBx9g7NixeOWVV/DgwQMxvpJV46TGmMj27t2LysrKPr8OU125cgUbN27Exx9/3Ombb9RqNaKionDz5k2sWrVKhAjNw9iSUEePHjUo9Onh4QEAaGxs1GvftGkTCgsLkZKSYqGo+y5OaoyZiIiQnJyMUaNGwcHBAW5ubpg7d67eeyc1Gg3s7e0xYMAAXdt7770HJycnCIKA6upqAEBUVBRWrlyJq1evQhAE+Pv7IzU1FXK5HP3798c777wDLy8vyOVyqNVqnD171izrAMxbgqg7UlNTQUSYPXt2l33i4uLw7LPPYs+ePThx4sRjl2fMfjGl9FFvl4TqzM2bN+Ho6AgfHx+9djc3N0ydOhUpKSn8E8WjxHmSwPwgoecsWO/pznNqMTExZG9vT/v27aOamhq6cOECjR8/njw8PKiiokLXb9GiReTp6ak3b1JSEgHQldAhIgoNDSU/Pz+9fpGRkeTk5ESXLl2i5uZmKi4upokTJ5JSqaTr16+bZR2mlCB6lDmON19fXwoICOh0mp+fH5WWlhIR0XfffUc2NjY0bNgwqq+vJyKi/Px8mjNnjt48xu4XY0ofEfV+SahHNTQ0kFKpJI1G0+n06Ohog1JZ3SWl8ydfqTFmgqamJiQnJ2PevHmIiIiAi4sLxowZg127dqG6utrgrTQ9YWdnp7vqCAgIQFpaGrRaLTIyMsyy/ODgYNTV1WHjxo1mWZ4pGhoaUFpaCj8/vyf2DQoKwvLly1FWVoa1a9d22qc7+0WtVkOlUqFfv34IDw9HQ0MDrl+/DgBobm5GWloaQkJCEBoaCldXV2zYsAEymazH2z8hIQHvv/8+bGwef/qNj4+Hl5cX4uLiOp0+fPhwAEBRUVGP4pEaTmqMmaC4uBj19fWYMGGCXvvEiRNhb2+vd3vQ3CZMmACFQtGjkkTWorKyEkQEhUJhVP+4uDiMGDECO3fuxOnTpw2m93S/PFr6SOySULm5ucjOzsZXX31l8Htch45td/v27W7HI0Wc1BgzQU1NDQDoPV/UwdXVFVqt1qLrd3BwQFVVlUXX0Ruam5sB/PZ9jCGXy5GRkQFBELB06VI0NTXpTTf3fnm4vNHDz5Vdu3bNYNCGKYwpCXXgwAEkJCSgoKAAw4YN67Kfo6MjgN+3JfsNJzXGTNDxXFBnJ8mamhoMHjzYYutubW21+Dp6S8cJ2ZSHiIOCgrBixQqUlJRg8+bNetPMvV8eLqFEjwzFP3PmjEnL6tBREuo///M/u+yzY8cOZGZm4tSpUxg4cOBjl9fS0gLg923JfsNJjTETBAYGwtnZGT/88INe+9mzZ9HS0oLnn39e12ZnZ6e7nWUOBQUFICK8+OKLFltHb+nfvz8EQTD5+bPNmzdj5MiROH/+vF67KfvFGL1dEoqIsGbNGhQVFSEvL6/TK85HdWw7T09Ps8UoBZzUGDOBXC7HypUrkZubi8zMTNTV1aGoqAjvvvsuvLy8EBkZqevr7++Pu3fvIi8vD62traiqqsK1a9cMlunu7o5bt26hrKwMWq1Wl6Ta29tx7949PHjwABcuXEBUVBS8vb31fo/pyTpMLUFkTgqFAr6+vroq9MbquA356PNcpuwXY9fzpPJG4eHh8PT0NOo1XU8qCXXp0iV88sknSE9Ph0wmM3id1rZt2wzm6dh2Y8aMMem7SZ6IIy/NChIaksp6T3eG9Le3t1NSUhINHz6cZDIZubm5UUhICF2+fFmv3507d2jatGkkl8vJx8eH3n//fVq9ejUBIH9/f93Q/HPnztHQoUPJ0dGRJk+eTBUVFRQZGUkymYwGDRpEdnZ2pFKpaO7cuXT16lWzrcOYEkRdMcfxptFoSCaTUWNjo64tNzeX/Pz8CAB5eHjQsmXLOp139erVBkP6jdkvxpY+InpyeaOQkBACQDExMU/8rk8qCVVUVEQAuvwkJSUZzBMcHEyDBg2i9vb2J67/SaR0/uSkxp5q1lpPLTIyktzd3cUOo0vmON5KSkrIzs6O9u3bZ6aoeldbWxtNmTKF9u7d2+vrrq6uJrlcTtu2bTPL8qR0/uTbj4xZKam/id3f3x+xsbGIjY3VexN9X9DW1oa8vDxotVpRqmRs2rQJ48aNg0aj6fV1WztOaowx0URHR2PBggUIDw/vUy8tLigowKFDh5Cfn2/0s3bmkpycjMLCQhw7dgwymaxX190XcFJjzMqsW7cOGRkZqK2thY+PD3JycsQOyaK2bNkCjUaDrVu3ih2K0WbMmIHPPvtM772bveHw4cO4f/8+CgoK4Obm1qvr7iu4nhpjViY+Ph7x8fFih9GrZs6ciZkzZ4odhtWbM2cO5syZI3YYVo2v1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmSGijS3ReNsqdXx7+Z7OxskSPpe/h4Y9ZIIJJGLXBBEMQOgTHG+qysrCyEhYWJHUaPSeZKTSK5mUmUIAiSOWkwZs34NzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSwUmNMcaYZHBSY4wxJhmc1BhjjEkGJzXGGGOSIRARiR0EY1ISGRmJy5cv67WdO3cOPj4+cHNz07XZ2triH//4BwYPHtzbITImWXZiB8CY1Hh6emL37t0G7RcuXND7f19fX05ojJkZ335kzMz+/Oc/P7GPvb09lixZYvlgGHvK8O1HxiwgMDAQly5dwuMOr8uXL+PZZ5/txagYkz6+UmPMAl5//XXY2tp2Ok0QBIwdO5YTGmMWwEmNMQt47bXX0NbW1uk0W1tb/OUvf+nliBh7OvDtR8YsRK1W4+zZs2hvb9drFwQB5eXlGDRokEiRMSZdfKXGmIUsXrwYgiDotdnY2GDy5Mmc0BizEE5qjFnIggULDNoEQcDrr78uQjSMPR04qTFmIR4eHpgxY4begBFBEBASEiJiVIxJGyc1xiwoIiJCN6zf1tYWL7/8Mp555hmRo2JMujipMWZB8+bNg729PQCAiBARESFyRIxJGyc1xizIyckJ//Ef/wHgt7eIvPrqqyJHxJi0cVJjzMIWLVoEAAgJCYGTk5PI0TAmbVb3nFp2djYWLlwodhiMMcaeYP78+Th48KDYYeix2rf0Z2VliR0CYz2yfft2AMDy5cuRmZmJ8PBw2NlZ7SFnFc6cOYOUlBQ+/vuAjn/f1sZqj7CwsDCxQ2CsRzr+gg0LC8Ps2bMhl8tFjqhvSElJ4eO/D7C2K7QO/JsaY72AExpjvYOTGmOMMcngpMYYY0wyOKkxxhiTDE5qjDHGJIOTGmNW7tixY3BxccGXX34pdihW78SJE4iOjsahQ4fg6+sLQRAgCAIWL15s0HfmzJlQKpWwtbXF6NGjce7cOREiNt5LL72k+z6PfpydnXX9YmNjERAQAJVKBQcHB/j7++PDDz9EfX29rs+RI0eQmJjYZSHbvoyTGmNWzsrej2C1PvroI6SmpmLdunUIDQ3FL7/8Aj8/PzzzzDPIzMzEP//5T73+X3/9NQ4ePIhXX30VxcXFGD9+vEiR99zkyZN1/33q1CksW7YMZWVlqK6uRnx8PFJSUvRKIXU8YjJjxgzU1NSIEbLFcFJjzMoFBwejtrbWKt4b2dTUBLVaLXYYBhISEnDgwAFkZ2dDqVTqTUtNTYWNjQ0iIyNRW1srUoQ9J5fLUVdXByLS+0RGRuLDDz/U9XN2dkZkZCTc3d2hVCoRFhaGkJAQHD9+HOXl5bp+H3zwAcaOHYtXXnkFDx48EOMrWQQnNcaY0fbu3YvKykqxw9Bz5coVbNy4ER9//HGnzwOq1WpERUXh5s2bWLVqlQgRmsfx48cNEnZ5eTkuXryI6dOn69qOHj2qV8MP+K22HwA0NjbqtW/atAmFhYVISUmxUNS9j5MaY1bs9OnT8Pb2hiAI+PTTTwEAaWlpcHJygkKhwOHDhzFr1iyoVCoMHjwY+/fv182bmpoKuVyO/v3745133oGXlxfkcjnUajXOnj2r66fRaGBvb48BAwbo2t577z04OTlBEARUV1cDAKKiorBy5UpcvXoVgiDA398fwG8nW5VKhS1btvTGJjGQmpoKIsLs2bO77BMXF4dnn30We/bswYkTJx67PCJCcnIyRo0aBQcHB7i5uWHu3Ln46aefdH2M3QcA0NbWhpiYGHh7e8PR0RHPPfec2V4DlpCQgA8++OCJ/W7evAlHR0f4+Pjotbu5uWHq1KlISUmRzm1usjJZWVlkhWExZrL58+fT/Pnze7yc8vJyAkA7duzQta1fv54A0MmTJ6m2tpYqKytpypQp5OTkRC0tLbp+kZGR5OTkRJcuXaLm5mYqLi6miRMnklKppOvXr+v6LVq0iDw9PfXWm5SURACoqqpK1xYaGkp+fn56/Y4ePUpKpZJiY2N7/F27c/z7+vpSQEBAp9P8/PyotLSUiIi+++47srGxoWHDhlF9fT0REeXn59OcOXP05omJiSF7e3vat28f1dTU0IULF2j8+PHk4eFBFRUVun7G7oNVq1aRg4MD5eTk0L1792jdunVkY2ND33//vUnf81E3btyggIAAamtre2y/hoYGUiqVpNFoOp0eHR1NAOj8+fMmrd9c/77Nja/UGOvD1Go1VCoV+vXrh/DwcDQ0NOD69et6fezs7HRXHQEBAUhLS4NWq0VGRoZZYggODkZdXR02btxoluWZoqGhAaWlpfDz83ti36CgICxfvhxlZWVYu3Ztp32ampqQnJyMefPmISIiAi4uLhgzZgx27dqF6upq7N6922Cex+2D5uZmpKWlISQkBKGhoXB1dcWGDRsgk8l6vP0TEhLw/vvvw8bm8afx+Ph4eHl5IS4urtPpw4cPBwAUFRX1KB5rwUmNMYnoqLDd2tr62H4TJkyAQqHQu53WV1VWVoKIoFAojOofFxeHESNGYOfOnTh9+rTB9OLiYtTX12PChAl67RMnToS9vb3ebdvOPLoPLl++jMbGRgQGBur6ODo6YsCAAT3a/rdu3cKRI0ewZMmSx/bLzc1FdnY2vvrqK4Pf4zp0bLvbt293Ox5rwkmNsaeQg4MDqqqqxA6jx5qbmwH89n2MIZfLkZGRAUEQsHTpUjQ1NelN7xje/vBzXx1cXV2h1WpNiq+hoQEAsGHDBr3nyq5du2YwaMMUiYmJ+Otf//rYF2UfOHAACQkJKCgowLBhw7rs5+joCOD3bdnXcVJj7CnT2tqKmpoaDB48WOxQeqzjhGzKQ8RBQUFYsWIFSkpKsHnzZr1prq6uAPD/27vzoKiufA/g34Zu6G5oNhWCCsriEhTimJAI0afGGhKlIqIYcUJeGZ95RCciisbgLiCCGqQwMk7QUPXccQk4jjhTxpApS82YUkbEiUHFBQ0gijayC7/3h9Md20bohl6vv08VVcm9595z+tzu/tn3nnN+HQav7vRZnz59ADzNPUbPDcU/c+aMXudSqaysxJ49ezBv3rwXltmyZQt27dqFkydPom/fvp2er6WlBcBvfWntOKgx9pIpKioCEWHUqFHqbWKxuMvblpbI3d0dIpFI7/lnycnJGDp0KC5cuKCxffjw4XB0dMRPP/2ksf3HH39ES0sLXn/9db3q8fLyglQqRXFxsV7HdSY9PR0xMTFwc3PT2kdEWLp0KUpKSpCfn9/hL87nqfrOw8PDYG00Jw5qjAlce3s7amtr8eTJE1y8eBHx8fHw9vbWeB7j7++PBw8eID8/H62trbh37x5u3rypdS43NzfcvXsXN27cQF1dHVpbW1FYWGi2If1yuRy+vr6oqKjQ6zjVbcjn53NJpVIkJCTg8OHD2LVrF5RKJUpKSjB37lx4enoiNjZW73o+/vhj7N27F9nZ2VAqlWhra0NFRQV+/fVXAEB0dDQ8PDx0WqarqqoK33zzDRYuXNjh/suXL2PDhg3IycmBRCLRWk5r06ZNWseo+i4wMFCv12apOKgxZsG++uorBAcHAwCWLl2KiIgIZGdnY/PmzQCAoKAgXL9+HTk5OUhISAAAvPfeeygrK1Ofo6mpCYGBgZDJZBgzZgwGDx6M77//XuM51Lx58zB+/HjMnDkTQ4YMQXJysvp2VEhIiHolirlz58Ld3R0BAQGYNGkSHjx4YJJ+6Ex4eDhKS0s1no99++238Pf3x7Vr1xAcHIz58+drHTdq1CgsWrRIa/vq1auRmpqKpKQk9O7dG2PHjsXAgQNRVFQEBwcHANDrGmRmZmLhwoVIT09Hr1694Onpifj4eNTW1gJ4evuvuroaBQUFXb7WDRs2YPLkyfD29u5wP3Vjrtm5c+fQr18/BAUF6X2sRTLfbIKO8Tw1JhSWMI8nNjaW3NzczNoGfXTn819WVkZisZh27txppFYZV1tbG40ZM4Z27Nhh8rprampIKpXSpk2b9D7WEt7fHeFfaowJnBBXYn+Wv78/kpKSkJSUpLESvTVoa2tDfn4+6urqEB0dbfL616xZgxEjRiAuLs7kdRuLIIPanDlzoFAoIBKJDPqA1hqlpKR0mKri2Xkzuno+nYfqz87ODu7u7hg3bhw2btyovq3CmKkkJiZi+vTpiI6OtqpFi4uKinDo0CEUFhbqPNfOUDIyMlBcXIxjx45BIpGYtG5jEmRQ2759O3JycszdDMF5Np2Hs7MziAjt7e2orq5GXl4efHx8sHTpUgwbNkxr9BgzvWXLliE3NxePHj2Cj48PDh48aO4mGdW6desQFxeH9evXm7spOpswYQJ2796tse6mKRQUFKC5uRlFRUVwdXU1ad3GJsigJjQ9Tfexc+dOrTkyly5dMkjbRCIRXFxcMG7cOOTm5iIvLw9VVVXqdCnWzlJTregiNTUVzc3NICKUl5cjKirK3E0yurCwMKSlpZm7GRYvIiICiYmJWqM/hUCwQU0kEpm7CQZjiek+XiQqKgqzZs1CdXU1tm3bZu7m9Jg19T1jTCBBjYiwceNGDBkyBPb29nB2dsaSJUs0ymzYsAFyuRwKhQLV1dVISEhAv379cOXKFZ1STeiaxkPVnq7O19N0H4ZmyPQhqvlPhYWFALjvGWMmZKZRly/UnSG9y5cvJ5FIRF9++SXV1tZSQ0MDbd26VSudgipVxIIFC2jLli00depU+ve//61zqgld03joer6epPvQVXJyMvXv359cXFxIIpHQwIEDKSIigv75z39qlNMnfYifnx85Ozu/cL9SqSQA5OXlpd72Mva9pQ55tmQ8pcd6WOr72+LePfq+qRsaGkgul9Pvf/97je179+59YVBrbGzUON7R0ZGio6M1jv/nP/9JADS+5GNjY7W+zM+dO0cAaO3atXqfzxRfrLdu3aLz589TXV0dNTc305kzZ+h3v/sdyWQyunTpUrfO2VVQIyISiUTk4uKi/v+Xse8t9UNvyTioWQ9LfX+LTfqz0AiuXr2KhoYGTJgwoVvH9zTVxPNpPHp6PkPz8vKCl5eX+v9HjRqF3NxcjBgxAlu3bkV2drbB66yvrwcRwcnJqdNyQu974OkSRHl5eSav11qpFvnlPrN8FRUVFrkottUHNdW6ZarVsPVliFQTz6bxMHTqCmMIDAwmgclPAAAgAElEQVSEra0tfvnlF6OcX3XeoUOHdlruZej7s2fPYsaMGSav19pxn1kHSxxRa/VBTZVPqLm5uVvH9zTVxPNpPAydusIY2tvb0d7ernMOKn0dP34cADBx4sROy70MfR8VFYUDBw6YvF5rlZeXhxkzZnRrDUNmWtOnTzd3Ezpk9aMfhw8fDhsbG/zwww/dPr4nqSaeT+Ohz/lMke7j3Xff1dp27tw5EBFCQkIMXl9lZSU2b96M/v37Y/bs2Z2WFXrfM8ZMz+qDWp8+fTBt2jQcPHgQO3bsgFKpxMWLF/H111/rdLy+qSa6SuOhz/l6ku5DV3fu3MG+ffvw8OFDtLa24syZM5gzZw68vb0xd+5cdTl904cQER4/foz29nYQEe7du4f9+/fj7bffhq2tLfLz87t8pib0vmeMmYEZB6l0qDujn+rq6mjOnDnUq1cvcnR0pNGjR9OqVasIAPXv35/+9a9/UXp6OslkMvVQ82dX9G5vb6eNGzfSoEGDSCKRkKurK0VGRtKVK1c06omNjSWJREL9+vUjsVhMTk5ONGXKFLp27ZpGOV3Pd//+fRo/fjxJpVLy8fGh+fPn05IlSwgA+fv7q4eqnz9/ngYMGEAymYxGjx6tMTS9KwkJCeTn50cODg4kFoupf//+9Mknn9Ddu3c1yh07dowUCgWlpKS88FxHjhyhoKAgksvlZGdnRzY2NgRAPdLxzTffpKSkJLp//77GcS9r31vq6DBLxqMfrYelvr9FRJZ189qS76l/+umnOHDgAO7fv2/uprx0rLHvVc8c+Jma7iz58880Wer72+pvP5qa0NN4WDLue8ZYVzioWZmff/65w1Qyz/+ZIzcTY4yZGwc1HVlKGo+hQ4dqrbjf0d++ffvM0j5jsJS+Z5bvxIkTSExM1Mr999FHH2mVDQsLg0KhgK2tLYYNG4bz58+bocW6Gzdu3Av/Efvs3MykpCQEBATAyckJ9vb28Pf3x+eff66RQPXIkSNIT08X5t0PszzJ6wQ/KGZCYakP0i1ZTz7/q1atovfff5+USqV6m5+fH/Xq1YsA0NGjR7WOKSwspIiIiG6315TGjh1LADr8e/fddzXKbd26le7fv09KpZL2799PEomE3nvvPY3zZWZm0tixY6m2trZb7bHU9zf/UmNMwEyRD84Scs6lpaVh3759yMvLg0Kh0NiXlZUFGxsbxMbGWnWOP6lUCqVSqXVXJjY2Fp9//rm6nKOjI2JjY+Hm5gaFQoEPPvgAkZGROH78OG7fvq0ut2DBArz22muYNGkSnjx5Yo6XZBQc1BgTMFPkgzN3zrmrV69i5cqVWLt2rXqFoWeFhoYiPj4ed+7cweLFi83QQsM4fvy4VsC+ffs2Ll26hHfeeUe97ejRo1rJP3v37g0AaGho0Ni+Zs0aFBcXIzMz00itNj0OaoxZEDJyPjhdc9P1NOecIfPzdSUrKwtEhMmTJ7+wTEpKCgYPHozt27fjxIkTnZ5Pl2uQnZ0NBwcHyOVyFBQUYOLEiXByckL//v2xd+9ejfO1tbVh1apV8Pb2hkwmQ1BQEPbv39+zF/0faWlpWLBgQZfl7ty5A5lMBh8fH43trq6uGDt2LDIzM4UzjcJ8dz47xs/UmFB055mDKfLB6Zqbrid16JOf71nd+fz7+vpSQEBAh/v8/PyovLyciIhOnz5NNjY2NHDgQHr8+DERdfxMTddroEqn9N1339GjR4+ourqaxowZQw4ODtTS0qIut3jxYrK3t6eDBw9SbW0tLVu2jGxsbOjcuXN6vc7nVVRUUEBAALW1tXVarr6+nhQKBcXFxXW4PzExUStNly74mRpjrFONjY3IyMjA1KlTERMTA2dnZwQGBmLbtm2oqanReek3XYjFYvUvkYCAAGRnZ6Ourg65ubkGOX94eDiUSiVWrlxpkPO9SH19PcrLy+Hn59dl2ZCQECxcuBA3btzAF1980WGZ7lyD0NBQODk5oU+fPoiOjkZ9fT1u3boFAGhqakJ2djYiIyMxbdo0uLi4YMWKFZBIJD3u67S0NMyfPx82Np1/jaempsLT0xMpKSkd7h80aBAAoKSkpEftsRQc1BizEObMB/d8bjprUV1dDSKCXC7XqXxKSgqGDBmCrVu34tSpU1r7e3oN7OzsAEC9RuiVK1fQ0NCA4cOHq8vIZDK88sorPerru3fv4siRI+p1T1/k8OHDyMvLw9/+9jet53Eqqr6rqqrqdnssCQc1xiyEufPBPZubzlo0NTUBgM5plKRSKXJzcyESiTB79mw0NjZq7Df0NaivrwcArFixQmNe2c2bN7UGbegjPT0dn3zySYcDY1T27duHtLQ0FBUVYeDAgS8sJ5PJAPzWl9aOgxpjFsKc+eCez01nLVRfyPpMIg4JCcGiRYtQVlaG5ORkjX2Gvgaq5MWbN2/WGoqvyvKtr8rKSuzZswfz5s17YZktW7Zg165dOHnyJPr27dvp+VpaWgD81pfWjoMaYxbCnPngns9NZ4w6jMHd3R0ikUjv+WfJyckYOnQoLly4oLG9pzn+nufl5QWpVIri4mK9jutMeno6YmJi4ObmprWPiLB06VKUlJQgPz+/w1+cz1P1nYeHh8HaaE4c1BizEKbMB9dVbrqe1qFvfr7uksvl8PX1RUVFhV7HqW5DPj+fS98cf7rU8/HHH2Pv3r3Izs6GUqlEW1sbKioq8OuvvwIAoqOj4eHhodMyXVVVVfjmm2+wcOHCDvdfvnwZGzZsQE5ODiQSidZyWps2bdI6RtV3gYGBer02S8VBjTELsnr1aqSmpiIpKQm9e/fG2LFjMXDgQBQVFcHBwUFdbt68eRg/fjxmzpyJIUOGIDk5WX37KCQkRL1yxNy5c+Hu7o6AgABMmjQJDx48APD0+UlgYCBkMhnGjBmDwYMH4/vvv9d4NtXTOkwlPDwcpaWlGs/Hvv32W/j7++PatWsIDg7G/PnztY4bNWoUFi1apLVdl2uQnZ2NzZs3AwCCgoJw/fp15OTkICEhAQDw3nvvoaysDACQmZmJhQsXIj09Hb169YKnpyfi4+NRW1sL4Ontv+rqahQUFHT5Wjds2IDJkyfD29u7w/3Ujblm586dQ79+/RAUFKT3sRbJfLMJOsbz1JhQWOo8ntjYWHJzczN3MzrUnc9/WVkZicVijeSz1qStrY3GjBlDO3bsMHndNTU1JJVKadOmTXofa6nvb/6lxthLSEirs/v7+yMpKQlJSUkaK9Fbg7a2NuTn56Ours4s6aLWrFmDESNGIC4uzuR1GwsHNcaY1UtMTMT06dMRHR1tVYsWFxUV4dChQygsLNR5rp2hZGRkoLi4GMeOHYNEIjFp3cbEQY2xl4iQc9OtW7cOcXFxWL9+vbmborMJEyZg9+7dGmtsmkJBQQGam5tRVFQEV1dXk9ZtbGJzN4AxZjqpqalITU01dzOMJiwsDGFhYeZuhsWLiIhARESEuZthFPxLjTHGmGBwUGOMMSYYHNQYY4wJBgc1xhhjgmGxA0WmT59u7iYw1iNnz54FwO9lfaiWbOI+s3xnz57VWCvUUoiILCuH95kzZ5CRkWHuZjBmUIWFhfjd735n8qHbjBmTKuOBJbG4oMaYEIlEIuzfvx8ffPCBuZvCmKDxMzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAmG2NwNYExoHj58CCLS2l5fX4/a2lqNbY6OjpBIJKZqGmOCJ6KOPn2MsW5755138P3333dZztbWFnfu3IGHh4cJWsXYy4FvPzJmYDNnzoRIJOq0jI2NDf7rv/6LAxpjBsZBjTEDi4qKgljc+Z19kUiE//7v/zZRixh7eXBQY8zAXF1dERYWBltb2xeWsbGxQWRkpAlbxdjLgYMaY0YQExOD9vb2DveJxWKEh4fD2dnZxK1iTPg4qDFmBJMnT4a9vX2H+9ra2hATE2PiFjH2cuCgxpgRyOVyREZGdjhcXyaTYdKkSWZoFWPCx0GNMSP5wx/+gNbWVo1tEokEUVFRkMlkZmoVY8LGQY0xI3n33Xe1npu1trbiD3/4g5laxJjwcVBjzEgkEgmio6NhZ2en3ubi4oIJEyaYsVWMCRsHNcaMaObMmWhpaQHwNMjFxMR0OYeNMdZ9vEwWY0bU3t6Ovn37oqqqCgBw6tQpvP3222ZuFWPCxb/UGDMiGxsbfPTRRwAAT09PhIaGmrlFjAmbYO+DVFRU4PTp0+ZuBmPo3bs3AOCtt97CgQMHzNwaxgAvLy+EhISYuxlGIdjbj3l5eZgxY4a5m8EYYxYnKipKsP/AEuwvNRWBxmxmRNOnTwcAg37oDx48iKioKIOdz9Ko/hHJnzfLp3p/CxU/U2PMBIQc0BizJBzUGGOMCQYHNcYYY4LBQY0xxphgcFBjjDEmGBzUGGOMCQYHNcaM5NixY3B2dsZf/vIXczfF4p04cQKJiYk4dOgQfH19IRKJIBKJ1KuxPCssLAwKhQK2trYYNmwYzp8/b4YW627cuHHq1/P8n6Ojo7pcUlISAgIC4OTkBHt7e/j7++Pzzz/H48eP1WWOHDmC9PR0tLW1meOlWAUOaowZCc/Z0s3q1auRlZWFZcuWYdq0abh+/Tr8/PzQq1cv7Nq1C3/96181yv/973/HgQMH8P7776O0tBQjR440U8t7bvTo0er/PnnyJD777DPcuHEDNTU1SE1NRWZmpsa8ssmTJ0MqlWLChAl4+PChOZps8TioMWYk4eHhePToEd5//31zNwWNjY0Wue5kWloa9u3bh7y8PCgUCo19WVlZsLGxQWxsLB49emSmFvacVCqFUqkEEWn8xcbG4vPPP1eXc3R0RGxsLNzc3KBQKPDBBx8gMjISx48fx+3bt9XlFixYgNdeew2TJk3CkydPzPGSLBoHNcZeAjt27EB1dbW5m6Hh6tWrWLlyJdauXQupVKq1PzQ0FPHx8bhz5w4WL15shhYaxvHjx7UC9u3bt3Hp0iW888476m1Hjx6Fra2tRjnVuqENDQ0a29esWYPi4mJkZmYaqdXWi4MaY0Zw6tQpeHt7QyQS4auvvgIAZGdnw8HBAXK5HAUFBZg4cSKcnJzQv39/7N27V31sVlYWpFIp3N3d8emnn8LT0xNSqRShoaH48ccf1eXi4uJgZ2eHV155Rb3tj3/8IxwcHCASiVBTUwMAiI+PR0JCAq5duwaRSAR/f38AT79snZycsG7dOlN0iZasrCwQESZPnvzCMikpKRg8eDC2b9+OEydOdHo+IkJGRgZeffVV2Nvbw9XVFVOmTMHPP/+sLqPrNQCAtrY2rFq1Ct7e3pDJZAgKCsL+/ft79qL/Iy0tDQsWLOiy3J07dyCTyeDj46Ox3dXVFWPHjkVmZibf5n4eCdT+/ftJwC+PGVFUVBRFRUX1+Dy3b98mALRlyxb1tuXLlxMA+u677+jRo0dUXV1NY8aMIQcHB2ppaVGXi42NJQcHB7p8+TI1NTVRaWkpBQcHk0KhoFu3bqnLffjhh+Th4aFR78aNGwkA3bt3T71t2rRp5Ofnp1Hu6NGjpFAoKCkpqcevtTufN19fXwoICOhwn5+fH5WXlxMR0enTp8nGxoYGDhxIjx8/JiKiwsJCioiI0Dhm1apVZGdnRzt37qSHDx/SxYsXaeTIkdS7d2+qrKxUl9P1GixevJjs7e3p4MGDVFtbS8uWLSMbGxs6d+6cXq/zeRUVFRQQEEBtbW2dlquvryeFQkFxcXEd7k9MTCQAdOHCBb3qN9T721LxLzXGzCA0NBROTk7o06cPoqOjUV9fj1u3bmmUEYvF6l8dAQEByM7ORl1dHXJzcw3ShvDwcCiVSqxcudIg59NHfX09ysvL4efn12XZkJAQLFy4EDdu3MAXX3zRYZnGxkZkZGRg6tSpiImJgbOzMwIDA7Ft2zbU1NTg66+/1jqms2vQ1NSE7OxsREZGYtq0aXBxccGKFSsgkUh63P9paWmYP38+bGw6//pNTU2Fp6cnUlJSOtw/aNAgAEBJSUmP2iM0HNQYMzM7OzsAQGtra6fl3njjDcjlco3badaquroaRAS5XK5T+ZSUFAwZMgRbt27FqVOntPaXlpbi8ePHeOONNzS2BwcHw87OTuO2bUeevwZXrlxBQ0MDhg8fri4jk8nwyiuv9Kj/7969iyNHjmDWrFmdljt8+DDy8vLwt7/9Tet5nIqq71RZ1dlTHNQYsyL29va4d++euZvRY01NTQCevh5dSKVS5ObmQiQSYfbs2WhsbNTYrxre/uy8LxUXFxfU1dXp1b76+noAwIoVKzTmld28eVNr0IY+0tPT8cknn3Q4MEZl3759SEtLQ1FREQYOHPjCcjKZDMBvfcme4qDGmJVobW3Fw4cP0b9/f3M3pcdUX8j6TCIOCQnBokWLUFZWhuTkZI19Li4uANBh8OpOn/Xp0wcAsHnzZq2h+GfOnNHrXCqVlZXYs2cP5s2b98IyW7Zswa5du3Dy5En07du30/O1tLQA+K0v2VMc1BizEkVFRSAijBo1Sr1NLBZ3edvSErm7u0MkEuk9/yw5ORlDhw7FhQsXNLYPHz4cjo6O+OmnnzS2//jjj2hpacHrr7+uVz1eXl6QSqUoLi7W67jOpKenIyYmBm5ublr7iAhLly5FSUkJ8vPzO/zF+TxV33l4eBisjULAQY0xC9Xe3o7a2lo8efIEFy9eRHx8PLy9vTWex/j7++PBgwfIz89Ha2sr7t27h5s3b2qdy83NDXfv3sWNGzdQV1eH1tZWFBYWmm1Iv1wuh6+vLyoqKvQ6TnUb8vn5XFKpFAkJCTh8+DB27doFpVKJkpISzJ07F56enoiNjdW7no8//hh79+5FdnY2lEol2traUFFRgV9//RUAEB0dDQ8PD52W6aqqqsI333yDhQsXdrj/8uXL2LBhA3JyciCRSLSW09q0aZPWMaq+CwwM1Ou1CR0HNcaM4KuvvkJwcDAAYOnSpYiIiEB2djY2b94MAAgKCsL169eRk5ODhIQEAMB7772HsrIy9TmampoQGBgImUyGMWPGYPDgwfj+++81nkPNmzcP48ePx8yZMzFkyBAkJyerb0eFhISoV6KYO3cu3N3dERAQgEmTJuHBgwcm6YfOhIeHo7S0VOP52Lfffgt/f39cu3YNwcHBmD9/vtZxo0aNwqJFi7S2r169GqmpqUhKSkLv3r0xduxYDBw4EEVFRXBwcAAAva5BZmYmFi5ciPT0dPTq1Quenp6Ij49HbW0tgKe3/6qrq1FQUNDla92wYQMmT54Mb2/vDvdTN+aanTt3Dv369UNQUJDexwqa+WYTGBfPU2PdZQnzeGJjY8nNzc2sbdBHdz5vZWVlJBaLaefOnUZqlXG1tbXRmDFjaMeOHSavu6amhqRSKW3atEnvYy3h/W1M/EuNMQsl9JXY/f39kZSUhKSkJI2V6K1BW1sb8vPzUVdXh+joaJPXv2bNGowYMQJxcXEmr9vScVDrxJw5c6BQKCASiQz6wNjU9uzZg+DgYCgUCgwYMAAff/wxKisr9T7P82lBVH92dnZwd3fHuHHjsHHjRvXtGca6kpiYiOnTpyM6OtqqFi0uKirCoUOHUFhYqPNcO0PJyMhAcXExjh07BolEYtK6rQEHtU5s374dOTk55m5Gj+zfvx8ffvghpk+fjoqKChQUFOAf//gHJk6cqPcK38+mBXF2dgYRob29HdXV1cjLy4OPjw+WLl2KYcOGaY1CY7pbtmwZcnNz8ejRI/j4+ODgwYPmbpJRrVu3DnFxcVi/fr25m6KzCRMmYPfu3RrrbppCQUEBmpubUVRUBFdXV5PWbS04qAncn//8Z/Tt2xdLliyBs7MzRowYgUWLFqG4uLjLVRZ0IRKJ4OLignHjxiE3Nxd5eXmoqqpSp11h+ktNTUVzczOICOXl5YiKijJ3k4wuLCwMaWlp5m6GxYuIiEBiYqLW6E/2Gw5qXRCJROZuQo/cvn0bnp6eGq/Dy8sLADoc+t1TUVFRmDVrFqqrq7Ft2zaDn58xxjrDQe0ZRISNGzdiyJAhsLe3h7OzM5YsWaJVrrOUFPqktvjhhx/w5ptvQi6Xw8nJCYGBgVAqlV3WoQ9fX1+tPFqq52m+vr7qbYZMQ6KaR1VYWKjeZk19xhizYmYefWk03RlivHz5chKJRPTll19SbW0tNTQ00NatW7XSO3SVkkKX1BaPHz8mJycnSk9Pp8bGRqqsrKSpU6eq04UYKu1FUVERSSQSysrKIqVSSZcuXaJXX32V3n33XY1y+qQh8fPzI2dn5xfuVyqVBIC8vLysss+EPuTZGHgKjfUQ+vtbsO9CfT9kDQ0NJJfL6fe//73G9r1792oEtcbGRpLL5RQdHa1xrL29Pc2bN4+IfvuCbmxsVJdRBcerV68SEdGlS5cIAB09elSrLbrUoY8VK1YQAPVf//796fbt23qfR6WroEZEJBKJyMXFhYisr8+E/qE3Bg5q1kPo72+xqX8ZWqqrV6+ioaEBEyZM6LRcd1NSPJ/awtfXF+7u7oiJicGCBQswa9Ys9Yrchkx7sXz5cmzfvh3fffcd3nrrLVRXV+OLL75ASEgITp8+rX6+Zkj19fUgIjg5OQGwvj4DgLNnz2L69Ol6H/eyUi3ZxH1m+c6ePauxfqjQ8DO1/1B9KFWrc7+IoVJSyGQynDx5EqNHj8a6devg6+uL6OhoNDY2GqyOX3/9Fenp6fjf//1fvPPOO3BwcICPjw9ycnJw9+5dbNy4Uedz6eOXX34BAAwdOhSAdfUZY8y68S+1/1DlN2pubu603LMpKeLj43tU57Bhw/CXv/wF9+7dQ0ZGBtLS0jBs2DD1CgU9raOsrAxtbW1aKSycnJzg5uaG0tLSHrX/RY4fPw4AmDhxIgDr6jOVUaNG4cCBAz0+z8siLy8PM2bM4D6zAkL/Nc2/1P5j+PDhsLGxwQ8//NBpOUOlpLh79y4uX74M4OmX/vr16zFy5EhcvnzZYHWockipVhVXqaurw4MHD4xy67GyshKbN29G//79MXv2bADW1WeMMevGQe0/+vTpg2nTpuHgwYPYsWMHlEolLl68iK+//lqjnC4pKXRx9+5dfPrpp/j555/R0tKCCxcu4ObNmxg1apTB6vDx8cH48eORk5ODf/zjH2hsbMTt27fVaTj+53/+R11W3zQkRITHjx+jvb0dRIR79+5h//79ePvtt2Fra4v8/Hz1MzVr6jPGmJUz80AVo+nOaKy6ujqaM2cO9erVixwdHWn06NG0atUq9YjBf/3rX0RE1NzcTEuXLiVvb28Si8XUp08fmjZtGpWWltLWrVtJLpcTABo0aBBdu3aNvv76a3JyciIANGDAAPrll1/oxo0bFBoaSq6urmRra0t9+/al5cuX05MnT7qsQx81NTUUHx9P/v7+ZG9vT46OjvT222/Tt99+q1Hu2LFjpFAoKCUl5YXnOnLkCAUFBZFcLic7OzuysbEhAOqRjm+++SYlJSXR/fv3tY61pj4T+ugwY+DRj9ZD6O9vEVE3EvlYAdU9foG+PGZEqmcO/HxId/x5sx5Cf3/z7UfGGGOCwUHNyvz8889aqV86+jNHjifGDOnEiRNITEzUSnn00UcfaZUNCwuDQqGAra0thg0bhvPnz5uhxfppbW1Famoq/P39YWdnBxcXFwwfPhw3btwAABw5cgTp6emCz6tnaBzUrMzQoUNBT1eC6fRv37595m4qY922evVqZGVlYdmyZRopj3r16oVdu3bhr3/9q0b5v//97zhw4ADef/99lJaWYuTIkWZque5mzJiB//u//8Pu3bvR0NCAf//73/Dz81MnTJ08eTKkUikmTJiAhw8fmrm11oODGmMWqLGxEaGhoVZfR3ekpaVh3759yMvLg0Kh0NiXlZUFGxsbxMbGWnVqo3379iE/Px8HDhzAW2+9BbFYDE9PTxQUFGisirNgwQK89tprmDRpkt75D19WHNQYs0A7duzQyq5gjXXo6+rVq1i5ciXWrl2rXhDhWaGhoYiPj8edO3ewePFiM7TQMP70pz9h5MiRCAwM7LLsmjVrUFxcjMzMTBO0zPpxUGPMAIgIGRkZePXVV2Fvbw9XV1dMmTJFY93JuLg42NnZaWRL/uMf/wgHBweIRCLU1NQAAOLj45GQkIBr165BJBLB398fWVlZkEqlcHd3x6effgpPT09IpVKEhoZqJHvtSR2AYVMQdUdWVhaICJMnT35hmZSUFAwePBjbt2/HiRMnOj2fLtdFn9RHhkhv1NLSgrNnz2LEiBE6lXd1dcXYsWORmZnJo0t1YfJJBCbC82ZYd3VnHs+qVavIzs6Odu7cSQ8fPqSLFy/SyJEjqXfv3lRZWaku9+GHH5KHh4fGsRs3biQA6hQ6RETTpk0jPz8/jXKxsbHk4OBAly9fpqamJiotLaXg4GBSKBR069Ytg9ShTwqiZxnq8+br60sBAQEd7vPz86Py8nIiIjp9+jTZ2NjQwIED6fHjx0REVFhYSBERERrH6HpddEl9RGSY9Ebl5eUEgEaMGEHjxo2jV155hezt7Wno0KH01VdfUXt7u9YxiYmJWimwukvo89T4lxpjPdTY2IiMjAxMnToVMTExcHZ2RmBgILZt24aamhqtVWl6QiwWq391BAQEIDs7G3V1dcjNzTXI+cPDw6FUKrFy5UqDnE8f9fX1KC8vh5+fX5dlQ0JCsHDhQty4cQNffPFFh2W6c11CQ0Ph5OSEPqfDwQ4AAAUuSURBVH36IDo6GvX19bh16xYAoKmpCdnZ2YiMjMS0adPg4uKCFStWQCKR6NX/qoEgffr0wbp161BaWoqqqipMmTIFn332Gfbs2aN1zKBBgwAAJSUlOtfzsuKgxlgPlZaW4vHjx3jjjTc0tgcHB8POzk7j9qChvfHGG5DL5d1Kr2NpqqurQUSQy+U6lU9JScGQIUOwdetWnDp1Smt/T6/L86mPDJXeyN7eHsDTxblDQ0Ph5uYGZ2dnrF27Fs7Ozh0GW1WfVFVV6VzPy4qDGmM9pBpu7ejoqLXPxcUFdXV1Rq3f3t4e9+7dM2odptDU1ATgty/9rkilUuTm5kIkEmH27NlobGzU2G/o62Ko9Eaenp4AoH6+qWJnZ4cBAwbg2rVrWsfIZDIAv/URezEOaoz1kIuLCwB0+CX58OFDdbYEY2htbTV6Haai+uLWZ7JxSEgIFi1ahLKyMiQnJ2vsM/R1eTaFEj03L/TMmTM6n8fR0RGDBg1SZ5x41pMnT+Ds7Ky1vaWlBcBvfcRejIMaYz00fPhwODo64qefftLY/uOPP6KlpQWvv/66eptYLFbfzjKEoqIiEJFGJmND12Eq7u7uEIlEes8/S05OxtChQ3HhwgWN7fpcF10YMr3RjBkzcOHCBVy/fl29raGhATdv3uxwmL+qTzw8PHpct9BxUGOsh6RSKRISEnD48GHs2rULSqUSJSUlmDt3Ljw9PdWpfgDA398fDx48QH5+PlpbW3Hv3j3cvHlT65xubm64e/cubty4gbq6OnWQam9vR21tLZ48eYKLFy8iPj4e3t7emDVrlkHq0DcFkSHJ5XL4+vqqs9DrSnUb0tbWVmu7rtdF13q6Sm8UHR0NDw+PLpfpWrRoEQYMGIBZs2bh1q1buH//PpYuXYrGxsYOB76o+kSXeW0vPXMOvTQmHtLPuqs7Q57b29tp48aNNGjQIJJIJOTq6kqRkZF05coVjXL379+n8ePHk1QqJR8fH5o/fz4tWbKEAJC/v796aP758+dpwIABJJPJaPTo0VRZWUmxsbEkkUioX79+JBaLycnJiaZMmULXrl0zWB26pCDqiKE+b3FxcSSRSKihoUG97fDhw+Tn50cAqHfv3vTZZ591eOySJUu0hvTrcl10TX1E1HV6o8jISAJAq1at6vK13r59m2bOnEmurq5kb29Pb775JhUWFnZYNjw8nPr169fhcH99CX1Iv2C/9Tmose6y1A99bGwsubm5mbsZHTLU562srIzEYjHt3LnTAK0yvba2NhozZgzt2LHDYOesqakhqVRKmzZtMsj5LPX9bSh8+5ExKyL0Fdv9/f2RlJSEpKQk9Xwua9HW1ob8/HzU1dUZNEvGmjVrMGLECMTFxRnsnELGQY0xZlESExMxffp0REdHW9WixUVFRTh06BAKCwt1nmvXlYyMDBQXF+PYsWOQSCQGOafQcVBjzAosW7YMubm5ePToEXx8fHDw4EFzN8mo1q1bh7i4OKxfv97cTdHZhAkTsHv3bo11N3uioKAAzc3NKCoqgqurq0HO+TIQm7sBjLGupaamIjU11dzNMKmwsDCEhYWZuxlmExERgYiICHM3w+rwLzXGGGOCwUGNMcaYYHBQY4wxJhgc1BhjjAkGBzXGGGOCIfjRjyKRyNxNYFaK3zv64z6zDlFRUeZugtGIiIjM3QhjqKiowOnTp83dDMYYszheXl4ICQkxdzOMQrBBjTHG2MuHn6kxxhgTDA5qjDHGBIODGmOMMcEQAzhg7kYwxhhjhvD/C6f/5ag9+msAAAAASUVORK5CYII=\n","text/plain":["<IPython.core.display.Image object>"]},"metadata":{"tags":[]},"execution_count":33}]},{"cell_type":"markdown","metadata":{"id":"agE43OxRhAAS"},"source":["#### Modèle 3"]},{"cell_type":"code","metadata":{"id":"soke0XOAhFjk","executionInfo":{"status":"ok","timestamp":1621426675869,"user_tz":-120,"elapsed":85706,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["X_train_pp_df = DateTimeImputer().transform(X_train)\n","X_train_pp_df = DateDissolver(replace=True).transform(X_train_pp_df)\n","X_train_pp_df = text_pipeline.fit_transform(X_train_pp_df)\n","\n","X_train_pp = full_pipeline.fit_transform(X_train_pp_df)"],"execution_count":34,"outputs":[]},{"cell_type":"code","metadata":{"id":"UMNg4Gg1ILVZ","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426699646,"user_tz":-120,"elapsed":109481,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"4297c343-ad18-442d-ae14-46c26bc0743b"},"source":["model_3 = keras.models.Sequential()\n","model_3.add(keras.layers.Dense(5000 , activation=\"relu\", input_shape=X_train_pp.shape[1:]))\n","model_3.add(keras.layers.Dense(1000, activation=\"relu\"))\n","model_3.add(keras.layers.Dense(2000, activation=\"relu\"))\n","model_3.add(keras.layers.Dense(100, activation=\"relu\"))\n","model_3.add(keras.layers.Dense(500, activation=\"relu\"))\n","model_3.add(keras.layers.Dense(X_train_pp.shape[1], activation=\"relu\"))\n","model_3.add(keras.layers.Dense(1))\n","\n","model_3.compile(loss=\"mean_squared_logarithmic_error\", optimizer=keras.optimizers.SGD(lr=1e-1))\n","history = model_3.fit(X_train_pp, y_train, epochs=100, batch_size=32, validation_split=0.2)"],"execution_count":35,"outputs":[{"output_type":"stream","text":["Epoch 1/100\n","75/75 [==============================] - 1s 5ms/step - loss: 0.9519 - val_loss: 0.4275\n","Epoch 2/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.4316 - val_loss: 0.4127\n","Epoch 3/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.4218 - val_loss: 0.3444\n","Epoch 4/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3959 - val_loss: 0.4382\n","Epoch 5/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3786 - val_loss: 0.3278\n","Epoch 6/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3577 - val_loss: 0.3251\n","Epoch 7/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3424 - val_loss: 0.3427\n","Epoch 8/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3619 - val_loss: 0.3504\n","Epoch 9/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3784 - val_loss: 0.3732\n","Epoch 10/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3491 - val_loss: 0.3144\n","Epoch 11/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3424 - val_loss: 0.3484\n","Epoch 12/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3348 - val_loss: 0.3144\n","Epoch 13/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3451 - val_loss: 0.3127\n","Epoch 14/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3315 - val_loss: 0.3413\n","Epoch 15/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3260 - val_loss: 0.3807\n","Epoch 16/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3395 - val_loss: 0.3118\n","Epoch 17/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3278 - val_loss: 0.3290\n","Epoch 18/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3523 - val_loss: 0.3461\n","Epoch 19/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3268 - val_loss: 0.3154\n","Epoch 20/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3217 - val_loss: 0.3029\n","Epoch 21/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3300 - val_loss: 0.3043\n","Epoch 22/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3419 - val_loss: 0.4492\n","Epoch 23/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3238 - val_loss: 0.3104\n","Epoch 24/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3253 - val_loss: 0.5641\n","Epoch 25/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3038 - val_loss: 0.3268\n","Epoch 26/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2939 - val_loss: 0.3059\n","Epoch 27/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3096 - val_loss: 0.3148\n","Epoch 28/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3208 - val_loss: 0.3130\n","Epoch 29/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3269 - val_loss: 0.4174\n","Epoch 30/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3226 - val_loss: 0.3199\n","Epoch 31/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2651 - val_loss: 0.3109\n","Epoch 32/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3184 - val_loss: 0.3045\n","Epoch 33/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3268 - val_loss: 0.3158\n","Epoch 34/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3083 - val_loss: 0.3060\n","Epoch 35/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2986 - val_loss: 0.3539\n","Epoch 36/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2798 - val_loss: 0.3034\n","Epoch 37/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3134 - val_loss: 0.2953\n","Epoch 38/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2895 - val_loss: 0.3337\n","Epoch 39/100\n","75/75 [==============================] - 0s 5ms/step - loss: 0.2845 - val_loss: 0.3331\n","Epoch 40/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3044 - val_loss: 0.3306\n","Epoch 41/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2858 - val_loss: 0.3371\n","Epoch 42/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.3145 - val_loss: 0.3190\n","Epoch 43/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2957 - val_loss: 0.3643\n","Epoch 44/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2751 - val_loss: 0.2997\n","Epoch 45/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2786 - val_loss: 0.2931\n","Epoch 46/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2992 - val_loss: 0.3145\n","Epoch 47/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2478 - val_loss: 0.2969\n","Epoch 48/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2873 - val_loss: 0.2962\n","Epoch 49/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2686 - val_loss: 0.2977\n","Epoch 50/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2728 - val_loss: 0.2977\n","Epoch 51/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2716 - val_loss: 0.2999\n","Epoch 52/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2905 - val_loss: 0.3359\n","Epoch 53/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2814 - val_loss: 0.2947\n","Epoch 54/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2938 - val_loss: 0.3294\n","Epoch 55/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2850 - val_loss: 0.3301\n","Epoch 56/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2741 - val_loss: 0.3092\n","Epoch 57/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2723 - val_loss: 0.3091\n","Epoch 58/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2633 - val_loss: 0.3438\n","Epoch 59/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2690 - val_loss: 0.2964\n","Epoch 60/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2806 - val_loss: 0.3141\n","Epoch 61/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2737 - val_loss: 0.3121\n","Epoch 62/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2557 - val_loss: 0.3004\n","Epoch 63/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2695 - val_loss: 0.3335\n","Epoch 64/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2735 - val_loss: 0.2975\n","Epoch 65/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2568 - val_loss: 0.2987\n","Epoch 66/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2792 - val_loss: 0.3232\n","Epoch 67/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2786 - val_loss: 0.3374\n","Epoch 68/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2513 - val_loss: 0.2915\n","Epoch 69/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2506 - val_loss: 0.2993\n","Epoch 70/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2673 - val_loss: 0.3039\n","Epoch 71/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2581 - val_loss: 0.2971\n","Epoch 72/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2349 - val_loss: 0.3018\n","Epoch 73/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2696 - val_loss: 0.2957\n","Epoch 74/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2843 - val_loss: 0.3133\n","Epoch 75/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2479 - val_loss: 0.2923\n","Epoch 76/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2703 - val_loss: 0.3056\n","Epoch 77/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2459 - val_loss: 0.3033\n","Epoch 78/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2547 - val_loss: 0.3630\n","Epoch 79/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2556 - val_loss: 0.3234\n","Epoch 80/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2479 - val_loss: 0.3501\n","Epoch 81/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2428 - val_loss: 0.3461\n","Epoch 82/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2698 - val_loss: 0.3219\n","Epoch 83/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2464 - val_loss: 0.4131\n","Epoch 84/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2452 - val_loss: 0.3020\n","Epoch 85/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2298 - val_loss: 0.3070\n","Epoch 86/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2498 - val_loss: 0.3086\n","Epoch 87/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2443 - val_loss: 0.3158\n","Epoch 88/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2642 - val_loss: 0.3066\n","Epoch 89/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2356 - val_loss: 0.3079\n","Epoch 90/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2389 - val_loss: 0.2982\n","Epoch 91/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2417 - val_loss: 0.3302\n","Epoch 92/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2361 - val_loss: 0.3032\n","Epoch 93/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2620 - val_loss: 0.2950\n","Epoch 94/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2414 - val_loss: 0.3177\n","Epoch 95/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2282 - val_loss: 0.3115\n","Epoch 96/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2538 - val_loss: 0.3059\n","Epoch 97/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2496 - val_loss: 0.3145\n","Epoch 98/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2775 - val_loss: 0.2998\n","Epoch 99/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2369 - val_loss: 0.3237\n","Epoch 100/100\n","75/75 [==============================] - 0s 3ms/step - loss: 0.2475 - val_loss: 0.3425\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"r8_Zl66Kil-R","colab":{"base_uri":"https://localhost:8080/","height":865},"executionInfo":{"status":"ok","timestamp":1621426699647,"user_tz":-120,"elapsed":109480,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"72b0790e-054b-486c-aab3-12c6e78c41fc"},"source":["keras.utils.plot_model(model_3, show_shapes=True)"],"execution_count":36,"outputs":[{"output_type":"execute_result","data":{"image/png":"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\n","text/plain":["<IPython.core.display.Image object>"]},"metadata":{"tags":[]},"execution_count":36}]},{"cell_type":"markdown","metadata":{"id":"W6NXdmSW3ZzJ"},"source":["## **Partie 4 : Prediction avec le test_set**"]},{"cell_type":"code","metadata":{"id":"caRkv4itithR","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426712627,"user_tz":-120,"elapsed":122458,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"9a02b76c-f78b-4a5e-850c-ffa782e659ae"},"source":["X_test = test_set.copy().drop(test_set.filter(regex=COLUMN_EXCLUDE_PATTERN), axis=1)\n","print(X_test.info())\n","X_test_pp_df = DateTimeImputer().transform(X_test)\n","X_test_pp_df = DateDissolver(replace=True).transform(X_test_pp_df)\n","X_test_pp_df = text_pipeline.fit_transform(X_test_pp_df)\n","print(X_test_pp_df.info())\n","\n","X_test_pp = full_pipeline.transform(X_test_pp_df)\n","X_test_pp[0]"],"execution_count":37,"outputs":[{"output_type":"stream","text":["<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 4398 entries, 0 to 4397\n","Data columns (total 16 columns):\n"," #   Column                Non-Null Count  Dtype  \n","---  ------                --------------  -----  \n"," 0   budget                4398 non-null   int64  \n"," 1   genres                4382 non-null   object \n"," 2   original_language     4398 non-null   object \n"," 3   original_title        4398 non-null   object \n"," 4   overview              4384 non-null   object \n"," 5   poster_path           4397 non-null   object \n"," 6   production_companies  4140 non-null   object \n"," 7   production_countries  4296 non-null   object \n"," 8   release_date          4397 non-null   object \n"," 9   runtime               4394 non-null   float64\n"," 10  spoken_languages      4356 non-null   object \n"," 11  status                4396 non-null   object \n"," 12  title                 4395 non-null   object \n"," 13  Keywords              4005 non-null   object \n"," 14  cast                  4385 non-null   object \n"," 15  crew                  4376 non-null   object \n","dtypes: float64(1), int64(1), object(14)\n","memory usage: 549.9+ KB\n","None\n","<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 4398 entries, 0 to 4397\n","Data columns (total 18 columns):\n"," #   Column                Non-Null Count  Dtype  \n","---  ------                --------------  -----  \n"," 0   budget                4398 non-null   int64  \n"," 1   genres                4398 non-null   float64\n"," 2   original_language     4398 non-null   float64\n"," 3   original_title        4398 non-null   float64\n"," 4   overview              4398 non-null   float64\n"," 5   poster_path           4398 non-null   float64\n"," 6   production_companies  4398 non-null   float64\n"," 7   production_countries  4398 non-null   float64\n"," 8   runtime               4394 non-null   float64\n"," 9   spoken_languages      4398 non-null   float64\n"," 10  status                4398 non-null   float64\n"," 11  title                 4398 non-null   float64\n"," 12  Keywords              4398 non-null   float64\n"," 13  cast                  4398 non-null   float64\n"," 14  crew                  4398 non-null   float64\n"," 15  release_date_Y        4397 non-null   float64\n"," 16  release_date_M        4397 non-null   float64\n"," 17  release_date_D        4397 non-null   float64\n","dtypes: float64(17), int64(1)\n","memory usage: 618.6 KB\n","None\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["array([-1.        , -0.12408246,  0.11283519, -0.92727275, -0.54887856,\n","        1.09549749, -1.        , -0.72091026, -0.46745562, -0.74620407,\n","       -0.99985613, -0.35059355, -0.95972814, -0.92252835, -0.97308113,\n","       -0.27272727,  0.09090909, -0.13333333])"]},"metadata":{"tags":[]},"execution_count":37}]},{"cell_type":"code","metadata":{"id":"kltpkMOaiwn_","executionInfo":{"status":"ok","timestamp":1621426733547,"user_tz":-120,"elapsed":555,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["from sklearn.metrics import mean_squared_log_error\n","\n","final_model = model_3\n","final_predictions = final_model.predict(X_test_pp)\n"],"execution_count":43,"outputs":[]},{"cell_type":"code","metadata":{"id":"DMGmoslHi01E","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1621426736906,"user_tz":-120,"elapsed":616,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"49728026-2cf2-4224-94fe-4e2485fa0ab5"},"source":["print(\"Predicts -> \", list(final_predictions[0:100]))"],"execution_count":44,"outputs":[{"output_type":"stream","text":["Predicts ->  [array([2.146637], dtype=float32), array([4.343375], dtype=float32), array([3.2421088], dtype=float32), array([9.988994], dtype=float32), array([0.48614928], dtype=float32), array([4.076198], dtype=float32), array([7.036892], dtype=float32), array([7.3835516], dtype=float32), array([8.251253], dtype=float32), array([21.099064], dtype=float32), array([4.560038], dtype=float32), array([6.240519], dtype=float32), array([6.5467963], dtype=float32), array([1.3017473], dtype=float32), array([9.123066], dtype=float32), array([5.091042], dtype=float32), array([9.173713], dtype=float32), array([14.839967], dtype=float32), array([10.557084], dtype=float32), array([27.874994], dtype=float32), array([8.741634], dtype=float32), array([9.58164], dtype=float32), array([7.309558], dtype=float32), array([5.8852158], dtype=float32), array([9.120153], dtype=float32), array([11.009326], dtype=float32), array([5.3465266], dtype=float32), array([12.496046], dtype=float32), array([1.0643876], dtype=float32), array([8.043134], dtype=float32), array([9.923597], dtype=float32), array([4.3907967], dtype=float32), array([1.4359467], dtype=float32), array([2.851491], dtype=float32), array([11.330215], dtype=float32), array([12.060323], dtype=float32), array([8.706669], dtype=float32), array([11.697313], dtype=float32), array([6.819201], dtype=float32), array([5.549999], dtype=float32), array([9.221496], dtype=float32), array([7.88814], dtype=float32), array([9.59614], dtype=float32), array([4.3303766], dtype=float32), array([21.615028], dtype=float32), array([2.752921], dtype=float32), array([7.4594994], dtype=float32), array([10.05572], dtype=float32), array([7.214434], dtype=float32), array([7.9482684], dtype=float32), array([5.90891], dtype=float32), array([8.083256], dtype=float32), array([6.0565066], dtype=float32), array([0.5153388], dtype=float32), array([8.8190565], dtype=float32), array([0.9129436], dtype=float32), array([8.558695], dtype=float32), array([16.963121], dtype=float32), array([4.5479503], dtype=float32), array([12.805654], dtype=float32), array([7.794913], dtype=float32), array([8.4412775], dtype=float32), array([4.7768226], dtype=float32), array([9.310807], dtype=float32), array([8.513081], dtype=float32), array([10.851182], dtype=float32), array([8.695085], dtype=float32), array([8.056407], dtype=float32), array([13.043227], dtype=float32), array([3.5966196], dtype=float32), array([6.3393197], dtype=float32), array([14.856055], dtype=float32), array([6.1188393], dtype=float32), array([0.90907955], dtype=float32), array([9.106612], dtype=float32), array([12.1709795], dtype=float32), array([13.652031], dtype=float32), array([12.122524], dtype=float32), array([8.563813], dtype=float32), array([9.79542], dtype=float32), array([5.9564524], dtype=float32), array([4.824466], dtype=float32), array([9.334288], dtype=float32), array([3.9939842], dtype=float32), array([1.8107365], dtype=float32), array([9.411311], dtype=float32), array([7.0703025], dtype=float32), array([0.46386424], dtype=float32), array([8.983081], dtype=float32), array([6.447975], dtype=float32), array([5.3498583], dtype=float32), array([6.959379], dtype=float32), array([12.60237], dtype=float32), array([12.696998], dtype=float32), array([12.186646], dtype=float32), array([4.979698], dtype=float32), array([25.011425], dtype=float32), array([3.5276313], dtype=float32), array([11.390969], dtype=float32), array([2.4308748], dtype=float32)]\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"pvVDAAuNi2fr","executionInfo":{"status":"ok","timestamp":1621426737578,"user_tz":-120,"elapsed":352,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}}},"source":["test_set['popularity_predicted'] = final_predictions\n","test_set[['id', 'popularity']].to_csv('./sample_submission.csv', header=True, index=False)"],"execution_count":45,"outputs":[]},{"cell_type":"code","metadata":{"id":"vvvSo9dwi52J","colab":{"base_uri":"https://localhost:8080/","height":1000},"executionInfo":{"status":"ok","timestamp":1621426738912,"user_tz":-120,"elapsed":583,"user":{"displayName":"Marina Delaunay","photoUrl":"","userId":"09507219299193984563"}},"outputId":"1afd99d0-fceb-4f15-a395-8ac32ba1a429"},"source":["test_set[['id', 'original_title', 'popularity', 'popularity_predicted']].head(n = 50)"],"execution_count":46,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>id</th>\n","      <th>original_title</th>\n","      <th>popularity</th>\n","      <th>popularity_predicted</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>3001</td>\n","      <td>ディアルガVSパルキアVSダークライ</td>\n","      <td>3.8515</td>\n","      <td>2.1466</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>3002</td>\n","      <td>Attack of the 50 Foot Woman</td>\n","      <td>3.5598</td>\n","      <td>4.3434</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>3003</td>\n","      <td>Addicted to Love</td>\n","      <td>8.0852</td>\n","      <td>3.2421</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>3004</td>\n","      <td>Incendies</td>\n","      <td>8.5960</td>\n","      <td>9.9890</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>3005</td>\n","      <td>Inside Deep Throat</td>\n","      <td>3.2177</td>\n","      <td>0.4861</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>3006</td>\n","      <td>SubUrbia</td>\n","      <td>8.6793</td>\n","      <td>4.0762</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>3007</td>\n","      <td>Drei</td>\n","      <td>4.8989</td>\n","      <td>7.0369</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>3008</td>\n","      <td>The Tigger Movie</td>\n","      <td>7.0234</td>\n","      <td>7.3836</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>3009</td>\n","      <td>Becoming Jane</td>\n","      <td>7.8297</td>\n","      <td>8.2513</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>3010</td>\n","      <td>Toy Story 2</td>\n","      <td>17.5477</td>\n","      <td>21.0991</td>\n","    </tr>\n","    <tr>\n","      <th>10</th>\n","      <td>3011</td>\n","      <td>Cruel World</td>\n","      <td>0.2624</td>\n","      <td>4.5600</td>\n","    </tr>\n","    <tr>\n","      <th>11</th>\n","      <td>3012</td>\n","      <td>Bande de filles</td>\n","      <td>4.2203</td>\n","      <td>6.2405</td>\n","    </tr>\n","    <tr>\n","      <th>12</th>\n","      <td>3013</td>\n","      <td>The Gods Must Be Crazy</td>\n","      <td>10.9735</td>\n","      <td>6.5468</td>\n","    </tr>\n","    <tr>\n","      <th>13</th>\n","      <td>3014</td>\n","      <td>Raising Victor Vargas</td>\n","      <td>1.1787</td>\n","      <td>1.3017</td>\n","    </tr>\n","    <tr>\n","      <th>14</th>\n","      <td>3015</td>\n","      <td>The Brothers Bloom</td>\n","      <td>7.9731</td>\n","      <td>9.1231</td>\n","    </tr>\n","    <tr>\n","      <th>15</th>\n","      <td>3016</td>\n","      <td>Beautiful Boy</td>\n","      <td>2.1148</td>\n","      <td>5.0910</td>\n","    </tr>\n","    <tr>\n","      <th>16</th>\n","      <td>3017</td>\n","      <td>Hot Tub Time Machine</td>\n","      <td>11.9677</td>\n","      <td>9.1737</td>\n","    </tr>\n","    <tr>\n","      <th>17</th>\n","      <td>3018</td>\n","      <td>Transcendence</td>\n","      <td>9.7302</td>\n","      <td>14.8400</td>\n","    </tr>\n","    <tr>\n","      <th>18</th>\n","      <td>3019</td>\n","      <td>All That Jazz</td>\n","      <td>5.6323</td>\n","      <td>10.5571</td>\n","    </tr>\n","    <tr>\n","      <th>19</th>\n","      <td>3020</td>\n","      <td>Titanic</td>\n","      <td>26.8891</td>\n","      <td>27.8750</td>\n","    </tr>\n","    <tr>\n","      <th>20</th>\n","      <td>3021</td>\n","      <td>Very Bad Things</td>\n","      <td>10.4323</td>\n","      <td>8.7416</td>\n","    </tr>\n","    <tr>\n","      <th>21</th>\n","      <td>3022</td>\n","      <td>My Best Friend's Girl</td>\n","      <td>14.1141</td>\n","      <td>9.5816</td>\n","    </tr>\n","    <tr>\n","      <th>22</th>\n","      <td>3023</td>\n","      <td>Broken Bridges</td>\n","      <td>1.1205</td>\n","      <td>7.3096</td>\n","    </tr>\n","    <tr>\n","      <th>23</th>\n","      <td>3024</td>\n","      <td>Suspect Zero</td>\n","      <td>7.5509</td>\n","      <td>5.8852</td>\n","    </tr>\n","    <tr>\n","      <th>24</th>\n","      <td>3025</td>\n","      <td>The Adderall Diaries</td>\n","      <td>4.1423</td>\n","      <td>9.1202</td>\n","    </tr>\n","    <tr>\n","      <th>25</th>\n","      <td>3026</td>\n","      <td>Cape Fear</td>\n","      <td>10.3027</td>\n","      <td>11.0093</td>\n","    </tr>\n","    <tr>\n","      <th>26</th>\n","      <td>3027</td>\n","      <td>The New Adventures of Pippi Longstocking</td>\n","      <td>3.0781</td>\n","      <td>5.3465</td>\n","    </tr>\n","    <tr>\n","      <th>27</th>\n","      <td>3028</td>\n","      <td>The Accountant</td>\n","      <td>13.4657</td>\n","      <td>12.4960</td>\n","    </tr>\n","    <tr>\n","      <th>28</th>\n","      <td>3029</td>\n","      <td>הסיפור של יוסי</td>\n","      <td>1.1620</td>\n","      <td>1.0644</td>\n","    </tr>\n","    <tr>\n","      <th>29</th>\n","      <td>3030</td>\n","      <td>American Outlaws</td>\n","      <td>10.8554</td>\n","      <td>8.0431</td>\n","    </tr>\n","    <tr>\n","      <th>30</th>\n","      <td>3031</td>\n","      <td>Shortbus</td>\n","      <td>11.5477</td>\n","      <td>9.9236</td>\n","    </tr>\n","    <tr>\n","      <th>31</th>\n","      <td>3032</td>\n","      <td>Rang De Basanti</td>\n","      <td>4.1021</td>\n","      <td>4.3908</td>\n","    </tr>\n","    <tr>\n","      <th>32</th>\n","      <td>3033</td>\n","      <td>I Married a Strange Person!</td>\n","      <td>0.8170</td>\n","      <td>1.4359</td>\n","    </tr>\n","    <tr>\n","      <th>33</th>\n","      <td>3034</td>\n","      <td>Ayurveda: Art of Being</td>\n","      <td>0.1130</td>\n","      <td>2.8515</td>\n","    </tr>\n","    <tr>\n","      <th>34</th>\n","      <td>3035</td>\n","      <td>Obvious Child</td>\n","      <td>6.4806</td>\n","      <td>11.3302</td>\n","    </tr>\n","    <tr>\n","      <th>35</th>\n","      <td>3036</td>\n","      <td>Predestination</td>\n","      <td>10.5242</td>\n","      <td>12.0603</td>\n","    </tr>\n","    <tr>\n","      <th>36</th>\n","      <td>3037</td>\n","      <td>Honey</td>\n","      <td>11.8640</td>\n","      <td>8.7067</td>\n","    </tr>\n","    <tr>\n","      <th>37</th>\n","      <td>3038</td>\n","      <td>The Boy</td>\n","      <td>17.0289</td>\n","      <td>11.6973</td>\n","    </tr>\n","    <tr>\n","      <th>38</th>\n","      <td>3039</td>\n","      <td>Ernest et Célestine</td>\n","      <td>7.2258</td>\n","      <td>6.8192</td>\n","    </tr>\n","    <tr>\n","      <th>39</th>\n","      <td>3040</td>\n","      <td>When the Wind Blows</td>\n","      <td>3.9331</td>\n","      <td>5.5500</td>\n","    </tr>\n","    <tr>\n","      <th>40</th>\n","      <td>3041</td>\n","      <td>Young Adult</td>\n","      <td>6.1429</td>\n","      <td>9.2215</td>\n","    </tr>\n","    <tr>\n","      <th>41</th>\n","      <td>3042</td>\n","      <td>Best Seller</td>\n","      <td>2.1738</td>\n","      <td>7.8881</td>\n","    </tr>\n","    <tr>\n","      <th>42</th>\n","      <td>3043</td>\n","      <td>Piranha 3DD</td>\n","      <td>5.6152</td>\n","      <td>9.5961</td>\n","    </tr>\n","    <tr>\n","      <th>43</th>\n","      <td>3044</td>\n","      <td>Barbecue</td>\n","      <td>4.7073</td>\n","      <td>4.3304</td>\n","    </tr>\n","    <tr>\n","      <th>44</th>\n","      <td>3045</td>\n","      <td>Captain America: The First Avenger</td>\n","      <td>19.3236</td>\n","      <td>21.6150</td>\n","    </tr>\n","    <tr>\n","      <th>45</th>\n","      <td>3046</td>\n","      <td>The Comfort of Strangers</td>\n","      <td>2.0695</td>\n","      <td>2.7529</td>\n","    </tr>\n","    <tr>\n","      <th>46</th>\n","      <td>3047</td>\n","      <td>To Have and Have Not</td>\n","      <td>10.0032</td>\n","      <td>7.4595</td>\n","    </tr>\n","    <tr>\n","      <th>47</th>\n","      <td>3048</td>\n","      <td>Running Scared</td>\n","      <td>7.4355</td>\n","      <td>10.0557</td>\n","    </tr>\n","    <tr>\n","      <th>48</th>\n","      <td>3049</td>\n","      <td>Short Circuit 2</td>\n","      <td>7.3468</td>\n","      <td>7.2144</td>\n","    </tr>\n","    <tr>\n","      <th>49</th>\n","      <td>3050</td>\n","      <td>Saturday Night Fever</td>\n","      <td>13.6572</td>\n","      <td>7.9483</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["      id  ... popularity_predicted\n","0   3001  ...               2.1466\n","1   3002  ...               4.3434\n","2   3003  ...               3.2421\n","3   3004  ...               9.9890\n","4   3005  ...               0.4861\n","5   3006  ...               4.0762\n","6   3007  ...               7.0369\n","7   3008  ...               7.3836\n","8   3009  ...               8.2513\n","9   3010  ...              21.0991\n","10  3011  ...               4.5600\n","11  3012  ...               6.2405\n","12  3013  ...               6.5468\n","13  3014  ...               1.3017\n","14  3015  ...               9.1231\n","15  3016  ...               5.0910\n","16  3017  ...               9.1737\n","17  3018  ...              14.8400\n","18  3019  ...              10.5571\n","19  3020  ...              27.8750\n","20  3021  ...               8.7416\n","21  3022  ...               9.5816\n","22  3023  ...               7.3096\n","23  3024  ...               5.8852\n","24  3025  ...               9.1202\n","25  3026  ...              11.0093\n","26  3027  ...               5.3465\n","27  3028  ...              12.4960\n","28  3029  ...               1.0644\n","29  3030  ...               8.0431\n","30  3031  ...               9.9236\n","31  3032  ...               4.3908\n","32  3033  ...               1.4359\n","33  3034  ...               2.8515\n","34  3035  ...              11.3302\n","35  3036  ...              12.0603\n","36  3037  ...               8.7067\n","37  3038  ...              11.6973\n","38  3039  ...               6.8192\n","39  3040  ...               5.5500\n","40  3041  ...               9.2215\n","41  3042  ...               7.8881\n","42  3043  ...               9.5961\n","43  3044  ...               4.3304\n","44  3045  ...              21.6150\n","45  3046  ...               2.7529\n","46  3047  ...               7.4595\n","47  3048  ...              10.0557\n","48  3049  ...               7.2144\n","49  3050  ...               7.9483\n","\n","[50 rows x 4 columns]"]},"metadata":{"tags":[]},"execution_count":46}]}]}