{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# Import libs\n\n\nimport pandas as pd\nimport numpy as np\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom matplotlib.colors import ListedColormap\n\nimport statsmodels.formula.api as smf\n\nfrom scipy import stats\n\nfrom sklearn import preprocessing\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.datasets import make_moons, make_circles, make_classification\n\nfrom sklearn.ensemble import RandomForestRegressor\nfrom sklearn.linear_model import LinearRegression\nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.neural_network import MLPClassifier\nfrom sklearn.neighbors import KNeighborsClassifier\nfrom sklearn.svm import SVC\nfrom sklearn import linear_model, svm, gaussian_process\nfrom sklearn.gaussian_process import GaussianProcessClassifier\nfrom sklearn.gaussian_process.kernels import RBF\nfrom sklearn.tree import DecisionTreeClassifier\nfrom sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\nfrom sklearn.naive_bayes import GaussianNB\nfrom sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\nfrom sklearn.metrics import accuracy_score\n\nfrom sklearn.linear_model import ElasticNet, Lasso,  BayesianRidge, LassoLarsIC\nfrom sklearn.kernel_ridge import KernelRidge\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.preprocessing import RobustScaler\nfrom sklearn.base import BaseEstimator, TransformerMixin, RegressorMixin, clone\nfrom sklearn.model_selection import KFold, cross_val_score, train_test_split\nimport lightgbm as lgb\n\nfrom sklearn import preprocessing\nfrom sklearn import utils\n\nimport statsmodels.formula.api as smf\nimport warnings\nwarnings.filterwarnings(\"ignore\")\n\nimport os\nprint(os.listdir(\"../input\"))","execution_count":1,"outputs":[{"output_type":"stream","text":"['train.csv', 'merchants.csv', 'sample_submission.csv', 'test.csv', 'historical_transactions.csv', 'Data_Dictionary.xlsx', 'new_merchant_transactions.csv']\n","name":"stdout"}]},{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"# Import models\n\n# 1) GradientBoostingRegressor Model\nfrom sklearn.ensemble import GradientBoostingRegressor\nimport statsmodels.formula.api as smf\nfrom sklearn.preprocessing import scale\nGBR = GradientBoostingRegressor()\n\n# 2) Logistic Regression Model\nfrom sklearn.linear_model import LogisticRegression\nLR = LogisticRegression()\n\n# 3) Aplly Random Forest Model\nfrom sklearn.ensemble import RandomForestClassifier\nRF = RandomForestClassifier(n_estimators=100)\n\n# 4) Aplly XGBOOST Model\nfrom xgboost import XGBClassifier\nXGB = XGBClassifier()\n\n# 5) Aplly KNeighbors Model\nfrom sklearn.neighbors import KNeighborsClassifier\nknn = KNeighborsClassifier()\n\n# 6) Aplly SVC Model\nSVC = SVC(probability=True)\n\n# 7) Aplly Decision Tree Model\nDTC = DecisionTreeClassifier()\n\n# 8) Aplly GaussianNB Model\nGNB = GaussianNB()\n\n# 9) Aplly Neural Model\nNN = MLPClassifier(hidden_layer_sizes=(100,100,50))\n\n# 10) Aplly lasso\nlasso = make_pipeline(RobustScaler(), Lasso(alpha =0.0005, random_state=1))\n\n# 11) Apply Elastic Net\nENet = make_pipeline(RobustScaler(), ElasticNet(alpha=0.0005, l1_ratio=.9, random_state=3))\n\n# 12) Apply Kernel Ridge\nKRR = KernelRidge(alpha=0.6, kernel='polynomial', degree=2, coef0=2.5)\n\n# 13) Apply LGBMRegressor\nmodel_lgb = lgb.LGBMRegressor(objective='regression',num_leaves=5,\n                              learning_rate=0.05, n_estimators=720,\n                              max_bin = 55, bagging_fraction = 0.8,\n                              bagging_freq = 5, feature_fraction = 0.2319,\n                              feature_fraction_seed=9, bagging_seed=9,\n                              min_data_in_leaf =6, min_sum_hessian_in_leaf = 11)\n","execution_count":2,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Read Dataframes\n\n# Read in the dataframes\ntrain = pd.read_csv('../input/train.csv')\ntest = pd.read_csv('../input/test.csv')\nmerch = pd.read_csv('../input/merchants.csv')\nhist = pd.read_csv('../input/historical_transactions.csv')\nsample = pd.read_csv('../input/sample_submission.csv')\n\ndf_train_original = train\ndf_test_original = test","execution_count":3,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Shapes\n\n# Print data shapes\nprint('Files Load!!!')\nprint('sample submission shape', sample.shape)\n\n# Train and Test\nprint('* * * Merchant File * * * ')\nprint('merchants shape', merch.shape)\n\n# Historical File\nprint('* * * Historical File * * * ')\nprint('historical_transactions', hist.shape)\n\n# Train and Test\nprint('* * * Train and Test File * * * ')\nprint('Train File - Original: ', train.shape)\nprint('Test File - Original: ', test.shape)","execution_count":4,"outputs":[{"output_type":"stream","text":"Files Load!!!\nsample submission shape (123623, 2)\n* * * Merchant File * * * \nmerchants shape (334696, 22)\n* * * Historical File * * * \nhistorical_transactions (29112361, 14)\n* * * Train and Test File * * * \nTrain File - Original:  (201917, 6)\nTest File - Original:  (123623, 5)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Drop columns\n\ntrain = train.fillna(0)\ntest = test.fillna(0)\n\ntrain['first1'] = train['first_active_month'].str[0:4]\ntrain['first2'] = train['first_active_month'].str[5:7]\ntrain = train.fillna(0)\ntrain['first1'] = train['first1'].astype(int)\ntrain['first2'] = train['first2'].astype(int)\n\ntest['first1'] = test['first_active_month'].str[0:4]\ntest['first2'] = test['first_active_month'].str[5:7]\ntest = test.fillna(0)\ntest['first1'] = test['first1'].astype(int)\ntest['first2'] = test['first2'].astype(int)\n\ntrain = train.drop(['first_active_month'], axis=1)\ntest = test.drop(['first_active_month'], axis=1)","execution_count":5,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Train and Test\nprint('Train File - Depois de todas Ops: ', train.shape)\nprint('Test File - Depois de todas Ops: ', test.shape)\n\nall_data = pd.concat((train.loc[:,'card_id':'first2'],\n                      test.loc[:,'card_id':'first2']))\n\nprint('Após concatenar os arquivos: ', all_data.shape)","execution_count":6,"outputs":[{"output_type":"stream","text":"Train File - Depois de todas Ops:  (201917, 7)\nTest File - Depois de todas Ops:  (123623, 6)\nApós concatenar os arquivos:  (325540, 7)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Substitui os campos nulos pelas médias da coluna em questão\nall_data = all_data.fillna(all_data.mean())","execution_count":7,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data = all_data.fillna(0)","execution_count":8,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data","execution_count":9,"outputs":[{"output_type":"execute_result","execution_count":9,"data":{"text/plain":"                card_id  feature_1    ...     first2    target\n0       C_ID_92a2005557          5    ...          6 -0.820283\n1       C_ID_3d0044924f          4    ...          1  0.392913\n2       C_ID_d639edf6cd          2    ...          8  0.688056\n3       C_ID_186d6a6901          4    ...          9  0.142495\n4       C_ID_cdbd2c0db2          1    ...         11 -0.159749\n5       C_ID_0894217f2f          4    ...          9  0.871585\n6       C_ID_7e63323c00          3    ...         12  0.230129\n7       C_ID_dfa21fc124          3    ...          9  2.135850\n8       C_ID_fe0fdac8ea          2    ...          8 -0.065406\n9       C_ID_bf62c0b49d          2    ...          8  0.300062\n10      C_ID_92853cdb2c          5    ...         10 -1.029562\n11      C_ID_269d816788          2    ...          3 -2.433761\n12      C_ID_61d50d7057          5    ...          9  1.304452\n13      C_ID_4e07413433          2    ...          4  0.093633\n14      C_ID_b6302b31c6          3    ...          8  1.597750\n15      C_ID_3b69154173          2    ...          2  1.010937\n16      C_ID_9feec11e78          2    ...          7  1.132236\n17      C_ID_f6658dbefe          4    ...         12  0.318062\n18      C_ID_4a7dda0f9e          2    ...         11  0.757280\n19      C_ID_6adae2a906          4    ...          1  0.334337\n20      C_ID_0b70ca7347          5    ...          9  1.442726\n21      C_ID_25b2509282          1    ...          6  4.747602\n22      C_ID_82fb11db22          2    ...          9  0.464404\n23      C_ID_e03db05bde          2    ...         10 -0.529658\n24      C_ID_dcb7c76747          5    ...         10 -0.084951\n25      C_ID_749cfc0c3a          2    ...          4  1.525862\n26      C_ID_4a97126360          2    ...          7 -0.090744\n27      C_ID_17dea8d446          3    ...          9 -1.361720\n28      C_ID_fcef0dfd67          5    ...          1  0.727093\n29      C_ID_82868156f0          5    ...          6  1.143863\n...                 ...        ...    ...        ...       ...\n123593  C_ID_cd42a5db89          2    ...          7 -0.393636\n123594  C_ID_715b9a183d          3    ...          8 -0.393636\n123595  C_ID_9b25df3309          2    ...         11 -0.393636\n123596  C_ID_043b3c2ccd          3    ...          2 -0.393636\n123597  C_ID_7685eae7b2          3    ...          1 -0.393636\n123598  C_ID_7ee1a21a95          4    ...         11 -0.393636\n123599  C_ID_2bf12fed6b          2    ...          9 -0.393636\n123600  C_ID_2932a6ea37          1    ...         12 -0.393636\n123601  C_ID_cc33d0d508          2    ...          9 -0.393636\n123602  C_ID_8852708dcb          4    ...          1 -0.393636\n123603  C_ID_c3e88ef62d          1    ...         11 -0.393636\n123604  C_ID_650557248c          2    ...         10 -0.393636\n123605  C_ID_f14af9cb56          1    ...          5 -0.393636\n123606  C_ID_068dcf31e8          5    ...          7 -0.393636\n123607  C_ID_6659f4625a          3    ...         12 -0.393636\n123608  C_ID_dc842bba27          3    ...         12 -0.393636\n123609  C_ID_bd625edce0          3    ...         12 -0.393636\n123610  C_ID_278446cc2f          3    ...          5 -0.393636\n123611  C_ID_ad358eb92e          5    ...          8 -0.393636\n123612  C_ID_7d9bcec574          3    ...          4 -0.393636\n123613  C_ID_b588b35e15          3    ...         11 -0.393636\n123614  C_ID_7147e348c3          5    ...          1 -0.393636\n123615  C_ID_727d3ec5da          5    ...         12 -0.393636\n123616  C_ID_107d93e20c          4    ...          5 -0.393636\n123617  C_ID_0e74bd4966          3    ...         10 -0.393636\n123618  C_ID_7a239d2eda          2    ...         10 -0.393636\n123619  C_ID_75ace375ae          3    ...          9 -0.393636\n123620  C_ID_21d56d950c          5    ...          9 -0.393636\n123621  C_ID_6c46fc5a9d          2    ...          6 -0.393636\n123622  C_ID_87e7979a5f          5    ...         10 -0.393636\n\n[325540 rows x 7 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>card_id</th>\n      <th>feature_1</th>\n      <th>feature_2</th>\n      <th>feature_3</th>\n      <th>first1</th>\n      <th>first2</th>\n      <th>target</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>C_ID_92a2005557</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>6</td>\n      <td>-0.820283</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>C_ID_3d0044924f</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>0.392913</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>C_ID_d639edf6cd</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>0.688056</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>C_ID_186d6a6901</td>\n      <td>4</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>0.142495</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>C_ID_cdbd2c0db2</td>\n      <td>1</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.159749</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>C_ID_0894217f2f</td>\n      <td>4</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>9</td>\n      <td>0.871585</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>C_ID_7e63323c00</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>0.230129</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>C_ID_dfa21fc124</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>2.135850</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>C_ID_fe0fdac8ea</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>-0.065406</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>C_ID_bf62c0b49d</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>0.300062</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>C_ID_92853cdb2c</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-1.029562</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>C_ID_269d816788</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>3</td>\n      <td>-2.433761</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>C_ID_61d50d7057</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>1.304452</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>C_ID_4e07413433</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>4</td>\n      <td>0.093633</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>C_ID_b6302b31c6</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>1.597750</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>C_ID_3b69154173</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>2</td>\n      <td>1.010937</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>C_ID_9feec11e78</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>7</td>\n      <td>1.132236</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>C_ID_f6658dbefe</td>\n      <td>4</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>0.318062</td>\n    </tr>\n    <tr>\n      <th>18</th>\n      <td>C_ID_4a7dda0f9e</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>0.757280</td>\n    </tr>\n    <tr>\n      <th>19</th>\n      <td>C_ID_6adae2a906</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>0.334337</td>\n    </tr>\n    <tr>\n      <th>20</th>\n      <td>C_ID_0b70ca7347</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>1.442726</td>\n    </tr>\n    <tr>\n      <th>21</th>\n      <td>C_ID_25b2509282</td>\n      <td>1</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>6</td>\n      <td>4.747602</td>\n    </tr>\n    <tr>\n      <th>22</th>\n      <td>C_ID_82fb11db22</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>0.464404</td>\n    </tr>\n    <tr>\n      <th>23</th>\n      <td>C_ID_e03db05bde</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.529658</td>\n    </tr>\n    <tr>\n      <th>24</th>\n      <td>C_ID_dcb7c76747</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2015</td>\n      <td>10</td>\n      <td>-0.084951</td>\n    </tr>\n    <tr>\n      <th>25</th>\n      <td>C_ID_749cfc0c3a</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>4</td>\n      <td>1.525862</td>\n    </tr>\n    <tr>\n      <th>26</th>\n      <td>C_ID_4a97126360</td>\n      <td>2</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>7</td>\n      <td>-0.090744</td>\n    </tr>\n    <tr>\n      <th>27</th>\n      <td>C_ID_17dea8d446</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-1.361720</td>\n    </tr>\n    <tr>\n      <th>28</th>\n      <td>C_ID_fcef0dfd67</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>1</td>\n      <td>0.727093</td>\n    </tr>\n    <tr>\n      <th>29</th>\n      <td>C_ID_82868156f0</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>6</td>\n      <td>1.143863</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>123593</th>\n      <td>C_ID_cd42a5db89</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>7</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123594</th>\n      <td>C_ID_715b9a183d</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123595</th>\n      <td>C_ID_9b25df3309</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123596</th>\n      <td>C_ID_043b3c2ccd</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>2</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123597</th>\n      <td>C_ID_7685eae7b2</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123598</th>\n      <td>C_ID_7ee1a21a95</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123599</th>\n      <td>C_ID_2bf12fed6b</td>\n      <td>2</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123600</th>\n      <td>C_ID_2932a6ea37</td>\n      <td>1</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123601</th>\n      <td>C_ID_cc33d0d508</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123602</th>\n      <td>C_ID_8852708dcb</td>\n      <td>4</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123603</th>\n      <td>C_ID_c3e88ef62d</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123604</th>\n      <td>C_ID_650557248c</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123605</th>\n      <td>C_ID_f14af9cb56</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123606</th>\n      <td>C_ID_068dcf31e8</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>7</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123607</th>\n      <td>C_ID_6659f4625a</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>12</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123608</th>\n      <td>C_ID_dc842bba27</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123609</th>\n      <td>C_ID_bd625edce0</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123610</th>\n      <td>C_ID_278446cc2f</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123611</th>\n      <td>C_ID_ad358eb92e</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123612</th>\n      <td>C_ID_7d9bcec574</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>4</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123613</th>\n      <td>C_ID_b588b35e15</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2015</td>\n      <td>11</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123614</th>\n      <td>C_ID_7147e348c3</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123615</th>\n      <td>C_ID_727d3ec5da</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2014</td>\n      <td>12</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123616</th>\n      <td>C_ID_107d93e20c</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123617</th>\n      <td>C_ID_0e74bd4966</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123618</th>\n      <td>C_ID_7a239d2eda</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123619</th>\n      <td>C_ID_75ace375ae</td>\n      <td>3</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123620</th>\n      <td>C_ID_21d56d950c</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>9</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123621</th>\n      <td>C_ID_6c46fc5a9d</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>6</td>\n      <td>-0.393636</td>\n    </tr>\n    <tr>\n      <th>123622</th>\n      <td>C_ID_87e7979a5f</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-0.393636</td>\n    </tr>\n  </tbody>\n</table>\n<p>325540 rows × 7 columns</p>\n</div>"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Substituindo\n\nhist = hist.fillna(0)\nDfhtGroup = pd.DataFrame()\n\nprint('# # # # # GERANDO CAMPOS DO HISTÓRICO # # # # #')\n\nprint('01 - Gerando o campo card_id')\nListGroup = hist.groupby(['card_id'])['card_id'].count()\nDfhtGroup['card_id'] = ListGroup.index\n\nprint('02 - Gerando o campo pur_mean')\nListGroup = pd.DataFrame(hist.groupby(['card_id'], as_index=False)['purchase_amount'].mean())\nDfhtGroup['pur_mean'] = ListGroup['purchase_amount']\n\nprint('03 - Gerando o campo Mon_mean')\nListGroup = pd.DataFrame(hist.groupby(['card_id'], as_index=False)['month_lag'].mean())\nDfhtGroup['Mon_mean'] = ListGroup['month_lag']\n\nprint('04 - Gerando o campo Count')\nListGroup = pd.DataFrame(hist.groupby(['card_id'], as_index=False)['authorized_flag'].count())\nDfhtGroup['Count'] = ListGroup['authorized_flag']\n\nprint('13 - Gerando o campo merchant_id - sera usado no Join')\nListGroup = pd.DataFrame(hist.groupby('card_id', as_index=False)['merchant_id'].agg(lambda x: x.value_counts().index[0]))\nDfhtGroup['merchant_id'] = ListGroup['merchant_id']\n\nprint ('Shape do Df principal - all_data: ', all_data.shape)\nprint ('Shape do Df com campos do histórico - DfhtGroup: ', DfhtGroup.shape)\nprint ('Shape do Df com campos de Mercados - merch: ', merch.shape)\n\ndfJoin  = pd.DataFrame()\ndfJoin  = pd.merge(all_data, DfhtGroup, on='card_id')\ndfJoin  = dfJoin.fillna(0)\n\nall_data = dfJoin\n\nprint ('All_data depois do primeiro (Historico): ', all_data.shape)\n\nmerch = merch.fillna(0)\nmerch2 = merch.drop_duplicates(subset=['merchant_id'], keep=False)\n\nmerged = pd.DataFrame()\nmerged = all_data.merge(merch2, indicator=True, how='outer')\nTot_merged = merged[merged['_merge'] != 'left_only']\nTot_merged = merged[merged['_merge'] != 'right_only']\n\nall_data_sel = all_data\n\n# Campos que não precisam de transformação - serão transformados no Get Dummies:\nall_data_sel['most_recent_sales_range'] = Tot_merged['most_recent_sales_range']\nall_data_sel['most_recent_purchases_range'] = Tot_merged['most_recent_purchases_range']\n\n# Campos que são númericos mas deveriam ser categoricos: tranformar em str e depois passar Get Dummies\n#all_data_sel['subsector_id'] = Tot_merged['subsector_id'].astype(str)\n#all_data_sel['city_id'] = Tot_merged['city_id'].astype(str)\n#all_data_sel['state_id'] = Tot_merged['state_id'].astype(str)\n\nall_data = all_data_sel\n\nprint ('All_data depois do segundo Join (Mercado): ', all_data.shape)","execution_count":10,"outputs":[{"output_type":"stream","text":"# # # # # GERANDO CAMPOS DO HISTÓRICO # # # # #\n01 - Gerando o campo card_id\n02 - Gerando o campo pur_mean\n03 - Gerando o campo Mon_mean\n04 - Gerando o campo Count\n13 - Gerando o campo merchant_id - sera usado no Join\nShape do Df principal - all_data:  (325540, 7)\nShape do Df com campos do histórico - DfhtGroup:  (325540, 5)\nShape do Df com campos de Mercados - merch:  (334696, 22)\nAll_data depois do primeiro (Historico):  (325540, 11)\nAll_data depois do segundo Join (Mercado):  (325540, 13)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data","execution_count":11,"outputs":[{"output_type":"execute_result","execution_count":11,"data":{"text/plain":"                card_id             ...               most_recent_purchases_range\n0       C_ID_92a2005557             ...                                         B\n1       C_ID_3d0044924f             ...                                         B\n2       C_ID_d639edf6cd             ...                                         B\n3       C_ID_186d6a6901             ...                                         B\n4       C_ID_cdbd2c0db2             ...                                         B\n5       C_ID_0894217f2f             ...                                         B\n6       C_ID_7e63323c00             ...                                         B\n7       C_ID_dfa21fc124             ...                                         B\n8       C_ID_fe0fdac8ea             ...                                         B\n9       C_ID_bf62c0b49d             ...                                         B\n10      C_ID_92853cdb2c             ...                                         B\n11      C_ID_269d816788             ...                                         A\n12      C_ID_61d50d7057             ...                                         A\n13      C_ID_4e07413433             ...                                         A\n14      C_ID_b6302b31c6             ...                                         A\n15      C_ID_3b69154173             ...                                         A\n16      C_ID_9feec11e78             ...                                         A\n17      C_ID_f6658dbefe             ...                                         A\n18      C_ID_4a7dda0f9e             ...                                         A\n19      C_ID_6adae2a906             ...                                         A\n20      C_ID_0b70ca7347             ...                                         A\n21      C_ID_25b2509282             ...                                         A\n22      C_ID_82fb11db22             ...                                         A\n23      C_ID_e03db05bde             ...                                         A\n24      C_ID_dcb7c76747             ...                                         A\n25      C_ID_749cfc0c3a             ...                                         A\n26      C_ID_4a97126360             ...                                         A\n27      C_ID_17dea8d446             ...                                         A\n28      C_ID_fcef0dfd67             ...                                         A\n29      C_ID_82868156f0             ...                                         A\n...                 ...             ...                                       ...\n325510  C_ID_cd42a5db89             ...                                         D\n325511  C_ID_715b9a183d             ...                                         D\n325512  C_ID_9b25df3309             ...                                         E\n325513  C_ID_043b3c2ccd             ...                                         D\n325514  C_ID_7685eae7b2             ...                                         C\n325515  C_ID_7ee1a21a95             ...                                         E\n325516  C_ID_2bf12fed6b             ...                                         D\n325517  C_ID_2932a6ea37             ...                                         D\n325518  C_ID_cc33d0d508             ...                                         E\n325519  C_ID_8852708dcb             ...                                         D\n325520  C_ID_c3e88ef62d             ...                                         D\n325521  C_ID_650557248c             ...                                         E\n325522  C_ID_f14af9cb56             ...                                         C\n325523  C_ID_068dcf31e8             ...                                         C\n325524  C_ID_6659f4625a             ...                                         D\n325525  C_ID_dc842bba27             ...                                         E\n325526  C_ID_bd625edce0             ...                                         D\n325527  C_ID_278446cc2f             ...                                         D\n325528  C_ID_ad358eb92e             ...                                         C\n325529  C_ID_7d9bcec574             ...                                         C\n325530  C_ID_b588b35e15             ...                                         E\n325531  C_ID_7147e348c3             ...                                         C\n325532  C_ID_727d3ec5da             ...                                         D\n325533  C_ID_107d93e20c             ...                                         C\n325534  C_ID_0e74bd4966             ...                                         D\n325535  C_ID_7a239d2eda             ...                                         E\n325536  C_ID_75ace375ae             ...                                         C\n325537  C_ID_21d56d950c             ...                                         C\n325538  C_ID_6c46fc5a9d             ...                                         E\n325539  C_ID_87e7979a5f             ...                                         D\n\n[325540 rows x 13 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>card_id</th>\n      <th>feature_1</th>\n      <th>feature_2</th>\n      <th>feature_3</th>\n      <th>first1</th>\n      <th>first2</th>\n      <th>target</th>\n      <th>pur_mean</th>\n      <th>Mon_mean</th>\n      <th>Count</th>\n      <th>merchant_id</th>\n      <th>most_recent_sales_range</th>\n      <th>most_recent_purchases_range</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>C_ID_92a2005557</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>6</td>\n      <td>-0.820283</td>\n      <td>-0.638341</td>\n      <td>-3.911538</td>\n      <td>260</td>\n      <td>M_ID_1a81c358a3</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>C_ID_3d0044924f</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>0.392913</td>\n      <td>-0.600018</td>\n      <td>-5.031429</td>\n      <td>350</td>\n      <td>M_ID_940fb4498f</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>C_ID_d639edf6cd</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>0.688056</td>\n      <td>-0.678311</td>\n      <td>-8.604651</td>\n      <td>43</td>\n      <td>M_ID_5634fd83e0</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>C_ID_186d6a6901</td>\n      <td>4</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>0.142495</td>\n      <td>-0.642745</td>\n      <td>-2.831169</td>\n      <td>77</td>\n      <td>M_ID_00a6ca8a8a</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>C_ID_cdbd2c0db2</td>\n      <td>1</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.159749</td>\n      <td>-0.366073</td>\n      <td>-1.285714</td>\n      <td>133</td>\n      <td>M_ID_cecefd9589</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>C_ID_0894217f2f</td>\n      <td>4</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>9</td>\n      <td>0.871585</td>\n      <td>-0.417462</td>\n      <td>-2.545455</td>\n      <td>33</td>\n      <td>M_ID_12637a44e9</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>C_ID_7e63323c00</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>0.230129</td>\n      <td>-0.713160</td>\n      <td>-6.646154</td>\n      <td>260</td>\n      <td>M_ID_548f18cdd2</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>C_ID_dfa21fc124</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>2.135850</td>\n      <td>-0.616629</td>\n      <td>-1.727273</td>\n      <td>22</td>\n      <td>M_ID_00a6ca8a8a</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>C_ID_fe0fdac8ea</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>-0.065406</td>\n      <td>-0.580558</td>\n      <td>-4.200000</td>\n      <td>15</td>\n      <td>M_ID_b9dcf28cb9</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>C_ID_bf62c0b49d</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>0.300062</td>\n      <td>-0.681390</td>\n      <td>-5.876106</td>\n      <td>113</td>\n      <td>M_ID_6cb0602e94</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>C_ID_92853cdb2c</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-1.029562</td>\n      <td>-0.632705</td>\n      <td>-5.195876</td>\n      <td>97</td>\n      <td>M_ID_a30f2f4685</td>\n      <td>C</td>\n      <td>B</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>C_ID_269d816788</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>3</td>\n      <td>-2.433761</td>\n      <td>-0.654233</td>\n      <td>-7.315789</td>\n      <td>19</td>\n      <td>M_ID_2c049b075c</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>C_ID_61d50d7057</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>1.304452</td>\n      <td>-0.640602</td>\n      <td>-2.216216</td>\n      <td>37</td>\n      <td>M_ID_1d56499c64</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>C_ID_4e07413433</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>4</td>\n      <td>0.093633</td>\n      <td>-0.625212</td>\n      <td>-3.050000</td>\n      <td>20</td>\n      <td>M_ID_ba886060ed</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>C_ID_b6302b31c6</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>1.597750</td>\n      <td>-0.570584</td>\n      <td>-0.784615</td>\n      <td>65</td>\n      <td>M_ID_50f575c681</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>C_ID_3b69154173</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>2</td>\n      <td>1.010937</td>\n      <td>-0.638843</td>\n      <td>-5.489474</td>\n      <td>190</td>\n      <td>M_ID_ae16956a28</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>C_ID_9feec11e78</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>7</td>\n      <td>1.132236</td>\n      <td>-0.633631</td>\n      <td>-4.191304</td>\n      <td>115</td>\n      <td>M_ID_57df19bf28</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>C_ID_f6658dbefe</td>\n      <td>4</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>0.318062</td>\n      <td>-0.612381</td>\n      <td>-6.658537</td>\n      <td>82</td>\n      <td>M_ID_6c86c6f9b9</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>18</th>\n      <td>C_ID_4a7dda0f9e</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>0.757280</td>\n      <td>-0.558578</td>\n      <td>-2.074074</td>\n      <td>27</td>\n      <td>M_ID_25ae9d5e52</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>19</th>\n      <td>C_ID_6adae2a906</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>0.334337</td>\n      <td>-0.699261</td>\n      <td>-7.109409</td>\n      <td>457</td>\n      <td>M_ID_e5374dabc0</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>20</th>\n      <td>C_ID_0b70ca7347</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>1.442726</td>\n      <td>-0.624178</td>\n      <td>-2.061111</td>\n      <td>180</td>\n      <td>M_ID_1d2ad7faef</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>21</th>\n      <td>C_ID_25b2509282</td>\n      <td>1</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>6</td>\n      <td>4.747602</td>\n      <td>0.016976</td>\n      <td>-8.214286</td>\n      <td>28</td>\n      <td>M_ID_5a0a412718</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>22</th>\n      <td>C_ID_82fb11db22</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>0.464404</td>\n      <td>-0.525673</td>\n      <td>-2.625000</td>\n      <td>16</td>\n      <td>M_ID_5c59739332</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>23</th>\n      <td>C_ID_e03db05bde</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.529658</td>\n      <td>-0.662277</td>\n      <td>-1.242424</td>\n      <td>132</td>\n      <td>M_ID_00a6ca8a8a</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>24</th>\n      <td>C_ID_dcb7c76747</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2015</td>\n      <td>10</td>\n      <td>-0.084951</td>\n      <td>-0.616409</td>\n      <td>-6.265060</td>\n      <td>166</td>\n      <td>M_ID_9e84cda3b1</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>25</th>\n      <td>C_ID_749cfc0c3a</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>4</td>\n      <td>1.525862</td>\n      <td>-0.707445</td>\n      <td>-6.355556</td>\n      <td>135</td>\n      <td>M_ID_1d8085cf5d</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>26</th>\n      <td>C_ID_4a97126360</td>\n      <td>2</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>7</td>\n      <td>-0.090744</td>\n      <td>-0.659593</td>\n      <td>-3.427083</td>\n      <td>96</td>\n      <td>M_ID_f7e47257c9</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>27</th>\n      <td>C_ID_17dea8d446</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-1.361720</td>\n      <td>-0.583317</td>\n      <td>-2.250000</td>\n      <td>20</td>\n      <td>M_ID_b680418f90</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>28</th>\n      <td>C_ID_fcef0dfd67</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>1</td>\n      <td>0.727093</td>\n      <td>-0.612910</td>\n      <td>-5.754545</td>\n      <td>110</td>\n      <td>M_ID_cfd42d87ed</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>29</th>\n      <td>C_ID_82868156f0</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>6</td>\n      <td>1.143863</td>\n      <td>-0.427538</td>\n      <td>-1.317365</td>\n      <td>167</td>\n      <td>M_ID_56e0c47cb1</td>\n      <td>A</td>\n      <td>A</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>325510</th>\n      <td>C_ID_cd42a5db89</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>7</td>\n      <td>-0.393636</td>\n      <td>-0.660766</td>\n      <td>-5.460843</td>\n      <td>332</td>\n      <td>M_ID_183cdc29e4</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325511</th>\n      <td>C_ID_715b9a183d</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>8</td>\n      <td>-0.393636</td>\n      <td>-0.661096</td>\n      <td>-1.882353</td>\n      <td>17</td>\n      <td>M_ID_00a6ca8a8a</td>\n      <td>D</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325512</th>\n      <td>C_ID_9b25df3309</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n      <td>-0.687154</td>\n      <td>-1.234375</td>\n      <td>64</td>\n      <td>M_ID_1f4773aa76</td>\n      <td>E</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325513</th>\n      <td>C_ID_043b3c2ccd</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>2</td>\n      <td>-0.393636</td>\n      <td>-0.727966</td>\n      <td>-5.441558</td>\n      <td>77</td>\n      <td>M_ID_a30d553adf</td>\n      <td>D</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325514</th>\n      <td>C_ID_7685eae7b2</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n      <td>-0.720692</td>\n      <td>-7.335938</td>\n      <td>128</td>\n      <td>M_ID_cbf911cbb0</td>\n      <td>C</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325515</th>\n      <td>C_ID_7ee1a21a95</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n      <td>-0.545686</td>\n      <td>-1.618182</td>\n      <td>55</td>\n      <td>M_ID_8f7b4791e1</td>\n      <td>D</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325516</th>\n      <td>C_ID_2bf12fed6b</td>\n      <td>2</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n      <td>-0.559791</td>\n      <td>-1.870968</td>\n      <td>31</td>\n      <td>M_ID_a16b37f080</td>\n      <td>C</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325517</th>\n      <td>C_ID_2932a6ea37</td>\n      <td>1</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n      <td>-0.522650</td>\n      <td>-6.000000</td>\n      <td>183</td>\n      <td>M_ID_9139332ccc</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325518</th>\n      <td>C_ID_cc33d0d508</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n      <td>-0.666776</td>\n      <td>-2.750000</td>\n      <td>36</td>\n      <td>M_ID_9139332ccc</td>\n      <td>D</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325519</th>\n      <td>C_ID_8852708dcb</td>\n      <td>4</td>\n      <td>3</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n      <td>-0.572501</td>\n      <td>-4.846304</td>\n      <td>514</td>\n      <td>M_ID_dd2beefc47</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325520</th>\n      <td>C_ID_c3e88ef62d</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>11</td>\n      <td>-0.393636</td>\n      <td>-0.598157</td>\n      <td>-1.013889</td>\n      <td>216</td>\n      <td>M_ID_e5374dabc0</td>\n      <td>C</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325521</th>\n      <td>C_ID_650557248c</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.393636</td>\n      <td>-0.463134</td>\n      <td>-2.055556</td>\n      <td>18</td>\n      <td>M_ID_883c24d43a</td>\n      <td>D</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325522</th>\n      <td>C_ID_f14af9cb56</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n      <td>-0.486030</td>\n      <td>-4.089888</td>\n      <td>89</td>\n      <td>M_ID_a44596d5a3</td>\n      <td>C</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325523</th>\n      <td>C_ID_068dcf31e8</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>7</td>\n      <td>-0.393636</td>\n      <td>0.521438</td>\n      <td>-5.111111</td>\n      <td>54</td>\n      <td>M_ID_6f274b9340</td>\n      <td>C</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325524</th>\n      <td>C_ID_6659f4625a</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>12</td>\n      <td>-0.393636</td>\n      <td>-0.696946</td>\n      <td>-1.000000</td>\n      <td>19</td>\n      <td>M_ID_4ef9297a6b</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325525</th>\n      <td>C_ID_dc842bba27</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n      <td>-0.652408</td>\n      <td>-6.216667</td>\n      <td>180</td>\n      <td>M_ID_8fc7b62fc5</td>\n      <td>D</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325526</th>\n      <td>C_ID_bd625edce0</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>12</td>\n      <td>-0.393636</td>\n      <td>-0.628545</td>\n      <td>-5.765766</td>\n      <td>444</td>\n      <td>M_ID_d5f8f015bc</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325527</th>\n      <td>C_ID_278446cc2f</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n      <td>-0.551558</td>\n      <td>-2.640625</td>\n      <td>64</td>\n      <td>M_ID_9ab926e24d</td>\n      <td>D</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325528</th>\n      <td>C_ID_ad358eb92e</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>8</td>\n      <td>-0.393636</td>\n      <td>-0.522126</td>\n      <td>-6.621622</td>\n      <td>37</td>\n      <td>M_ID_16f8c3bb43</td>\n      <td>D</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325529</th>\n      <td>C_ID_7d9bcec574</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>4</td>\n      <td>-0.393636</td>\n      <td>-0.709905</td>\n      <td>-5.925620</td>\n      <td>847</td>\n      <td>M_ID_1dc093fd3c</td>\n      <td>C</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325530</th>\n      <td>C_ID_b588b35e15</td>\n      <td>3</td>\n      <td>3</td>\n      <td>1</td>\n      <td>2015</td>\n      <td>11</td>\n      <td>-0.393636</td>\n      <td>-0.607496</td>\n      <td>-5.777778</td>\n      <td>36</td>\n      <td>M_ID_9139332ccc</td>\n      <td>E</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325531</th>\n      <td>C_ID_7147e348c3</td>\n      <td>5</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>1</td>\n      <td>-0.393636</td>\n      <td>0.119077</td>\n      <td>-5.066667</td>\n      <td>15</td>\n      <td>M_ID_21c509b90e</td>\n      <td>D</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325532</th>\n      <td>C_ID_727d3ec5da</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2014</td>\n      <td>12</td>\n      <td>-0.393636</td>\n      <td>-0.195756</td>\n      <td>-7.000000</td>\n      <td>13</td>\n      <td>M_ID_b9dcf28cb9</td>\n      <td>C</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325533</th>\n      <td>C_ID_107d93e20c</td>\n      <td>4</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>5</td>\n      <td>-0.393636</td>\n      <td>-0.536897</td>\n      <td>-4.703704</td>\n      <td>27</td>\n      <td>M_ID_fc7d7969c3</td>\n      <td>D</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325534</th>\n      <td>C_ID_0e74bd4966</td>\n      <td>3</td>\n      <td>2</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-0.393636</td>\n      <td>-0.693708</td>\n      <td>-4.136000</td>\n      <td>125</td>\n      <td>M_ID_b7b6e86df1</td>\n      <td>E</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>325535</th>\n      <td>C_ID_7a239d2eda</td>\n      <td>2</td>\n      <td>2</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>10</td>\n      <td>-0.393636</td>\n      <td>-0.676778</td>\n      <td>-1.767123</td>\n      <td>73</td>\n      <td>M_ID_1556d09f7f</td>\n      <td>E</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325536</th>\n      <td>C_ID_75ace375ae</td>\n      <td>3</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2017</td>\n      <td>9</td>\n      <td>-0.393636</td>\n      <td>-0.702088</td>\n      <td>-2.714286</td>\n      <td>7</td>\n      <td>M_ID_8980c12692</td>\n      <td>D</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325537</th>\n      <td>C_ID_21d56d950c</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>9</td>\n      <td>-0.393636</td>\n      <td>-0.177938</td>\n      <td>-3.837838</td>\n      <td>37</td>\n      <td>M_ID_92bf4d10e9</td>\n      <td>D</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>325538</th>\n      <td>C_ID_6c46fc5a9d</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>2017</td>\n      <td>6</td>\n      <td>-0.393636</td>\n      <td>-0.631026</td>\n      <td>-2.145161</td>\n      <td>62</td>\n      <td>M_ID_9fccc8b78f</td>\n      <td>E</td>\n      <td>E</td>\n    </tr>\n    <tr>\n      <th>325539</th>\n      <td>C_ID_87e7979a5f</td>\n      <td>5</td>\n      <td>1</td>\n      <td>1</td>\n      <td>2016</td>\n      <td>10</td>\n      <td>-0.393636</td>\n      <td>-0.602719</td>\n      <td>-6.976190</td>\n      <td>42</td>\n      <td>M_ID_508b6c1301</td>\n      <td>D</td>\n      <td>D</td>\n    </tr>\n  </tbody>\n</table>\n<p>325540 rows × 13 columns</p>\n</div>"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"fig = plt.figure()\nres = stats.probplot(all_data['Count'] , plot=plt)\nplt.show()\nall_data.plot.scatter(x='target', y='Count');\nall_data.shape","execution_count":12,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"execute_result","execution_count":12,"data":{"text/plain":"(325540, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data_temp = all_data[all_data['Count'] < 1500]","execution_count":13,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"fig = plt.figure()\nres = stats.probplot(all_data_temp['Count'] , plot=plt)\nplt.show()\nall_data_temp.plot.scatter(x='target', y='Count');\nall_data_temp.shape","execution_count":14,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"execute_result","execution_count":14,"data":{"text/plain":"(325527, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data_temp = all_data_temp[all_data_temp['target'] < 12]\nall_data_temp = all_data_temp[all_data_temp['target'] > -12]","execution_count":15,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"fig = plt.figure()\nres = stats.probplot(all_data_temp['Count'] , plot=plt)\nplt.show()\nall_data_temp.plot.scatter(x='target', y='Count');\nall_data_temp.shape","execution_count":16,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAZEAAAEWCAYAAACnlKo3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3XmYlNWd9vHvDaKCGkVRgyCgDuqgce1xj4qi0h2N0ZiJDgajJkSN0UySSXQw++ibPRMnEzM446hJJ9GMWXzn7ZbNfbchasQlEhEFFRHEDUSW3/vHeZouke6ubqrqqaq+P9fVV1edeqrq1yX23eec5zlHEYGZmVlv9Mu7ADMzq10OETMz6zWHiJmZ9ZpDxMzMes0hYmZmveYQMTOzXnOImG2ApG9I+mUvn/tJSXd38XirpLM2dKykNyXt2pv37WGNt0v6VLnfx+qfQ8TqhqRnJa3IfhEvknStpC3zrmt9EdEYEdd18tiWEfEMQFb/v/T2fUrxeUgaJSkkbdLbOqy+OUSs3pwUEVsCBwANwGXrH6Ckr/zb7/bzMNsYfeV/JOtjImIh0ArsDeuGby6XdA+wHNhV0k6Sbpa0VNJcSZ9e72U2l3SDpDckzZa0b/sDki6R9NfsscclnbLecyXpp5Jek/SkpGMLHuh0KCn7q/9vJE0CJgBfznoS/1fSP0m6ab3jr5T0k55+Huu9Rj9Jl0maL+llSddL2jp7+M7s+7KsjkO7ey/rWxwiVpck7Qw0AX8qaP4EMAnYCpgP/AZYAOwEnAZcIemYguNPBn4LbAv8CviDpAHZY38FPghsDXwT+KWkoQXPPTg7ZgjwdeB3krYttv6ImAI0A9/LhrhOAn4JjJe0TfYzbgKcDlzf3et18nm0+2T2NRbYFdgS+Gn22JHZ922yOu4r9mewvsEhYvXmD5KWAXcDdwBXFDx2bUTMiYjVwPuBw4GvRMTbEfEw8J/AxILjZ0XE/0TEKuBHwObAIQAR8duIeCEi1kbEDcDTwEEFz30Z+NeIWJU9/hTwoY35wSLiRVLP4GNZ03jglYiY1cXTuvo82k0AfhQRz0TEm8ClwOmeB7Fi+B+J1ZuPRMSMTh57vuD2TsDSiHijoG0+ad7gPcdHxFpJ7b0WJE0EvgCMyg7ZktTraLcw3r266fz2526k64DzgauBM4FfdHN8V59Hu51I9bWbT/rdsGNvi7S+wz0R60sKf6m/AGwraauCthHAwoL7O7ffyCbihwMvSBpJ+iV+IbBdRGwDPAao4LnDJBXeH5G9Z2/rbfcHYB9JewMnkoa8NtYLwMiC+yOA1cCiTmowW8chYn1SRDwP3Av8H0mbS9oHOJc079DuQEmnZsM6nwdWAvcDW5B+uS4GkHQ2752w3gG4SNIASR8D/hZo6WGZi0hzFIV1vw38D2mO5sGIeK6Hr7khvwb+UdIu2SnAVwA3ZMN+i4G169dh1s4hYn3ZGaThqBeA3wNfX2/o54/Ax4FXSZPyp2ZzHI8DPwTuI/2i/wBwz3qv/QAwGngFuBw4LSKW9LC+/wLGSFom6Q8F7ddl79ndUFaxrsle605gHvA28DmAiFhOqv+erI5DSvSeVifkTanMaoukEcCTwPsj4vW867G+zT0RsxqSzc18AfiNA8Sqgc/OMqsRkrYgDZ/NJ53ea5Y7D2eZmVmveTjLzMx6re6Hs4YMGRKjRo3Kuwwzs5oxa9asVyJi+2KOrfsQGTVqFG1tbXmXYWZWMyTN7/6oxMNZZmbWaw4RMzPrNYeImZn1mkPEzMx6rewhIumabLe0xwraviFpoaSHs6+mgscuzXaZe0rSCQXt47O2uZIuKXfdZmbWvUr0RK5lw1fX/jgi9su+WgAkjSHt1LZX9pyfSeovqT/w70AjMAY4IzvWzMwKNDfDqFHQr1/63lyKzQK6UPZTfCPiTkmjijz8ZNKaQCuBeZLm0rFb3NyIeAZA0m+yYx8vcblmZjWruRkmTYLly9P9+fPTfYAJE8rznnnOiVwo6dFsuGtw1jaMd+8+tyBr66x9gyRNktQmqW3x4sWlrtvMLHcb6nFMntwRIO2WL0/t5ZJXiFwF7AbsB7xI2puhZCJiSkQ0RETD9tsXddGlmVnVaw8OCT7xidTTiOjocczv5BLB50qxdVkncrliPSIWtd+WdDXwv9ndhRRsSUrajrR9u9LO2s3M6lZ7D2P+/BQe7Wvmrr927vLl0L8/rFnz3tcYMaJ89eXSE5E0tODuKaT9qQFuBk6XtJmkXUg7wz0IPASMzrbv3JQ0+X5zJWs2M6uk5mYYMgTOPLOjh9Hdoutr1sCgQe9uGzQILr+8PDVCZU7x/TVpG9E9JC2QdC7wPUl/lvQoMBb4R4CImAPcSJowvwX4bESsyfZ6vhCYCjwB3Jgda2ZWdy64IA1XLenhhsojR8KUKem71HG/XJPq0Af2E2loaAgvwGhmtaC5GT7zGXjrrZ4/d9Cg0gWGpFkR0VDMsb5i3cysCjQ3w9ln9yxApPS9Ej2OztT9UvBmZrXg4oth1aruj2ufXB85Ms115BEchRwiZmY5u+CC4uY/ttsOfvKT/IOjkEPEzCwnPZkDOf98+NnPyl9TTzlEzMxy0D4HUswQVrUGCDhEzMxyMXly9wHSrx9cf311DV+tz2dnmZnloLulSKTqDxBwiJiZ5WLbbbt+/Lzzqj9AwCFiZlZxzc1dn41VzXMg63OImJlVUHMzTJzY9TG1EiDgEDEzq6jJk2Ht2s4fHzmycrWUgkPEzKxCmps73/OjXTlX3C0Hh4iZWQU0N6eVebsi1cZkeiFfJ2JmVmbNzWlfkO6cd175ayk190TMzMqo2ACB2ppQb+cQMTMro4svLu64WptQb+cQMTMro2JW591kk9qbUG/nEDEzK5Pm5u6P2XxzuPba2ptQb+eJdTOzMhg3DmbO7PqYMWNgzpzK1FMuZe+JSLpG0suSHito+76kJyU9Kun3krbJ2kdJWiHp4ezr5wXPOVDSnyXNlXSl1L4xpJlZdRk2rPsA6d+/9gMEKjOcdS0wfr226cDeEbEP8Bfg0oLH/hoR+2VfhSe8XQV8Ghidfa3/mmZmubvgAnjhhe6Pu+668tdSCWUPkYi4E1i6Xtu0iFid3b0fGN7Va0gaCrwvIu6PiACuBz5SjnrNzDbGf/xHccfV6hzI+qphYv0coLXg/i6S/iTpDkkfzNqGAQsKjlmQtW2QpEmS2iS1LV68uPQVm5l1oqt1sdode2z566iUXENE0mRgNdB+DsOLwIiI2B/4AvArSe/r6etGxJSIaIiIhu233750BZuZdWHcuO6PGTgQZswofy2VktvZWZI+CZwIHJsNURERK4GV2e1Zkv4K7A4s5N1DXsOzNjOzqtHdZPpOO8HCOvvNlUtPRNJ44MvAhyNieUH79pL6Z7d3JU2gPxMRLwKvSzokOytrIvDHHEo3M9ugwYO7P6beAgQq0BOR9GvgaGCIpAXA10lnY20GTM/O1L0/OxPrSOBbklYBa4HzIqJ9Uv4C0pleA0lzKIXzKGZmuVq2rOvH+/evTB2VVvYQiYgzNtD8X50cexNwUyePtQF7l7A0M7OSKGYuZNKk8teRh2o4O8vMrKYVc2V6La7QWwyHiJlZGQ0YUB9XpnfGIWJmthEuuKDrx995pzJ15MUhYma2Ea66Ku8K8uUQMTPrpe56IX2BQ8TMrJe6Wyfr/PMrU0eeHCJmZr3U3TpZ9XpGViGHiJlZL9TrxYM95RAxM+uhceO674XstFNlasmbQ8TMrIe6u7gQ6nOdrA1xiJiZlVhf6YWAQ8TMrEfSmrFd6yu9EHCImJkVba+9uj+mnnYtLIZDxMysSI8/3v0x9bRrYTEcImZmRdh00+6PKWaoq944RMzMunHBBbBqVffH/eIX5a+l2jhEzMy6Ucwii+efDxMmlL+WauMQMTPrQrFDVH1hiZMNqUiISLpG0suSHito21bSdElPZ98HZ+2SdKWkuZIelXRAwXPOyo5/WtJZlajdzPquYpc26WtnZBWqVE/kWmD8em2XADMjYjQwM7sP0AiMzr4mAVdBCh3g68DBwEHA19uDx8ys1Jqbu1/apF1fOyOrUEVCJCLuBJau13wycF12+zrgIwXt10dyP7CNpKHACcD0iFgaEa8C03lvMJmZlcSZZxZ3XER566h2ec6J7BgRL2a3XwJ2zG4PA54vOG5B1tZZ+3tImiSpTVLb4sWLS1u1mVlmzJi8K8hfVUysR0QAJcvziJgSEQ0R0bD99tuX6mXNrI8YXORA+Zw55a2jFuQZIouyYSqy7y9n7QuBnQuOG561ddZuZlZSy5Z1f0xfH8Zql2eI3Ay0n2F1FvDHgvaJ2VlahwCvZcNeU4HjJQ3OJtSPz9rMzEqmmPWxHCAdNqnEm0j6NXA0METSAtJZVt8BbpR0LjAf+Pvs8BagCZgLLAfOBoiIpZK+DTyUHfetiFh/st7MbKMUsz6WdahIiETEGZ089J6zq7P5kc928jrXANeUsDQzs3UGDer+mL58TciGVMXEuplZ3pqbYcWK7o/ry9eEbIhDxMyM4q4LGTCg/HXUmh6FSDapvU+5ijEzy0Nzc3HHvfNOeeuoRd2GiKTbJb0vW3ZkNnC1pB+VvzQzs8pwL6T3iumJbB0RrwOnkpYjORgYV96yzMyqi3shG1ZMiGySXQz498D/lrkeM7OKKuaMLF8X0rliQuRbpIv6/hoRD0naFXi6vGWZmVVGMWdkWee6vU4kIn4L/Lbg/jPAR8tZlJlZJRSz4ZR7IV0rZmJ9d0kz2zeUkrSPpMvKX5qZmVW7YoazrgYuBVYBRMSjwOnlLMrMrNyK6YX085V03SrmIxoUEQ+u17a6HMWYmVWTNWvyrqD6FRMir0jajWy/D0mnAS92/RQzs+pVTC/EilPMAoyfBaYAe0paCMwDitw40sysNnlCvTjFnJ31DDBO0hZAv4h4o/xlmZmVh+dCSqvbEJH0tfXuAxAR3ypTTWZmufJcSPGKGc56q+D25sCJwBPlKcfMrHw8F1J6xQxn/bDwvqQf4G1pzaxOeS6kZ3oz8jcIGF7qQszMysm9kPIoZk7kz2Sn9wL9ge1J62ltFEl7ADcUNO0KfA3YBvg0sDhr/+eIaMmecylwLrAGuCgi3CMys5JxL6TnipkTObHg9mpgUURs9MWGEfEUsB+ApP7AQuD3wNnAjyPiB4XHSxpDulJ+L2AnYIak3SPCU2Bm1iX3Qsqn0+EsSdtmG1G9UfC1AmjfoKqUjiWtEjy/i2NOBn4TESsjYh4wFzioxHWYWR/lXkjvdNUTmUUaxtpQhgdp+KlUTgd+XXD/QkkTgTbgixHxKjAMuL/gmAVZm5lZp9wLKa9OeyIRsUtE7Jp9X/+rZAEiaVPgw3QsN38VsBtpqOtF4IedPLWr15wkqU1S2+LFi7t/gpnVpWL3TncvpPeKmRNB0mBgNOk6EQAi4s4S1dAIzI6IRdnrLip436vp2E1xIbBzwfOGZ23vERFTSEu10NDQ4H8eZn1UMXun28YpZj+RTwF3kq4N+Wb2/RslrOEMCoaysq14250CPJbdvhk4XdJmknYhhdr6qwubmQHFD2O5F7JxiumJXAz8HXB/RIyVtCdwRSnePFuP6zjgMwXN35O0H2ne5dn2xyJijqQbgcdJZ4l91mdmmdnGGDgw7wpqXzEh8nZEvC0JSZtFxJPZNR4bLSLeArZbr+0TXRx/OXB5Kd7bzOpXsb2Q5cvLW0dfUEyILJC0DfAHYLqkV4GuTsU1M8vN4MHFHedhrNIoZu2sU7Kb35B0G7A1cEtZqzIz66Vly/KuoG/pNEQktQC/Av4QEW8CRMQdlSrMzKynPJleeV2dnfUfwIeAeZJulHRKdk2HmVnVKfaaEG84VVpdXWz4x4g4AxgJ3ARMBJ6T9N+SjqtUgWZmxSj2mhBvOFVa3WZyRCyPiBuyuZHjSVeSe07EzKqGh7HyU8zFhjtK+pyke0hnaE0FDih7ZWZmRfDaWPnqamL906SryfcgDWf9U0TcW6nCzMxKyb2Q8ujqFN9Dgf8DzIyItRWqx8ysaB7Gyl+nIRIR51SyEDOznhjmjSCqgk92M7Oa9MILxR3nXkh5OUTMrOZ4GKt6dDWx3uUWuBGxtPTlmJl1zWdjVZdit8cdAbya3d4GeA7YpezVmZkVKPaqdHAvpFK63R4XmAGcFBFDImI74ERgWqUKNDNrV+xV6Q6QyilmTuSQiGhpvxMRrcBh5SvJzOy9PIxVnYrZT+QFSZcBv8zuTwCKPC/CzGzj9SRA3AuprGJ6ImcA2wO/B36X3T6jnEWZmfWGA6TyitmUailwsaQtsu1szcwqxqfzVrdiFmA8TNLjwBPZ/X0l/axUBUh6VtKfJT0sqS1r21bSdElPZ98HZ+2SdKWkuZIeleSFIM3qmOdBql8xw1k/Bk4AlgBExCPAkSWuY2xE7BcRDdn9S0hrdo0GZmb3ARqB0dnXJOCqEtdhZlXC8yC1oagr1iPi+fWayr2ty8nAddnt64CPFLRfH8n9wDaShpa5FjOrYg6QfBUTIs9LOgwISQMkfYlsaKtEApgmaZakSVnbjhHxYnb7JWDH7PYwoDDQFmRt7yJpkqQ2SW2LFy8uYalmVgnF9kIGDixvHTVp9Wq46y647rrujy2BYk7xPQ/4CemX9ULShYafLWENR0TEQkk7ANMlPVn4YESEpB79rRERU4ApAA0NDf47xayG9GQYa/ny8tVRU156CW65BVpaYNo0eO012GYbmDABNinm13zvdfnqkvoDn4iICeUqICIWZt9flvR74CBgkaShEfFiNlz1cnb4QmDngqcPz9rMrA54HqRIa9bAAw+k0GhthdmzU/vQoXDaadDUBOPGlT1AoJsQiYg1kv6BNLlecpK2APpFxBvZ7eOBbwE3A2cB38m+/zF7ys3AhZJ+AxwMvFYw7GVmNax//+KP7ZMB8vLLMHVqR29j6dL0oR16KFxxRQqOffap+CltxcTU3ZJ+CtwArLtOJCJml+D9dwR+r/RDbwL8KiJukfQQcKOkc4H5wN9nx7cATcBcYDlwdglqMLMqsNb7p77b2rXQ1pZCo6Ul3Y6AHXeEk05KoXHccTB4cK5lKrqJdEm3baA5IuKY8pRUWg0NDdHW1pZ3GWbWBQ9jZZYsSb2N1tY0x/HKK+nDOeQQaGxMwbH//tCvvFtBSZpVcMlFl4q5Yn3sxpdkZrZhfTpA1q6FP/2pY27jgQdS25AhMH58Co3jj4fttsu70k51GyKSdgSuAHaKiEZJY4BDI+K/yl6dmdW1Phkgr74K06en4LjlFli0KH0QDQ3w1a+m4DjwwJ5NEuWomDmRa4H/BiZn9/9Cmh9xiJhZr/WZAImARx/tmNu47750dtXgwXDCCSk0TjgBdtgh70p7pZgQGRIRN0q6FCAiVksq9xXrZlbH6n5NrNdfhxkzOoapXsh2zzjgALjkkhQcBx1UkVNwy62Yn+AtSduRrixH0iHAa2WtyszqVk8DpCZ6IREwZ04KjJYWuPvudOX41lunOY3GxjTHMbT+VmkqJkS+QLo+YzdJ95D2EzmtrFWZWV2qqwB5802YObOjt/F8tiLTPvvAl76UguPQQ2HAgHzrLLNizs6aLekoYA9AwFMRsarslZlZXan5AImAp57qmNu46y545x3Yaqt0dfjXvpZ6G8OH511pRXUaIpJO7eSh3SUREb8rU01mVmdqNkCWL4fbbusIjmefTe177QUXXZTmNg4/HDbdNNcy89RVT+Sk7PsOwGHArdn9scC9pK1yzcxKKvcAefrpjiGq22+HlSth0KDU2/jKV9Iw1ciRORdZPToNkYg4G0DSNGBM+xpV2YKI11akOjOraTXRA1mxAu64o2NSfO7c1L7HHnDBBam38cEPwmab5VBc9StmYn3n9RY5XASMKFM9ZlYnqjpA5s3r6G3cemsKkoEDYexY+PznU29j110rWFDtKiZEZkqaCvw6u/9xYEb5SjKzWld1AbJyZZoIb5/beOqp1L7bbvCpT6XexlFHeZerXijm7KwLJZ1Cx77qUyLi9+Uty8xqVdUEyHPPdQxRzZwJb72VhqSOPhrOPz8Fx+jRZXrzvqOYTalmZIswOjjMrEu5Xon+zjtwzz0dwTFnTmofNQrOOisNUY0dC1tskWOR9aeYTanWSto6InyVupl1qjcBstG9kIULO0Jjxgx44410cd+RR8I556Tg2HPPPrDOSn6KmRN5E/izpOm8e1Oqi8pWlZnVlIoFyOrVaQHD9rmNRx9N7TvvDGeckYaojjkmXQBoFVFMiPwOXxNiZp0o+xzISy+lJdPbt4V97bW0cOERR8B3v5uCY6+93NvISTEhcgPwN9ntuRHxdhnrMbMaUbbex5o1aXOm9mGq2dlO3EOHwmmnpSGqcePS4oaWu66WPdmEtBnVOaR9zgXsLOm/gckbu36WpJ2B60n7rAfprK+fSPoG8GlgcXboP0dES/acS4FzgTXARRExdWNqMLPeKXmALF6cehutrWl72KVL0xawhx0GV1yRgmPffd3bqEJd9US+D2wF7BIRbwBIeh/wg+zr4o1879XAF7MFHrcCZmXzLgA/jogfFB6c7ah4OrAXsBMwQ9LuEeG9TcwqqCQBsnYttLV1XPD30EPpoB12gJNOSkNUxx2XNm6yqtZViJwI7B7R8Z8/Il6XdD7wJBsZItlV8C9mt9+Q9AQwrIunnAz8JiJWAvMkzQUOAu7bmDrMrDjNzXDmmT1/3rrfIEuWpDmN9m1hX3klJdIhh8A3v5mCY//9Uw/EakZXIRKFAVLQuEZSSS8PkjQK2B94ADgcuFDSRKCN1Ft5lRQw9xc8bQFdh46ZlUhveh9iLWvb/gT/ks1tPPBA6oEMGZKWTG9sTBs2DRlS+oKtYroKkcclTYyI6wsbJZ1J6omUhKQtgZuAz2c9nauAb5PmSb4N/JA0L9OT15wETAIYMcLLfJltjJ4EyDa8ynFMp5FWzt6xFRoWpQf+7u/gq19NwdHQAP37l6dYq7iuQuSzwO8knQPMytoagIHAKaV4c0kDSAHS3L4/SUQsKnj8auB/s7sLgZ0Lnj48a3uPiJgCTAFoaGjIe2Fps5pUXHgE+/AoTbTQSCuHcS+bsCbNZYw9IQ1RnXBCmuuwutTVUvALgYMlHUOazAZoiYiZpXhjSQL+C3giIn5U0D60YNXgU4DHsts3A7+S9CPSxPpo4MFS1GJm79ZVgGzF64xjxrrgGMYLAMxmfzaZfEkKjoMOStdyWN0rZgHGW+nYkKqUDgc+Qboa/uGs7Z+BMyTtRxrOehb4TFbHHEk3Ao+Tzuz6rM/MMiutDYdHsBdzaKSVJlo4grsZwGqWsTXTOY4WmriF8bwYQytdrlUBbWDuvK40NDREW1tb3mWYVb3CANmCNzmWmeuCYwTPA/AI+9BCE600ch+HspoBQBXsRmglJWlWRDQUc6z7m2Z9XAqPYA+eWjdEdSR3shnv8AZbMp3j+BZf4xbGs5Dh73quw8McImZ91CAtZyy38dMsOHZlHgBzGMOVXEQLTdzD4axi0w0+3wFi4BAx61uefpqLdk9DVEu5nc1ZyVsMYgbj+B5fppVGnmNkty/jALF2DhGzerZiBdxxB7S28vSVLYxmLlcCT7IHP+MCWmnkTo7kHTYr6uUcHrY+h4hZvZk3r2NNqltvhRUrWMHm/IVj+AkX00ojz7Bbj1/WAWIb4hAxq3UrV8Jdd3UEx5NpQYm57EYLn6KFJu7gKN5mYK9e3uFhXXGImNWi557r2G9j5kx46y3YbDOmrjyKFs6jhSbmMnqj38YBYt1xiJjVgnfegXvu6QiOOXNS+8iR/OytibTQxG0rx7KcLUrydg4PK5ZDxKxaLVyYQqO1FaZPhzfegAED4Mgj4Zxz+NsvNvLk/D1J+8WVhsPDesohYlYtVq+G++7rmNt45JHUPnw4nHEGJ09p4tZVx/DmzK2gJCvYdXB4WG85RMzy9NJLaYOmlpa0YdNrr6WFCw8/HL77Xfb+ShNzFuwFU8qzLazDwzaWQ8SsktasgQcfTKHR0gKzZ6f2oUPhox/lo9c0MWP1OF6/Y2u4o3xlODysVBwiZuW2eDFMnZpCY+pUWLoU+vXjrrWH0crltNDEIy/uC9eUp7dRyOFhpeYQMSu1tWuhra1jbuOhhyCCRexAKyfRSiPT1h7PMgZXpBwHh5WTQ8SsFJYsgWnT+MU/tHACU9mBxaxFPMDBtPBNWmjiT+xP0K8i5Tg4rFIcImY90L7nhljLfjxMEy000cLBPEB/1jKeIUzlBFpoYhrHs4QhFa3P4WGV5hAxo7j9xLdmGR9jGo200kgr72cRAA/yd/wLl9FCE200sJb+Za62g0PD8uYQsZpXTAD0TrAPj67bqOkw7mUT1rCUwUzlBFppZCon8DI7lquADVfl4LAq4hCxLpXvF3R12orXGceMdcExjBcAmM3+fIdLaKWRBziYNRX8X8ehYdWsMrN8JSRpvKSnJM2VdEk53qO5GUaNgn790vfm5ve2DxmSvgpvS+m+VD9f9S8Ywxy+xPe5lbEsYTt+x0f5GL/lXg7jbK5hKC9wILP5Kv/CvRxe1gCJeO+XWTWrqZ6IpP7AvwPHAQuAhyTdHBGPl+o9mpth0iRYvjzdnz8/3b/nHrjuuo72JUs6nlN42//TV78teJNjuHXdpPgIngfgEfbhh3yRFpq4j0NZzYCy1uF/K1YPaipEgIOAuRHxDICk3wAnAyULkcmTO4Ki3fLlMGVKutjYalGwB0/RSNoW9kjuZDPe4Q22ZDrH8W2+SiuNLGR4ed7dYWF1rNZCZBhkfzYmC4CD1z9I0iRgEsCIESN69AbPPbfhdgdIbRnIcsZy27q5jV2ZB8AcxnAlF9FKI3dzBKvYdKPfyyFhfVmthUhRImIKMAWgoaGhR/+LjxiRhrDW17+/g6Ta7cbcdUNUR3M7m7OStxjETI7le3yZVhp5jpFFv57Dwax7tRYiC4GdC+4Pz9pK5vLL3z0nAjBoEJx11rvnRCx/m/E2R3HHuuAYzVwAnmJ3Nv/8+dDUxBYf/CAf3nxzPpxzrWb1qtZC5CFgtKRdSOFxOvAPpXyDCRPS98mT09DWiBEpWCZMSKuerMExAAAMeUlEQVRzt7dvu206bunSjttLlqQzmur9L9hcf7558zp297v1VlixAjbfHI45BhovhsZG9thttxwLNOtbFDX2G09SE/CvQH/gmoi4vKvjGxoaoq2trSK1WRmsXAl33dURHE8+mdp33RU+9CFobISjj4aBA3Mt06yeSJoVEQ3FHFtrPREiogVoybsOK6PnnuvYFnbGDHjrLdh00xQW552XgmP06L5yIYtZVau5ELE6tGpVuhCnfaOmOXNS+8iRMHEiNDXB2LGwxRb51mlm7+EQsXwsXNixLez06fDGGzBgABx5JJx9dgqOPfd0b8OsyjlErDJWr4b77uuY23jkkdQ+fDiccUYaojr2WNhqq3zrNLMecYhY+bz00rt7G8uWpQtujjgCvvvdFBx77+3ehlkNc4hY6axZAw8+2DG3MXt2ah86FE49NQ1RjRsHW2+db51mVjIOEds4ixfD1KkpNKZOTRfO9OsHhx6aLrBpaoJ993Vvw6xOOUSsZ9auhba2jrmNhx5KVx/usAOcdFIaojruuI4rMM2srjlErHtLlsC0aSk4brkl9T4kOPhg+OY3U3AccEDqgZhZn+IQsfdauxYefjj1NFpb4f77U9t228H48WmI6vjj005cZtanOUQsWbYsnUHVHhyLFqX2hga47LIUHA0N6ewqM7OMQ6SvioBHH+2Y27j33nR21eDBqZfR1AQnnAA77ph3pWZWxRwifcnrr6e1qNrXpVqYraK///7wla+k4Dj4YNjE/yzMrDj+bVHPIuDxxzuGqO66K105/r73dfQ2xo9P13GYmfWCQ6TevPlm2mejPTja9/v9wAfgi19MwXHooWmdKjOzjeQQqXUR8NRTHXMbd94J77wDW26Zrg6/7LJ0Cu7w4XlXamZ1yCFSi5Yvh9tu6wiOefNS+5gx8LnPpd7GEUekPTjMzMrIIVIr5s7tGKK67ba049+gQWnl23/6p9TbGDUq7yrNrI9xiFSrt9+GO+7oCI6nn07tu+8O55+fQuPII9P+4mZmOcklRCR9HzgJeAf4K3B2RCyTNAp4AngqO/T+iDgve86BwLXAQNL2uBdHrW0Q35158zqGqG69FVasSCExdixcdFEKjt12y7tKM7N18uqJTAcujYjVkr4LXAp8JXvsrxGx3waecxXwaeABUoiMB1orUWzZrFwJd9/dsXT6k0+m9l13hXPPTXMbRx8NAwfmWqaZWWdyCZGImFZw937gtK6OlzQUeF9E3J/dvx74CLUYIs8913Gx34wZ8NZbaQL86KPhM59JwTF6tJdON7OaUA1zIucANxTc30XSn4DXgcsi4i5gGLCg4JgFWdsGSZoETAIYMWJEyQvukVWr4J57OuY2HnsstY8cCRMnptAYOxa22CLfOs3MeqFsISJpBvD+DTw0OSL+mB0zGVgNNGePvQiMiIgl2RzIHyTt1dP3jogpwBSAhoaGys+bvPBCx9zGjBlpuZEBA+CDH4Qf/CAFx557urdhZjWvbCESEeO6elzSJ4ETgWPbJ8gjYiWwMrs9S9Jfgd2BhUDh1XLDs7bqsHo13HdfR3A88khqHz4cPv7xFBrHHgtbbZVvnWZmJZbX2VnjgS8DR0XE8oL27YGlEbFG0q7AaOCZiFgq6XVJh5Am1icC/5ZH7eu89FLaoKm1NW3YtGxZWib9iCPgO99JwbH33u5tmFldy2tO5KfAZsB0pV+y7afyHgl8S9IqYC1wXkQszZ5zAR2n+LZS6Un1NWvgwQc75jZmzUrt738/nHpqOv123DjYZpuKlmVmlqe8zs76m07abwJu6uSxNmDvctb1HosXw9SpKTimToWlS9MWsIceCpdfnoJjv/3c2zCzPqsazs6qPsuXwzHHpJ5HBOywA5x4YhqiOu442HbbvCs0M6sKDpENGTQoXavR1JS+Djgg9UDMzOxdHCKd+cUv8q7AzKzq+c9rMzPrNYeImZn1mkPEzMx6zSFiZma95hAxM7Nec4iYmVmvOUTMzKzXHCJmZtZrqrdtytcnaTEwP+86MkOAV/Iuogr4c0j8OST+HJJq+hxGRsT2xRxY9yFSTSS1RURD3nXkzZ9D4s8h8eeQ1Orn4OEsMzPrNYeImZn1mkOksqbkXUCV8OeQ+HNI/DkkNfk5eE7EzMx6zT0RMzPrNYeImZn1mkMkB5K+KCkkDcm7ljxI+r6kJyU9Kun3krbJu6ZKkjRe0lOS5kq6JO968iJpZ0m3SXpc0hxJF+ddU54k9Zf0J0n/m3ctPeEQqTBJOwPHA8/lXUuOpgN7R8Q+wF+AS3Oup2Ik9Qf+HWgExgBnSBqTb1W5WQ18MSLGAIcAn+3DnwXAxcATeRfRUw6Ryvsx8GWgz57REBHTImJ1dvd+YHie9VTYQcDciHgmIt4BfgOcnHNNuYiIFyNidnb7DdIv0GH5VpUPScOBDwH/mXctPeUQqSBJJwMLI+KRvGupIucArXkXUUHDgOcL7i+gj/7iLCRpFLA/8EC+leTmX0l/XK7Nu5Ce2iTvAuqNpBnA+zfw0GTgn0lDWXWvq88hIv6YHTOZNKTRXMnarLpI2hK4Cfh8RLyedz2VJulE4OWImCXp6Lzr6SmHSIlFxLgNtUv6ALAL8IgkSEM4syUdFBEvVbDEiujsc2gn6ZPAicCx0bcuVloI7Fxwf3jW1idJGkAKkOaI+F3e9eTkcODDkpqAzYH3SfplRJyZc11F8cWGOZH0LNAQEdWyamfFSBoP/Ag4KiIW511PJUnahHQywbGk8HgI+IeImJNrYTlQ+mvqOmBpRHw+73qqQdYT+VJEnJh3LcXynIjl4afAVsB0SQ9L+nneBVVKdkLBhcBU0kTyjX0xQDKHA58Ajsn+HTyc/TVuNcQ9ETMz6zX3RMzMrNccImZm1msOETMz6zWHiJmZ9ZpDxMzMes0hYjVB0nYFp4G+JGlhdnuZpMcrXMt+haeiSvpwb1fjlfRsXqs5S/qkpJ0K7v9n+wKIedZltcUhYjUhIpZExH4RsR/wc+DH2e39KMN6Q9lFgZ3ZD1gXIhFxc0R8p9Q1VMAngXUhEhGfioiKBrLVPoeI1YP+kq7O9qSYJmkggKTdJN0iaZakuyTtmbWPknRrtp/JTEkjsvZrJf1c0gPA9yRtIekaSQ9m+zycLGlT4FvAx7Oe0Mezv+h/mr3GjtkeKY9kX4dl7X/I6pgjaVJ3P5CksyX9JXvvqwte/1pJpxUc92b2fcvsZ5kt6c/ZYp/tP+sT638+2Ws0AM3ZzzFQ0u2SGjZQy5lZHQ9L+o9s34v+WS2PZe/3jxvx389qmEPE6sFo4N8jYi9gGfDRrH0K8LmIOBD4EvCzrP3fgOuy/UyagSsLXms4cFhEfIG0aOatEXEQMBb4PjAA+BpwQ9YzumG9Wq4E7oiIfYEDgPar0c/J6mgALpK0XWc/jKShwDdJV3QfQdp3pDtvA6dExAFZrT/MlhXZ4OcTEf8DtAETsp9jRSe1/C3wceDwrOe3BphA6o0Ni4i9I+IDwH8XUaPVIS/AaPVgXkQ8nN2eBYzKVoY9DPhtx+9SNsu+Hwqcmt3+BfC9gtf6bUSsyW4fT1oY70vZ/c2BEd3UcgwwESB7ndey9osknZLd3pn0i31JJ69xMHB7+7pikm4Adu/mfQVcIelI0vDeMGDH7LH3fD7dvFahY4EDgYeyz3Eg8DLwf4FdJf0b8P+AaT14TasjDhGrBysLbq8h/aLrByzL/nruibcKbov0V/tThQdIOrgnL5gtqjcOODQilku6nRRIvbGabARBUj9g06x9ArA9cGBErMoW+Gx/jw19PkWXT+q1vWf3SUn7AicA5wF/T9obxvoYD2dZXcr2pZgn6WOQVozNfukB3Aucnt2eANzVyctMBT7XPiwkaf+s/Q3SApIbMhM4Pzu+v6Stga2BV7MA2ZO0FWxXHgCOys5IGwB8rOCxZ0k9A4APk4bXyN7j5SxAxgIju3mP7n6Owp/nNEk7ZD/TtpJGZmdu9YuIm4DLSEN31gc5RKyeTQDOlfQIaW6ifRvazwFnS3qUtIrsxZ08/9ukX9KPSpqT3Qe4DRjTPrG+3nMuBsZK+jNp6GgMcAuwiaQngO+QtgTuVES8CHwDuA+4h3fvu301KWAeIQ3LtfecmoGG7H0nAk929R6Za4Gft0+sd1LL46SQmJZ9XtOBoaThstslPQz8EnhPT8X6Bq/ia1bllDbwaoiIC/OuxWx97omYmVmvuSdiZma95p6ImZn1mkPEzMx6zSFiZma95hAxM7Nec4iYmVmv/X8aFlClub9HNgAAAABJRU5ErkJggg==\n"},"metadata":{}},{"output_type":"execute_result","execution_count":16,"data":{"text/plain":"(323273, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAY4AAAEKCAYAAAAFJbKyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzsvXmcVPWV9/8+99bCvtggyiaa1hiaCKMkaDBO1CxGUTIjmkQTsxmfzCNOZuJCMj4Gkd9kEtEsRmMeTHwSIsYgTgTUTOKCIkRQVJrQqNhBgQZBaaHZq7ru/f7+uFVFLbeqq7pr6+7zfr0aum/duvWt7Z57ts8RYwyKoiiKUihWtRegKIqidC/UcCiKoihFoYZDURRFKQo1HIqiKEpRqOFQFEVRikINh6IoilIUajgURVGUolDDoSiKohSFGg5FURSlKALVXkA5GDZsmBk3bly1l6EoitKtePnll3cbY4Z3tF+PNBzjxo1j7dq11V6GoihKt0JEthSyn4aqFEVRlKJQw6EoiqIUhRoORVEUpSjUcCiKoihFoYZDURRFKQo1HIpSBloPRGjctpfWA5FqL0VRSk6PLMdVlGqyZN12Zj2ynqBl0e663H7paVwyaVS1l6UoJUM9DkUpIa0HIsx6ZD1H2l32R2IcaXe56ZH16nkoPQo1HIpSQlr2HC5qu6J0R9RwKEoJ6R+yOdLupm070u7SP2Rr3kPpMWiOQ1FKyMGoQ9gWIo5JbgvbwhMbdvKLZ5s176H0CNTjUJQukupJjB7aFzfjdhe4Z/mbmvdQegzqcShKF8isoLrlovEYY9L2cV1DKGgTwUlus0Ro2tHGOaccW+klK0qXUY9DUTqJXwXVnGVN9Anaafv1Cdm0O+l+yKGowzcXrGXpuu2VXLKilAQ1HIrSSVr2HCZopX+FgrbF4aiTti3S7jD74gbCgfR9IzGjISulW6KGQ1E6yeihfWl30z2JmOtiMvZzDVww4Tjuu2oy/TK8kaBlaamu0u1Qw6EonaRuQJjLzxidtu3sk4fhZlgO10DTjn00jBxEpllpd11GD+1b7qUqSkkpm+EQkftF5F0R2eBz2/UiYkRkWPxvEZG7RKRZRNaLyOkp+35FRN6M/3ylXOtVlHz49WC0Hoiw6OWWtP1WbHovxxEMdQPC3H7pafQJWgwMB+gTtLj90tOoGxAu48oVpfSUs6rqN8DdwILUjSIyBvg0sDVl82eBk+M/U4B7gSkicgwwG5gMGOBlEVlqjNlTxnUrShq5tKcSOY4jKQW4IdvGGJdYitsRtIWGkYMBuGTSKKbWD6Nlz2FGD+2rRkPplpTN4zDGrADe97npJ8BNkOazTwcWGI/VwBAROR74DPCkMeb9uLF4ErigXGtWlEzyaU/55zgMc6Z7ifB+IZtwwOLOyyamGYi6AWEmjhlSkNHQbnOlFqloH4eITAe2G2MaRST1plHAtpS/W+Lbcm33O/Y1wDUAY8eOLeGqld6Mn1eRSGhPHDOE2y89jesfbqQ93inuuC4DwwH++t3zuuxVqMquUqtULDkuIv2A/wC+X47jG2PmG2MmG2MmDx8+vBwPofRC/LyK1IT21PphWCnXQDEXbly8HqBgr8IPVdlVaplKVlV9ADgRaBSRt4HRwCsichywHRiTsu/o+LZc2xWlInSU0G7Zc5iQnV5iG4m5PLhmq9/hCsa3R0RLd5UaoWKhKmPM34CkvkLceEw2xuwWkaXATBF5CC853maMeUdE/gz8QESGxu/2aeB7lVqzokD+hPbooX2JOpnqVHD38je5YsrYTnscHXk6ilJNylmO+3vgBeCDItIiIt/Is/sTwGagGbgP+N8Axpj3gbnAS/Gf2+LbFKWi5Epo1w0IM/Pc+qz9bRF++9e3ad61v9OPp6W7Sq0imYJsPYHJkyebtWvXVnsZSi+h9UCEj/3waSIx/+/SVWeN5bbpH+70sbV0V6kUIvKyMWZyR/tp57iidJG6AWHmzZhIn6BFv2D2V2rBC1u75Hl0JcmuKOVADYeilIBLJo1i1azzuPLME3xvX9mcq6NcUbofOo9DUVJIDQ0BRYWJ6gaEmTh6iO9twwb0Kek6FaWaqOFQlDipDXdHYg7GGPoGA0U137UdbvfdftYH6ny3aw5D6Y6o4VAU0hvuUrvE90diANz0yHqm1g/Le3JvPRDhtseasrYHbfHZWzvDle6L5jgUBf+Gu1QKab5r2XMYW7KP4bqGph1tadu0M1zpzqjhUBT8G+5SKaT5bvTQvsRcJ2u7Y+Dq376UNiZWO8OV7owaDkUhu+EuaAsBi6Ka7+oGhPnn00f73hZ1SPMoiu0MV5VcpZbQHIeixMmUFoHiqqoABvYJ5rwt4VHUDQgnDdVNGTkOv8fRXIhSa6jhUJQUEif11L+L4TPjR3Df82/53pbpURQy1MkvaV9Iol5RyokaDkUpIScOH4AlZM0dDwfE16PINFSZtOw5TMBKr8pK9VwUpRqo4VCUEtKy5zB+8m93XjaJMcf0o/VAJO2E33ogEq+4EhpGDsoyBhu2t3Egkp5wV5Vcpdqo4VCUDiimSe9v2/bgJ3V43e9fZUA4vZlwybrt3JAyPTBgwY8vn5TMX7QeiDD38Y1Zx7pl2nj1NpSqooZDUfLQUWI606g8vmGn73EM6c2E448fxE2L1yeNBiSmBzYm8xd+Y2v7h2wmjBxcnierKAWihkNRctBRYtrPqIyr68cLm/OPjAlaFuu27cW2sjvKbTmav/Ar2XWM0TCVUnW0j0NRcpCvSc+v8/vGxY2clkPkMJV212XSmCE4mRl0IOY6tB2OJnMhOsxJqUXU41CUHORr0vMLI0VihrmPv9bhcS+fPJr6EQOZN+M0rk/JcQheNda3fvcKjnGZN2NiQSW7ilJpyjk69n4ReVdENqRsmycir4vIehH5o4gMSbnteyLSLCJviMhnUrZfEN/WLCLfLdd6FSWTfFf8uSRKDkWzJUcyWbS2hdYDES6ZNIrV3zufBV//CHd/8R+wxMtzHGp3iMQM//6HdTTv2q/DnJSao5yhqt8AF2RsexKYYIw5DdgEfA9ARMYDXwAa4vf5hYjYImID9wCfBcYDX4zvqygVITGg6YGrp7Bq1nnJxHiqUekXtIs6pi2S1KSqGxDmnFOOZVDfIE5G5Mox8NmfrUjTuFKUWqBshsMYswJ4P2PbX4wxsfifq4GEsM904CFjTMQY8xbQDHw0/tNsjNlsjIkCD8X3VZSKkeuKP2FUfvnlMwgHCv8qHYw6bIir5SY0qPblmOPR7qKquUrNUc0cx9eBP8R/H4VnSBK0xLcBbMvYPqX8S1OUwvA8huHMm3EaNy5ejzEuBUSruG1ZExiY+/hGgpZFJJb7TtoprtQaVamqEpGbgRiwsITHvEZE1orI2vfe0/nOSukoRJnWxP/1m8fhh+MY5jy2MVmVFc2MU6WgneJKrVFxj0NEvgpMA843JinOsB0Yk7Lb6Pg28mxPwxgzH5gPMHny5NzfQkUpgkKUaROluZGYgZS+8aAtuK7Jyl0AxAz08R8MmEY4oCW4Su1RUY9DRC4AbgIuMcYcSrlpKfAFEQmLyInAycCLwEvAySJyooiE8BLoSyu5ZqX3UuiUvnzTAy0Rzqn3nzd+pD29KitoC+GAMDAcIBwQrv/UKfz1u+ephLpSc5TN4xCR3wOfAIaJSAswG6+KKgw8KSIAq40x3zLGNInIImAjXgjrWmOMEz/OTODPgA3cb4zJHuqsKGXAr1fDL9+QqzQ30Z+xormVgOWV2uYiHLCYN+M07dlQugVlMxzGmC/6bP51nv3/E/hPn+1PAE+UcGmKUhCFTulLHcokBg77WAjX9YyDJXA4w9PoF7T55ZfP4JxThiePpyi1jEqOKL2SfAnvxG1AwZIfidLcmefV+z5e0Ba+PnWcr8yIYwwNIwd18RkpSuVQyRGl15Ev4e1326pZ5xUUPqobEObDObSqHGO4f9XbvtVTjuuyqnm35jKUboN6HEq3pZAyWb/75Ep457oNKFjyoz1HP8anPnQsPmK4gJf70CY/pTuhHofSLSmkTNaPfAnvxO8dJcPz0djS5rv9zxvfzRonm4qF0LRjXzLPkUkxw6QUpdyox6F0Owotk/UjX8K70GR4vnXl8iryGQ3whA2/uWCtry7VknXbmfqjZ/jSr9Yw9UfPqHaVUnXUcCjdjnxzMjoin+JtZ+ZfNO/az+K12/i/z/2dqT96hl88+/dOP69ILNsAdsVIKkq50FCV0u3oqmeQb8ZFMfMvvv/o31iwemvnnkQOMkNjhfaSKEolUY9D6XaUYjJevhkXhcy/aN61v+RGAyDqpBvArhpJRSkH6nEo3ZJKT8bLTE6vbN5dlseZeW592nNJbS5MLQRQb0OpJmo4lG5LIi9Rbpas285NixuxxcIxLp+fPIY/rG0py2MNGxBixaZ3aRg5OPncdHysUmvIUYHansPkyZPN2rVrq70MpQfQeiDClB88lVdnqhwEbeHOyyZqU6BSUUTkZWPM5I720xyHouSg9UCEZY07Km40wBNIvHGxVk8ptYmGqpReQbENdIkGwwJGZqRxwjF92fJ+x2XBhWBbotVTSk2ihkPpUfgZiGK7zFN7J/ywLSFgQci2aXddPjN+BI9v2EnYlpIZDQDHNVo9pdQkajiUHoOfgZhaPyxpBBK9EDc9sp6p9cNyXsn79U6Al3ewRNLmZvQP2Uy7eyUxxxDLM/61WIK29zjqbSi1iBoOpUeQ6iWkGoj5Xz6j6Aa60UP7EnX8vA3D49d9nPoRAwGvqqtx215fI9MZ/uufJjBqaD/2HW5nUN+gSq0rNYsaDqVHkKvDGqToBrq6AWFmnlvPnU9uStveJxDgYPSo+m3rgQhth9s5kkMRt1gE2Pb+IW5d1kTAElxjmDdDK6uU2kOrqpQeQa4O64aRgzrVZX7FlLGEA+mp8VSDkxAe/JcHXk6OiO0q3/3jBm5+dAPtjuFwu0skZvjOonVFV1Z1Rm5eUYqhbIZDRO4XkXdFZEPKtmNE5EkReTP+/9D4dhGRu0SkWUTWi8jpKff5Snz/N0XkK+Var9K9ySdDkpjO98DVU1g167yCruDrBoSZN2Oi7/FSw2KpHkg5iLnwwt8L71JXJV2lEpStAVBEzgEOAAuMMRPi224H3jfG/FBEvgsMNcbMEpELgeuAC4EpwM+MMVNE5BhgLTAZMMDLwBnGmD35HlsbAHsvpZ5bkXm81gMRlr/+LrOXbuBgtDINHkEb7rxsUofd460HIkz90TNp1WB9gharZp2nSXalIAptACxbjsMYs0JExmVsng58Iv77b4FngVnx7QuMZ8VWi8gQETk+vu+Txpj3AUTkSeAC4PflWrfSvSmVDEmqwZg4xhsHu3D1FuY8tpGgJRUzGgDtDnxn0TpsyyJk5y4prmUlXR1E1bOodHJ8hDHmnfjvO4ER8d9HAdtS9muJb8u1PQsRuQa4BmDs2LElXLLS2/Ar691/JMbNj3pR12iZHtcCEP+hTzEXYq5LJJa7pLhWlXQ7O61RqV2qlhyPexcli5MZY+YbYyYbYyYPH+4/flNROsJvcNKNixu5dVlTh/ft6pfJsuC6c+sJ5BojmILf4KpSyM2XGh1E1TOptMexS0SON8a8Ew9FvRvfvh0Yk7Lf6Pi27RwNbSW2P1uBdSq9FL9wjy0Wlpis6qmAJVhiCFg2jnG5/lMf5Ad/er3Tjx1z4WfPNBe0by5PotaUdGs5fKZ0nkobjqXAV4Afxv9fkrJ9pog8hJccb4sblz8DP0hUXwGfBr5X4TUrvQi/cI9jXPBRrZozvYELGo5LnqQLGV3bVfqHbRzXpFV4ZRqJSsnNF0Kths+UrlE2wyEiv8fzFoaJSAswG89gLBKRbwBbgMvjuz+BV1HVDBwCvgZgjHlfROYCL8X3uy2RKFeUrpArWZtrcBJ4eQVbhHbHZfbFDVw55YTkfSpB/5DNnIsbOPfUY6kbEO4WuQMdRNUz0XkcSq+jkBOun2HJZWxSt59z+9McjJbnO5VaWtuZ0ttqVjZpVVX3oOrluIrSWcp5ksmlaZVZoeQX7vHblmqEoo7X7V0qLppwHE+/8a7vlXqxuYNMY3nLReOZMGpwxU7ktRQ+U7qOGg6lpih3+KWUyVo/I1QqwgHhts9N4Lb4mjNP8MXkDvzWefOjGxgQtonF8yW1FuJSahvVqlJqhtYDEW5a3FjW0s1SJWsTHeS2FDvqKT/9gjZ9ghbzZkxMXqVPHDMkzWg079rP8tff5TufOqWg0tuEsczkQMTR8lilU6jHodQMC9dszQr1lLp0sxTJ2oRXFLCk5FpVX/nYCVz98ZPYczDK4rXbmDRmSFLGHeD7j/6NBau3Jv++fPJorpxyQt6Qk5+xTEXLY5ViUcOh1AStByLcszy7hyHqOCUv3exKr0NH0wG7yq9Xvc2OvYdZ0vhOcttVZ43ltukfpnnX/jSjAbBobQvXfPykvM8h1Vjakm3stDxWKRY1HEpN0LLnMCHbSkpqJJh57slluRLubLI213TAUhGNuWlGA2DBC1u56sxxrNu21/c+67btTfNK/Eg1lht2tDH3sY1aHqt0GjUcSk3gF04JByyumFJbumP9QzaHorGKP+66bXuZFBdbzCTX9kxScyapjYvlMBpaftuz0eS4UhP46Sx1duZ2uQYZLVm3nYt+vpISjhYvmPcPRqkfMZCrzko3pFedNbZDb8MPv6R7qdCZID0fbQBUaoquXqmWq5y39UCEj/3wmaxQWqVIbe5r3rU/6YF0xmiUE50J0r0ptAFQPQ6lrBR79d+VK+FyKrF6FV/VMRoAxjVJLayh/UOcPGIgQ/uHqraeXPiV/vop+SrdG81xKGWj0lpK5VBibT0QoWnHPn7+9KZSLbNTRBxD/5Bdkte0nPkHFTXsHajhUMpCodIepaTUJ63ESToWcymhkkinCNkWr+/c3+XXtNzGXEUNewdqOJSyUI05DKU8aZW7X6NYoo7Lv/3hVayMTvViXtNKGfNamwmilB41HEpZqETIwi/kUqqTVsuewyWXE+kqXool3fUp5jWtpDFXUcOejRoOpSyUO2SRL+RSipPW6KF9ORIrrZxIObhl2viCn6vmH5RSUZDhEJGpxphVHW1TlFTKFbKoRMhlz8EopjaiVDnpG7CYMHJwwftr/kEpFYV6HD8HTi9gm6KkUY6QRb6QS+L2rhiqJeu2c+Pi9WUSFSkdLqZobyHTmAM0bturuQilKPIaDhE5C/gYMFxEvpNy0yDA7uyDisi/A1fjBWz/hjcq9njgIaAOeBn4sjEmKiJhYAFwBtAKfN4Y83ZnH1vp/uQKuWzY3sbn57/Q5VLVWY+sJ1rFno1CCFgkpdeLJWHMu8PoWaU26agBMAQMwDMwA1N+9gEzOvOAIjIK+FdgsjFmAp4B+gLwI+Anxph6YA/wjfhdvgHsiW//SXw/pRfjJ09yy0Xjmfv4xi43/+WaXVEr9A1a3HrxeNb8xydznuQLabr0a5a8cXGjzuVQCiKvx2GMeQ54TkR+Y4zZUuLH7Ssi7UA/4B3gPOCK+O2/BW4F7gWmx38HWAzcLSJieqJWilIwU+uHMf/LZwBCw8hBJasY6mh2RbWJuYaLJ47M+ZwK9SL8Xq9IzPDgmq1cd/7JZVu/0jMo9NIqLCLzReQvIvJM4qczD2iM2Q7cAWzFMxhteKGpvcaYhOxoC5D4tI8CtsXvG4vvX9eZx1Zqm0LlSRIietcufJVrfreWVc2781YMFSN7kurN9At1OhpbNlzXsOdg1Pe2jiRXUl+H0UP7EnWyq8buXt6sXofSIYUmxx8Gfgn8CuhSjaKIDMXzIk4E9saPfUFXjhk/7jXANQBjx9aWFLfSMYVeKeeqqFo16zzfiqGVzbtzHje1DwSOJtUTCeSmHW18c8HarKmE1cQxcOFdzzP74gYmjBqcltT28yIshKYd+9hzKJr1Osw892TufDJdSiVkl66vQ6XVey6FGo6YMebeEj3mJ4G3jDHvAYjIfwNTgSEiEoh7FaOBhBbzdmAM0CIiAWAwXpI8DWPMfGA+eOq4JVqrUgGKKa/NF5LyqxhKKLVmHjfVoBxujyEi9AnYtLsut0wbz5ih/QD4/rQG5j6+0XdyXrWIOoabH93AgLBNzDVJY+jndR1qd/jmgrU4rkvMJe11eGzm2dy9vDlNvLFUfR2aeO/ZFBqqWiYi/1tEjheRYxI/nXzMrcCZItJPRAQ4H9gILOdowv0rwJL470vjfxO//RnNb/QsilFU7aiJLVVdN9dxm3bsSwvpxFxod0wyvHPzHzdw1f0vctX9L3HrsiZumTaeOZc0MCBcW6GrAxEnLRyVCLOFA+kd75GYS2aRWNCyOBh1mDcjvcigFH0d5VQpVmqDQj2OxIn7xpRtBjip2Ac0xqwRkcXAK0AMeBXPU3gceEhE/r/4tl/H7/Jr4Hci0gy8j1eBpfQgiulorhsQ9k7kyzYStAUnfsXtd7LLdVwwBY9/bXcMc5Y28eDVU2h3ajNpnloEcMmkUQzpF+RbD7zCoTweUuL1nThmSMmbNKuhU6ZUloIMhzHmxFI+qDFmNjA7Y/Nm4KM++x4BLivl4yuFU4k4dTEdzUvWbY/PyxbaYy6zL27IGQLJddyGkYOLqpyKOobPz18NNaBd1TdgcTiW38g2jByMm8Mp7xP0PLDU17fUTZoqbdLzKWgCoIhc5bfdGLOg5CsqAToBsDRUOk7dkZHq7HQ5v+MuXbc9aVAORWNFj4MNWELMrWzENBywuO+qyWzbcyhuPHO/L0vj3e+Zw6dCtvDEv3687JMDU19fzXF0HwqdAFhoqOojKb/3wctLvILX0a30QDqrB9UVD6WjK9/OhkD8jpuaSO8fsrnwrhUUk/sOBy1OHtqP13buL/xOXSTmuGx7/xBXnnkCFzQcl/d19kJWIb71u5c51H70iYUDdkWS/Cqt3rMpNFR1XerfIjIETx5E6aF05iRdbg+lsyGQXMYsVXrDICQky22BL595Agte2JIzC3Iw4lTUaIBXinvzoxtA4MopJ2S9D5nPs2HkINwuyLB3FZVW77l0VlvhIF4fhtJDKfYkXYlKGj+pkY6qgBLNgl/61Rqm/ugZlq7bnnZ7Yt3tKbEq27KYdtrx2Hb1cxp+zFm2Met19XuenXm9FKUQCpVVX8bRCTI28CFgUbkWpVSfYiW4W/YcJmB1fjpdoRQTAmnetZ8bH24k6pic4TY/zyrquHzhV2uoTbMBQVvSXtd8YcVCXi9t1FOKpdAcxx0pv8eALcaYljKsR6khijlJb9jexoFIeuy8XGGRQkIgCWn0aEbWO1N+vX/I9q2wihWbLa8gR6IO294/lHxPmnbswyK30c73emmjntIZCs1xPCciIziaJH+zfEtSaolCTtKtByLMfXxj1vZiptOVknzS6O2uy5rNrVz2lzcI2hYx1+WTp47gydfeJVqjfRqZOAZm/v5VgrbwxY+M4Q9rW7KqpwrN/VRiBrnS8ygoxyEilwMv4vVTXA6sEZFOyaorPQ+/Du3+Ibuo6XSlXo8fIVu4ZOJIfvCn14k6hoNRh0jM8PiGnWTO8u4OtDuGBau3ZhmNcEAKymUU07GvKKkUGqq6GfiIMeZdABEZDjyFJ3Ou9HL8EumOKW46XSnj7P1DdlqvR4J7rzydf3nwVd/7ZIa0ujN3XjaJaRNHdviaaqOe0lkKNRxWwmjEaaXzFVlKD6Ors6xLHWc/GHWwhbSmPgtobGkjaAn+ouQ9h0df3Y5jTIevqc4gVzpLoZ3j84DTgN/HN30eWG+MmVXGtXUa7RyvDp3xGvy6wRMd0g0jB3XqJNa8az+f/MmKrO1Bgfae41jkJRyw0kJYfh32iferf8hrCtSqKqUkneMiUg+MMMbcKCL/DJwdv+kFYGHXl6n0dvwn0bl863cv42KK8j4SJ8K2w+2+kiCZRkMASwSnG4ktW5DWlGhbntBjJpmlxJml0X5e3sQxQ8q2bqVn0VGo6qfA9wCMMf8N/DeAiHw4ftvFZV2dUrNkehedDTflGtWakMkotMon9fGjjluQjlTQFowxRetUVZV4g3tCrPA7nzyFH/zp9azdMp9Sau5Cq6mUrtJRnmKEMeZvmRvj28aVZUU9lGLGl9Y6mV3KC1dvKbprPPF6AHlHtaZW+eR6DTO71jOrjHIRsC0s6V6puoQ9bHdcHpt5NiMG98HOaLy86qyxeedsaDWV0lU68jjy+a5aelEgPanJyu9qdc6yJkKB9BORbQnLX3+Xc089Nusq1u/1WDXrPN9RrYkr5cR9bBHaHU9O/cozTwD8w12F4LgGU+R9agXHhZfefp9bl21MC1WFAxbfPv8U6gaE00QcD0ad5LAnraZSukpHl1trReSbmRtF5Grg5fIsqWfR06ah+V6t2lZWOevBiMPspU1Z+lC5Xg+Ac045lnkzJmZdKQPJ+xyMOsnRqQvXbAG8cFdHzXvfPq+e6z91CuHA0WPPm3Ead1w2iUD3cjqS/J9Hm3xk0496DnUDwmzY3saFdz3PFfetTnqHLXsOc8tF41XDSuk0HXkc/wb8UUSu5KihmAyEgH8q58J6Cj1tGlquno3ZF49n7mMbsS3hYFx6JCHfnRo/7+j18JM5ady2F9tniNKcZRu5oOE46gaEmXluPXc+uSnnus8YdwznnDKcK6aMzar8mlo/jO8/+jce37Cry69PJfFL6kedo57DwtVbPDVdIOp470XqrPJbpo1nwsjByf0bt+3VyiqlIPJeaxljdhljPgbMAd6O/8wxxpxljNlZ/uV1f3paWCCX4uqVU05g1azzmHNxA/0zchWp8fNCXo/UueHJ+/h4FAmxP4ArpozNmrWdjsk6diJnsvjllm5nNHIx89z65HObsyxbBgaOziqf+9hGRg/ty8rm3R0qCPeU/JxSGgrVqloOLC/Vg8bnefwKmID3jf468AbwB7yk+9vA5caYPSIiwM+AC4FDwFeNMa+Uai3lpic2WeUSP6wbEObcU4/l/yzZkLZ/6lVwZ16PugFhZl/ckLx6TuC4Ju2482ZM5MbFjWk5EvAMTEOG/ElqziTXYKPM0tfuwGcnHAfAwjVbOwzfBS2Lph378lZY9aT8nFI6Cu0cLzU/A/5ABz28AAAgAElEQVTHGDNDREJAP+A/gKeNMT8Uke8C3wVmAZ8FTo7/TAHujf/fbeiJ09ByiR8mDMP1DzcmZ1w4rsuq5t3JE05nXo8LJhxHy55D/HrlWwQDFo5rsgxO4rgPrtnK3cubkz0O82ak75eaZ8nHuLq+bG7tXpVGtz3WxE8+/w/cs7xjHVIvfGVyhg4BLdtVfKm44RCRwcA5wFcBjDFRICoi04FPxHf7LfAsnuGYDiwwXov7ahEZIiLHG2PeqfDSu0RvmoY2tX4YqRWiMTf7hFPM65F61QuGz00cxdemjvOdm103IMx155/MZyccx7pte5k0Zkhyv6MNgtGCqrC6m9EAWPFmK39p2oktFpB/RKxrYNuewzlDh/nyUUBOw5/a45NvP6X7Ug2P40TgPeD/ichEvKT7t/F6RhLGYCcwIv77KGBbyv1b4tu6leHoTbTsOUzItonEYsltnS0I8Cv/XfjiVha/0sK8Gf5hE7/wioGUBkHHt9u6p3DLkqaCGiDbHcPcxzZyy0Xjmfv4Rt/QoZ9R2bC9jc/Pf8E3fJX62h+JORhj6BsMaJirh1ENwxEATgeuM8asEZGf4YWlkhhjjIgU9c0WkWuAawDGjh1bqrX2aorVnkrVPipVQUCuHo1IzOWGhxsZf/wghvYPpV3hZhqaGx5eB0jaJMBane5XCgoxGgmClsWEUYNZNes835xVZj4qYWT8wleQ/doD7I/E0vZTz6P7Uw3D0QK0GGPWxP9ejGc4diVCUCJyPJBQ490OjEm5/+j4tjSMMfOB+eCJHJZr8b2FYpOimftfPnk0i9a2dLkgIJckCXhS6Bf87Hng6FXttZ+o9xkFC5kiHD31A1JsQj9h0HOFDjPzUR2Fr/KFALtzGbqSTsUNhzFmp4hsE5EPGmPeAM4HNsZ/vgL8MP7/kvhdlgIzReQhvKR4W3fLb9QShXgRxWoZ+e2/aG0Lj808O011tTPquYmrXr9qKTh6dZ24qr17eTPGdLdaqM5jCdz35TNobGnjby1tLN/0Xt79+wQtXNcQDtgFG/RMo5LPm8xl5DP3U7o31aqqug5YGK+o2gx8De9iaZGIfAPYgjdpEOAJvFLcZrxy3K9Vfrk9g0K9iI6a9DINQK79D0adpOJqV8o6E1e9v3p+M/c+tznvviJeMj4fIVtwjelwv+7Ctxa+gi1wxMewZuK6hh9fPolBfYOdkq3vqJw69bZEjiMcsGl3vDBXZ3JcmlyvPapiOIwx6/A60DM532dfA1xb9kX1cIrxIvI16fkZgKn1w/JehZZCjbVuQJgLJhzPb//6FofyDNXoqMQ2ZAtP/OvHeeb1d31VZbsbrgHXMbQXuH/UMdywuBGg08nqfOXUmbf9z4adzHlsI6GAxdzHNzKwT6Dgx9Qektqlm6r0dE+q2YFbjCJqru5wIKfOlN/+pVZjHT20LzHjn9buF7IJBSzCdu60d9CG2Rc3UD9iIO1O/lLVnsyRdrfLmmmZ3f1+twHMfXwj0Zib7Fa/4eFGmnft7/D4PU3jradRrVBVr6PaV0/FSp/k0ozKFcLKdxVaKtkVr4N8PDf/Mb2DPByQuGEzfGdRY877ByzvqveNXfv43Qtbi3rsnki5k9V+IcyoY7jw5yu5I0cpdb771lJyvbeH0NTjqAC1cPWUy4voSOojVdep7XA0KZaXINUA5LoK9XvsWy4aT8uew2mvQSEe2ZVTTuA/PzeBkC30D9n0CVpMnzSK7zzcyKxH1uftzzgcv9Je8MLWHltVBTB2aJ+C9os6Lm2H2wv+HBbrMeeqiIvGOv78+9232PWWi8x5NJnaXr2BgmaOdzdqbeZ447a9fOlXa5KVPwADwwEeuHpKxcd1duZKqVRNXYnH3rC9LavhLLVBr5DjJo61ZnNrj8hVlJr4oMCc2JYgFP4+dtZjXrpuOzc83Jglu1/I53/puu1ZifZqNxO2Hogw9UfPpOXS/Oa5d1dKMnNcKQ2VUMhNbb5LLYHNpFjpE7/Edjhgcc+VpxddlZPY9/PzX0g73o2L1wOGSMwUXP6beK53/OWNgh+/N2Hwwgl9QzYx18U1JLXDgKRnVkhzXleKGy6ZNIrxxw/iwp+vJJpSxlbI5z8R/kwO+HKOrveGxesZf/wgX9mZclLrIbRKoYajApRLITfzCt64hohjkvOoS3FV5vdFCdkWg/sGO7V+v+PZloARUrWVEsqtg/sG04zhyubdySvfSAfqr70dFy+8c/XZJ/L//vp2muHIJN/Jr6sny/oRA7ljRuc+/3UDwgzuG8qSsInGXC6863nuuGxizs94OfIQPW1MQmdRw1EhSq2QmwgdBCzhQCQ975Bwo0sh8dA/ZBOJpR8/4rhZMzcKxXcQlGvIDKwcbo/xzQVrwXjGMBywMMZT2nUMRY+J7a20O4b7nn+rQxmSfCe/Upwsu/L5z5krcUzOz3i5ilF64piEzqDJ8QqSr4SxGFJDB5lGI5XOlLymsmTddqbdvRIrLnWbGLEqxjDt7pWdSgpmJsrDAWHmufV8f1pD2jYRIRJzicSvkiMxl6jjGY1U+gQtQrZ02/GvlSDmGoI5Xp+wLR0WSnSmsCLXcTrz+U88fuZce/D/jJe7GOWSSaNYNes8Hrh6Cqtmndcre0vU4+iG5BL+y6QrLrTfzIpEiDriGMhztZe4f66ry8TV58I1W7lneTPzV2z2BPTio0zbDke5duGrtDuxrOP68cS/fpwdbUe8OHhPaQcvMbmiVEaEx2eeTf2IgQW9Z9UqQU3mSu56Pi3R7vcZr0QeojeNSfBDr9NqjEJKHnO57qH4u9knaHX6qjCBX9NeJrk8mkLKFfccjHL3M28SiR29KkyMMm0YObjD6XX94qW4t196GvUjBnLOKcP5/sXji3uSvQjbEl+vI2gLO9qOcNfTb/KxHz5d0yWm9SMGcsdlEzv0fDQPUX7U46ghCo3L+spdx6/WO6qqykfqFWc+VdoEfl/GQipwlqzbzo2L12eVaCYM0dutB3HyPHbQFm6/9MNxvaWjI2HHDO1X1PPtTfQJ2PzXP3+Y7zzcmFbddDDi8PXfvJTMgSQS0Jnv2cLVW5izrImgLbS7MPvi8Vw55YSKP49CPB/NQ5QfNRw1QrElj+VKtqd+0TK/fAmpdFskp2hdIQKJNy1en3byStDuekn3WY+szytAeOZJx3D9w+sJ2ZkGtuf1JJWKqONy1gfqfDvv/RLnqe/ZwtVbkvPeE8b+5j9uoOX9Q1z98ZMqfkIuJExU7dBaT0cNR43QmbhsqeKsuYzWqlnnZQ34+eCIQXlF63J1/CY8k/ue3+ybh7Atr3z4YNTpMH/z/JutAMnjJAxsw8jB3hVxnrLT3soXPzKGugHhNK80H4n3rPVAhDmPbfTd597nNnP/qrdzTmKsNr09D1FONMdRI1QzLptPhDBTdiRTtC6zWiURJgimiA06rsuq5t0sXL2FX+aQRbfFm1VeSIgs+77C8tffZc/BKP963smEbMvrDVGSDB/oaY21xxzaC+h/Sbxn3hjg3K9lpAD5kFSqKfSplA71OGqEugFhbpk2njnLNhK0Bcc1FYvLFmq0CvWKptYPI/W8HXPhxsWN5FO3CQcCtOw5zMQxQ7j8jNEsWF24COHBqMPNj24gEnOxJXcFUW/mzr9s4q6n3yTimDSjnouY63lyj808u8MekEIrlqot9KmUDvU4aoQl67Yz97GNBC2hPeblDyr1pcpXp596hViMgQnZ6Q2CtljYeaq0Esdp3rWfh9a2FP0cEmGrQoxGb/RFXEj2xBQayksM40p8NsI5mmUK8YxrQehTKR3qcdQAfj0Tcx/fyAUTjqtYjNYvmVhIwrzQckjHuOQ7ZV8+eTQrm3dzo48gnh+2wBc+MpZH123vMF6fiTokhZEoVjihrj+PzTybHW2Hufq3a7Pen1umeWXQjdv25kxEd7a3orfLl9cqVTMcImIDa4HtxphpInIi8BBQB7wMfNkYExWRMLAAOANoBT5vjHm7SssuC9UQTvP7QqYmEztKmDft2AeYtHLYBH7lkJefMZrf5Qk//eGlFv7w0raCjEbYFu77ykdoGDmIP6zd1olnr+QjZIFlW1x+xmim3b0y+R5e+4l6wgGbaEpjZv+QTeuBKFN/9EzeEFRncnga2qpdqulxfBt4DRgU//tHwE+MMQ+JyC+BbwD3xv/fY4ypF5EvxPf7fDUWXC4qnRgv5AuZz5i93Xqww/unejD9QzYX3rWiQ5nvTKHDXBigYaT3semJYwGqTdSFfzv3JH65YnPahcPdy98k02uMuS73LG8mEstfRp56MZGvnDtBKcYNK+WjKjkOERkNXAT8Kv63AOcBi+O7/Bb4XPz36fG/id9+fnz/HkOptICg46qVQmPN/mW1TrLPopBYdaIi64kNO+komuS4Jh7O6ph/+ccP0LLnMC/8vZWgilSVhXue/Tt2xtcsZNvMPLc+TVNsxuljCGRUsOVSFLhk0ihuuWg87a5JlnPn6lAv1bhhpTxUy+P4KXATkBDTrwP2GmMSPnALkLiEHQVsAzDGxESkLb7/7sott/yUomEp05O45aLxTBg1OO14hYbFEsbsO4vWJZvxXAN/2rAz62rDtrxy2HNPPTZr3a0HItyzvLnDtX/xo2PoG7S5f9VbhGw72XD4+xe3pSVzLeBnzzTzi+f+rv0aZSRoW1llu+2uyxVTxnLFlLFJnbEljdk5plzecmo5dzT+Tc/lRVRbNkRzK/mp+OWaiEwD3jXGvFzi414jImtFZO17771XykNXjK6o5/p5Ejc/uoErf7U6TXuomC/k1PphaZVQ7Y7hp0+9yaH29PsfjDjMXtrkq3HkVVh1/DH7zV+3cO9zm4k5hmvOOYlVs87jjBOOwRIvEZ4g8chqNMqLaww3fPqDaSN6E+N+9xyM8otnvfCUnzrz5ZNH502Qp5LLiyjEC8/nXXelX0RHw3ZMNTyOqcAlInIh0Acvx/EzYIiIBOJex2gg8W5tB8YALSISAAbjJcnTMMbMB+aDNzq27M+ixsilmJv4Yt/0iDcx7WDU4ZZp4+Olv/l1fBIn/dRObydHTiFx1Zn6OAnNq44EC1NxDNz1zCbOOukYbnx4XYchLqU0fPWsE1j44lZCAQvHNVw+eTQ/fmoToYBF1DF8btJI5j6+kYDlyd3nuxRYtLaFq84cl6WZVqwXkc8Lz5en60pSXXMrhVHVmeMi8gnghnhV1cPAIynJ8fXGmF+IyLXAh40x34onx//ZGHN5vuPW2szxSuA3CzmVPkEL1zWEA3bOMFaxx/QjbAtGhHCKjtSW1kPc+eSmop6PJV5oTKkMX5oyls9NGkkwYNM/ZDPt7pVFve+p+H0GEifu1DniiduKDdE279qfJa+emPsNdGkmeOO2vXzpV2uSI2qhsPnoPYVCZ47XUmZxFvAdEWnGy2H8Or7910BdfPt3gO9WaX01Tapr7zed70i7S9QxRyXMH9/Y4Rc1e+hSxx+XiGOIxtIT52eddEzRz0eNRmV5YM1WZvzf1TzyyjZe37kvb5d/Kv1C2Z8Jv89AImSUOQTJQFFhoSXrtnvzy3MoK3c1qV7t3Ep3oaoNgMaYZ4Fn479vBj7qs88R4LKKLqybkurab9jRlgxHRRwXiY9gTVBon0hmuOBnT29iwQtH+zE+OKI/b+w6mPP+Qcvi7dZDvlIgtsCVZ47lwTVb86rhKpVjwQtb097ffAQtuO2SCUQcN/lZOxIfM9ye57OW6BcqNiyU2D+XsnLi5B51CkvW+1FN6Z/uhHaO9zASX8qJY4ZwQcNxyT6KaXevTDtzF/tlSnxxbpv+Ya46cxzrtu1lXF0/vnT/i3nveyTmMK6un68UiG1ZfPv8U/jkh0bEZ0IU/jyV6tPuwvsHo1zzjx/ggobjuOXRv/HEhl1Z+yVm1GdWKuXyAnJd0OTK4wUtuPYT9QCsbN6d5q0G4qrLhZ74M6V/Zl/coE2HPqjh6MGknvBzSYV0puywfsRA6kcMpHHb3g4l0I0xHGp3CNmSFV6wLVjWuIPX39mnRqObMu8vb3DpGaPZczDqazTAm1H/2buex3UNfYI2jvGu4scfPygrj3Kk3fUNtYJ/GMkWsCyL+Ss2c/fyZhzXTfss2ZbF1PphBT2XWpD+6S6o4egGlKKmvFAtqmKurgqRQO8bDLDvcAzHJ2lxuN3l1mX+sx6U7oEVl7Q/EGnPuU9qiDS1+m72NP9Rv4+80sKsz34oa3vdgDC3XDQ+PonQIua6uMYTuMw1az5kFy7dU0iPk/Z3eFS1qqpc9KSqqnLp9bQeiPCxHz6T9oUrpvokQaJKBvCtwimnzLlWXtUGIcuTgSmmCMsWEBH/6YM2/Olfz8kq5018FwKW571+4+xxPPDC1rQKqEyK+Uz7VRGm3r83aGcVWlWlHkcN05Wa8sSVUf+QzY62I+w73B6f0T3IGwe6ZmvWVVpnhBUzNan+tGEndy9vxhhD1DFlnY2hRqM2iHYizOgYyFW6ZYlw4V3PJ0vHEyW7mWGk+1e+lfUZsC0hYJFUHygmv5FvVrn2d6SjhqPCFOPq+rnOFkLTjn2cc8rwnMdLXBlBthcQsGDO9Ancs/zNrMdLHfFa7JrbDkdpOyxcMWUsMcflZ890LDOiKH5EYp41SKjw3vTIeuZ/+QyfPQU3I1QqGB6/LttbKZRcTYfVULCuZdRwlJHME26xrq5fDuFQu8M3F6xl3ozTMJB1PL8rs1RiLty6tMnrCM9Qop15bn2WpENCkyjk08wFsHD1Fr6/dAOJ5nBB510onaOPLTh4zWWZpeP7DseyPtORmEv/kMXBFJenbzDAwajTpWY9v1nl2t+RTi01APYoMvVuFq7eUvQEtITrHA6kq49GYi43Lm7kpsXZx2vasS+rASoT25IsAbuQ7XkLqes/67+e4sdPbiKSo5lr4eot3PzoUaMBajSUbMIBi4+OG5rzdtsSrzrKtjAm3WgAROK9IWE7U603u1Iv82ReqhnnpVSw7gmox1EG/OKhc5Y1EQr4d7Tm+/BdMmkUQ/oF+dYDr3AoRbjJFitroJ5nMEyHlU7GGG74zAe5/X/eSCYnXWNY1bybSyaNovVAhOtTVHH91gww5zGtiFJyE7TgUw0jePq1d9n4zr6c+yUq7g7lECYTEa5/eF1W0bebkSPJ7Nko1MMvNHxcCgXrnoJ6HAVSzJWLr+yBbXV4dZSLhpGDs74kjnGzSlzbXZeGkYO5/dLTCNn+I0tsgc9/ZAx3/mVTWkVLzCXpTTTtyN1XkciDeM+xsLEogTy7FXgIpZtxzsl1/Onb5/DM6+8RiRlfFd1C8cptDcYYwoGjs0BcQ9rn1BiSPRt+atE3Lm5kxaZ3077DxSrhdkXBuiehhqMAiv1w+c/cNsy+eHynXF0/N3nejInMm+HvOl8yaRT/cWF2HTzAdz51CovWtvjWvR/1JnIHnL4+dVyygqpQafNYnt20MqpnsuLNVn74p9c6DJsWQ99ggPuumswDV0/h1osbsj47jiE+0tj/4i0SM3zrgVeS3+F8Q81KFeLqqWioqgM6U4aXq6zvkkmjkjIgxbq6mWWviaqRVbPO8z3e2Tm6ZY/pH8LKjHHFiToObYejjBzcl6AtvobhVyvf4nertxCJuToTQ8nLU6+/l9PzzUUimhsO2L4DohpGDmJl825mL23yvf++w+00btvrXdj4hGwPpTQgzv/yZN9KqYVrtvKLZ5t7dL9GV1HD0QGdLcPLFQ/1q9golLoBYVY27y4obls/YiBXnTU2TbDu4/V13Lpso6+3kWimu3bhq0RiDh894RhWv9Wa1YfR7hjaHR2SoRTGxNFDWLdtT0HNgSHbQgS+f/F4JowczJq3Wrnjz28QsLxGwVsu8jrNZz2yPivsC16S/fqH16VNkFy0tgUL4VB7+mc2Vz4w6jgFzVDv7WioqgO6UoZX6nhosa71bdM/zFP/fg53zDiNX191Bmve3pNlNOz4hD3XeEZhfyRG1DGs2uwZjX8YMxgd6610lpe27MHqIFyVyHNFHU86ZO5jG9mwvY0fP7kJEeFQu4vg6UY9uGarb/grYAkYQyR2dHTAorUtPDbzbH755TOyRgKk5gNTw70zzz05a2KlzjrPRj2ODsjXTVppcnk/+Vzr+hEDaXpnH/+y8JWsq7S+QW/am9/VW4JXt7WV58kovYZcOlIAQVsIWlaaR2BbwpxlTWmfy4hjwDHcvfzNrIZzW7xKwcyPcdCyOBh1OOeU4cyb4f8dzowMANzzbHrzai31a9SKVpYajgKoRhme3wfEz/uJOi73LH+TSMxkudYATTvauGlxo69xcFwTHw2qoSelOlgiOCbDo3dMvArR53Npsgsqcl33RGJOUmnXz0A0btub/H6lfqdr5UIxk1rSylKRwxok3wckc/TmtZ+oZ/6KzVmjLr95zkn84tlmLBHf+vhQwGL2tPHMfXxjp0eEKkpXCVhw1kl1PN/cmtz2yVOHs+LN3Xk94ULoE/RCTpkn2EJOwLVyZZ+6nq6MxC0UFTnspnRUxZV55bTnYJSfP5OuO5Wa4PMjZAtPXHc29SMGgpCcdhZzXD50/EDWbfNv1rKAj59cx3NvtvrerijFEnNJMxrgVWPlwq9bPBeJk2zq9yff9wtIMxa1YDAS1JpWVsXTniIyRkSWi8hGEWkSkW/Htx8jIk+KyJvx/4fGt4uI3CUizSKyXkROr/SaK0khM5MTSfeVzbuZdvdKrHh2MWwL4YAw4/QxXrIwB7MvbqB+xECWrNvOnKUbwBiORBy+PvVELjtjLMEcnwoX1GgoVcPLb+cv7/Wr/k39/uT6fi1cs9W3V6tW+jlqTSurGvUyMeB6Y8x44EzgWhEZD3wXeNoYczLwdPxvgM8CJ8d/rgHurfySK0ehH5DWA5GkVlXiysrrBBeWNG7PqoFP0D9sM2HUYFoPRPi3h9YRdfDkz4F7n9vMzY9uKGqugqJUkhs+fYrv9nDA4l/+8SRsnzNaquqzf57Q89AzqxUXrt5SVONvOak1rayKGw5jzDvGmFfiv+8HXgNGAdOB38Z3+y3wufjv04EFxmM1MEREjq/wsitGoR8Qv3kaTnwaWj55h1j8S/TC31tVkFDpVoSDNlNOquM/PzeBoA19AxYhG/7lH0/ivqsm0zByMOLjkVx6+tH8hd/3y68EN1HZVYwoabm5ZNIoVs06jweunsKqWedVtSmxqjkOERkH/AOwBhhhjHknftNOYET891HAtpS7tcS3vUMPJV8Vl6cl1cbdz2zq8Dj9wzafmziK37+4NRkZdQ2sat7NGztzi84pSi1yMOKw5q1WRgzqgyUWYgnGEe57fjP/769v5yzyWLJuO//96vZkErygElyfyi5LhKYdbZxzyrHle5IdUCu5l6q1donIAOAR4N+MMWlnMeOVehV1QSwi14jIWhFZ+957uZNr5aAccVC/5sGEZta3HngFv0hUZl7DcQ3DBoTSVEXbHZMc9aoo3Y0fPPE6NzzcSCTmcijq0O4YYq7/2OIEB6NuUuSwedd+GrftBUh+v/y8kNkXj8fJqDg9FPVm4VQzZFUrVMXjEJEgntFYaIz57/jmXSJyvDHmnXgo6t349u3AmJS7j45vS8MYMx+YD145btkWn0GlaqtTq0Fy8c+nj2Jp447kWi6ZONJ3Ep8twoeOH1TyNSpKJeisRlokZvjMT1fQLxSg3XW5ZZonbTJ6aN+kF9K0ow0QRg7uw7WfqOfujOrESMyoBAlVMBwiIsCvgdeMMT9OuWkp8BXgh/H/l6RsnykiDwFTgLaUkFZVqeQcYr9yvEyWNu7gsZlnczDqNT5deNfzvvtFYg6D+ga5fPIoFq3Vqyel9+AYkj1PN/9xA/1DNo4x3H7p0Yma4HkwYVtwjckqAe7NI2MTVMPjmAp8GfibiKyLb/sPPIOxSES+AWwBLo/f9gRwIdAMHAK+Vtnl5qaStdV+1SCZJCQWJo4ZwopN72GJ/yDXmAvfXPAyIvDVs07ggTVb02ZzKEqtE7QgGLCJxlxcYzotz5+oPrz+4XUIki1zAmR+hwotg621JsJSUnHDYYxZSe5i7PN99jfAtWVdVCepZG11Ig574+L1ORv72l2X/iGbu55+M28DIBzVD3rwxS10VBuvKLWGa7ycQ8AiS7uqM3hSWf4HCtuCESFs55YgyTQS5Qxh14JB0s7xLlBpAURvjGyIb/3u5SyZ6JAtXH7GaC76+fNE8k1OyiCa5wujKLVKwhnIc21UMsQSHo+HgP1O1plG4pZp45n72MayhLBrRa9KDUcXKbUAYkdXEw0jB+FmnOhDAYsHv/FRvnT/i0UZDUXprdiWV1Kar9k1bAtiCbdfeponz+ODX55zzrKNWWOVSxHCrmROtSPUcJSAUtVWp15NRB2HmeeezBVTxiaPnejh+NrHxnH/qrcJpbjOwYBNTxSsVJRy0Cdg86NLT+PfF63LqtLqGx/NPOaYfskLuNQLOjiqaeWb57SF9ljpQ9i1pFelhqNG8LuauPPJTdy9vJl5M7yKj+sXrUu65rbAzHPrk4aledd+9TYUpUAcYzj1uIFcffaJ3Pvc5vTbXO+2+hEDaT0QSeYMQ7bFkZiD6xr6hmzaHZcvfmQsR2JO1v1nX9zA3Mc3ljSEXUt6VSqrXiOs2PSeb+4CPB2emONmzR2wBV68+ZMALGvczq3LXqvEUhWl2yJ4od3zTz2Wp17bScCys8fK2oJtCZdPHs0fXtpW8AVZpoy7n5fSP2Qny+Vz5UzykTlWodQ5jkJl1dVwVJnWAxEWrtmatwqqb9DicI5g7FfPOoEHX9yKCOpxKEoH2OKNqi2HkKct8IdrzuTE4QN8K6zA6w8JWF5SPzWHUszJv5xVVWo4atRwpL7pK5t3c+PD63zlQ1IJ2tLpbllFUSqHJ/tj6BuMd6df1PGwtHIMZOosOsipBlmybjs3LW7EFouY69LumIIKYR3HYEGennFFUWqBRDd1IXoAAA16SURBVCNtojt9zrImQoH8koDdsRO9aiKHvY3WAxGuX7SOSMxwqN0hWqDRAM9gqNFQlOphARd+eAThgNA/ZGOLp/nWEZ7Cbv5veq55O7UwQCoX6nGUiI7ijk079lWkWUlRlNJj28Lc6R9m7nTSEt4/f/pNfvPClpz3c4xh9sVeQyCk5ziCtmCJZFVc1UqTXz7UcJSAxBtti9DuuNzwmQ8y5cS6DCPSsX/hryylKEq1scWrXDxp+EAaRg5KlsA/+NI2330TCryJk/4FDcclq6qe2LCTu5/ZRMCycEz61aRfWf4NDzcy/vhBySZElRzpRuR6s/zkzn/wxOsELbBtK/nBGTm4LwFL8ooJfu+zpzLvL29oIlxRaowjMZMsd7eAM8YN4eW392aFkPsGLb4/bTxB22LSmCHJk32iSbj1QIRfPNscH9nsVcXcuLgx2f3t1+QXdQwX/nwld8w4quBbbW9EDUcGfgYi4VEELE89c/bF47lyygkANO1o85UIbHe92OVNj6xnZ9sR7vjLG+ST8PzqWSdwzT9+gJa9h1jwwtZyPDVFUUqAC7z09l7f2w63u3zvjxvoE7QwJr1JF/y7vyMxw4NrtnLd+SfnVMGOxlxuXLweMERiRiVHagm/2OLU+mFZHsXNf9wABgb0CXDT4sa8/ROO6/KDP73e4WNPHncMzbv2s3C1Gg1F6e4kzhee+sObzJsxkUsmjWL00L5p42gT/PSpTUkDc/ulp3HDw41ZSXXbEjACHL1/tSqytKoqTmrIKXU4fdOONt/qiVuXbeCmxes7bLrzaQT35brfv8qnfrIiqztcUZTuTSRmuHFxIys2eUNNvz71xKx9HOMV0ABMrR/Gjy+fSMhOPz07rsnKiVRLckQ9jjh+LqRtCY++uiM57CWVgFiI1XE5Xub0sFyovVCUnkskZvhfv3sZA3xt6rgce5m0qIdrXIK20CdgJyMgQJrkyC0Xjadlz2GALDHGcnohajji+MUWD0Yc/vtV/9GqUcclkOGwWXhle0FbiLlww6dP4Y6/bELNgqIoCdmgXz+/OatQxhZ4Y+d+bv/z67Q7JC9gQzbcc+U/0DBycNIQJMY4bNjexm2PNWGLV531+cljWPRyS0US5yo5kkJCQMy2hIOR/DGmoC380z+MTJvZbVvCbdMbmDBycNJ9nPXIep567d2i16IoigLwhY+MZvIJx6RVabUeiDDlB0/l7Q3rjJRJoZIj3SbHISIXiMgbItIsIt8tx2NcMmkUq2adx42fPqXDfUMBiz9meCOOa7ht2cakDtXHfviMGg1FUbrEQy+1cMPi9XzyJyv4/pK/AYU1FCcS5+WgWxgOEbGBe4DPAuOBL4rI+HI8Vt2AMCcN95/2lUp7zBCQ7JfPtoSmHfuY9Uju2eCKoiidYcELW2netZ99h9s73LecifNuYTiAjwLNxpjNxpgo8BAwvVwP1jByEB3okvGNs8dljXAFz+sAQ9Aqz0trd5yPVxSlB7Num38PSSqhgFWS4VG56C6GYxSQ2tvfEt+WRESuEZG1IrL2vffe69KD1Q0I8+PLJxGy/W8P2cLVHz+JeTMmphmYoC3Mm3EaDSMH+zbxdJWgLdw2fQIFFHMpitJN6ej7PWnMEAb1zV3XFLKFJ647u6wd5T2mqsoYMx+YD15yvKvHu2TSKKbWD+PBNVv56VObkv0VQVu447KJ1A0IJ/fx6q9NWuXD7ZeeliybizouH6+v47k3dxO0hUPRbKNii1fLnfg/aIFlWXz/4vGMGdov7fgHozF+8ETHTYVhGxwjjKvry+bdh8Coyq6idBVbYNYFp3LHk5u8CkrH8PV4ie2vV74FeDIhiVEI4YDgGnBdQziQPXEwwTn1dVx9zkk0jBzM/zTtZM6yjTgZkz+vOmss9SMGMrR/yHdOTzggzJsxMZlELxfdoqpKRM4CbjXGfCb+9/cAjDH/5bd/qQc5tR6I+BqHQu6XWlOd+HvD9jbmPr7xqCjipz/IlJPqihoredPiRhatbUn+feaJQ1nX0na0vnva+GR1V+pjt8ccNuzYxzt7D7Pl/UMMHxDGYBgQCnAg/ti7D0YY3CfIyCF9icQc2g7HeG3HPg63x3ANDBsQ5txTj+XYgWF+98LbrN/eRnvMJRiweP9QrASvuKJ0jfHHDeTN9w54QoKuyxVTxjL5hKGA0LSjjftWbEbEUwG6bfoEppx4DCubdzNsQIhTjxvEnzbs5O7lzQRs4VDEwQAhC8SSZBe4X89EYpvfdxmOKuvuORhlZfNuYo5LwLY4u35Y1sk+9Tv7duuhtKoq8KpAb1zsVYE6rsmSN+kMPWoCoIgEgE3A+cB24CXgCmNMk9/+tTwBMEEpGnWad+1n3ba9yQ9ULahmZn7Yx9X1Y+e+CLsPHOHs+uHsPRTlzxt3MWJgmOMGe1+md9oOs37bHiIxw8kjBnD62KG8vnMfuw9E6ReyCdoWz77xLpt27sM1wsA+NqOP6c+YIX3Z+I6XKAwFvSu/xNXdkXZD1HHoE7AI2sKew+0ciuafgWIDItCTJ/DapApWQN+AFw+POQbHMdgBOKZfmKH9QrzTdphgwGZQOMC+SIz+oQD9wzYB22JA0OLVljYsgWP69+GjJw7l0tNHc6jd4Y2d+3lj5z6CtsXoof049biB7DnUTsCCpnf2c9Kw/hw7MMyfNuwEDF/4yFhOHD6AF/7eyu4DEcYM7cu2PYcYNqAPZ32gDjg6r3tH25F4YtgwqG+I9phDY0sb55w8jCH9QgV/Hwr5rvjNDK/md8uPUn/ne5ThABCRC4Gf4n327zfG/GeufbuD4VAURak1etzoWGPME8AT1V6HoihKb6e7VFUpiqIoNYIaDkVRFKUo1HAoiqIoRaGGQ1EURSmKblNVVQwi8h6wpdrryMEwYHe1F1FF9Pn37ucP+hrU8vM/wRgzvKOdeqThqGVEZG0h5W49FX3+vfv5g74GPeH5a6hKURRFKQo1HIqiKEpRqOGoPPOrvYAqo89f6e2vQbd//prjUBRFUYpCPQ5FURSlKNRwVAARuUxEmkTEFZHJGbd9Lz5H/Q0R+Uy11lhJRORWEdkuIuviPxdWe02VQEQuiL/PzSLy3Wqvp9KIyNsi8rf4e94rVEhF5H4ReVdENqRsO0ZEnhSRN+P/D63mGjuDGo7KsAH4Z2BF6sb43PQvAA3ABcAv4vPVewM/McZMiv/0ePHK+Pt6D/BZYDzwxfj739s4N/6ed+ty1CL4Dd53O5XvAk8bY04Gno7/3a1Qw1EBjDGvGWPe8LlpOvCQMSZijHkLaMabr670PD4KNBtjNhtjosBDeO+/0oMxxqwA3s/YPB34bfz33wKfq+iiSoAajurS4Sz1HsxMEVkfd+W7naveCXrze53AAH8RkZdF5JpqL6aKjDDGvBP/fScwopqL6QzdZh5HrSMiTwHH+dx0szFmSaXXU23yvR7AvcBcvBPJXOBO4OuVW51SJc42xmwXkWOBJ0Xk9fgVea/FGGNEpNuVtqrhKBHGmE924m7bgTEpf4+Ob+v2FPp6iMh9wGNlXk4t0GPf60IxxmyP//+uiPwRL3zXGw3HLhE53hjzjogcD7xb7QUVi4aqqstS4AsiEhaRE4GTgRervKayE/+yJPgnvOKBns5LwMkicqKIhPCKIpZWeU0VQ0T6i8jAxO/Ap+kd77sfS4GvxH//CtDtIhLqcVQAEfkn4OfAcOBxEVlnjPmMMaZJRBYBG4EYcK0xxqnmWivE7SIyCS9U9Tbwv6q7nPJjjImJyEzgz4AN3G+MaarysirJCOCPIgLeeedBY8z/VHdJ5UdEfg98AhgmIi3AbOCHwCIR+Qaeivfl1Vth59DOcUVRFKUoNFSlKIqiFIUaDkVRFKUo1HAoiqIoRaGGQ1EURSkKNRyKoihKUajhUJROICJDROR/V+BxPiEiHyv34yhKMajhUJTOMQQo2HCIR2e+b58A1HAoNYX2cShKJxCRhLrtG8By4DRgKBAE/o8xZomIjMNr+FsDnAFcCHwSmAXsBRqBiDFmpogMB34JjI0/xL/hSZKsBhzgPeA6Y8zzlXh+ipIPNRyK0gniRuExY8wEEQkA/Ywx+0RkGN7J/mTgBGAz8DFjzGoRGQn8FTgd2A88AzTGDceDwC+MMStFZCzwZ2PMh0TkVuCAMeaOSj9HRcmFSo4oStcR4Acicg7g4smlJ6SytxhjVsd//yjwnDHmfQAReRj4/9u7W5UKgjgM489rMviBN2EwCRosIoI3IEbBbNFrMHk7R7DZT7IcVMSmxSIYRAXBMIYd4SAKjuLxhOeXdmfY2dn0MjPw3/natwEs1JIcADNJpkYxeamVwSH93jZdHbKlUsprkhtgsvY9f3OMCWCllPIy3DgUJNLY8HBc+plHYLpezwJ3NTTW6baoPnMKrCWZq9tbW0N9J8De+00tAvnxPdJYMDikHyil3AP9JBfAIrCc5BzYAa6+eOYWOKQrnd+nqwz8ULv36xhnSS6B3dp+DGwmGSRZ/avvkVp4OC6NUJKpUspTXXH06Mqr9/57XlILVxzSaB0kGdD9xOgaOPrn+UjNXHFIkpq44pAkNTE4JElNDA5JUhODQ5LUxOCQJDUxOCRJTd4AKsRAhupeFj8AAAAASUVORK5CYII=\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"fig = plt.figure()\nres = stats.probplot(all_data_temp['pur_mean'] , plot=plt)\nplt.show()\nall_data_temp.plot.scatter(x='target', y='pur_mean');\nall_data_temp.shape","execution_count":17,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"execute_result","execution_count":17,"data":{"text/plain":"(323273, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data_temp = all_data_temp[all_data_temp['pur_mean'] < 20]\n\nfig = plt.figure()\nres = stats.probplot(all_data_temp['pur_mean'] , plot=plt)\nplt.show()\nall_data_temp.plot.scatter(x='target', y='pur_mean');\nall_data_temp.shape","execution_count":18,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"execute_result","execution_count":18,"data":{"text/plain":"(322698, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data_temp = all_data_temp[all_data_temp['Mon_mean'] > -12]\n\nfig = plt.figure()\nres = stats.probplot(all_data_temp['Mon_mean'] , plot=plt)\nplt.show()\nall_data_temp.plot.scatter(x='target', y='Mon_mean');\nall_data_temp.shape","execution_count":19,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"execute_result","execution_count":19,"data":{"text/plain":"(322670, 13)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"all_data = all_data.drop(['card_id'], axis=1)\nall_data = all_data.drop(['target'], axis=1)\nall_data = all_data.drop(['merchant_id'], axis=1)\n#all_data = all_data.drop(['merchant_group_id'], axis=1)\n\n# Get_Dummies para transformar categoricos em Numéricos \nall_data  = all_data.fillna(0)\nall_data = pd.get_dummies(all_data)\n\nprint('all_data - Depois de todos processamentos: ', all_data.shape)\n\n#Cria Matriz X_train utilizando a Matriz com todos os dados all_data: do inicio da matriz (:) até o fim  da matriz df_train.shape[0]\nX_train = all_data[:train.shape[0]]\n\n#Cria Matriz X_test utilizando a Matriz com todos os dados all_data: a partir do último registro matriz df_train.shape[0], ou seja, todos os registros que não estiverem em df_train\nX_test = all_data[train.shape[0]:]\n\n# Cria o y, ou seja, o que será previsto, apenas com o campo \"Target\"\ny = df_train_original.target","execution_count":20,"outputs":[{"output_type":"stream","text":"all_data - Depois de todos processamentos:  (325540, 20)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Function declaration \n\ndef RunModel (ModelName, Model, Df_Test_Original, X_train, X_test, y):\n\n    print ('# # # # Prediction for Model:  ', ModelName, '# # # #')\n    print ('Shape X_train: ', X_train.shape)\n    print ('Shape X_test: ', X_test.shape)\n    print ('Shape y: ', y.shape)\n    print ('Model: ', Model)\n    \n    print ('.FIT Model: ', Model)\n\n    Model.fit(X_train, y)\n\n    print ('PREDICT TRAIN: ', Model)\n\n    yhat_Train = Model.predict(X_train)\n    \n    print ('PREDICT TEST: ', Model)\n\n    yhat_test = Model.predict(X_test)\n    \n    print ('# # # # Tamanho do Df_Test_Original:', Df_Test_Original.shape)\n    print ('# # # # Prediction:', yhat_test.shape, yhat_test)\n\n    Filename = 'Output_' + ModelName + '.csv'\n    \n    df_Output= pd.DataFrame()\n    df_Output['card_id'] = Df_Test_Original['card_id']\n    df_Output['target'] = yhat_test\n    df_Output.to_csv(Filename, index = False)\n    \n    return yhat_test\n\ndef Bagging (ModelName, Model, Df_Test_Original, X_train, X_test, y):\n\n    from sklearn.ensemble import BaggingClassifier\n    n_estimators = [1,2,3,4,5,10,15,20]\n    dt_bag_scores = []   \n    lr_bag_scores = []\n    for ne in n_estimators:\n        dt = DecisionTreeClassifier(max_depth=15, random_state=1)\n        lr = LogisticRegression(random_state=1)\n    \n        dt_bag = BaggingClassifier(dt, n_estimators=ne)\n        lr_bag = BaggingClassifier(lr, n_estimators=ne)\n\n        dt_bag.fit(X_train, y)\n        lr_bag.fit(X_train, y)\n\n        dt_bag_scores.append(eval_auc(dt_bag, X_test, yhat_test))\n        lr_bag_scores.append(eval_auc(lr_bag, X_test, yhat_test))\n\n        print(ne, dt_bag_scores[-1], lr_bag_scores[-1])\n\ndef GridSearch (ModelName, Model, Df_Test_Original, X_train, X_test, y):\n    from sklearn.model_selection import GridSearchCV\n    params = {\"max_depth\": [3,10,None],\"max_features\": [None, \"auto\"],\"min_samples_leaf\": [1, 5, 10],\"min_samples_split\": [2, 10]         }\n    gsdt = GridSearchCV(dt, params, n_jobs=-1, cv=5)\n    gsdt.fit(X_train, y)\n    display ('Best Parameters!!: ',  gsdt.best_params_)\n    display ('Best Score     !!: ',  gsdt.best_score_)\n    bdt.get_params()\n    params = {\"base_estimator__max_depth\": [3,10,None],\"base_estimator__max_features\": [None, \"auto\"],\"base_estimator__min_samples_leaf\": [1, 5, 10],\"base_estimator__min_samples_split\": [2, 10],'bootstrap_features': [False, True],'max_features': [0.5, 1.0],'max_samples': [0.5, 1.0],'n_estimators': [5, 15, 40],         }\n    gsbdt = GridSearchCV(bdt, params, n_jobs=3, cv=5)\n    gsbdt.fit(X_train, y)\n    display ('Best Parameters!!: ',  gsdt.best_params_)\n    display ('Best Score     !!: ',  gsdt.best_score_)\n#\n# ! ! ! ! ! ! ! \n#\ndef Runlightgbm  (ModelName, Model, Df_Test_Original, X_train, X_test, y):\n\n    print ('# # # # Prediction for Model:  ', ModelName, '# # # #')\n    print ('Shape X_train: ', X_train.shape)\n    print ('Shape X_test: ', X_test.shape)\n    print ('Shape y: ', y.shape)\n    print ('Model: ', Model)\n\n    import lightgbm as lgb\n    d_train = lgb.Dataset(X_train, label=y)\n    \n    params ={\n                'task': 'train',\n                'boosting': 'goss',\n                'objective': 'regression',\n                'metric': 'rmse',\n                'learning_rate': 0.01,\n                'subsample': 0.9855232997390695,\n                'max_depth': 7,\n                'top_rate': 0.9064148448434349,\n                'num_leaves': 63,\n                'min_child_weight': 41.9612869171337,\n                'other_rate': 0.0721768246018207,\n                'reg_alpha': 9.677537745007898,\n                'colsample_bytree': 0.5665320670155495,\n                'min_split_gain': 9.820197773625843,\n                'reg_lambda': 8.2532317400459,\n                'min_data_in_leaf': 21,\n                'verbose': -1,\n                'seed':int(2),\n                'bagging_seed':int(2),\n                'drop_seed':int(2)\n                }\n    \n    print ('.FIT Model (Nesse caso, lgb.train: ', Model)\n\n    clf = lgb.train(params, d_train)\n\n    print ('PREDICT TEST: ', Model)\n\n    yhat_test=clf.predict(X_test)\n    \n    print ('# # # # Tamanho do Df_Test_Original:', Df_Test_Original.shape)\n    print ('# # # # Prediction:', yhat_test.shape, yhat_test)\n\n    Filename = 'Output_' + ModelName + '.csv'\n    \n    df_Output= pd.DataFrame()\n    df_Output['card_id'] = Df_Test_Original['card_id']\n    df_Output['target'] = yhat_test\n    df_Output.to_csv(Filename, index = False)\n    \n    return yhat_test","execution_count":21,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 0) Let's Bagging:\nSel_Model = 'Bagging'\nNameM = 'Bagging'\n#Bagging(NameM, Sel_Model, df_test_original, X_train, X_test, df_train_original['target'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 0) Grid Search:\nSel_Model = 'GridS'\nNameM = 'GridS'\n#GridSearch(NameM, Sel_Model, df_test_original, X_train, X_test, df_train_original['target'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# -1) Run Runlightgbm:\nSel_Model = lgb\nNameM = 'lgb'\n#Runlightgbm(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":23,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 1) Gradiente Boost Model\nSel_Model = GBR\nNameM = 'GBR'\nMGBR = RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":24,"outputs":[{"output_type":"stream","text":"# # # # Prediction for Model:   GBR # # # #\nShape X_train:  (201917, 20)\nShape X_test:  (123623, 20)\nShape y:  (201917,)\nModel:  GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n             learning_rate=0.1, loss='ls', max_depth=3, max_features=None,\n             max_leaf_nodes=None, min_impurity_decrease=0.0,\n             min_impurity_split=None, min_samples_leaf=1,\n             min_samples_split=2, min_weight_fraction_leaf=0.0,\n             n_estimators=100, n_iter_no_change=None, presort='auto',\n             random_state=None, subsample=1.0, tol=0.0001,\n             validation_fraction=0.1, verbose=0, warm_start=False)\n.FIT Model:  GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n             learning_rate=0.1, loss='ls', max_depth=3, max_features=None,\n             max_leaf_nodes=None, min_impurity_decrease=0.0,\n             min_impurity_split=None, min_samples_leaf=1,\n             min_samples_split=2, min_weight_fraction_leaf=0.0,\n             n_estimators=100, n_iter_no_change=None, presort='auto',\n             random_state=None, subsample=1.0, tol=0.0001,\n             validation_fraction=0.1, verbose=0, warm_start=False)\nPREDICT TRAIN:  GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n             learning_rate=0.1, loss='ls', max_depth=3, max_features=None,\n             max_leaf_nodes=None, min_impurity_decrease=0.0,\n             min_impurity_split=None, min_samples_leaf=1,\n             min_samples_split=2, min_weight_fraction_leaf=0.0,\n             n_estimators=100, n_iter_no_change=None, presort='auto',\n             random_state=None, subsample=1.0, tol=0.0001,\n             validation_fraction=0.1, verbose=0, warm_start=False)\nPREDICT TEST:  GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n             learning_rate=0.1, loss='ls', max_depth=3, max_features=None,\n             max_leaf_nodes=None, min_impurity_decrease=0.0,\n             min_impurity_split=None, min_samples_leaf=1,\n             min_samples_split=2, min_weight_fraction_leaf=0.0,\n             n_estimators=100, n_iter_no_change=None, presort='auto',\n             random_state=None, subsample=1.0, tol=0.0001,\n             validation_fraction=0.1, verbose=0, warm_start=False)\n# # # # Tamanho do Df_Test_Original: (123623, 5)\n# # # # Prediction: (123623,) [-0.83611589  0.29803942 -0.25530564 ... -1.15522376 -1.01066509\n -0.40555532]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 4) xgboost Model (Megazord) XD - Didn't run, Kernel died\nSel_Model = XGB\nNameM = 'XgBoost'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 8) Gaussian is not possible - Unknown label type\nSel_Model = GNB\nNameM = 'Gaussian'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 8) Gaussian is not possible - Unknown label type\nSel_Model = GNB\nNameM = 'Gaussian'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 9) Neural Model is not possible - Unknown label type\nSel_Model = NN\nNameM = 'NeuralModel'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 10) lasso\nSel_Model = lasso\nNameM = 'lasso'\n#MLASSO = RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":31,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 11) Elastic Net\nSel_Model = ENet\nNameM = 'ElasticNet'\n#MENET = RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":32,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 12) Kernel Ridge - Error Unknow\nSel_Model = KRR\nNameM = 'NeuralModel'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 13) LGBMRegressor\nSel_Model = model_lgb\nNameM = 'LGBMRegressor'\nLGBMR = RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":28,"outputs":[{"output_type":"stream","text":"# # # # Prediction for Model:   LGBMRegressor # # # #\nShape X_train:  (201917, 20)\nShape X_test:  (123623, 20)\nShape y:  (201917,)\nModel:  LGBMRegressor(bagging_fraction=0.8, bagging_freq=5, bagging_seed=9,\n       boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n       feature_fraction=0.2319, feature_fraction_seed=9,\n       importance_type='split', learning_rate=0.05, max_bin=55,\n       max_depth=-1, min_child_samples=20, min_child_weight=0.001,\n       min_data_in_leaf=6, min_split_gain=0.0, min_sum_hessian_in_leaf=11,\n       n_estimators=720, n_jobs=-1, num_leaves=5, objective='regression',\n       random_state=None, reg_alpha=0.0, reg_lambda=0.0, silent=True,\n       subsample=1.0, subsample_for_bin=200000, subsample_freq=0)\n.FIT Model:  LGBMRegressor(bagging_fraction=0.8, bagging_freq=5, bagging_seed=9,\n       boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n       feature_fraction=0.2319, feature_fraction_seed=9,\n       importance_type='split', learning_rate=0.05, max_bin=55,\n       max_depth=-1, min_child_samples=20, min_child_weight=0.001,\n       min_data_in_leaf=6, min_split_gain=0.0, min_sum_hessian_in_leaf=11,\n       n_estimators=720, n_jobs=-1, num_leaves=5, objective='regression',\n       random_state=None, reg_alpha=0.0, reg_lambda=0.0, silent=True,\n       subsample=1.0, subsample_for_bin=200000, subsample_freq=0)\nPREDICT TRAIN:  LGBMRegressor(bagging_fraction=0.8, bagging_freq=5, bagging_seed=9,\n       boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n       feature_fraction=0.2319, feature_fraction_seed=9,\n       importance_type='split', learning_rate=0.05, max_bin=55,\n       max_depth=-1, min_child_samples=20, min_child_weight=0.001,\n       min_data_in_leaf=6, min_split_gain=0.0, min_sum_hessian_in_leaf=11,\n       n_estimators=720, n_jobs=-1, num_leaves=5, objective='regression',\n       random_state=None, reg_alpha=0.0, reg_lambda=0.0, silent=True,\n       subsample=1.0, subsample_for_bin=200000, subsample_freq=0)\nPREDICT TEST:  LGBMRegressor(bagging_fraction=0.8, bagging_freq=5, bagging_seed=9,\n       boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n       feature_fraction=0.2319, feature_fraction_seed=9,\n       importance_type='split', learning_rate=0.05, max_bin=55,\n       max_depth=-1, min_child_samples=20, min_child_weight=0.001,\n       min_data_in_leaf=6, min_split_gain=0.0, min_sum_hessian_in_leaf=11,\n       n_estimators=720, n_jobs=-1, num_leaves=5, objective='regression',\n       random_state=None, reg_alpha=0.0, reg_lambda=0.0, silent=True,\n       subsample=1.0, subsample_for_bin=200000, subsample_freq=0)\n# # # # Tamanho do Df_Test_Original: (123623, 5)\n# # # # Prediction: (123623,) [-0.85167928 -0.4426249  -0.16059523 ... -1.06481923 -1.10803855\n -0.3208102 ]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 14) Linear Regression\nSel_Model = LR\nNameM = 'LinearRegression'\n#RunModel(NameM, Sel_Model, df_test_original, X_train, X_test, y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#15) Ensemble with best values\nensemble = MGBR*0.50 + LGBMR*0.50\n\nModelName = 'Ensemble'\nFilename = 'Output_' + ModelName + '.csv'\n    \ndf_Output= pd.DataFrame()\ndf_Output['card_id'] = df_test_original['card_id']\ndf_Output['target'] = ensemble","execution_count":34,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.4","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}