credit-part3-modeling-2-selectmodels-pipe1-part2_fs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b6a3b465",
   "metadata": {
    "papermill": {
     "duration": 0.032929,
     "end_time": "2021-12-28T11:23:59.136891",
     "exception": false,
     "start_time": "2021-12-28T11:23:59.103962",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Feature engineering and feature selection is an iterative process that starts with engineering new features, training a model, and then testing the model predictions against the target labels. The goal is to determine which features improve the model’s prediction accuracy. You repeat this process, along with hyperparameter tuning, until you are satisfied with the model’s accuracy. \n",
    "\n",
    "# Load Lib "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "855d6056",
   "metadata": {
    "_cell_guid": "ddedf45c-cdd2-4013-973f-6ccf6c78347a",
    "_uuid": "20c9b28e-4c93-447d-8210-cad89f404354",
    "collapsed": false,
    "execution": {
     "iopub.execute_input": "2021-12-28T11:23:59.203333Z",
     "iopub.status.busy": "2021-12-28T11:23:59.201726Z",
     "iopub.status.idle": "2021-12-28T11:24:02.055804Z",
     "shell.execute_reply": "2021-12-28T11:24:02.056269Z"
    },
    "jupyter": {
     "outputs_hidden": false
    },
    "papermill": {
     "duration": 2.889469,
     "end_time": "2021-12-28T11:24:02.056617",
     "exception": false,
     "start_time": "2021-12-28T11:23:59.167148",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type='text/css'>\n",
       ".datatable table.frame { margin-bottom: 0; }\n",
       ".datatable table.frame thead { border-bottom: none; }\n",
       ".datatable table.frame tr.coltypes td {  color: #FFFFFF;  line-height: 6px;  padding: 0 0.5em;}\n",
       ".datatable .bool    { background: #DDDD99; }\n",
       ".datatable .object  { background: #565656; }\n",
       ".datatable .int     { background: #5D9E5D; }\n",
       ".datatable .float   { background: #4040CC; }\n",
       ".datatable .str     { background: #CC4040; }\n",
       ".datatable .time    { background: #40CC40; }\n",
       ".datatable .row_index {  background: var(--jp-border-color3);  border-right: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  font-size: 9px;}\n",
       ".datatable .frame tbody td { text-align: left; }\n",
       ".datatable .frame tr.coltypes .row_index {  background: var(--jp-border-color0);}\n",
       ".datatable th:nth-child(2) { padding-left: 12px; }\n",
       ".datatable .hellipsis {  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .vellipsis {  background: var(--jp-layout-color0);  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .na {  color: var(--jp-cell-editor-border-color);  font-size: 80%;}\n",
       ".datatable .sp {  opacity: 0.25;}\n",
       ".datatable .footer { font-size: 9px; }\n",
       ".datatable .frame_dimensions {  background: var(--jp-border-color3);  border-top: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  display: inline-block;  opacity: 0.6;  padding: 1px 10px 1px 5px;}\n",
       ".datatable .frame thead tr.colnames {  background-image: url('data:image/png;base64,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');  background-repeat: repeat-x;  background-size: 14px;  height: 28px;}\n",
       "</style>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Load the librarys\n",
    "import pandas as pd #To work with dataset\n",
    "import numpy as np #Math library\n",
    "import matplotlib.gridspec as gridspec\n",
    "import seaborn as sns #Graph library that use matplot in background\n",
    "import matplotlib.pyplot as plt #to plot some parameters in seaborn\n",
    "import warnings\n",
    "# Preparation  \n",
    "from sklearn.preprocessing import LabelEncoder, OrdinalEncoder\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.preprocessing import PowerTransformer, StandardScaler,Normalizer,RobustScaler,MaxAbsScaler,MinMaxScaler,QuantileTransformer\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.preprocessing import KBinsDiscretizer\n",
    "# Import StandardScaler from scikit-learn\n",
    "from sklearn.experimental import enable_iterative_imputer\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.impute import KNNImputer,IterativeImputer\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.compose import make_column_transformer,ColumnTransformer\n",
    "from sklearn.pipeline import make_pipeline, Pipeline,FeatureUnion\n",
    "from sklearn.manifold import TSNE\n",
    "# Import train_test_split()\n",
    "# Metrics\n",
    "from sklearn.metrics import roc_auc_score, average_precision_score,recall_score\n",
    "from sklearn.metrics import make_scorer,mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_error,classification_report,f1_score\n",
    "from sklearn.metrics import roc_curve,confusion_matrix\n",
    "from datetime import datetime, date\n",
    "from sklearn.linear_model import ElasticNet, Lasso,  BayesianRidge, LassoLarsIC\n",
    "from sklearn.linear_model import LinearRegression, RidgeCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "#import tensorflow as tf \n",
    "#from tensorflow.keras import layers\n",
    "#from tensorflow.keras.callbacks import EarlyStopping\n",
    "#from tensorflow.keras.callbacks import LearningRateScheduler\n",
    "#import smogn\n",
    "from sklearn.experimental import enable_hist_gradient_boosting\n",
    "from sklearn.base import BaseEstimator, TransformerMixin, RegressorMixin, clone\n",
    "from sklearn.kernel_ridge import KernelRidge\n",
    "from sklearn.ensemble import GradientBoostingRegressor,RandomForestRegressor\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import HistGradientBoostingClassifier\n",
    "# For training random forest model\n",
    "import lightgbm as lgb\n",
    "from scipy import sparse\n",
    "from sklearn.neighbors import KNeighborsRegressor \n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from yellowbrick.cluster import KElbowVisualizer\n",
    "# Model selection\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import f_regression,f_classif,chi2\n",
    "from sklearn.feature_selection import mutual_info_regression\n",
    "from sklearn.feature_selection import SelectPercentile\n",
    "from sklearn.feature_selection import mutual_info_classif,VarianceThreshold\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from lightgbm import LGBMClassifier\n",
    "import lightgbm as lgbm\n",
    "from catboost import CatBoostRegressor, CatBoostClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn import svm\n",
    "from xgboost import XGBClassifier,XGBRegressor\n",
    "from sklearn import set_config\n",
    "from itertools import combinations\n",
    "# Cluster :\n",
    "from sklearn.cluster import MiniBatchKMeans\n",
    "#from yellowbrick.cluster import KElbowVisualizer\n",
    "#import smong \n",
    "import category_encoders as ce\n",
    "import warnings\n",
    "#import optuna \n",
    "from joblib import Parallel, delayed\n",
    "import joblib \n",
    "from sklearn import set_config\n",
    "from typing import List, Optional, Union\n",
    "set_config(display='diagram')\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc069e7b",
   "metadata": {
    "_cell_guid": "38c6befb-892d-4ef3-9c63-13be5959b671",
    "_uuid": "2953f9f3-02f8-4fa9-a7fc-cb93565c2183",
    "papermill": {
     "duration": 0.029784,
     "end_time": "2021-12-28T11:24:02.118291",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.088507",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "\n",
    "# Load the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "37e61098",
   "metadata": {
    "_cell_guid": "98a5b4a6-90e9-449d-96cc-0d85c9ff1c77",
    "_uuid": "e086592f-9f6a-4592-80b5-c1e7ad227e02",
    "collapsed": false,
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:02.186940Z",
     "iopub.status.busy": "2021-12-28T11:24:02.186215Z",
     "iopub.status.idle": "2021-12-28T11:24:02.598859Z",
     "shell.execute_reply": "2021-12-28T11:24:02.599307Z"
    },
    "jupyter": {
     "outputs_hidden": false
    },
    "papermill": {
     "duration": 0.448529,
     "end_time": "2021-12-28T11:24:02.599467",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.150938",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 177 ms, sys: 83 ms, total: 260 ms\n",
      "Wall time: 395 ms\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>SeriousDlqin2yrs</th>\n",
       "      <th>RevolvingUtilizationOfUnsecuredLines</th>\n",
       "      <th>age</th>\n",
       "      <th>NumberOfTime30-59DaysPastDueNotWorse</th>\n",
       "      <th>DebtRatio</th>\n",
       "      <th>MonthlyIncome</th>\n",
       "      <th>NumberOfOpenCreditLinesAndLoans</th>\n",
       "      <th>NumberOfTimes90DaysLate</th>\n",
       "      <th>NumberRealEstateLoansOrLines</th>\n",
       "      <th>NumberOfTime60-89DaysPastDueNotWorse</th>\n",
       "      <th>NumberOfDependents</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.766127</td>\n",
       "      <td>45</td>\n",
       "      <td>2</td>\n",
       "      <td>0.802982</td>\n",
       "      <td>9120.0</td>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.957151</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>0.121876</td>\n",
       "      <td>2600.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0.658180</td>\n",
       "      <td>38</td>\n",
       "      <td>1</td>\n",
       "      <td>0.085113</td>\n",
       "      <td>3042.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  SeriousDlqin2yrs  RevolvingUtilizationOfUnsecuredLines  age  \\\n",
       "0           1                 1                              0.766127   45   \n",
       "1           2                 0                              0.957151   40   \n",
       "2           3                 0                              0.658180   38   \n",
       "\n",
       "   NumberOfTime30-59DaysPastDueNotWorse  DebtRatio  MonthlyIncome  \\\n",
       "0                                     2   0.802982         9120.0   \n",
       "1                                     0   0.121876         2600.0   \n",
       "2                                     1   0.085113         3042.0   \n",
       "\n",
       "   NumberOfOpenCreditLinesAndLoans  NumberOfTimes90DaysLate  \\\n",
       "0                               13                        0   \n",
       "1                                4                        0   \n",
       "2                                2                        1   \n",
       "\n",
       "   NumberRealEstateLoansOrLines  NumberOfTime60-89DaysPastDueNotWorse  \\\n",
       "0                             6                                     0   \n",
       "1                             0                                     0   \n",
       "2                             0                                     0   \n",
       "\n",
       "   NumberOfDependents  \n",
       "0                 2.0  \n",
       "1                 1.0  \n",
       "2                 0.0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time \n",
    "train = pd.read_csv('../input/GiveMeSomeCredit/cs-training.csv')\n",
    "test = pd.read_csv('../input/GiveMeSomeCredit/cs-test.csv')\n",
    "train.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2cfeea9f",
   "metadata": {
    "papermill": {
     "duration": 0.030558,
     "end_time": "2021-12-28T11:24:02.661755",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.631197",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# 1-Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb183be5",
   "metadata": {
    "papermill": {
     "duration": 0.030591,
     "end_time": "2021-12-28T11:24:02.723211",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.692620",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Convert Dtypes :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4c4833ae",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:02.788357Z",
     "iopub.status.busy": "2021-12-28T11:24:02.787784Z",
     "iopub.status.idle": "2021-12-28T11:24:02.872784Z",
     "shell.execute_reply": "2021-12-28T11:24:02.873273Z"
    },
    "papermill": {
     "duration": 0.118975,
     "end_time": "2021-12-28T11:24:02.873438",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.754463",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Convert Dtypes :\n",
    "train[train.select_dtypes(['int64','int16','float16','float32','float64','int8']).columns] = train[train.select_dtypes(['int64','int16','float16','float32','float64','int8']).columns].apply(pd.to_numeric)\n",
    "train[train.select_dtypes(['object','category']).columns] = train.select_dtypes(['object','category']).apply(lambda x: x.astype('category'))\n",
    "# Convert Dtypes :\n",
    "test[test.select_dtypes(['int64','int16','float16','float32','float64','int8']).columns] = test[test.select_dtypes(['int64','int16','float16','float32','float64','int8']).columns].apply(pd.to_numeric)\n",
    "test[test.select_dtypes(['object','category']).columns] = test.select_dtypes(['object','category']).apply(lambda x: x.astype('category'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45dbcb84",
   "metadata": {
    "papermill": {
     "duration": 0.031449,
     "end_time": "2021-12-28T11:24:02.936900",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.905451",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Reduce Memory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b6b8e113",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:03.003198Z",
     "iopub.status.busy": "2021-12-28T11:24:03.002574Z",
     "iopub.status.idle": "2021-12-28T11:24:03.067403Z",
     "shell.execute_reply": "2021-12-28T11:24:03.066458Z"
    },
    "papermill": {
     "duration": 0.099437,
     "end_time": "2021-12-28T11:24:03.067633",
     "exception": false,
     "start_time": "2021-12-28T11:24:02.968196",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Memory usage of dataframe is 13.73 MB\n",
      "Memory usage after optimization is: 3.29 MB\n",
      "Decreased by 76.0%\n",
      "Memory usage of dataframe is 9.29 MB\n",
      "Memory usage after optimization is: 2.90 MB\n",
      "Decreased by 68.7%\n"
     ]
    }
   ],
   "source": [
    "# Author : https://www.kaggle.com/gemartin/load-data-reduce-memory-usage\n",
    "def reduce_mem_usage(df):\n",
    "    \"\"\" iterate through all the columns of a dataframe and modify the data type\n",
    "        to reduce memory usage.        \n",
    "    \"\"\"\n",
    "    start_mem = df.memory_usage().sum() / 1024**2\n",
    "    print('Memory usage of dataframe is {:.2f} MB'.format(start_mem))\n",
    "    \n",
    "    for col in df.columns:\n",
    "        col_type = df[col].dtype\n",
    "        name =df[col].dtype.name \n",
    "        \n",
    "        if col_type != object and col_type.name != 'category':\n",
    "        #if name != \"category\":    \n",
    "            c_min = df[col].min()\n",
    "            c_max = df[col].max()\n",
    "            if str(col_type)[:3] == 'int':\n",
    "                if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:\n",
    "                    df[col] = df[col].astype(np.int8)\n",
    "                elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:\n",
    "                    df[col] = df[col].astype(np.int16)\n",
    "                elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:\n",
    "                    df[col] = df[col].astype(np.int32)\n",
    "                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:\n",
    "                    df[col] = df[col].astype(np.int64)  \n",
    "            else:\n",
    "                if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:\n",
    "                    df[col] = df[col].astype(np.float16)\n",
    "                elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:\n",
    "                    df[col] = df[col].astype(np.float32)\n",
    "                else:\n",
    "                    df[col] = df[col].astype(np.float64)\n",
    "        else:\n",
    "            df[col] = df[col].astype('category')\n",
    "\n",
    "    end_mem = df.memory_usage().sum() / 1024**2\n",
    "    print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))\n",
    "    print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))\n",
    "    \n",
    "    return df\n",
    "train= reduce_mem_usage(train)\n",
    "test= reduce_mem_usage(test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de6eadfc",
   "metadata": {
    "papermill": {
     "duration": 0.031153,
     "end_time": "2021-12-28T11:24:03.132002",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.100849",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## X and Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9be77d16",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:03.203215Z",
     "iopub.status.busy": "2021-12-28T11:24:03.202267Z",
     "iopub.status.idle": "2021-12-28T11:24:03.204736Z",
     "shell.execute_reply": "2021-12-28T11:24:03.205151Z"
    },
    "papermill": {
     "duration": 0.041203,
     "end_time": "2021-12-28T11:24:03.205314",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.164111",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Cardinality : \n",
    "# - RevolvingUtilizationOfUnsecuredLines :125728, high Outlier\n",
    "# - DebtRatio :114194 , high Outlier \n",
    "# deal with outlier + bin \n",
    "PERCENTAGE = [\"RevolvingUtilizationOfUnsecuredLines\", \"DebtRatio\"]\n",
    "# MonthlyIncome:13594 , high outlier +bin \n",
    "REAL= [\"MonthlyIncome\"]\n",
    "# Can be considred as cat \n",
    "NUMERIC_DISCRET_low = [\"NumberOfDependents\",\n",
    "                       \"NumberOfTime60-89DaysPastDueNotWorse\",\n",
    "                       \"NumberRealEstateLoansOrLines\",\n",
    "                       \"NumberOfTimes90DaysLate\",\n",
    "                       \"NumberOfOpenCreditLinesAndLoans\",\n",
    "                       \"NumberOfTime30-59DaysPastDueNotWorse\",\n",
    "                       \"age\"]\n",
    "Late_Pay_Cols = ['NumberOfTime30-59DaysPastDueNotWorse',\n",
    "                 'NumberOfTimes90DaysLate',\n",
    "                 'NumberOfTime60-89DaysPastDueNotWorse']\n",
    "TARGET = [\"SeriousDlqin2yrs\"]\n",
    "\n",
    "#also change the type for TARGET to categorical\n",
    "#df[TARGET] = df[TARGET].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "57219a3e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:03.273428Z",
     "iopub.status.busy": "2021-12-28T11:24:03.272765Z",
     "iopub.status.idle": "2021-12-28T11:24:03.275510Z",
     "shell.execute_reply": "2021-12-28T11:24:03.275961Z"
    },
    "papermill": {
     "duration": 0.03882,
     "end_time": "2021-12-28T11:24:03.276132",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.237312",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "y = train['SeriousDlqin2yrs']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec559971",
   "metadata": {
    "papermill": {
     "duration": 0.032003,
     "end_time": "2021-12-28T11:24:03.340029",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.308026",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Imput before Descritaziation : \n",
    "\n",
    "### IterativeImputer\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c7a8a1fa",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:03.412179Z",
     "iopub.status.busy": "2021-12-28T11:24:03.411449Z",
     "iopub.status.idle": "2021-12-28T11:24:03.415013Z",
     "shell.execute_reply": "2021-12-28T11:24:03.414424Z"
    },
    "papermill": {
     "duration": 0.043559,
     "end_time": "2021-12-28T11:24:03.415144",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.371585",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 331 µs, sys: 0 ns, total: 331 µs\n",
      "Wall time: 337 µs\n"
     ]
    }
   ],
   "source": [
    "%%time \n",
    "param =  {   \"verbosity\": 0,\n",
    "            #\"objective\": \"binary:logistic\",\n",
    "            #\"eval_metric\": \"auc\",\n",
    "            'random_state': 42,\n",
    "            # regression\n",
    "            'objective':'reg:squarederror', \n",
    "             'eval_metric': 'mae',\n",
    "            #early_stopping_rounds=100 ,\n",
    "            #'gpu_id':0, \n",
    "            #'predictor':\"gpu_predictor\",\n",
    "            # use exact for small dataset.\n",
    "            #\"tree_method\": \"exact\",\n",
    "            # big data :\n",
    "            # 'tree_method': 'gpu_hist',\n",
    "            # defines booster, gblinear for linear functions.\n",
    "             'booster': 'gbtree', \n",
    "            'lambda': 8.544792472633987e-07,\n",
    "            'alpha': 0.31141671752487043,\n",
    "            'subsample': 0.8779467596981366, \n",
    "            'colsample_bytree': 0.9759532762677546,\n",
    "            'learning_rate': 0.008686087328805853, \n",
    "            'n_estimators': 6988,\n",
    "            'max_depth': 9,\n",
    "            'min_child_weight': 2, \n",
    "            'eta': 3.7603213457541647e-06,\n",
    "            'gamma': 2.1478058456847449e-07,\n",
    "            'grow_policy': 'lossguide'}\n",
    "                \n",
    "\n",
    "#model_xgb = XGBRegressor(\n",
    "       #objective=\"mae\",\n",
    " #   **xgb_params2)\n",
    "\n",
    "numeric_transformer1 = Pipeline(\n",
    "                            steps=[\n",
    "                            ('imputer', SimpleImputer(strategy='median'\n",
    "                                                      ,add_indicator=True)),\n",
    "                            ('scaler', PowerTransformer()),#(Numerical Input, Numerical Output)\n",
    "                            # Create an SelectKBest object to select features with two best ANOVA F-Values\n",
    "                            #The F-value scores examine if, when we group the numerical feature by the target vector, the means for each group are significantly different\n",
    "                           # ('reducedim',  SelectPercentile(f_classif,percentile=90))\n",
    "                            ]\n",
    "                            )\n",
    "numeric_transformer2 = Pipeline(\n",
    "                            steps=[\n",
    "                            #('imputer', SimpleImputer(strategy='median'\n",
    "                             #                         ,add_indicator=True)),\n",
    "                            ('scaler', PowerTransformer()),#(Numerical Input, Numerical Output)\n",
    "                            # Create an SelectKBest object to select features with two best ANOVA F-Values\n",
    "                            #The F-value scores examine if, when we group the numerical feature by the target vector, the means for each group are significantly different\n",
    "                           # ('reducedim',  SelectPercentile(f_classif,percentile=90))\n",
    "                            ]\n",
    "                            )\n",
    "\n",
    "pipe_xgbr1 = Pipeline(\n",
    "                    steps=[\n",
    "                        ('preprocessor', numeric_transformer1),\n",
    "                        ('classifier', XGBRegressor(\n",
    "                      #objective=\"mae\",\n",
    "                       **param))\n",
    "                    ]\n",
    "                )\n",
    "pipe_xgbr2 = Pipeline(\n",
    "                    steps=[\n",
    "                        ('preprocessor', numeric_transformer2),\n",
    "                        ('classifier', XGBRegressor(\n",
    "                      #objective=\"mae\",\n",
    "                       **param))\n",
    "                    ]\n",
    "                )\n",
    "\n",
    "#imputer=IterativeImputer(estimator=XGBRegressor(\n",
    "#                      #objective=\"mae\",\n",
    " #                     **param),max_iter=2,\n",
    "  #                       add_indicator=True,\n",
    "   #                      random_state=42)\n",
    "#train=train.drop(['Unnamed: 0','SeriousDlqin2yrs'], axis=1)\n",
    "#test=test.drop(['Unnamed: 0','SeriousDlqin2yrs'], axis=1)\n",
    "#imputer.fit(train)\n",
    "#train = imputer.transform(train)\n",
    "#test = imputer.transform(test)\n",
    "#features=['RevolvingUtilizationOfUnsecuredLines', 'age',\n",
    "#      'NumberOfTime30-59DaysPastDueNotWorse', 'DebtRatio', 'MonthlyIncome',\n",
    " #      'NumberOfOpenCreditLinesAndLoans', 'NumberOfTimes90DaysLate',\n",
    " #     'NumberRealEstateLoansOrLines', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
    " #     'NumberOfDependents']\n",
    "#train=pd.DataFrame(train,columns=features)\n",
    "#test=pd.DataFrame(test,columns=features)\n",
    "#train= reduce_mem_usage(train)\n",
    "#test= reduce_mem_usage(test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc647005",
   "metadata": {
    "papermill": {
     "duration": 0.031707,
     "end_time": "2021-12-28T11:24:03.478972",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.447265",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Rgression Imputer "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7784565e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:24:03.550745Z",
     "iopub.status.busy": "2021-12-28T11:24:03.549695Z",
     "iopub.status.idle": "2021-12-28T11:43:13.428404Z",
     "shell.execute_reply": "2021-12-28T11:43:13.428960Z"
    },
    "papermill": {
     "duration": 1149.918278,
     "end_time": "2021-12-28T11:43:13.429147",
     "exception": false,
     "start_time": "2021-12-28T11:24:03.510869",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 14min 21s, sys: 6.06 s, total: 1h 14min 27s\n",
      "Wall time: 19min 9s\n"
     ]
    }
   ],
   "source": [
    "%%time \n",
    "train=train.drop(['Unnamed: 0','SeriousDlqin2yrs'], axis=1)\n",
    "test=test.drop(['Unnamed: 0','SeriousDlqin2yrs'], axis=1)\n",
    "train_final= pd.concat( [train, test], axis=0) \n",
    "#testdf_income= train_final[train_final['MonthlyIncome'].isnull()==True]\n",
    "traindf_income = train_final[train_final['MonthlyIncome'].isnull()==False]\n",
    "y_income = traindf_income['MonthlyIncome']\n",
    "X_income=traindf_income.drop([\"MonthlyIncome\"],axis=1)\n",
    "pipe_xgbr1.fit(X_income, y_income)\n",
    "train_income_missing=train[train['MonthlyIncome'].isnull()==True].drop([\"MonthlyIncome\"],axis=1)\n",
    "test_income_missing=test[test['MonthlyIncome'].isnull()==True].drop([\"MonthlyIncome\"],axis=1)\n",
    "train_predicted = pipe_xgbr1.predict(train_income_missing)\n",
    "test_predicted = pipe_xgbr1.predict(test_income_missing)\n",
    "train.loc[(train.MonthlyIncome.isnull()), 'MonthlyIncome'] = train_predicted\n",
    "test.loc[(test.MonthlyIncome.isnull()), 'MonthlyIncome'] = test_predicted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f71c2f77",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T11:43:13.501351Z",
     "iopub.status.busy": "2021-12-28T11:43:13.500645Z",
     "iopub.status.idle": "2021-12-28T12:06:20.899228Z",
     "shell.execute_reply": "2021-12-28T12:06:20.899716Z"
    },
    "papermill": {
     "duration": 1387.438306,
     "end_time": "2021-12-28T12:06:20.899891",
     "exception": false,
     "start_time": "2021-12-28T11:43:13.461585",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 29min 35s, sys: 5.54 s, total: 1h 29min 41s\n",
      "Wall time: 23min 7s\n"
     ]
    }
   ],
   "source": [
    "%%time \n",
    "traindf_NumberOfDependents = train_final[train_final['NumberOfDependents'].isnull()==False]\n",
    "y_NumberOfDependents = traindf_NumberOfDependents['NumberOfDependents']\n",
    "X_NumberOfDependents=traindf_NumberOfDependents.drop([\"NumberOfDependents\"],axis=1)\n",
    "pipe_xgbr2.fit(X_NumberOfDependents, y_NumberOfDependents)\n",
    "train_NumberOfDependents_missing=train[train['NumberOfDependents'].isnull()==True].drop([\"NumberOfDependents\"],axis=1)\n",
    "test_NumberOfDependents_missing=test[test['NumberOfDependents'].isnull()==True].drop([\"NumberOfDependents\"],axis=1)\n",
    "train_predicted = pipe_xgbr2.predict(train_NumberOfDependents_missing)\n",
    "test_predicted = pipe_xgbr2.predict(test_NumberOfDependents_missing)\n",
    "train.loc[(train.NumberOfDependents.isnull()), 'NumberOfDependents'] = train_predicted\n",
    "test.loc[(test.NumberOfDependents.isnull()), 'NumberOfDependents'] = test_predicted"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd2eb238",
   "metadata": {
    "papermill": {
     "duration": 0.033521,
     "end_time": "2021-12-28T12:06:20.967021",
     "exception": false,
     "start_time": "2021-12-28T12:06:20.933500",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Features Engineer\n",
    "### Add Bin \n",
    "It's always good to bin data when we have lot of outlier and skweness "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5a0f9925",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:21.056488Z",
     "iopub.status.busy": "2021-12-28T12:06:21.055872Z",
     "iopub.status.idle": "2021-12-28T12:06:21.741701Z",
     "shell.execute_reply": "2021-12-28T12:06:21.742145Z"
    },
    "papermill": {
     "duration": 0.736688,
     "end_time": "2021-12-28T12:06:21.742366",
     "exception": false,
     "start_time": "2021-12-28T12:06:21.005678",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Add bin data \n",
    "# initializing append_str\n",
    "append_str = 'cat_'\n",
    "# Append suffix / prefix to strings in list\n",
    "num_features1=[\"RevolvingUtilizationOfUnsecuredLines\", \"DebtRatio\",\"MonthlyIncome\"]\n",
    "num_features2=[\"NumberOfDependents\",\n",
    "                       \"NumberOfTime60-89DaysPastDueNotWorse\",\n",
    "                       \"NumberRealEstateLoansOrLines\",\n",
    "                       \"NumberOfTimes90DaysLate\",\n",
    "                       \"NumberOfOpenCreditLinesAndLoans\",\n",
    "                       \"NumberOfTime30-59DaysPastDueNotWorse\",\n",
    "                       \"age\"]\n",
    "cat_features1 = [append_str + sub for sub in num_features1]\n",
    "cat_features2 = [append_str + sub for sub in num_features2]\n",
    "\n",
    "# create the discretizer object with strategy quantile and 1000 bins\n",
    "discretizer1 = KBinsDiscretizer(n_bins=40, encode='ordinal',strategy='quantile')\n",
    "discretizer2 = KBinsDiscretizer(n_bins=4, encode='ordinal',strategy='quantile')\n",
    "\n",
    "pipeline1 = Pipeline([\n",
    "        ('imputer', SimpleImputer( strategy='median')),\n",
    "        ('bin', discretizer1)\n",
    "    ])\n",
    "# fit the discretizer to the train set\n",
    "pipeline1.fit(train.loc[:,num_features1])\n",
    "# apply the discretisation\n",
    "train_cat1 = pipeline1.transform(train.loc[:,num_features1])\n",
    "test_cat1 = pipeline1.transform(test.loc[:,num_features1])\n",
    "train_df1=pd.DataFrame(train_cat1,columns=cat_features1).astype('category')\n",
    "test_df1=pd.DataFrame(test_cat1,columns=cat_features1).astype('category')\n",
    "train_final1= pd.concat( [train.loc[:,num_features1], train_df1], axis=1) \n",
    "test_final1= pd.concat( [test.loc[:,num_features1], test_df1], axis=1) \n",
    "\n",
    "pipeline2 = Pipeline([\n",
    "        ('imputer', SimpleImputer( strategy='median')),\n",
    "        ('bin', discretizer2)\n",
    "    ])\n",
    "# fit the discretizer to the train set\n",
    "pipeline2.fit(train.loc[:,num_features2])\n",
    "# apply the discretisation\n",
    "train_cat2 = pipeline2.transform(train.loc[:,num_features2])\n",
    "test_cat2 = pipeline2.transform(test.loc[:,num_features2])\n",
    "train_df2=pd.DataFrame(train_cat2,columns=cat_features2).astype('category')\n",
    "test_df2=pd.DataFrame(test_cat2,columns=cat_features2).astype('category')\n",
    "train_final2= pd.concat( [train.loc[:,num_features2], train_df2], axis=1) \n",
    "test_final2= pd.concat( [test.loc[:,num_features2], test_df2], axis=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f74bdcf3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:21.815464Z",
     "iopub.status.busy": "2021-12-28T12:06:21.814870Z",
     "iopub.status.idle": "2021-12-28T12:06:21.821336Z",
     "shell.execute_reply": "2021-12-28T12:06:21.821911Z"
    },
    "papermill": {
     "duration": 0.045219,
     "end_time": "2021-12-28T12:06:21.822069",
     "exception": false,
     "start_time": "2021-12-28T12:06:21.776850",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_final= pd.concat( [train_final1, train_final2], axis=1) \n",
    "test_final= pd.concat( [test_final1, test_final2], axis=1) "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "417a8d93",
   "metadata": {
    "papermill": {
     "duration": 0.032659,
     "end_time": "2021-12-28T12:06:21.887683",
     "exception": false,
     "start_time": "2021-12-28T12:06:21.855024",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Final X and y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "3fbc0fa8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:21.956988Z",
     "iopub.status.busy": "2021-12-28T12:06:21.956084Z",
     "iopub.status.idle": "2021-12-28T12:06:21.959627Z",
     "shell.execute_reply": "2021-12-28T12:06:21.960150Z"
    },
    "papermill": {
     "duration": 0.039806,
     "end_time": "2021-12-28T12:06:21.960311",
     "exception": false,
     "start_time": "2021-12-28T12:06:21.920505",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Pour le train test\n",
    "target= \"SeriousDlqin2yrs\"\n",
    "X = train_final# axis=1\n",
    "X_test_final =test_final# axis=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "453c1bea",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.029884Z",
     "iopub.status.busy": "2021-12-28T12:06:22.028992Z",
     "iopub.status.idle": "2021-12-28T12:06:22.052467Z",
     "shell.execute_reply": "2021-12-28T12:06:22.053229Z"
    },
    "papermill": {
     "duration": 0.060125,
     "end_time": "2021-12-28T12:06:22.053447",
     "exception": false,
     "start_time": "2021-12-28T12:06:21.993322",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 101503 entries, 0 to 101502\n",
      "Data columns (total 20 columns):\n",
      " #   Column                                    Non-Null Count   Dtype   \n",
      "---  ------                                    --------------   -----   \n",
      " 0   RevolvingUtilizationOfUnsecuredLines      101503 non-null  float16 \n",
      " 1   DebtRatio                                 101503 non-null  float32 \n",
      " 2   MonthlyIncome                             101503 non-null  float32 \n",
      " 3   cat_RevolvingUtilizationOfUnsecuredLines  101503 non-null  category\n",
      " 4   cat_DebtRatio                             101503 non-null  category\n",
      " 5   cat_MonthlyIncome                         101503 non-null  category\n",
      " 6   NumberOfDependents                        101503 non-null  float16 \n",
      " 7   NumberOfTime60-89DaysPastDueNotWorse      101503 non-null  int8    \n",
      " 8   NumberRealEstateLoansOrLines              101503 non-null  int8    \n",
      " 9   NumberOfTimes90DaysLate                   101503 non-null  int8    \n",
      " 10  NumberOfOpenCreditLinesAndLoans           101503 non-null  int8    \n",
      " 11  NumberOfTime30-59DaysPastDueNotWorse      101503 non-null  int8    \n",
      " 12  age                                       101503 non-null  int8    \n",
      " 13  cat_NumberOfDependents                    101503 non-null  category\n",
      " 14  cat_NumberOfTime60-89DaysPastDueNotWorse  101503 non-null  category\n",
      " 15  cat_NumberRealEstateLoansOrLines          101503 non-null  category\n",
      " 16  cat_NumberOfTimes90DaysLate               101503 non-null  category\n",
      " 17  cat_NumberOfOpenCreditLinesAndLoans       101503 non-null  category\n",
      " 18  cat_NumberOfTime30-59DaysPastDueNotWorse  101503 non-null  category\n",
      " 19  cat_age                                   101503 non-null  category\n",
      "dtypes: category(10), float16(2), float32(2), int8(6)\n",
      "memory usage: 2.7 MB\n"
     ]
    }
   ],
   "source": [
    "X_test_final.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "53394384",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.126371Z",
     "iopub.status.busy": "2021-12-28T12:06:22.125481Z",
     "iopub.status.idle": "2021-12-28T12:06:22.148175Z",
     "shell.execute_reply": "2021-12-28T12:06:22.148623Z"
    },
    "papermill": {
     "duration": 0.059993,
     "end_time": "2021-12-28T12:06:22.148794",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.088801",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 150000 entries, 0 to 149999\n",
      "Data columns (total 20 columns):\n",
      " #   Column                                    Non-Null Count   Dtype   \n",
      "---  ------                                    --------------   -----   \n",
      " 0   RevolvingUtilizationOfUnsecuredLines      150000 non-null  float16 \n",
      " 1   DebtRatio                                 150000 non-null  float32 \n",
      " 2   MonthlyIncome                             150000 non-null  float32 \n",
      " 3   cat_RevolvingUtilizationOfUnsecuredLines  150000 non-null  category\n",
      " 4   cat_DebtRatio                             150000 non-null  category\n",
      " 5   cat_MonthlyIncome                         150000 non-null  category\n",
      " 6   NumberOfDependents                        150000 non-null  float16 \n",
      " 7   NumberOfTime60-89DaysPastDueNotWorse      150000 non-null  int8    \n",
      " 8   NumberRealEstateLoansOrLines              150000 non-null  int8    \n",
      " 9   NumberOfTimes90DaysLate                   150000 non-null  int8    \n",
      " 10  NumberOfOpenCreditLinesAndLoans           150000 non-null  int8    \n",
      " 11  NumberOfTime30-59DaysPastDueNotWorse      150000 non-null  int8    \n",
      " 12  age                                       150000 non-null  int8    \n",
      " 13  cat_NumberOfDependents                    150000 non-null  category\n",
      " 14  cat_NumberOfTime60-89DaysPastDueNotWorse  150000 non-null  category\n",
      " 15  cat_NumberRealEstateLoansOrLines          150000 non-null  category\n",
      " 16  cat_NumberOfTimes90DaysLate               150000 non-null  category\n",
      " 17  cat_NumberOfOpenCreditLinesAndLoans       150000 non-null  category\n",
      " 18  cat_NumberOfTime30-59DaysPastDueNotWorse  150000 non-null  category\n",
      " 19  cat_age                                   150000 non-null  category\n",
      "dtypes: category(10), float16(2), float32(2), int8(6)\n",
      "memory usage: 4.0 MB\n"
     ]
    }
   ],
   "source": [
    "X.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "69496dfc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.220068Z",
     "iopub.status.busy": "2021-12-28T12:06:22.219118Z",
     "iopub.status.idle": "2021-12-28T12:06:22.222558Z",
     "shell.execute_reply": "2021-12-28T12:06:22.223088Z"
    },
    "papermill": {
     "duration": 0.040908,
     "end_time": "2021-12-28T12:06:22.223243",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.182335",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "del train\n",
    "del test \n",
    "del train_final\n",
    "del test_final"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "304b463c",
   "metadata": {
    "papermill": {
     "duration": 0.033496,
     "end_time": "2021-12-28T12:06:22.290463",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.256967",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Extract final cat  and num Features \n",
    "**What should we do for each colmun**\n",
    "\n",
    "Separate features by dtype:\n",
    "\n",
    "Next we’ll separate the features in the dataframe by their datatype. There are a few different ways to achieve this. I’ve used the select_dtypes() function to obtain specific data types by passing in np.number to obtain the numeric data and exclude=['np.number'] to return the categorical data. Appending .columns to the end returns an Index list containing the column names. For the categorical features, we don’t want to include the target  column, so I’ve dropped that.\n",
    "### Cat Features "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "eafadc11",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.366824Z",
     "iopub.status.busy": "2021-12-28T12:06:22.365919Z",
     "iopub.status.idle": "2021-12-28T12:06:22.370969Z",
     "shell.execute_reply": "2021-12-28T12:06:22.371480Z"
    },
    "papermill": {
     "duration": 0.047619,
     "end_time": "2021-12-28T12:06:22.371664",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.324045",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['cat_RevolvingUtilizationOfUnsecuredLines', 'cat_DebtRatio',\n",
       "       'cat_MonthlyIncome', 'cat_NumberOfDependents',\n",
       "       'cat_NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'cat_NumberRealEstateLoansOrLines', 'cat_NumberOfTimes90DaysLate',\n",
       "       'cat_NumberOfOpenCreditLinesAndLoans',\n",
       "       'cat_NumberOfTime30-59DaysPastDueNotWorse', 'cat_age'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# select non-numeric columns\n",
    "cat_columns = X.select_dtypes(exclude=['int64','int16','float16','float32','float64','int8']).columns\n",
    "cat_columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45015b1e",
   "metadata": {
    "papermill": {
     "duration": 0.033949,
     "end_time": "2021-12-28T12:06:22.439759",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.405810",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Num Features\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6dad3149",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.511320Z",
     "iopub.status.busy": "2021-12-28T12:06:22.510438Z",
     "iopub.status.idle": "2021-12-28T12:06:22.522732Z",
     "shell.execute_reply": "2021-12-28T12:06:22.522073Z"
    },
    "papermill": {
     "duration": 0.049174,
     "end_time": "2021-12-28T12:06:22.522865",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.473691",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# select the float columns\n",
    "num_columns = X.select_dtypes(include=['int64','int16','float16','float32','float64','int8']).columns\n",
    "num_columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c54256ee",
   "metadata": {
    "papermill": {
     "duration": 0.034626,
     "end_time": "2021-12-28T12:06:22.592479",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.557853",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Extracted features "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "7a63c665",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.664723Z",
     "iopub.status.busy": "2021-12-28T12:06:22.664146Z",
     "iopub.status.idle": "2021-12-28T12:06:22.673658Z",
     "shell.execute_reply": "2021-12-28T12:06:22.674132Z"
    },
    "papermill": {
     "duration": 0.047249,
     "end_time": "2021-12-28T12:06:22.674303",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.627054",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class FeaturesEngineer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    def transform(self, X, y=None):\n",
    "        X=X.copy()\n",
    "        # Calculate some metrics across rows\n",
    "        X[\"num_missing\"]  = X.isnull().sum(axis=1)\n",
    "        X[\"std_row\"] = X.std(axis=1)\n",
    "        X[\"sem_row\"] = X.sem(axis=1)\n",
    "        X[\"abs_sum_row\"] = X.abs().sum(axis=1)\n",
    "        X[\"mean_row\"] = X.mean(axis=1)\n",
    "        X[\"max_row\"]= X.max(axis=1)\n",
    "        X[\"min_row\"]= X.min(axis=1)\n",
    "        X['Weighted_Sum_PastDue'] = 2 * X['NumberOfTime30-59DaysPastDueNotWorse'] + 3 * X['NumberOfTime60-89DaysPastDueNotWorse'] + 6 * X['NumberOfTimes90DaysLate']\n",
    "        X['90days_out_of_TotalPastDue'] = X['NumberOfTimes90DaysLate'] / (X['Weighted_Sum_PastDue'])\n",
    "        X.loc[X['Weighted_Sum_PastDue']==0, '90days_out_of_TotalPastDue'] = 0\n",
    "        X['RemainingLines'] = X['NumberOfOpenCreditLinesAndLoans'] - X['NumberRealEstateLoansOrLines']\n",
    "        X['Loans_vs_Other_Lines'] = X['RemainingLines'] / (1 + X['NumberRealEstateLoansOrLines'])\n",
    "        X['Debt_per_Real_Estate_Loan'] = X['DebtRatio'] / X['NumberRealEstateLoansOrLines']\n",
    "        X.loc[X['NumberRealEstateLoansOrLines']==0, 'Debt_per_Real_Estate_Loan'] = 0\n",
    "        X['Disposable_Income_per_person'] = (X['MonthlyIncome'] - X['DebtRatio']) / (X['NumberOfDependents'] + 1)\n",
    "        X.loc[X['NumberOfDependents']==0, 'Disposable_Income_per_person'] = 0\n",
    "        X['RemainingLines_per_person'] = X['RemainingLines'] / (X['NumberOfDependents'] + 1)\n",
    "        X['NumberRE_X_DebtRatio_X_age'] = X['NumberRealEstateLoansOrLines'] * X['DebtRatio'] / X['age']\n",
    "        X['RevolvingUtilization_per_age'] = X['RevolvingUtilizationOfUnsecuredLines'] / X['age']\n",
    "        X[X==np.inf]=np.nan\n",
    "        X=reduce_mem_usage(X)\n",
    "        return X  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dab0d458",
   "metadata": {
    "papermill": {
     "duration": 0.034703,
     "end_time": "2021-12-28T12:06:22.743821",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.709118",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Kmeans + encoder "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "976990e2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.816809Z",
     "iopub.status.busy": "2021-12-28T12:06:22.816017Z",
     "iopub.status.idle": "2021-12-28T12:06:22.823034Z",
     "shell.execute_reply": "2021-12-28T12:06:22.823514Z"
    },
    "papermill": {
     "duration": 0.045062,
     "end_time": "2021-12-28T12:06:22.823710",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.778648",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class MiniKmeansTransformerEncoder(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, num_clusters = 11, encoder=ce.woe.WOEEncoder()):\n",
    "        self.num_clusters = num_clusters\n",
    "        self.encoder= encoder\n",
    "        if self.num_clusters > 0:\n",
    "            self.kmeans = MiniBatchKMeans(n_clusters=self.num_clusters, random_state=0)\n",
    "    \n",
    "    def fit(self, X, y=None):\n",
    "        if self.num_clusters > 0:\n",
    "            self.kmeans.fit(X)\n",
    "            preds=self.kmeans.predict(X)\n",
    "            preds=pd.DataFrame(preds, columns=['kmeans']).astype('category')\n",
    "            self.encoder.fit(preds,y)\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X, y=None):\n",
    "        pred_classes = self.kmeans.predict(X)\n",
    "        pred_classes=pd.DataFrame(pred_classes, columns=['kmeans']).astype('category')\n",
    "        pred_encoded = self.encoder.transform(pred_classes)\n",
    "        return np.hstack((X, pred_encoded))\n",
    "        #return pred_encoded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5e68e407",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:22.899323Z",
     "iopub.status.busy": "2021-12-28T12:06:22.898731Z",
     "iopub.status.idle": "2021-12-28T12:06:22.907744Z",
     "shell.execute_reply": "2021-12-28T12:06:22.908420Z"
    },
    "papermill": {
     "duration": 0.047914,
     "end_time": "2021-12-28T12:06:22.908639",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.860725",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class SparseInteractions(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, degree=2, feature_name_separator=\"_\"):\n",
    "        self.degree = degree\n",
    "        self.feature_name_separator = feature_name_separator\n",
    "    \n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "        \n",
    "    def transform(self, X):\n",
    "        if not sparse.isspmatrix_csc(X):\n",
    "            X = sparse.csc_matrix(X)\n",
    "            \n",
    "        if hasattr(X, \"columns\"):\n",
    "            self.orig_col_names = X.columns\n",
    "        else:\n",
    "            self.orig_col_names = np.array([str(i) for i in range(X.shape[1])])\n",
    "            \n",
    "        spi = self._create_sparse_interactions(X)\n",
    "        return spi\n",
    "    \n",
    "    \n",
    "    def get_feature_names(self):\n",
    "        return self.feature_names\n",
    "    \n",
    "    def _create_sparse_interactions(self, X):\n",
    "        out_mat = []\n",
    "        self.feature_names = self.orig_col_names.tolist()\n",
    "        \n",
    "        for sub_degree in range(2, self.degree + 1):\n",
    "            for col_ixs in combinations(range(X.shape[1]), sub_degree):\n",
    "                # add name for new column\n",
    "                name = self.feature_name_separator.join(self.orig_col_names[list(col_ixs)])\n",
    "                self.feature_names.append(name)\n",
    "                \n",
    "                # get column multiplications value\n",
    "                out = X[:, col_ixs[0]]    \n",
    "                for j in col_ixs[1:]:\n",
    "                    out = out.multiply(X[:, j])\n",
    "\n",
    "                out_mat.append(out)\n",
    "\n",
    "        return sparse.hstack([X] + out_mat)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e97321a1",
   "metadata": {
    "papermill": {
     "duration": 0.035592,
     "end_time": "2021-12-28T12:06:22.981026",
     "exception": false,
     "start_time": "2021-12-28T12:06:22.945434",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Complete Pipe \n",
    "\n",
    "1. Features Engineer\n",
    "\n",
    "1. Kmeans\n",
    "\n",
    "1. Sparse\n",
    "\n",
    "1. Poly\n",
    "\n",
    "1. Scaler/Transformer\n",
    "\n",
    "1. imput"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "fe857b65",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:23.055709Z",
     "iopub.status.busy": "2021-12-28T12:06:23.055032Z",
     "iopub.status.idle": "2021-12-28T12:06:23.062294Z",
     "shell.execute_reply": "2021-12-28T12:06:23.061777Z"
    },
    "papermill": {
     "duration": 0.045083,
     "end_time": "2021-12-28T12:06:23.062425",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.017342",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Cat pipeline\n",
    "categorical_transformer = Pipeline(\n",
    "                    steps=[\n",
    "                        ('imputer', SimpleImputer(strategy='most_frequent',\n",
    "                                                  fill_value='missing',\n",
    "                                                  add_indicator=True)),\n",
    "                        ('encoder',  ce.woe.WOEEncoder()),#(Numerical Input, Categorical Output)\n",
    "                        #('sparse_features', SparseInteractions(degree=2)),\n",
    "                        #('reducedim',  SelectPercentile( mutual_info_classif, percentile=90))\n",
    "\n",
    "                    ]\n",
    "                    ) \n",
    "#Define vnum pipeline\n",
    "numeric_transformer = Pipeline(\n",
    "                            steps=[\n",
    "                            ('imputer', SimpleImputer(strategy='median'\n",
    "                                                      ,add_indicator=True)),\n",
    "                            #('general_features',FeaturesEngineer()),\n",
    "                            ('scaler', PowerTransformer()),#(Numerical Input, Numerical Output)\n",
    "                            # Create an SelectKBest object to select features with two best ANOVA F-Values\n",
    "                            #The F-value scores examine if, when we group the numerical feature by the target vector, the means for each group are significantly different\n",
    "                            ('kmeans',MiniKmeansTransformerEncoder()),\n",
    "                            ('polynominal_features', PolynomialFeatures(degree=2)),\n",
    "                            #('reducedim',  SelectPercentile(f_classif,percentile=90))\n",
    "                            ]\n",
    "                            )\n",
    "# Features union cat + num \n",
    "# WOE+PowerTransformer\n",
    "preprocessor_woe_powertransformer = ColumnTransformer(\n",
    "            transformers=[\n",
    "                ('numerical', numeric_transformer, num_columns),\n",
    "               # ('categorical', categorical_transformer, cat_columns)\n",
    "            ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "58fb1fe6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:23.138004Z",
     "iopub.status.busy": "2021-12-28T12:06:23.137396Z",
     "iopub.status.idle": "2021-12-28T12:06:23.202355Z",
     "shell.execute_reply": "2021-12-28T12:06:23.201769Z"
    },
    "papermill": {
     "duration": 0.103177,
     "end_time": "2021-12-28T12:06:23.202482",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.099305",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"0af52731-ab74-4ec7-9a88-ae0632546e73\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"0af52731-ab74-4ec7-9a88-ae0632546e73\">ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e7df5902-dbaf-4419-998f-71b9de373d98\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"e7df5902-dbaf-4419-998f-71b9de373d98\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1ccfbc3c-f4ad-43c7-a5be-808ea7e61568\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1ccfbc3c-f4ad-43c7-a5be-808ea7e61568\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"06cfa903-5526-4dbb-a081-bb794c888711\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"06cfa903-5526-4dbb-a081-bb794c888711\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"983fbb83-15f3-4084-b50c-2cfc5d025c42\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"983fbb83-15f3-4084-b50c-2cfc5d025c42\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ae0d3bb0-7664-4a21-bd74-a1815ffd537c\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"ae0d3bb0-7664-4a21-bd74-a1815ffd537c\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"0e9d3252-a658-4c59-b85c-2c4fc71acbe8\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"0e9d3252-a658-4c59-b85c-2c4fc71acbe8\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocessor_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d348ceb",
   "metadata": {
    "papermill": {
     "duration": 0.035601,
     "end_time": "2021-12-28T12:06:23.274037",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.238436",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Modeling\n",
    "Modeling is the part of the Cross-Industry Standard Process for Data Mining (CRISP-DM) process model that i like best. Our data is already in good shape, and now we can search for useful patterns in our data.\n",
    "\n",
    "Tasks\n",
    "\n",
    "    Select modeling technique Select technique\n",
    "\n",
    "    Generate test design\n",
    "\n",
    "    Build model\n",
    "\n",
    "    Assess model\n",
    "\n",
    "Define CrossValidation Strategie :\n",
    "\n",
    "                skf = StratifiedKFold(n_splits=FOLDS, \n",
    "                shuffle=True, \n",
    "                random_state=SEED)\n",
    "\n",
    "We will use StratifiedCV in order to keep track of our imblanced data .\n",
    "\n",
    "\n",
    "\n",
    "## Chose Best model\n",
    "\n",
    "We can combine CV and hyper param tuning like this :\n",
    "\n",
    "    RF_MODEL = {}\n",
    "    # Définir la pipe\n",
    "    RF_MODEL['pipeline']=Pipeline([\n",
    "                                   ('data_cleaning', data_cleaning),\n",
    "                                   ('rf', RandomForest(random_state=7))])\n",
    "    # Définir la grille d'hyperparams\n",
    "    RF_MODEL['hyperparams'] = {}\n",
    "    RF_MODEL['hyperparams']['rf__n_estimators'] = [10, 30, 50, 100, 150]\n",
    "    RF_MODEL['hyperparams']['rf__max_features'] = [None, 'sqrt', 'log2', .1, .25, .50, .75, .85]\n",
    "    RF_MODEL['hyperparams']['rf__max_depth'] = [None, 4, 7, 10, 20]\n",
    "    # Recherche d'HP\n",
    "    RF_MODEL['gridsearch'] = GridSearchCV(\n",
    "        estimator=RF_MODEL['pipeline'],\n",
    "        param_grid=RF_MODEL['hyperparams'],\n",
    "        scoring='accuracy',\n",
    "        cv=cross_validation_design\n",
    "    )\n",
    "    RF_MODEL['gridsearch'].fit(X, Y)\n",
    "\n",
    "\n",
    "But this is so computing expensive.\n",
    "\n",
    "For this reason i used optuna to tune param and CV separatly .\n",
    "\n",
    "## List of models tuned for pipe1:\n",
    "**1-HGBC:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "fe567d17",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:23.349880Z",
     "iopub.status.busy": "2021-12-28T12:06:23.348950Z",
     "iopub.status.idle": "2021-12-28T12:06:23.718153Z",
     "shell.execute_reply": "2021-12-28T12:06:23.718717Z"
    },
    "papermill": {
     "duration": 0.409072,
     "end_time": "2021-12-28T12:06:23.718891",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.309819",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"7f7b8893-cee6-4a46-b7d6-3db1120f24f3\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"7f7b8893-cee6-4a46-b7d6-3db1120f24f3\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 HistGradientBoostingClassifier(early_stopping='False',\n",
       "                                                l2_regularization=2.940296779699346e-10,\n",
       "                                                learning_rate=0.015318591316397998,\n",
       "                                                max_bins=137, max_depth=19,\n",
       "                                                max_iter=1537,\n",
       "                                                max_leaf_nodes=23,\n",
       "                                                min_samples_leaf=140))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2b4beb7f-7840-4912-bd96-db8eb50058f7\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"2b4beb7f-7840-4912-bd96-db8eb50058f7\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"376942a4-7836-42e4-8693-e18c4a758b43\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"376942a4-7836-42e4-8693-e18c4a758b43\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4f455960-9d86-47fa-bd3d-80dd305cb443\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4f455960-9d86-47fa-bd3d-80dd305cb443\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c2dab2e2-2a5a-4474-a83a-2a12ae2f2f62\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"c2dab2e2-2a5a-4474-a83a-2a12ae2f2f62\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"de9b110f-a706-44a3-9446-4eddcaf11ffd\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"de9b110f-a706-44a3-9446-4eddcaf11ffd\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b81c65d5-188e-4b70-b742-cfdd429beb4c\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"b81c65d5-188e-4b70-b742-cfdd429beb4c\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4997b5a0-4be0-495d-9e5c-a4ca9eaadac0\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4997b5a0-4be0-495d-9e5c-a4ca9eaadac0\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e53fecca-1daa-439b-8581-c740c969e403\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"e53fecca-1daa-439b-8581-c740c969e403\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ed0be830-b0df-4522-8552-26f70d7bc85b\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"ed0be830-b0df-4522-8552-26f70d7bc85b\">HistGradientBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>HistGradientBoostingClassifier(early_stopping='False',\n",
       "                               l2_regularization=2.940296779699346e-10,\n",
       "                               learning_rate=0.015318591316397998, max_bins=137,\n",
       "                               max_depth=19, max_iter=1537, max_leaf_nodes=23,\n",
       "                               min_samples_leaf=140)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 HistGradientBoostingClassifier(early_stopping='False',\n",
       "                                                l2_regularization=2.940296779699346e-10,\n",
       "                                                learning_rate=0.015318591316397998,\n",
       "                                                max_bins=137, max_depth=19,\n",
       "                                                max_iter=1537,\n",
       "                                                max_leaf_nodes=23,\n",
       "                                                min_samples_leaf=140))])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_hgbc = {'l2_regularization': 2.940296779699346e-10, \n",
    "         'early_stopping': 'False',\n",
    "         'learning_rate': 0.015318591316397998, \n",
    "         'max_iter': 1537, \n",
    "         'max_depth': 19, \n",
    "         'max_bins': 137, \n",
    "         'min_samples_leaf': 140,\n",
    "         'max_leaf_nodes': 23}\n",
    "model_hgbc = HistGradientBoostingClassifier(**params_hgbc)\n",
    "pipe1_model_hgbc_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        #('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "        #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_hgbc)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_hgbc_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31c452fb",
   "metadata": {
    "papermill": {
     "duration": 0.037178,
     "end_time": "2021-12-28T12:06:23.792379",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.755201",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**2-CatBoost**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "2f6c8f10",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:23.873020Z",
     "iopub.status.busy": "2021-12-28T12:06:23.872004Z",
     "iopub.status.idle": "2021-12-28T12:06:24.217701Z",
     "shell.execute_reply": "2021-12-28T12:06:24.216930Z"
    },
    "papermill": {
     "duration": 0.384981,
     "end_time": "2021-12-28T12:06:24.217884",
     "exception": false,
     "start_time": "2021-12-28T12:06:23.832903",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"72ec76c5-84e1-4573-bc30-857a5e0ebdef\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"72ec76c5-84e1-4573-bc30-857a5e0ebdef\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 <catboost.core.CatBoostClassifier object at 0x7fa8e553b8d0>)])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1c76a4d9-e080-47ae-a0fe-2e8a56ded91e\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1c76a4d9-e080-47ae-a0fe-2e8a56ded91e\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4d663dfc-276f-404b-9f6a-625a272d5673\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4d663dfc-276f-404b-9f6a-625a272d5673\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1bbed385-216b-4fca-b03d-8e8b8bf84e7d\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1bbed385-216b-4fca-b03d-8e8b8bf84e7d\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"122a11bd-840f-4d52-b4c5-64a71fd31972\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"122a11bd-840f-4d52-b4c5-64a71fd31972\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"8383db38-a068-4645-ad61-7d960880d712\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"8383db38-a068-4645-ad61-7d960880d712\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"50814add-accb-45cc-af85-63e6abfc3a09\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"50814add-accb-45cc-af85-63e6abfc3a09\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f42d4398-f02d-49e6-89d2-d30e09d8ec6c\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"f42d4398-f02d-49e6-89d2-d30e09d8ec6c\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3d2c696b-3724-41f6-ad22-52612e8a8554\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"3d2c696b-3724-41f6-ad22-52612e8a8554\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3b17f6a1-066f-44c6-ae48-e108da477ea4\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"3b17f6a1-066f-44c6-ae48-e108da477ea4\">CatBoostClassifier</label><div class=\"sk-toggleable__content\"><pre><catboost.core.CatBoostClassifier object at 0x7fa8e553b8d0></pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 <catboost.core.CatBoostClassifier object at 0x7fa8e553b8d0>)])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_catboost = { \"random_state\": 42,\n",
    "                    \"loss_function\":\"Logloss\",\n",
    "                    \"eval_metric\":\"AUC\",\n",
    "                    #\"task_type\" : \"GPU\",\n",
    "                   'learning_rate': 0.03174251209279996,\n",
    "                   'iterations': 1957,\n",
    "                   'objective': 'CrossEntropy',\n",
    "                   'colsample_bylevel': 0.05457006237382004,\n",
    "                   'depth': 4,\n",
    "                   'boosting_type': 'Ordered',\n",
    "                   'bootstrap_type': 'MVS'}\n",
    "model_catboost = CatBoostClassifier(\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             **params_catboost)\n",
    "pipe1_model_catboost_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_catboost)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_catboost_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e093020",
   "metadata": {
    "papermill": {
     "duration": 0.036374,
     "end_time": "2021-12-28T12:06:24.297138",
     "exception": false,
     "start_time": "2021-12-28T12:06:24.260764",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**-3 XGBR**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f8d3cbb7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:24.409242Z",
     "iopub.status.busy": "2021-12-28T12:06:24.408218Z",
     "iopub.status.idle": "2021-12-28T12:06:24.711137Z",
     "shell.execute_reply": "2021-12-28T12:06:24.711577Z"
    },
    "papermill": {
     "duration": 0.377339,
     "end_time": "2021-12-28T12:06:24.711748",
     "exception": false,
     "start_time": "2021-12-28T12:06:24.334409",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a5e52e72-5de0-4047-99b5-2899e253a056\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a5e52e72-5de0-4047-99b5-2899e253a056\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "                               interaction_constraints=None,\n",
       "                               lambda=0.05718879852112006,\n",
       "                               learning_rate=0.04028462871213848,\n",
       "                               max_delta_step=None, max_depth=3,\n",
       "                               min_child_weight=9, missing=nan,\n",
       "                               monotone_constraints=None, n_estimators=6258,\n",
       "                               n_jobs=None, num_parallel_tree=None,\n",
       "                               predictor=None, random_state=42, reg_alpha=None,\n",
       "                               reg_lambda=None, scale_pos_weight=None,\n",
       "                               subsample=0.4600950043573827, ...))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"93fda12d-fa31-433b-bff6-9f494646b868\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"93fda12d-fa31-433b-bff6-9f494646b868\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"43a0a41e-ee8f-406f-944d-2c8d864a02fa\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"43a0a41e-ee8f-406f-944d-2c8d864a02fa\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6383a279-ea2c-489c-860d-4a4af4498f75\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"6383a279-ea2c-489c-860d-4a4af4498f75\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"eaf5851b-b1c0-413c-a531-5efbac0f5c56\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"eaf5851b-b1c0-413c-a531-5efbac0f5c56\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"64b97512-808b-465a-a9e5-0536cfdc8533\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"64b97512-808b-465a-a9e5-0536cfdc8533\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"21b94b01-a77a-442a-8f6d-d838129fc62e\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"21b94b01-a77a-442a-8f6d-d838129fc62e\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c71d7055-3fec-41c3-be19-fa58aa5549e2\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"c71d7055-3fec-41c3-be19-fa58aa5549e2\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f1ab4b66-2060-455e-af54-4768b2088250\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"f1ab4b66-2060-455e-af54-4768b2088250\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"89a139d1-fcbc-47c2-a4be-955833fcae06\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"89a139d1-fcbc-47c2-a4be-955833fcae06\">XGBClassifier</label><div class=\"sk-toggleable__content\"><pre>XGBClassifier(alpha=0.3744455966091751, base_score=None, booster='gbtree',\n",
       "              colsample_bylevel=None, colsample_bynode=None,\n",
       "              colsample_bytree=0.4826795171274075, enable_categorical=False,\n",
       "              eta=1.0243140309668855e-07, eval_metric='auc',\n",
       "              gamma=0.0002882867942209271, gpu_id=None, grow_policy='lossguide',\n",
       "              importance_type=None, interaction_constraints=None,\n",
       "              lambda=0.05718879852112006, learning_rate=0.04028462871213848,\n",
       "              max_delta_step=None, max_depth=3, min_child_weight=9, missing=nan,\n",
       "              monotone_constraints=None, n_estimators=6258, n_jobs=None,\n",
       "              num_parallel_tree=None, predictor=None, random_state=42,\n",
       "              reg_alpha=None, reg_lambda=None, scale_pos_weight=None,\n",
       "              subsample=0.4600950043573827, ...)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "                               interaction_constraints=None,\n",
       "                               lambda=0.05718879852112006,\n",
       "                               learning_rate=0.04028462871213848,\n",
       "                               max_delta_step=None, max_depth=3,\n",
       "                               min_child_weight=9, missing=nan,\n",
       "                               monotone_constraints=None, n_estimators=6258,\n",
       "                               n_jobs=None, num_parallel_tree=None,\n",
       "                               predictor=None, random_state=42, reg_alpha=None,\n",
       "                               reg_lambda=None, scale_pos_weight=None,\n",
       "                               subsample=0.4600950043573827, ...))])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_xgbr = { \"verbosity\": 0,\n",
    "                \"objective\": \"binary:logistic\",\n",
    "                \"eval_metric\": \"auc\",\n",
    "                'random_state': 42,\n",
    "                # regression\n",
    "                #'objective':'reg:squarederror', \n",
    "                # 'eval_metric': 'mae',\n",
    "                #early_stopping_rounds=100 ,\n",
    "               # 'gpu_id':0, \n",
    "               # 'predictor':\"gpu_predictor\",\n",
    "                # use exact for small dataset.\n",
    "                #\"tree_method\": \"exact\",\n",
    "                # big data :\n",
    "                # 'tree_method': 'gpu_hist',\n",
    "                # defines booster, gblinear for linear functions.\n",
    "               'booster': 'gbtree',\n",
    "               'lambda': 0.05718879852112006,\n",
    "               'alpha': 0.3744455966091751, \n",
    "               'subsample': 0.4600950043573827, \n",
    "               'colsample_bytree': 0.4826795171274075, \n",
    "               'learning_rate': 0.04028462871213848, \n",
    "               'n_estimators': 6258, \n",
    "               'max_depth': 3,\n",
    "               'min_child_weight': 9,\n",
    "               'eta': 1.0243140309668855e-07,\n",
    "               'gamma': 0.0002882867942209271, \n",
    "               'grow_policy': 'lossguide'\n",
    "              }\n",
    "model_xgbr = XGBClassifier(\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             **params_xgbr)\n",
    "pipe1_model_xgbr_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_xgbr)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_xgbr_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae35da6d",
   "metadata": {
    "papermill": {
     "duration": 0.037313,
     "end_time": "2021-12-28T12:06:24.786521",
     "exception": false,
     "start_time": "2021-12-28T12:06:24.749208",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**4-LR:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "c41e40a1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:24.865124Z",
     "iopub.status.busy": "2021-12-28T12:06:24.864193Z",
     "iopub.status.idle": "2021-12-28T12:06:25.187186Z",
     "shell.execute_reply": "2021-12-28T12:06:25.187636Z"
    },
    "papermill": {
     "duration": 0.363816,
     "end_time": "2021-12-28T12:06:25.187797",
     "exception": false,
     "start_time": "2021-12-28T12:06:24.823981",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"8438f7e1-9122-4591-9d77-b8f1e82230cd\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"8438f7e1-9122-4591-9d77-b8f1e82230cd\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LogisticRegression(C=6805.270095210636, random_state=42,\n",
       "                                    solver='liblinear'))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"0aa048e4-4433-4dc8-8e3f-fc9fe2c5e955\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"0aa048e4-4433-4dc8-8e3f-fc9fe2c5e955\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"426a132c-8343-4bdc-9dbe-7285af31ab54\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"426a132c-8343-4bdc-9dbe-7285af31ab54\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b0e28e60-e775-460d-be6d-458b7f993e09\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"b0e28e60-e775-460d-be6d-458b7f993e09\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"aa5b9e9b-f3d5-47ad-ace4-d936072f639b\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"aa5b9e9b-f3d5-47ad-ace4-d936072f639b\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"44a8e75f-f696-4e2a-9403-942e5e7eca94\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"44a8e75f-f696-4e2a-9403-942e5e7eca94\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f63df2fa-06e8-4405-bd55-5ded7868e215\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"f63df2fa-06e8-4405-bd55-5ded7868e215\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ed7c9e83-7978-43e7-9760-cf7dfef05524\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"ed7c9e83-7978-43e7-9760-cf7dfef05524\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"66181070-e019-45f5-a1af-fb046d9677ac\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"66181070-e019-45f5-a1af-fb046d9677ac\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"342f227d-d3f9-4c86-b117-a1101a9a9532\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"342f227d-d3f9-4c86-b117-a1101a9a9532\">LogisticRegression</label><div class=\"sk-toggleable__content\"><pre>LogisticRegression(C=6805.270095210636, random_state=42, solver='liblinear')</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LogisticRegression(C=6805.270095210636, random_state=42,\n",
       "                                    solver='liblinear'))])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_lr = {\n",
    "             \"random_state\": 42,\n",
    "             'solver':'liblinear',\n",
    "             'C': 6805.270095210636,\n",
    "             'penalty': 'l2'\n",
    "             }\n",
    "model_lr = LogisticRegression(**params_lr,\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             )\n",
    "pipe1_model_lr_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_lr)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_lr_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed97d12f",
   "metadata": {
    "papermill": {
     "duration": 0.038276,
     "end_time": "2021-12-28T12:06:25.264261",
     "exception": false,
     "start_time": "2021-12-28T12:06:25.225985",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**5-SVM:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "a472c202",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:25.356402Z",
     "iopub.status.busy": "2021-12-28T12:06:25.351572Z",
     "iopub.status.idle": "2021-12-28T12:06:25.663080Z",
     "shell.execute_reply": "2021-12-28T12:06:25.662580Z"
    },
    "papermill": {
     "duration": 0.36029,
     "end_time": "2021-12-28T12:06:25.663208",
     "exception": false,
     "start_time": "2021-12-28T12:06:25.302918",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2d850ce1-f73a-4fca-b75e-742014f50b9e\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"2d850ce1-f73a-4fca-b75e-742014f50b9e\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier', SVC())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5e9a2b6d-fc79-4a18-b0f4-c6ec6739cf2c\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"5e9a2b6d-fc79-4a18-b0f4-c6ec6739cf2c\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"d1ee91f2-dfd8-492a-9a83-b4bc9987b4ae\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"d1ee91f2-dfd8-492a-9a83-b4bc9987b4ae\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4a0e0247-4dd8-4e75-bfe0-a9d74973a6cd\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4a0e0247-4dd8-4e75-bfe0-a9d74973a6cd\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2a1583f3-ae40-44bb-8403-0dea7bc9e6fe\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"2a1583f3-ae40-44bb-8403-0dea7bc9e6fe\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2efac088-581a-4310-b094-edf4c0578966\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"2efac088-581a-4310-b094-edf4c0578966\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"d76ba103-2884-4e75-861f-7800d8c8ce0a\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"d76ba103-2884-4e75-861f-7800d8c8ce0a\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9026264b-bd4a-483e-8afb-c6a410e87699\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"9026264b-bd4a-483e-8afb-c6a410e87699\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1b699be3-392e-4341-8d58-35caca45c910\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1b699be3-392e-4341-8d58-35caca45c910\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"64b451a4-b76e-49c4-b60d-a1fc414fdf59\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"64b451a4-b76e-49c4-b60d-a1fc414fdf59\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC()</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier', SVC())])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_svm = {\"random_state\": 42,\n",
    "             'C': 244.5231198228094, \n",
    "             'penalty': 'l2'\n",
    "            }\n",
    "model_svm = svm.SVC(\n",
    "     #**params_svm,\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             )\n",
    "pipe1_model_svm_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_svm)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_svm_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aff4f732",
   "metadata": {
    "papermill": {
     "duration": 0.040126,
     "end_time": "2021-12-28T12:06:25.742949",
     "exception": false,
     "start_time": "2021-12-28T12:06:25.702823",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**6-RF:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "7d73dd44",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:25.839987Z",
     "iopub.status.busy": "2021-12-28T12:06:25.834829Z",
     "iopub.status.idle": "2021-12-28T12:06:26.151355Z",
     "shell.execute_reply": "2021-12-28T12:06:26.150865Z"
    },
    "papermill": {
     "duration": 0.369215,
     "end_time": "2021-12-28T12:06:26.151485",
     "exception": false,
     "start_time": "2021-12-28T12:06:25.782270",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"830c138d-7023-46a8-9742-e1d1f9b97aff\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"830c138d-7023-46a8-9742-e1d1f9b97aff\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 RandomForestClassifier(max_depth=12, min_samples_leaf=4,\n",
       "                                        min_samples_split=3, n_estimators=2975,\n",
       "                                        random_state=42))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a01be104-2e40-45df-94b1-9d5a42d6d641\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a01be104-2e40-45df-94b1-9d5a42d6d641\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"17d9d255-47ba-4ff9-8215-582814cdae06\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"17d9d255-47ba-4ff9-8215-582814cdae06\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b772aa25-b8d5-43b7-bdba-9d0bb62a2f15\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"b772aa25-b8d5-43b7-bdba-9d0bb62a2f15\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"aea7e02d-163c-4841-a61e-5d046d357909\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"aea7e02d-163c-4841-a61e-5d046d357909\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2ead18f1-3b97-4a61-ac84-ba80ea558d01\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"2ead18f1-3b97-4a61-ac84-ba80ea558d01\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e05d095d-684f-4281-8d78-8b66ce53b768\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"e05d095d-684f-4281-8d78-8b66ce53b768\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"610476c4-2f61-433b-948d-5b7a695582ff\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"610476c4-2f61-433b-948d-5b7a695582ff\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"28403598-7837-4acf-bd69-b674e4eee7fb\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"28403598-7837-4acf-bd69-b674e4eee7fb\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"83c363a7-cc24-4677-947e-de45fc22c188\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"83c363a7-cc24-4677-947e-de45fc22c188\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(max_depth=12, min_samples_leaf=4, min_samples_split=3,\n",
       "                       n_estimators=2975, random_state=42)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 RandomForestClassifier(max_depth=12, min_samples_leaf=4,\n",
       "                                        min_samples_split=3, n_estimators=2975,\n",
       "                                        random_state=42))])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_rf =  {\"random_state\": 42,\n",
    "              'n_estimators': 2975,\n",
    "              'max_features': 'auto', \n",
    "              'max_depth': 12,\n",
    "              'min_samples_split': 3, \n",
    "              'min_samples_leaf': 4, \n",
    "              'bootstrap': True\n",
    "             }\n",
    "model_rf = RandomForestClassifier(**params_rf,\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             )\n",
    "pipe1_model_rf_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_rf)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_rf_woe_powertransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed98891a",
   "metadata": {
    "papermill": {
     "duration": 0.040896,
     "end_time": "2021-12-28T12:06:26.232222",
     "exception": false,
     "start_time": "2021-12-28T12:06:26.191326",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**7-lgbm:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "87a09653",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:26.315292Z",
     "iopub.status.busy": "2021-12-28T12:06:26.314370Z",
     "iopub.status.idle": "2021-12-28T12:06:26.644080Z",
     "shell.execute_reply": "2021-12-28T12:06:26.644495Z"
    },
    "papermill": {
     "duration": 0.372341,
     "end_time": "2021-12-28T12:06:26.644681",
     "exception": false,
     "start_time": "2021-12-28T12:06:26.272340",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"dad44e4c-1a02-40af-a4be-f974577c816e\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"dad44e4c-1a02-40af-a4be-f974577c816e\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LGBMClassifier(bagging_fraction=0.8225459856082522,\n",
       "                                bagging_freq=3,\n",
       "                                feature_fraction=0.7099515479738928,\n",
       "                                lambda_l1=2.5825901359648176e-05,\n",
       "                                lambda_l2=4.490094872871969,\n",
       "                                learning_rate=0.005282912028625591,\n",
       "                                metric='auc', min_child_samples=35,\n",
       "                                n_estimators=1491, num_leaves=76,\n",
       "                                objective='binary', random_state=42,\n",
       "                                verbosity=-1))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"35531cf7-434b-4202-b907-d482b7da3534\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"35531cf7-434b-4202-b907-d482b7da3534\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a0e7c19b-bebf-48cf-bada-6a48ecdef2bd\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a0e7c19b-bebf-48cf-bada-6a48ecdef2bd\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e6bde4e4-f65c-4d89-9e9d-7c87501cc2d1\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"e6bde4e4-f65c-4d89-9e9d-7c87501cc2d1\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ed1d5965-1bc9-47bd-b5a3-71090f89da72\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"ed1d5965-1bc9-47bd-b5a3-71090f89da72\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1ea57056-f48d-4ba4-a9a9-a77f49970af7\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1ea57056-f48d-4ba4-a9a9-a77f49970af7\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4a68497d-fd90-4426-a68b-bdb8255ac305\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4a68497d-fd90-4426-a68b-bdb8255ac305\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"602c6347-ac6c-44d8-91b3-843da45dc1aa\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"602c6347-ac6c-44d8-91b3-843da45dc1aa\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"34e2cef4-a1bf-4843-b320-b7130782b483\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"34e2cef4-a1bf-4843-b320-b7130782b483\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a6015262-e581-4fd1-bab4-55d0b1244fa7\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a6015262-e581-4fd1-bab4-55d0b1244fa7\">LGBMClassifier</label><div class=\"sk-toggleable__content\"><pre>LGBMClassifier(bagging_fraction=0.8225459856082522, bagging_freq=3,\n",
       "               feature_fraction=0.7099515479738928,\n",
       "               lambda_l1=2.5825901359648176e-05, lambda_l2=4.490094872871969,\n",
       "               learning_rate=0.005282912028625591, metric='auc',\n",
       "               min_child_samples=35, n_estimators=1491, num_leaves=76,\n",
       "               objective='binary', random_state=42, verbosity=-1)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LGBMClassifier(bagging_fraction=0.8225459856082522,\n",
       "                                bagging_freq=3,\n",
       "                                feature_fraction=0.7099515479738928,\n",
       "                                lambda_l1=2.5825901359648176e-05,\n",
       "                                lambda_l2=4.490094872871969,\n",
       "                                learning_rate=0.005282912028625591,\n",
       "                                metric='auc', min_child_samples=35,\n",
       "                                n_estimators=1491, num_leaves=76,\n",
       "                                objective='binary', random_state=42,\n",
       "                                verbosity=-1))])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_lgbm =  {\n",
    "               \"objective\": \"binary\",\n",
    "                'random_state': 42,\n",
    "                'metric': 'auc',\n",
    "                \"verbosity\": -1,\n",
    "                \"boosting_type\": \"gbdt\",\n",
    "                #\"device\": \"gpu\",\n",
    "                #\"early_stopping_rounds\":100,\n",
    "                'learning_rate': 0.005282912028625591, \n",
    "                'n_estimators': 1491,\n",
    "                'lambda_l1': 2.5825901359648176e-05, \n",
    "                'lambda_l2': 4.490094872871969,\n",
    "                'num_leaves': 76,\n",
    "                'feature_fraction': 0.7099515479738928,\n",
    "                'bagging_fraction': 0.8225459856082522,\n",
    "                'bagging_freq': 3,\n",
    "                'min_child_samples': 35\n",
    "               }\n",
    "model_lgbm = lgbm.LGBMClassifier(**params_lgbm,\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             )\n",
    "pipe1_model_lgbm_woe_powertransformer = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_lgbm)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_lgbm_woe_powertransformer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ebbcb15e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:26.734103Z",
     "iopub.status.busy": "2021-12-28T12:06:26.733122Z",
     "iopub.status.idle": "2021-12-28T12:06:27.057916Z",
     "shell.execute_reply": "2021-12-28T12:06:27.057440Z"
    },
    "papermill": {
     "duration": 0.373393,
     "end_time": "2021-12-28T12:06:27.058056",
     "exception": false,
     "start_time": "2021-12-28T12:06:26.684663",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>div.sk-top-container {color: black;background-color: white;}div.sk-toggleable {background-color: white;}label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.2em 0.3em;box-sizing: border-box;text-align: center;}div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}div.sk-estimator {font-family: monospace;background-color: #f0f8ff;margin: 0.25em 0.25em;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;}div.sk-estimator:hover {background-color: #d4ebff;}div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;}div.sk-item {z-index: 1;}div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}div.sk-parallel-item:only-child::after {width: 0;}div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0.2em;box-sizing: border-box;padding-bottom: 0.1em;background-color: white;position: relative;}div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}div.sk-label-container {position: relative;z-index: 2;text-align: center;}div.sk-container {display: inline-block;position: relative;}</style><div class=\"sk-top-container\"><div class=\"sk-container\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"48853374-a0c4-499c-979d-53a40c486e4f\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"48853374-a0c4-499c-979d-53a40c486e4f\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LGBMClassifier(learning_rate=0.0018069834369607075,\n",
       "                                max_depth=8, metric='auc', min_samples_leaf=47,\n",
       "                                n_estimators=5000, random_state=42,\n",
       "                                subsample=0.8573598985000007))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4eb8b4ca-f486-438b-a538-9cc624a24b07\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"4eb8b4ca-f486-438b-a538-9cc624a24b07\">preprocessor: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[('numerical',\n",
       "                                 Pipeline(steps=[('imputer',\n",
       "                                                  SimpleImputer(add_indicator=True,\n",
       "                                                                strategy='median')),\n",
       "                                                 ('scaler', PowerTransformer()),\n",
       "                                                 ('kmeans',\n",
       "                                                  MiniKmeansTransformerEncoder()),\n",
       "                                                 ('polynominal_features',\n",
       "                                                  PolynomialFeatures())]),\n",
       "                                 Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5ccc4350-eb72-43c3-a9b6-582a3eb4ce16\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"5ccc4350-eb72-43c3-a9b6-582a3eb4ce16\">numerical</label><div class=\"sk-toggleable__content\"><pre>Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome',\n",
       "       'NumberOfDependents', 'NumberOfTime60-89DaysPastDueNotWorse',\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object')</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"bf8da0d6-2757-4925-a464-71dc82e60f9c\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"bf8da0d6-2757-4925-a464-71dc82e60f9c\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(add_indicator=True, strategy='median')</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b441cfba-fc67-4df8-baa7-61a4559d2552\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"b441cfba-fc67-4df8-baa7-61a4559d2552\">PowerTransformer</label><div class=\"sk-toggleable__content\"><pre>PowerTransformer()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1b11e264-22a6-4a3c-af84-2d71284466dd\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"1b11e264-22a6-4a3c-af84-2d71284466dd\">kmeans: MiniKmeansTransformerEncoder</label><div class=\"sk-toggleable__content\"><pre>MiniKmeansTransformerEncoder()</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a4cc58e4-62c5-4d22-8c91-92daac9c0a97\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"a4cc58e4-62c5-4d22-8c91-92daac9c0a97\">WOEEncoder</label><div class=\"sk-toggleable__content\"><pre>WOEEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"764eda3c-a5ee-4150-bd62-b7e261804242\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"764eda3c-a5ee-4150-bd62-b7e261804242\">PolynomialFeatures</label><div class=\"sk-toggleable__content\"><pre>PolynomialFeatures()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ad33e24b-974e-4ac5-9a34-75678ae900c3\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"ad33e24b-974e-4ac5-9a34-75678ae900c3\">SelectPercentile</label><div class=\"sk-toggleable__content\"><pre>SelectPercentile(percentile=95)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"33f828c3-ebfe-40d8-8f2d-f4ae5971d8ba\" type=\"checkbox\" ><label class=\"sk-toggleable__label\" for=\"33f828c3-ebfe-40d8-8f2d-f4ae5971d8ba\">LGBMClassifier</label><div class=\"sk-toggleable__content\"><pre>LGBMClassifier(learning_rate=0.0018069834369607075, max_depth=8, metric='auc',\n",
       "               min_samples_leaf=47, n_estimators=5000, random_state=42,\n",
       "               subsample=0.8573598985000007)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('preprocessor',\n",
       "                 ColumnTransformer(transformers=[('numerical',\n",
       "                                                  Pipeline(steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=True,\n",
       "                                                                                 strategy='median')),\n",
       "                                                                  ('scaler',\n",
       "                                                                   PowerTransformer()),\n",
       "                                                                  ('kmeans',\n",
       "                                                                   MiniKmeansTransformerEncoder()),\n",
       "                                                                  ('polynominal_features',\n",
       "                                                                   PolynomialFeatures())]),\n",
       "                                                  Index(['RevolvingUtilizationOfUnsecuredLines', 'DebtRatio', 'MonthlyIncome'...\n",
       "       'NumberRealEstateLoansOrLines', 'NumberOfTimes90DaysLate',\n",
       "       'NumberOfOpenCreditLinesAndLoans',\n",
       "       'NumberOfTime30-59DaysPastDueNotWorse', 'age'],\n",
       "      dtype='object'))])),\n",
       "                ('reducedim', SelectPercentile(percentile=95)),\n",
       "                ('classifier',\n",
       "                 LGBMClassifier(learning_rate=0.0018069834369607075,\n",
       "                                max_depth=8, metric='auc', min_samples_leaf=47,\n",
       "                                n_estimators=5000, random_state=42,\n",
       "                                subsample=0.8573598985000007))])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgbm_param2={'learning_rate': 0.0018069834369607075,\n",
    "                             'max_depth': 8,\n",
    "                             #'max_features': 4,\n",
    "                             'min_samples_leaf': 47,\n",
    "                             #'min_samples_split': 389,\n",
    "                             'subsample': 0.8573598985000007,\n",
    "                             #'n_iter_no_change': 300,\n",
    "                             'n_estimators': 5000,\n",
    "                            # 'verbose': 0,\n",
    "                             'random_state': 42,\n",
    "                             'metric': 'auc',\n",
    "                            #\"device_type\" : \"gpu\",\n",
    "                            'boosting_type': 'gbdt',\n",
    "                           # 'tree_method': \"gpu_hist\"\n",
    "                           }\n",
    "model_lgbm2 = lgbm.LGBMClassifier(**lgbm_param2,\n",
    "                              # use it  only on gpu\n",
    "                            # rsm=np.nan\n",
    "                              #cat_features=cat_features\n",
    "                             )\n",
    "pipe1_model_lgbm_woe_powertransformer2 = Pipeline(\n",
    "    steps=[\n",
    "        ('preprocessor', preprocessor_woe_powertransformer),\n",
    "         ('reducedim',  SelectPercentile( f_classif, percentile=95)),\n",
    "        # ('dim_red', SelectKBest(f_regression, k=20)),\n",
    "        #('Sparseinteractions', SparseInteractions(degree=2)),\n",
    "         #('PolynomialFeatures', PolynomialFeatures(degree=2)),\n",
    "        #('PCAreducer', PCA(n_components=0.8)),\n",
    "        ('classifier',model_lgbm2)\n",
    "    ]\n",
    ")\n",
    "pipe1_model_lgbm_woe_powertransformer2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1b69319",
   "metadata": {
    "papermill": {
     "duration": 0.041075,
     "end_time": "2021-12-28T12:06:27.142962",
     "exception": false,
     "start_time": "2021-12-28T12:06:27.101887",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Test models pipe1  with CV "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "d385e8dc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:27.230118Z",
     "iopub.status.busy": "2021-12-28T12:06:27.229205Z",
     "iopub.status.idle": "2021-12-28T12:06:27.235046Z",
     "shell.execute_reply": "2021-12-28T12:06:27.235538Z"
    },
    "papermill": {
     "duration": 0.050362,
     "end_time": "2021-12-28T12:06:27.235720",
     "exception": false,
     "start_time": "2021-12-28T12:06:27.185358",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def plot_roc(y,y_pred):\n",
    "    from sklearn.metrics import auc\n",
    "    fpr, tpr, thresholds = roc_curve(y_true=y, y_score=y_pred)\n",
    "    auc = auc(fpr, tpr)\n",
    "    plt.figure(1)\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.plot(fpr, tpr, label='Model(area = {:.3f})'.format(auc))\n",
    "    plt.xlabel('False positive rate')\n",
    "    plt.ylabel('True positive rate')\n",
    "    plt.title('ROC curve')\n",
    "    plt.legend(loc='best')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "c72908c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:06:27.321417Z",
     "iopub.status.busy": "2021-12-28T12:06:27.320481Z",
     "iopub.status.idle": "2021-12-28T12:30:11.155988Z",
     "shell.execute_reply": "2021-12-28T12:30:11.156439Z"
    },
    "papermill": {
     "duration": 1423.880031,
     "end_time": "2021-12-28T12:30:11.156645",
     "exception": false,
     "start_time": "2021-12-28T12:06:27.276614",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 0 | AUC: 0.866285974540478\n",
      "Fold 0 | Avergae_precesion: 0.40153176937358215\n",
      "Fold 0 | recall: 0.1864406779661017\n",
      "Fold 0 | F1: 0.28462709284627097\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.99      0.97     13997\n",
      "           1       0.60      0.19      0.28      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.77      0.59      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13873   124]\n",
      " [  816   187]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1 | AUC: 0.8609375132443635\n",
      "Fold 1 | Avergae_precesion: 0.38271210643004555\n",
      "Fold 1 | recall: 0.19441674975074777\n",
      "Fold 1 | F1: 0.2916978309648467\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.99      0.97     13997\n",
      "           1       0.58      0.19      0.29      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.76      0.59      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13858   139]\n",
      " [  808   195]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 2 | AUC: 0.8660398030029366\n",
      "Fold 2 | Avergae_precesion: 0.40260659529955894\n",
      "Fold 2 | recall: 0.19641076769690927\n",
      "Fold 2 | F1: 0.2916358253145818\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.99      0.97     13997\n",
      "           1       0.57      0.20      0.29      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.76      0.59      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13846   151]\n",
      " [  806   197]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 3 | AUC: 0.863946917552693\n",
      "Fold 3 | Avergae_precesion: 0.3967393184748874\n",
      "Fold 3 | recall: 0.1984047856430708\n",
      "Fold 3 | F1: 0.295910780669145\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.99      0.97     13997\n",
      "           1       0.58      0.20      0.30      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.76      0.59      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13854   143]\n",
      " [  804   199]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 4 | AUC: 0.8657162398636769\n",
      "Fold 4 | Avergae_precesion: 0.4015988489452065\n",
      "Fold 4 | recall: 0.2023928215353938\n",
      "Fold 4 | F1: 0.30185873605947955\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.99      0.97     13997\n",
      "           1       0.59      0.20      0.30      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.77      0.60      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13858   139]\n",
      " [  800   203]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 5 | AUC: 0.874196194014228\n",
      "Fold 5 | Avergae_precesion: 0.44035609669675335\n",
      "Fold 5 | recall: 0.20039880358923232\n",
      "Fold 5 | F1: 0.30994602929838083\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.99      0.97     13997\n",
      "           1       0.68      0.20      0.31      1003\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.81      0.60      0.64     15000\n",
      "weighted avg       0.93      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13904    93]\n",
      " [  802   201]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfEAAAFoCAYAAABHW8L2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAABbrklEQVR4nO3deXhM5/vH8fdk32OPtGKrWNooag8SQmoX+760qBZBlSpKiipqX6I0Py2tUm3VHpRSW20VKvY9miAjJLLLMnN+f0TnK7UkJJOTSe7XdfXqnDNnznzymOSe5yzPo1EURUEIIYQQJsdM7QBCCCGEeDlSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFEWagcQQrwYHx8f7t27h7m5OXZ2djRp0oTJkydjb29v2ObkyZMsXLiQM2fOYGZmRt26dRk7diyVKlUybJOQkMCiRYvYvXs3sbGxFC9enGbNmjF06FCKFSumxo8mhHhB0hMXwgQtX76cU6dOsWnTJs6fP09QUJDhuVOnTjFo0CCaN2/OwYMH2bNnD1WqVKFXr16Eh4cDkJqayoABA7h69SorVqwgJCSEn376iSJFinDmzBmj5U5PTzfavoUojKSIC2HCSpYsSePGjblw4YJh3Zw5c/Dz82PAgAE4ODhQpEgRRo8eTY0aNViyZAkAmzdv5s6dOwQGBlKpUiXMzMwoXrw4w4cPx9vb+6nvdeXKFd59913q1auHp6cny5cvB2D8+PEsWLDAsN2xY8fw8vIyLPv4+BAUFET79u2pWbMmQUFBjBw5MtO+p0+fzvTp0wGIj49n4sSJNG7cmCZNmrBgwQJ0Ol3uNJgQBYwUcSFMWGRkJAcPHqRs2bIAJCcnc+rUKVq1avXEtq1bt+bw4cMAHD58mCZNmmQ6BP88CQkJvPvuuzRp0oSDBw+ya9cuGjZsmO2cwcHBBAUFceLECdq2bcv+/ftJSEgAQKfTsXPnTtq1awdkfCmwsLBg165dbNq0iT///JNffvkl2+8lRGEiRVwIEzR8+HBq1aqFt7c3xYoVM/RsY2Nj0ev1lCxZ8onXlCxZkpiYGAAePHjw1G2eZd++fZQoUYKBAwdibW2Ng4MDNWrUyPbr+/Xrh6urKzY2Nrz66qu8/vrr/P777wAcPXoUGxsbatasyb1799i/fz8TJ07Ezs6O4sWL88477xAcHJzt9xKiMJEL24QwQUuXLsXT05Pjx48zZswYYmJicHJywsnJCTMzM6KionjttdcyvSYqKoqiRYsCUKRIEaKiorL9fnfu3DH09l+Gq6trpuV27dqxbds2OnbsyLZt2wy98Nu3b5Oenk7jxo0N2+r1+ideL4TIID1xIUxYvXr16Ny5M19++SUAdnZ21KxZk507dz6x7Y4dO2jQoAEAnp6eHDp0iKSkpGy9j6urq+GiuP+ytbXl4cOHhuV79+49sY1Go8m03Lp1a44fP05kZCS7d++mffv2AJQuXRorKyuOHj3KiRMnOHHiBCdPnpSeuBDPIEVcCBM3YMAADh8+zMWLFwEYM2YMmzZt4vvvvychIYHY2FgWLFjA33//jb+/PwB+fn6ULl2aESNGcO3aNfR6PTExMSxfvpz9+/c/8R5NmzYlKiqKVatWkZqaSkJCAqdPnwagWrVq7N+/nwcPHhAVFcV3332XZeZixYpRr149JkyYQJkyZQxHDUqVKkWjRo2YNWsWCQkJ6PV6/vnnH44fP55bzSVEgSJFXAgTV6xYMfz8/Fi6dCkAderUYcWKFezevZsmTZrQrFkzLly4wNq1aylfvjwAVlZWrFq1iooVKzJw4EBq165Nt27diImJ4c0333ziPRwcHPj222/5448/aNSoES1btuTYsWNAxheCqlWr4uPjw8CBA2nTpk22crdr147Dhw8bDqX/a/bs2aSlpdGmTRvq1q3LyJEjX+jQvxCFiUZRFEXtEEIIIYR4cdITF0IIIUyUFHEhhBDCREkRF0IIIUyUFHEhhBDCRJnUYC96vZ7ExEQsLS2fuO9UCCGEKGgURSEtLQ17e3vMzJ7sd5tUEU9MTOTy5ctqxxBCCCHyVOXKlXF0dHxivUkVcUtLSyDjh7GyssqVfZ49exYPD49c2VdhJu2Yc9KGOSdtmHPShjmXm22YmprK5cuXDfXvv0yqiP97CN3Kygpra+tc229u7qswk3bMOWnDnJM2zDlpw5zL7TZ81ilkubBNCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFFGK+ITJkygYcOGtGvX7qnPK4rC9OnT8fX1pX379pw7d85YUfJElSpVGDt2rGE5PT2dBg0a8P7777/Qfnx8fIiOjs72Ng8fPqRv377odLoXD51Hvv76a3x9fWnZsiUHDx586jZHjhyhU6dO+Pn50atXL27evAnAhg0baNCgAX5+fvj5+fHLL78YXjNnzhzatWtHu3bt2L59u2H96NGjCQsLM+rPJIQQ+YHRinjnzp1ZsWLFM58/cOAAYWFh7Nq1i88//5wpU6YYK0qesLOz48qVKzx8+BCAP//8ExcXF6O/76+//oqvry/m5ubZ2l5RFPR6vZFT/c/Vq1cJDg4mODiYFStWMHXq1Kd+4ZgyZQpz585l8+bNtGvXjmXLlhmea9OmDZs3b2bz5s1069YNgH379nH+/Hk2bdrEzz//zDfffENCQgIAvXr1eu5nTwghCgqjjdhWt25dIiIinvn8nj176NixIxqNhpo1axIXF8fdu3cpVapUjt533NYQ1p++me3tU1NTsdrx/O271ijH7Pa1s9yXt7c3+/bto1WrVgQHB9O2bVtCQkIAePDgARMnTiQ8PBxbW1umTZtG1apViYmJYcyYMWi1WmrWrImiKIb9bd68mdWrV5OWlkaNGjX47LPPnijWW7duZe7cuUDG2PLDhg0jLi6O9PR0Ro0aRYsWLYiIiGDQoEHUqFGDc+fOERQUxI4dO9ixYwepqan4+voycuRIAIYNG0ZkZCQpKSn079+fHj16ZLstn2bPnj20bdsWKysr3NzcKFeuHKGhodSqVeuJbf8twgkJCVl+Dq5evUqdOnWwsLDAwsKCKlWqcODAAdq0aUOdOnUYP3486enpWFiY1KCEQgiVPUhO5fr9+Ezr0nR6dl26g7VF1v3e2Ng4SpilUjvrkpErVPsLp9VqKV26tGG5dOnSaLXabBXxs2fPPne/qampL5Qlq+21Wq2hGD+LTqejQoUKrFmzBmdnZ/7++2+qV69ObGwsISEhrFq1iiJFijBo0CDOnTvHyJEjmTlzJt999x2urq4MHz6cU6dOsX79ek6fPk18fDw//vgjH3/8MRYWFnz77bcsXrwYLy8vUlJSOH36NHZ2dly/fh2tVotWq0Wn0zF48GDs7OyIi4vjs88+o0iRIty7d4+bN2/y7rvv0qNHD3bt2sWJEyeYMGECiqIwb948ihUrRrVq1ejRowcODg6kpqYyadIkXFxcnhh0f/Xq1Zw/f/6JNmjYsCEdOnTItO7MmTO4u7sb2s/CwoIjR448cTSgf//+DBw4ECsrK2xtbZk6dSohISGEhYURHBzMgQMHcHV1pV+/fhQvXhxzc3N27txJrVq1SElJ4cCBA9jY2BiOfhQtWpSNGzdSsWLF5/675TdZfc5E1qQNc85U2/BechrrLkUTFpuCTTYK7uN23YzDycqcuNScn5q0MVNo9MoJzM2MP9umSXZTPDw8njku7Xcv+O0nJCSE2rnwlcnc3JxOnTqxevVqbt++TatWrahcuTKHDh2idu3afP755yxZsgQ3Nzdq167NN998Q5UqVbh58yaBgYGG9UFBQdSoUYPt27cTERHBjBkzgIxz31WrVqV27dpYW1tTo0YN0tLSKFasmCF/WloaM2fO5K+//sLMzIwHDx5Qrlw5SpcuzSuvvELPnj0B+P3337l8+TKff/45AElJSVhbW1O7dm2WLFnC7t27gYyjB0WKFKFmzZqZftantdez2jE4OJjy5csbnitevDgVK1Z8YtuVK1fy7bffUqNGDVasWMHOnTv54osvqFixIv7+/lhZWbFu3Tp++OEHvv/+e2rXrs3Dhw/58ssvKVasGPXr16ds2bKG/ZYvX57ixYvnyr9tXsmtz2JhJm2Ycy/ahqnpmYteYmo6Oy/eZvflOzjZPH3Sjn1XI0lK1WH1goX2eS5oY3O8j7hUHa+7OJOQmk7n6mUzPZeQmoZn+VKUcrB54nXXr19n6VdLCf8nnFKlSvJe947Uq1snx3kAUlJSnttxVa2Iu7i4EBkZaViOjIzMk3PIxubj48Ps2bP5/vvvefDgwUvvR1EUOnXqxJgxY565jY2NTaajCFu3biU6OpoNGzZgaWmJj48PKSkpQMY5+8f3PWTIEENR/9exY8c4fPgwP/30E7a2tvTr18/w+sfNmDGDY8eOZVqXnJxM165dGTJkSKb1//131mq1T/w7R0dHc/HiRWrUqAFknAMfPHgwkNGj/le3bt2YM2eOYXno0KEMHToUgDFjxlChQgXDc6mpqdjYPPnLJoTIcFEby/XoBI6GRWFj+b/TdLduRfFbzBnD8o8nb1DC/umdpgPX7+Yog5W5Gc62Ty/0L6qEvTX3ElN4o7QzQxpUpvObZXnGnCHPVNzOGiuL7F1fBBl/Z+bMmcPChQvR6XQMefddpkyZkqdTZqtWxH18fPjhhx9o27Ytp0+fxtHRMcfnw/ODrl274uTkRJUqVTIVujp16rBlyxaGDx/OsWPHKFq0KA4ODtStW5etW7cybNgw9u/fT2xsxrfJhg0bMmzYMN555x2KFy/OgwcPSExM5NVXXzXs09nZGZ1OR0pKCtbW1sTHx1O8eHEsLS05evQot27demrGxo0bs2jRItq3b4+9vT1arRYLCwvi4+NxdnbG1taWa9eu8ffffz/19RMnTnxi3bO+vfv4+DBmzBjeffddtFotYWFhvPnmm5m2cXJyIj4+nhs3blChQgX+/PNPXnvtNYBM10ns3bvXsF6n0xEXF0fRokW5ePEily5dolGjRoZ9hoWF4e7u/tT8QpiaqISHJKamA/AwTcf+61p+OhX2zOL6X6k6PVvPReBkY4kGiH2Y9vwXhEa9UL4WlV0zLd+OTaJnrfL4uLtiZ/X0oujiYEtpJ9sXep/85p9//iEwMBBXV1cWL15M06ZN8zyD0Yr4Rx99xPHjx4mJicHLy4sRI0aQnp7xIezVqxfe3t7s378fX19fbG1tDYeNTV3p0qXp37//E+v9/f2ZOHEi7du3x9bWllmzZgEwfPhwxowZQ9u2balVqxavvPIKAJUqVeLDDz9k4MCB6PV6LC0tCQgIyFTEARo1akRISAienp60b9+eoUOH0r59ezw8PJ55Prhx48Zcu3bN0BO3s7Njzpw5eHl5sW7dOlq3bk2FChWeOIz+Mtzd3WndujVt2rTB3NycgIAAw8V57733HtOnT8fFxYXp06czcuRINBoNzs7Ohs/D6tWr2bt3L+bm5jg7OzNz5kwg4xa+Pn36AODg4MCcOXMMF7Hdu3cPa2trSpYsmeP8QuS1lHQde65E8uPJG/xy+iZputy7myTuYRo1Xsk4uhWTnErNV4pS2smW9m+4Yfno/O2VK1ee+AJcx604Re1kelLI6H3fvXuXMmXKUKlSJdasWUOdOnVwcnJSJY9Gefxy6Hzu33MDzzsn/qJM/RzauXPnWLVqVabDzGrIT+24atUq7O3tDbejmYr81IamyhTbcN/VSDaE/sP2C7dwcbTh6M17T2xjY2FOddciVHNxBjIKcHP30rxd5RWKZbO4mmk0FM9Gz90U2zCvnD17lmHDhqHX69m7dy9WVlZP3S432zCrumeSF7aJ/3njjTeoX78+Op0u2/eKF3SOjo74+fmpHUMUcnq9wo3oBBT+10+KTkrl2M0ozDQafgi5zvF/7md6zY3oBDQaUBQY1+wNWlR2xfs1FyzMZXBNNaWlpbFgwQLmzp1Leno6/fr1Iy0t7ZlFPC9JES8AunbtqnaEfKVLly5qRxCFTEJKGmO3hHArNonfL9/hFWdbwqITs/36KiWd+LaXJ/XcSmCWB7cliew7f/48w4YNIzQ0FFdXVxYtWkSLFi3UjmUgRVwIIR75IeQ647aGUK6o/Qu97r896rDoRNyK2BH+IIl36r6G2WOXSd9PSqFbjXKYaTQ0LF+Ssi/4XiLv6HQ6+vXrx40bN+jduzdffPEFzs7OasfKRIq4EKJASNfpWX8lmsNJF9h3TcvDNN1zB9v4JyaRM3ceYGGmwUyjIfWxC8i08Q+xeYFbjSzNzUjT6VnTtzGe5UtJYTZxSUlJ2NnZYW5uzoIFC0hJScHX11ftWE8lRVwIka8pikJMciop6ToW7r/A3YSHmYpzdFIqm8+G/+8Ff0U+ZS/Plq5XaFCu+KP3gqaVXJjWqqachy6E0tPTWbJkCV9//TV//PEHrq6ueHl5qR3ruaSICyHypS1nw/lg/VG08Q9f6HUBb79JddeivOJsyxsuRZ67rYW5BltL+TMo4OLFi/j7+3Py5ElKly5NeHg4rq6uWb9QZfLpFULkC+ExicQ+TCUy/iHdv9v/xIAkHau7cS8hhe41y9GqaubxEqzMzXArav/o1p4aeRlbmLj09HSWLl3KzJkzSU1NpXv37sycOTPTaJH5mRRxIUSeOXMnhr9vxbDr0m3MNBp0ej0/ngp75vbF7ay5GdBZesvCaCZNmkRQUBAuLi7Mnz+f1q1bqx3phchvhhAiV1yOiiMsOsGwfPC6lp9O3cTOyhwNGq7ejyPpOTNElXa0xb2kI9VdixL7MJVPW1SnSqn8dSWwKBgURUHz6I6BDz74gISEBKZNm0axYsVUTvbipIgLIbIlJV3H+cj/zRT1f0evcOZODLaW5iSn6Tgc9uzxtovYWmFpZoa5mZ6S9jZMaVWDckXtqVrKGTMNvOpsZ/ijKoQxXb16lZEjRxIQEECDBg0oX748gYGBasd6aVLEhRAAxD1MJT4l3bCcmq7j9S+34GhtiUYD9xKfnNHuv9xLODKg7muG5SJ2VnSvUT5bw30KYUw6nY7ly5fzxRdf8PDhQ3bu3EmDBg3UjpVjUsSFKITO3Ikh/EESM3afIfxBIg8eppLwWAF/3P2kFF53caaUgw23YpN4p15GkVYUKFfUnuGNqwKgAbktS+RL165dw9/fn2PHjlGiRAmWL19Ohw4d1I6VK6SIC1GI3IlLoszUX5/6nK2lOS6ONjQs97/Z3+JT0vms5Zu8VaZ4XkUUIlcdOnSIHj16kJycjJ+fH3PmzKFEiRJqx8o1UsSFKMCu3otj0Lojhnmb15++aXjurTLF6PJmWeqVLYGPe/6/H1aIl1GrVi3eeOMNhg4dSqdOndSOk+ukiAtRAKXr9Gw5F0G37/Y/9fn703tQxFb9GZiEyG16vZ4VK1Zgb29Pnz59sLe357fffiuwF05KERfChO2+dJvDYRlTWwafj8DS3IyY5FQuaGMzbXd5QkccrTN+3YvYWmH1AuOCC2EqwsLCGDFiBH/++Sdly5ale/fuWFpaFtgCDlLEhTBJiqJQf+F2QiKin3ju379XZhoNzSq5sGlgM+ys5FddFFx6vZ5vv/2WqVOnkpiYSNu2bZk7dy6WlpZqRzM6+c0WIh9LTdeRkv6/2bVqzw9Gp+ifmKt69wcZ8xtXLulEmSIyg5YoPOLi4ujbty+HDh2iSJEiBAUF0aVLlwLd+36cFHEh8oEHyalsvBpDj50bKeVgA8Ct2CQiYpOeur1bETtSdXq+bFebfnUq5mVUIfIVR0dHrK2tadWqFfPnz6d06dJqR8pTUsSFUElMUgrv/3KUX0P/ybT+RnQC1hZm6JWMZScbS5pULAVkTJvZuuorjGhSLa/jCpFvhIeH89tvvzF48GA0Gg2rVq3Czq5wjvonRVyIPPQwTcfFu7HUXbAdvaJkes7NwYp+DSoztukbOMuV40I8QVEUvvvuOwICAkhISKBu3brUqFEDe/vCewpJirgQeSA8JpERG4+z9VxEpvVFbK2Y1qoGwxtXfTSNZi2VEgqRv0VERDBy5Ej27duHk5MTX331FW+++abasVQnRVwIIwk+H8HU305zOy6ZO3HJmZ6rV7Y4v77TlFec7VRKJ4TpWLNmDRMmTCAhIQFfX18WLFjAK6+8onasfEGKuBBGMGLDcb7689IT6w/6t6RBuZKYmRW+c3dCvKxr166h0WhYsmQJvXv3LpTnvp9FirgQuSQhJY2D1+/y7fGrbHh0sdqrznacHdceGwtzGWBFiGxSFIXt27fTqlUrzM3N+eSTTxg0aBCvvvqq2tHyHSniQrwknV5Pj+8PEp2Uwv5r2ieer+bizNlxBWOmJCHyyu3bt/nwww/5/fffmTFjBh988AHW1tZSwJ9BirgQWUhJ17HtfATxD9NZd+oGt2KTuHIvnjSd/olt3Us40qNWeXzcXfF6dFuYECJriqKwbt06JkyYQFxcHE2bNqVdu3Zqx8r3pIgL8RxxD1Mp+ulPz91mdZ/GdH2zLJbmZnKuToiXcOfOHUaPHs2uXbtwcHBg/vz5DBgwQH6fskGKuBDPoChKpgI+vFEVarxalLpuJShb1F5mARMilxw5coRdu3bh7e3N4sWLcXNzUzuSyZAiLsR/JKSksebkDYatP2ZYd3G8H+4lnVRMJUTBotVqsbGxwdnZmU6dOuHo6EiLFi2k9/2CzNQOIER+kpquw3niukwFfPOgZlLAhcgliqLwyy+/0LBhQyZMmACARqPB19dXCvhLkJ64KPQURSH0TgwTg0+x8+Jtw/rFnerSt3ZFGQJViFxy9+5dxowZQ3BwMHZ2drz11lsoiiLFOwekiItCKzVdx7fHrzH812OZ1pd0sObHfl40q1S4ZkMSwlgURWHDhg188sknREdH06hRI5YsWUL58uXVjmbypIiLQkevVxgffJJ5+85nWv9O3deoX64EQxpWVimZEAVTeHg4w4YNw9LSklmzZjF48GDMzORsbm6QIi4KnSG/HGHl8WuG5WGNqjC7/VvYWsqvgxC5KS4uDicnJ8qWLUtgYCB16tShQoUKascqUOSvlig0mi39jQPX7xqWfSqVZtcHcjWsELnt3r17fPzxx4SFhbFr1y4sLS3p1q2b2rEKJDmeIQo0vV5h1p4zmI9ZbSjgrzrbUa9scSngQhjBli1b8PT0ZPPmzdjY2PDgwQO1IxVo0hMXBdaG0H/o9t3+TOuWda0v57yFMIL79+8zbtw4Nm7ciI2NDZ9//jkffPAB5uYy8Y8xSREXBVJKui5TAV/erQH961TEWmYSEyLXKYpCp06dOHv2LHXq1GHp0qW4u7urHatQkCIuCpSjN6NotHhnpnXpc/vKYXMhjODfe7w1Gg2ffvoply9fZtiwYdL7zkNyTlwUCDq9nvYr9mYq4FVKOrHrfTnvLYQxbN++nSZNmhAVFQVAy5YtGTFihBTwPCY9cWGybsUm8cvfYYzbdhKdXjGst7U0J+KzrjJBiRBGEBMTw4QJE/j555+xsrLir7/+ok2bNmrHKrSkiAuTVXbar0+s+2PY23i95qJCGiEKvt9++43Ro0cTGRlJrVq1CAwMpFq1amrHKtSkiAuTtPjABcPjX9/xpnGFUpRwsFExkRAF28KFC5k2bRqWlpZMnjyZESNGYGEhJURt8i8gTMa+q5EcCYti1t6zJKSkA9C3dkU6Vi+rcjIhCr62bduya9cu5s6dy+uvv652HPGIUS9sO3DgAC1btsTX15egoKAnnr99+zb9+vWjY8eOtG/fnv379z9lL6Kw0+n1fLI1hObLdjNpx9+GAl7HrTj/172ByumEKJhiY2MZOXIkp06dAsDd3Z3t27dLAc9njNYT1+l0TJs2jZUrV+Li4kLXrl3x8fGhUqVKhm2WLVtG69at6d27N1evXmXIkCHs3bvXWJGEiUlOS+edHw+z/vTNTOuD3/OhQjEHqpRyVimZEAXb77//zqhRo7hz5w4PHz58aidM5A9GK+KhoaGUK1cONzc3IONQzJ49ezIVcY1GQ0JCAgDx8fGUKlXKWHGECQrYcTpTAZ/euibjm3vILWNCGElcXBzz5s1j586dWFhYMH78eEaPHq12LPEcRiviWq2W0qX/Nx+zi4sLoaGhmbbx9/dn0KBB/PDDDyQnJ7Ny5UpjxREmZszmEyx8dPHaAr86jPSSK2CFMKZz587Rs2dPbt26RfXq1Vm6dCkeHh5qxxJZUPXCtuDgYDp16sTAgQM5deoU48aNY9u2bVnOM3v27NlczRESEpKr+yusXrYdFUVhzolIFODC/WT+iU8lIU0PQFFrczztEgvNv1Fh+TmNSdrw5SQmJqIoCv369aNXr16kpKRIW+ZAXrWd0Yq4i4sLkZGRhmWtVouLS+b7d9evX8+KFSsAqFWrFikpKcTExFC8ePHn7tvDwwNra+tcyRkSEkLt2rVzZV+F2cu247LDl/D/9fhTn/vAszKBnesVmsPn8lnMOWnDF7Nv3z7i4uLo0KEDACdOnODcuXPShjmUm5/DlJSU53ZcjVbEq1evTlhYGOHh4bi4uBAcHMy8efMybePq6sqRI0fo3Lkz165dIyUlhWLFihkrksgnYpNTWX3iOlfvx7Pk4EXD+tnt3qJ1tVcpYmvFK852KiYUomCLj49nypQprFy5kuLFi9OiRQvs7OywsZGxFkyN0Yq4hYUFAQEBDB48GJ1OR5cuXXB3d2fRokV4eHjQvHlzxo8fz6RJk1i1ahUajYZZs2YVml5XYfXp9lPM2pP5W2X5YvZc+7SzSomEKFwOHDjAiBEjCA8Pp1q1aixduhQ7O/nSbKqMek7c29sbb2/vTOtGjRpleFypUiXWrVtnzAgiHzl4XZupgH/Tw5NSjja8XdlVxVRCFA4pKSlMmjSJb775BnNzc8aMGcPYsWNz7dSkUIeM2CbyxKmIaJou3WVY1s3rp2IaIQofS0tLLl++TJUqVVi6dClvvfWW2pFELpAiLoxOURTqLAg2LCfM6qViGiEKj8TERP744w/atWuHmZkZK1aswNHRUc59FyAyn7gwCp1ez94rdxj+6zEsxv5gWB/yUVtsLeW7oxDGdvjwYZo0acKAAQM4fjzjDpCSJUtKAS9g5K+pyHWKomA9bg2Kknn9Z2+/Sc1X5e4DIYwpKSmJzz//nKCgIDQaDSNGjODNN99UO5YwEiniItf5ffuHoYD7ebjxfsPKNK3kgrWFubrBhCjgjh49ir+/P9evX8fd3Z3AwEDq1q2rdixhRFLERa65n5hC9+/2s++aFoChnpUJ7FJf5VRCFB5btmzhxo0bDB8+nIkTJ2Jra6t2JGFkUsRFrkjXK5QK+DnTOingQhjfuXPnqFatGmZmZkyaNIlOnTpJ77sQkQvbRI7p9Ho8110wLP8x7G3S5/ZVMZEQBV9ycjIBAQF4e3uzatUqAOzs7KSAFzLSExc5kq7TYz1ujWH5gH9LGlWQKWWFMKa//voLf39/rly5QoUKFahWTWb5K6ykJy5eSnJaOiuOXslUwLcN9pECLoQRPXz4kClTptC6dWuuXLnCkCFDOHDgAA0bNlQ7mlCJ9MTFC0lJ1/Hez0dYE3Ij0/rFzcrSutqrKqUSonDYuXMnixcvpnz58ixZsoRGjRqpHUmoTIq4yJbElDSafrWLkxHRmdaPa/YGI5pU5c7VC894pRAiJ1JSUkhPT8fe3h4/Pz/mzZtH9+7dsbe3VzuayAekiIssKYqC08TME9UcHdWa2mWKY2aWMevcHTWCCVHAnTp1iuHDh1O/fn0WLFiARqPh3XffVTuWyEekiIvnUhQl07CpG95tip+Hm3qBhCgEUlJSmDt3LgsXLkSn09GoUSP0ej1mZnIZk8hMirh4pv87eoUPfjlqWP66WwMp4EIY2enTpxk2bBgXLlzAzc2NJUuW4OXlpXYskU9JERdPNX13KJ/tPG1Y/r/uDRlYv5KKiYQo+O7evUurVq1ISUnh3XffZcqUKTg6OqodS+RjUsTFU8354xwA9lYWxM7oiUajUTmREAWXTqfD3NycUqVKERAQQLVq1WjatKnasYQJkCIuMknX6am/cDsJKelULunEhfF+akcSosBKS0tj/vz5HDx4kM2bN2Nubs7QoUPVjiVMiBRxAUDQkct8sfsMEbFJhnUNypVQMZEQBdvZs2cZPnw4Z86c4ZVXXiE8PJzy5curHUuYGLnUUdB55T6Grj+WqYBP9n2Tlb1kIAkhcltaWhpz5syhefPmnDlzhr59+3L48GEp4OKlSE+8EFMUBfvxa0lJ1wNQv2wJ/hzZSs5/C2FEvXv3Zs+ePbi6urJw4UJ8fX3VjiRMmBTxQupURDR1FgQblr/qWp/3G1ZWMZEQhUOfPn0oVaoUM2bMwNnZWe04wsTJ4fRC5lRENJ1X7stUwNf2bSIFXAgjOX/+PD179iQmJgaAjh07snTpUingIldIT7wQuR2blKl4A0RO7UZJBxuVEglRcKWnpxMYGMisWbNITU1l8+bNvPPOO2rHEgWMFPFCpOXXvxse7xnqS+MKpbAwl4MxQuS2S5cuMXz4cE6ePImLiwsLFiygVatWascSBZAU8UIgXafn6yOXOa+NBeD02HZ4uBZVOZUQBdMvv/zCyJEjSUlJoXv37sycOZOiReX3TRiHFPFCwG3ar9xNeGhYlgIuhPG88cYblCxZklmzZtGmTRu144gCTop4AafXK4YCPtn3TcY0fV3lREIULDqdjmXLltGsWTPeeOMNXn/9dUJCQrC0tFQ7migEpIgXYGfvxFBj7jYAzDQaprSqoXIiIQqWq1ev4u/vz/Hjx2nWrBm//vorgBRwkWfkqqYCau+VO4YCDrCkcz0V0whRsOh0Or766iu8vLw4fvw4nTp1IigoSO1YohDKsiceFhbGhAkT0Gq17N27l3PnzrF3715GjBiRF/nESwgJv4/v8v9diR43oyf21tIzECI3RERE8N5773Hs2DGKFy/OsmXL8POTiYKEOrLsiU+ZMoWhQ4ca5rStVq0aO3fuNHow8XIURaHewu2G5fS5faWAC5GLHBwc+Oeff/Dz8+PIkSNSwIWqsizi8fHxeHl5GcbTNjMzk/M9+VRqug6LsT8YlhNn9ZZx0IXIBTdu3GDfvn0AFClShL1797Jy5UpKlJCZ/oS6sizi5ubmpKWlGYqBVqvFzExOpecnqek6Bq07jO0naw3rlndrgI2luYqphDB9er2eoKAgmjRpwqBBg3jw4AEALi4u6gYT4pEsz4n37t0bf39/YmJiWLJkCZs2bWL06NF5kU1k4YI2lk7f/sGVe/GZ1m8a2JT2b7iplEqIgiEsLIwRI0bw559/UrRoUWbPni3jnYt8J8si3rFjR8qUKcMff/xBcnIyX375JXXq1MmLbOI5/rxxF6/A3zKt+7pbAwbUfQ1LGUpViJem1+tZuXIlU6ZMITExkTZt2jBv3jzpfYt8KcsivnnzZvz8/DIV7n/XCXUoipKpgJ/5uD2vly6iXiAhChBFUVi/fj2WlpZ8/fXXdO3aVa4tEflWll22VatWZWudyDtTfjtteKyd2k0KuBA5pCgKJ06cADKuA/r666/5888/6datmxRwka89syd+5swZQkNDiYmJYc2aNYb1CQkJpKWl5Uk48SRFUZi++wwAn/i8QQmZRlSIHAkPD2fkyJEcPHiQ3377jdq1a1O2bFm1YwmRLc8s4lqtlrNnz5KcnMzZs2cN6+3t7Zk5c2aehBOZ3Ut4iMtnvxiWv2hTS8U0Qpg2RVH4/vvvmTx5MgkJCfj6+uLq6qp2LCFeyDOLeIsWLWjRogWHDh2icePGeZlJPEVMUkqmAh78no8c5hPiJUVERDBq1Cj++OMPnJycCAwMpFevXvI7JUxOlhe2NW7cmOvXr3Px4kVSU1MN6zt27GjMXOI/pu0KNTy+NMGPSiWcVEwjhGlbunQpf/zxB82bN2fhwoW8+uqrakcS4qVkWcS///57fvrpJ6KioqhevTonTpygbt26UsTz2NdHLgPwU38vKeBCvISoqChKlCiBRqNh4sSJ1KpVSy5cEyYvy6vTf/75Z3755RdcXV355ptv+OWXX7C3t8+LbOIR32W7SUnXA9D2dekxCPEiFEVh7dq11K1b13CRrqOjI927d5cCLkxelj1xKysr7Ozs0Ov1KIpC5cqVCQsLy4NoAqDLqn3svRoJwCivqthayhTwQmTX7du3+eijj9i1axcODg5YWMjvjyhYsvxE29rakpaWRtWqVZkzZw6urq7o9fq8yFbobT4bzqYz4QB0rO7GfL+6KicSwjQoisJPP/3EhAkTiI2NxdvbmyVLllCmTBm1owmRq7I8nP7ZZ5+RlpbG+PHjiY2N5a+//mL27NnZ2vmBAwdo2bIlvr6+BAUFPXWb7du306ZNG9q2bcuYMWNeLH0BpY1PxnzMajqv3AeAewlHfn2nqaqZhDAlu3fvZtiwYeh0OubPn8+GDRukgIsC6bk9cZ1Ox86dOxk5ciR2dnZ88cUX2d6xTqdj2rRprFy5EhcXF7p27YqPjw+VKlUybBMWFkZQUBA//vgjzs7O3L9//+V/kgLiSlQcVWdtNixXdy3CAf+WKiYSwjQoikJ6ejoAvr6+jB49mgEDBsjALaJAe25P3NzcnAMHDrzUjkNDQylXrhxubm5YWVnRtm1b9uzZk2mbn3/+mT59+hhmBipevPhLvVdB4rt8t+Hx+U868PfY9jjZWKmYSIj8T6vV0q9fP77++msANBoNkydPlgIuCrwsz4k3bdqUb775ho4dO2JnZ2dYb2tr+9zXabVaSpcubVh2cXEhNDQ00zb/XiDXs2dP9Ho9/v7+eHl5ZRn68RHkckNISEiu7u9lLTmlJfxBEgDbO7mTEH6VkHCVQ72A/NKOpkza8MUoisIff/xBYGAg8fHx1KxZk2PHjskFbDkkn8Ocy6s2zPKTHhgYCMCcOXPQaDQoioJGo+HChQs5fnOdTsfNmzdZvXo1kZGR9O3bl61bt+Lk9Pz7oD08PLC2ts7x+0NGQ9euXTtX9pUTe6/cYfWF8wC0e70MLRs3UDnRi8kv7WjKpA1fzN27dxk7dizbtm3Dzs6O2bNnU6NGDerWlQtAc0I+hzmXm22YkpLy3I5rlkX84sWLL/XGLi4uREZGGpa1Wu0T8/G6uLhQo0YNLC0tcXNzo3z58oSFhfHmm2++1HuaMt/lvwOg0cCmgU3VDSNEPhcXF0eTJk2IiorC09OTJUuWUKFCBelBikIny6vTX1b16tUJCwsjPDyc1NRUgoOD8fHxybRNixYtOH78OADR0dGEhYXh5uZmrEj51u5Ltw2PU2f3lQEohMiCk5MT/fv3Z+bMmWzZsoUKFSqoHUkIVRjtxJGFhQUBAQEMHjwYnU5Hly5dcHd3Z9GiRXh4eNC8eXOaNGnCn3/+SZs2bTA3N2fcuHEULVrUWJHyrXd+PJzx/7qvYWYmBVyIp9m0aRObNm3i22+/xczMjE8//VTtSEKozqhXf3h7e+Pt7Z1p3ahRowyPNRoNEyZMYMKECcaMka9tv3CLyPhkAGa2lalFhfiv+/fv8/HHH7Np0yZsbGw4d+4c1atXVzuWEPmC0Q6ni+z5bOffANQvW4JSjs+/4l+Iwmbr1q14enqyadMm6tWrx4EDB6SAC/GYLIv4/fv3GTt2LH369AEyLnT78ccfjR6sMDh84y4nI6IB2C8DugiRydixYxkwYADx8fF8/vnnBAcHZxosSgiRjSI+adIkateuTVxcHAAVK1Zk7dq1Rg9WGHyw/qjhsaW5HBQR4nF169alTp067N+/n+HDh2Nubq52JCHynSwrh1arpVevXoZfICsrK8zMpODk1K3YJM5FxgJw7/PuKqcRQn0xMTGMHz/e0GHo3r07O3bswN3dXeVkQuRfWVbj/458FBcXh6IoRgtUWPT54aDhcVG73Bm4RghTtWPHDjw9PQkKCjJMlqTRaKT3LUQWsrw63dfXl4CAABITE9mwYQNr166lS5cueZGtwErX6Tl4/S4AIR+1VTmNEOp58OABEyZM4KeffsLKyoqAgAD8/f3VjiWEyciyiL/33nts2bKFuLg49u/fT79+/fDz88uLbAVWm//730QwNV8tpmISIdRz8OBB3n//fSIjI6lVqxaBgYFUq1ZN7VhCmJQsi/itW7fo0KEDHTp0yIs8BV5Kuo49VzKGo905pLnKaYRQj5OTE3FxcUyaNImRI0fKpCVCvIQsf2t69uzJa6+9RufOnWnZsmWuTTxSWNWYs9Xw2LfKKyomESLv7d69m7Jly1KlShVq1KhBaGgoxYrJ0SghXlaWF7bt27ePAQMG8Pvvv+Pt7c3kyZM5depUXmQrkK7ciwdgaZf6KicRIu/ExcUxYsQIevTowejRow0Xx0oBFyJnsizi5ubmNGvWjMWLF7Nz5040Gg29e/fOi2wFzsmI+4bHH3hWVjGJEHlnz549eHp6smbNGqpXr26Y1lgIkXPZOgn14MEDtm3bxsaNG0lISGDkyJHGzlUgbT4bDkBz99IqJxHC+OLi4pg8eTKrV6/GwsKC8ePHM3r0aCwtLdWOJkSBkWUR9/f3JyQkhBYtWjBx4kSZLP4lKYrCnD/OAdCvzmsqpxHC+NLS0ti5cyceHh4sXbpUxjwXwgiyLOJvv/02c+fOxcbGJi/yFFgtv/6dlHQ9AH4eZVROI4RxxMfHc+3aNWrWrEnx4sXZtGkTr732GlZWVmpHE6JAeuY58dTUVCBjsBdFUUhOTs70n8g+RVEMt5V937sRTjbyB00UPPv376dx48b06NGD6OiMiX2qVasmBVwII3pmT7xHjx5s3LiRWrVqodFoUBQl0/8vXLiQlzlN2q3YJMPjPrUrqphEiNyXkJDA1KlT+eabbzA3N+fDDz/EwcFB7VhCFArPLOIbN24EMqYeFTkTeucBAC0qu6obRIhcdujQIUaMGMHNmzepWrUqS5cupVatWmrHEqLQyPIWsy+++CJb68Sz/fXPPUCuShcFi6IoTJkyhfDwcEaPHs0ff/whBVyIPJblhW0nTpx4Yt1ff/1llDAF1ddHLgPg6mSnchIhcu7WrVu8+uqraDQali5dSkJCgty1IoRKnlnEd+zYwY4dO7h16xajRo0yrE9ISJAr1V9A/MM0tPEPAWhVVYZZFaYrMTGRzz//nFWrVrF7926qV69OlSpV1I4lRKH2zCJeoUIFmjZtypkzZ2jatKlhvYODAw0bNsyLbAVC66D/zVhW0kG+/AjTdOTIEfz9/blx4wbu7u7o9Xq1IwkheE4Rr1q1KlWrVsXHx4ciRYrkYaSCI12n58jNKAAO+rdUOY0QLy4pKYnp06fz9ddfAzBixAjGjx+Pra2tysmEEPCcIv7dd98xYMAAgoKCnvr8uHHjjBaqoPho8/+uJ/CsUErFJEK8nDlz5rB8+XIqVapEYGAg9erVUzuSEOIxzyzi/045amcnF2O9rFRdxiHHz95+U+UkQmRfSkoKVlZWaDQaPvzwQ6ysrPjwww+l9y1EPvTMIt6zZ08gY+x08XL2X9MC8E69SionESJ7jh8/jr+/P+PGjaNr1644OzszYcIEtWMJIZ4hy/vEV65cSXx8xhzYH3/8Ma1ateLQoUNGD2bqHqbpuBwVB4BbETmaIfK3hw8f8tlnn9GmTRuuXbvGlStX1I4khMiGLIv4hg0bcHR05OjRo0RHRzNjxgzmz5+fF9lMmu/y3YbHMneyyM9OnDiBt7c3S5YsoVy5cmzbtk1630KYiCwHezE3Nwfg2LFjtG/fnrfeegtFUYwezJTp9QqHwzKuSl/Vq5HKaYR4tqNHj9KuXTv0ej1Dhgxh8uTJ2Nvbqx1LCJFNWRZxGxsbgoKCCA4OZs2aNSiKQlpaWl5kM1lDfjlieNyvjkx4IvKvevXq0b17d/r06UOjRvKFUwhTk+Xh9JkzZxIVFcXYsWMpWbIk4eHhtG/fPi+ymaSw6ARWHr8GwLwOMhSlyF9SUlL4/PPP+fzzzwEwMzPjq6++kgIuhInKsideoUIFPv30U5KSkkhKSqJs2bK8//77eZHNJH3wy1HD4w+9X1cxiRCZnTp1iuHDh3Px4kUqVKjAmDFj5BZSIUxclj3xf/75h+7du1O/fn0aNGhAz549CQ8Pz4tsJumPq5EAnBrTTuUkQmRISUnhiy++4O233+bixYu8++677Nu3Twq4EAVAlj3xzz77jO7du9OlSxcg42r1gIAAVq5cafRwpkZRFNL1GRf9vflKUZXTCAHJycm8/fbbnDt3jjJlyrB48eJMcyEIIUxblj3x6OhounbtikajQaPR0KVLF6Kjo/Mim8m5GZOodgQhMrG1tcXT05MBAwZw6NAhKeBCFDBZFnEzMzOuX79uWL5x44bhtjORWcdv/wBkylGhrrNnzzJ+/HjDraAzZ85kwYIFODk5qZxMCJHbsjycPnr0aPr06UO1atVQFIVLly4xe/bsvMhmcq7fTwBgZJNqKicRhVFaWhrz589n3rx5pKen0759exo1aoSZWZbf1YUQJirLIu7l5UVwcDCnT58GoEaNGhQrVszowUxRYmo6AM3dS6ucRBQ2586dY/jw4YSGhuLq6sqiRYvktjEhCgH5ip5Lkh4VcAALc2lWkXeWLl2Kj48PoaGh9O7dm8OHD9OiRQu1Ywkh8kCW1WbXrl20bt2aH374gdWrV9O2bVt+//33vMhmUipM3wDAq85y247IWw4ODpQoUYKffvqJwMBAnJ2d1Y4khMgjWR5OX7BgAevWraNChQoAhIWFMXToUPmm/5hbsUncS0wBYGHHuiqnEQVdeno63377Lb1798bBwYH+/fvTqVMnuXBNiEIoy564tbW1oYADlC9fHhsbG6OGMjXjt500PO78ZlkVk4iC7uLFi7Rq1Yrx48czb948IGOWPCngQhROWRbx5s2bs2zZMqKiorh79y7Lly+nefPmPHz4kOTk5LzImO+tPXkDgK+7NVA5iSio0tPTWbRoEU2bNuXkyZP06NGDUaNGqR1LCKGyLA+nL126FIBFixZlWh8YGIhGo+HChQvGSWYiHqbpDI8HN3BXMYkoqK5cucLQoUM5efIkLi4uzJ8/n9atW6sdSwiRD2RZxC9evJgXOUzWmTsxANQvW0LlJKKgevDgAX///TfdunVj1qxZFC0qQ/oKITJkWcTF853XxgLgVtRe5SSiILly5QqWlpaUL1+eunXrcujQIapWrap2LCFEPmPUG5oPHDhAy5Yt8fX1JSgo6Jnb/fbbb1SpUoUzZ84YM45RHA67C0hPXOQOnU5HYGAg3t7eDB8+HL1eDyAFXAjxVEYr4jqdjmnTprFixQqCg4PZtm0bV69efWK7hIQEvv/+e2rUqGGsKEZ1Jy7j4j6ZtUzkVEREBG3btiUgIAAHBwfef/99GTJVCPFcRvsLERoaSrly5XBzc8PKyoq2bduyZ8+eJ7ZbtGgR7733HtbW1saKYjT3Eh4SfP4WAB6li6gbRpgsnU7HV199xfvvv8/x48fp2LEjhw8fpkOHDmpHE0Lkc1meE79//z4zZ87kzp07rFmzhosXL3Lq1Cl69er13NdptVpKl/7fGOIuLi6EhoZm2ubcuXNERkbStGlTvvnmm2yHPnv2bLa3zY6QkJCXet135+4ZHt+6cp5buRXIRL1sOxZ2MTExfPnll9ja2jJu3Di8vb25efMmN2/eVDuaSZLPYc5JG+ZcXrVhlkV80qRJeHl5sXbtWgAqVqzIxx9/nGURz4per2fWrFnMnDnzhV/r4eGRaz33kJAQateu/VKvfX9/MACbBjal9htuuZLHVOWkHQsjvV5PREQEZctmDA60du1akpOTZSTEHJLPYc5JG+ZcbrZhSkrKczuuWR5O12q19OrVyzCHuJWVVbbO07m4uBAZGZlpPy4uLoblxMRELl++TP/+/fHx8eHvv/9m6NChJnVxW1h0xtSjTSq6ZLGlEP8TFhZGhw4daN26NbGxGXc3NGrUSG4dE0K8sCyrsYVF5s56XFwciqJkuePq1asTFhZGeHg4qampBAcH4+PjY3je0dGRY8eOsXfvXvbu3UvNmjVZtmwZ1atXf4kfQx0xyakAFLG1UjmJMAV6vZ4VK1bQuHFjDh8+TO3atUlPT8/6hUII8QxZHk739fUlICCAxMRENmzYwNq1a+nSpUvWO7awICAggMGDB6PT6ejSpQvu7u4sWrQIDw8Pmjdvnis/gFruxCWpHUGYkJs3bzJixAgOHTpE0aJFWbhwIV26dEGj0agdTQhhwrIs4u+99x5btmwhLi6O/fv3069fP/z8/LK1c29vb7y9vTOte9Z4z6tXr87WPvOL7/66BkCrqq+onESYgg8++IBjx47Rpk0b5s2bl+nUkhBCvKxsjdjWoUMHud3lP2KSMg6lt6n2qspJRH6VlJSEnV3G/PKzZ8/mwoULdOvWTXrfQohck2URHzly5FP/6Px3QpTCRFEU5u47D8hFbeJJiqLw3Xff8cUXX7B161aqVq1K9erVTep6DyGEaciyiDdr1szwOCUlhd9++43XXnvNqKHyu9tx/5uCVQZ5EY+LiIhg5MiR7Nu3DycnJ/755x8ZMlUIYTRZFvFOnTplWu7cuTODBg0yWiBTcPC6FoAub5bFzEwOjYqM3vf333/P5MmTSUhIwNfXlwULFvDKK3LNhBDCeF542FWNRoNWqzVGFpMxfXfGvexuRWTmMpFhyZIljB49Go1Gw5IlS1i3bp0UcCGE0b3QOXFFUbh06RKenp5GD5ZfJaSkceHR9KMjm8hh0sJMURTD70bfvn05d+4ckydPpkyZMionE0IUFi90Ttzc3JxBgwaZ7IxjuSHoyBXD43LFHFRMItR0+/ZtRo8eTZ8+fejQoQPFihXj66+/VjuWEKKQeW4R1+l0hISEMH369LzKk++tP50xKcWyrvVVTiLUoCgKP/74IxMnTiQuLo4iRYrI7ZdCCNU8t4ibm5tz6dKlvMpiEkIi7gPQ560KKicRee3OnTuMHj2aXbt24eDgwIIFC+jfv7/asYQQhViWh9MbNGjAtGnT6Nixo2HgCoBKlSoZNVh+ZWtpQXxKGvbWlmpHEXno/PnztG3bltjYWLy9vVm8eDFuboV75johhPqeWcQnTpzIjBkzCA7OmG5z3759huc0Gg179uwxerj8RqfXE5+SxltliqkdReSxypUrU7NmTTp06MA777wjo64JIfKFZxbxCxcuALB37948C5PfzXs0Stv1+wkqJxHGpigK69ev586dO4wcORILCws2bNggxVsIka9ka+x0kWHV8YxJTwLeflPlJMKY7t69y5gxYwgODsbJyYkBAwbg7OwsBVwIke88s4hfvnyZhg0bPrH+33tjjxw5YtRg+Y2iKFyKigPgnbqFe9jZgkpRFDZs2MAnn3xCdHQ0jRo1YsmSJTg7O6sdTQghnuqZRbx8+fIEBQXlZZZ8bemh/12l72xrpWISYQzp6ekMHjyYLVu2YGdnx5dffsmgQYMwM3vhQQ2FECLPPLOIW1lZ8eqrMs3mvwIPXQRgTNPXVU4ijMHCwgInJycaNmxIYGAgFSrILYRCiPzvmUXc0lJuoXrclXvxAExrVVPdICLX3Lt3j7Vr1zJixAg0Gg1ffvkl1tbW0vsWQpiMZxbxn3/+OS9z5GuKohge21iaq5hE5JbNmzfz8ccfc+/ePapUqULLli2xtbVVO5YQQrwQuTo9G2KSUwGo9arcH27q7t+/z7hx49i4cSM2NjZMnz6dFi1aqB1LCCFeihTxbDgVEQ2Ag7U0lynbsWMHH374IVFRUdStW5fAwEDc3d3VjiWEEC9NqlI2HA6LAsCzfEmVk4iciIiIIC4ujqlTpzJs2DDMzeXUiBDCtEkRz4Zdl24D0KhCKZWTiBe1e/duGjdujK2tLYMGDcLX15fy5curHUsIIXKFXIabDWcjHwDQUHriJiMmJob333+fHj16MHPmTADMzMykgAshChTpiWdD3MM0AIrZWaucRGTHzp07GT16NFqtlrfeeovevXurHUkIIYxCeuLZZG8l33fyuwcPHjBs2DB69+5NTEwMAQEB7Ny5k6pVq6odTQghjEIqUxbCojNmLHvTtajKSURWLly4wLp166hZsyZLly6lWrVqakcSQgijkiKehe/+ypi5LFWnUzmJeJrY2FiSk5MpXbo0DRs2ZP369Xh5eWFhIR9tIUTBJ4fTs3BeGwvAAJm5LN/ZvXs3np6eDB061DCqno+PjxRwIUShIUU8CzfuZ4yZ3qaaTAaTX8TFxTFixAh69OhBVFQUnp6e6ORIiRCiEJIuy3M8TNMR8mi0tled7VROIwD27t3LyJEjuX37NtWrV2fp0qV4eHioHUsIIVQhRfw5fL7aZXhsZSGje6ktNjaWgQMHkpSUxLhx4/joo4+wspK53YUQhZcU8ecoV8yeY//cY+O7TdWOUqjFxcXh5OSEs7MzX331FWXKlOHNN99UO5YQQqhOzok/x/F/7gEyZrpa4uPj+eijj/Dy8iI+/tG1CW3aSAEXQohHpIg/Q3RSCmHRiQA4WFuqnKbwOXDgAI0bN2bVqlXY29tz7949tSMJIUS+I0X8GXZezJj0pIitFTaWcj48ryQkJPDxxx/TsWNHbt++zZgxY9i7dy8VKlRQO5oQQuQ7ck78GX4NvQnA+w1lvum8NGjQIHbv3k2VKlVYunQpb731ltqRhBAi35Ii/gw37mcMtzqovhRxY1MUBY1GA8C4ceN44403GDduHDY2NionE0KI/E0Opz9DTHIqAGWKyP3hxnT48GGaNGnCtWsZw9vWrl2bgIAAKeBCCJENUsSfQlEU/olJpIitFdZyf7hRJCUlMWHCBNq3b8/Fixc5ePCg2pGEEMLkyOH0p7gcFQdkjNgmct/Ro0fx9/fn+vXruLu7ExgYSN26ddWOJYQQJkd64k8R/iAJgDavy3jpuW3dunW0bduWGzdu4O/vz759+6SACyHES5Ke+FMcvK4FoJ5bCZWTFDw+Pj7UrVuXqVOnUr9+fbXjCCGESZOe+FP8O/2os60M8pJTycnJTJ48mV27MsahL1WqFDt37pQCLoQQuUB64k/x962Mmcu61yyvbhAT99dff+Hv78+VK1f4+++/efvtt9WOJIQQBYr0xP9DURSuP7pH3EmGW30pDx8+ZMqUKbRu3ZqrV6/y/vvv89NPP6kdSwghChyj9sQPHDjAF198gV6vp1u3bgwZMiTT8ytXruSXX37B3NycYsWKMWPGDF59Vd2LyZIfXZFuptFgZqZRNYspioiIoGvXrly+fJny5csTGBiIp6en2rGEEKJAMlpPXKfTMW3aNFasWEFwcDDbtm3j6tWrmbapVq0av/76K1u3bqVly5bMmTPHWHGy7eq9jNmy2smV6S/FxcUFW1tbhgwZwsGDB6WACyGEERmtiIeGhlKuXDnc3NywsrKibdu27NmzJ9M2DRo0wNbWFoCaNWsSGRlprDjZdiL8PoBhGFCRtZMnT7Jz504ALC0t2bFjB7NmzcLe3l7lZEIIUbAZrYhrtVpKly5tWHZxcUGr1T5z+/Xr1+Pl5WWsONmWptcD0LCczCGelZSUFKZPn07Lli1ZvHix4UuYDJkqhBB5I19cnb5582bOnj3LDz/8kK3tz549m6vvHxISYnj86dZLAJjF3SUk5GGuvk9BcuXKFWbPnk1YWBguLi6MGTOGW7ducevWLbWjmbTHP4vi5Ugb5py0Yc7lVRsarYi7uLhkOjyu1WpxcXF5YrvDhw+zfPlyfvjhB6ysrLK1bw8PD6ytrXMlZ0hICLVr1zYs1zkRw+7Ld+jfoiElHaRH+V+KojBz5kwWLFiATqfj3XffZcqUKVy+fDlTO4oX99/Ponhx0oY5J22Yc7nZhikpKc/tuBrtcHr16tUJCwsjPDyc1NRUgoOD8fHxybTN+fPnCQgIYNmyZRQvXtxYUV7I7st30GighH3ufEkoaDQaDTdv3sTV1ZUNGzYwb948HB0d1Y4lhBCFktF64hYWFgQEBDB48GB0Oh1dunTB3d2dRYsW4eHhQfPmzZk9ezZJSUmMGjUKAFdXV5YvX26sSFlK12WcD1cUubDtcf9+CevUqRMAs2fPRqPR4OTkpHIyIYQo3Ix6Ttzb2xtvb+9M6/4t2ACrVq0y5tu/sNtxyQCUdJBe+L/Onj3L8OHDOXPmDBYWFrRv3x5nZ2e1YwkhhEBGbMvkXOQDANq/7qZukHwgLS2NOXPm4OPjw5kzZ+jbt+8TX8iEEEKoK19cnZ5fnIy4r3aEfOH8+fMMHz6c06dP4+rqysKFC/H19VU7lhBCiP+Qnvhjzj7qiTet9ORV9IXJnj17OH36NL169eLw4cNSwIUQIp+SnvhjopNSAajxSlGVk+S9S5cuUb58eaytrRk2bBhvvfUWjRo1UjuWEEKI55Ce+GN+v3wHAGeb7N2vXhCkp6ezYMECvL29DWPXm5ubSwEXQggTID3xx1R3LcKZOw9wK1o4xvy+ePEi/v7+nDx5EhcXF+rUqaN2JCGEEC9AeuKPSU3X4+JY8EdpS09PZ9GiRTRt2pSTJ0/SvXt3Dh8+TKtWrdSOJoQQ4gVIT/wxl6LicCtip3YMoztx4gRTp06lVKlSzJ8/nzZt2qgdSQghxEuQIv4fkfEFc9ITnU5HYmIiTk5ONGjQgCVLltC6dWuKFSumdjQhhBAvSQ6nPxIekwhAtVIFbzSyq1ev0qZNG4YMGYKiKAD06dNHCrgQQpg4KeKPrDx+FShYE5/odDq++uorvLy8+Ouvv7C3t+fhw4J5pEEIIQojOZz+SFRiCgANypdQOUnuuHbtGv7+/hw7dowSJUqwfPlyOnTooHYsIYQQuUiK+CP/d/QKAG9XeUXlJDmXnJxM69atuXfvHn5+fsyZM4cSJQrGlxMhhBD/I0X8kbRH05A2Kl9K5SQvT6fTYW5ujq2tLVOnTsXGxsYwfagQQoiCR86JP8bS3AwzM9ObR1yv1xMUFESzZs1ISkoCoFevXlLAhRCigJMiDkQ+mkf8FSdblZO8uLCwMPz8/Bg/fjy3bt3i0qVLakcSQgiRR6SIA0dvRgHweuki6gZ5AXq9nhUrVtCkSRP+/PNP2rZty+HDh6lVq5ba0YQQQuQROScOaBMybruq9arpzF724Ycf8sMPP1CkSBGCgoLo0qULGo3pnQoQQgjx8qSIA1vPRQDwWnEnlZNkX+/evYmJiWHOnDmULl1a7ThCCCFUIIfTAWcbSwDcSzqqnOTZwsPD6du3Lzdv3gSgQYMGrF69Wgq4EEIUYlLEAfNHV6S7Fcl/U5AqisKqVato1KgR27dvZ82aNWpHEkIIkU/I4XTg71vRANhY5K/vNBEREYwcOZJ9+/bh5OREYGAgvXr1UjuWEEKIfCJ/VS2VuDplTD/q+Oiwen6wZ88ePD092bdvHy1atODw4cP07t1bLl4TQghhID1xIDVdB4CNhbnKSf6nWrVqFC1alJkzZ0rxFkII8VRSxIHQOw+wNDdTtVAqisLatWtxc3PDy8uLV155hRMnTmBpmX+ODgghhMhfpIiT0QN/kJyq2vvfvn2b0aNHs3v3bqpVq8ahQ4fQaDRSwIUQQjxXoT8nrigKkfHJlC2a91emK4rCjz/+iKenJ7t378bb25uffvpJDp0LIYTIlkLfE9cpGf+PzeOeeExMDMOGDeO3337DwcGB+fPnM2DAACngQgghsq3QF/F0fUYVr1+uZJ6+r52dHTdv3sTLy4vFixdTtmzZPH1/IYQQpk+KuJJRxC3Njd8D1mq1nDx5ktatW2Ntbc2mTZsoUaIEZmaF/qyGEEKIl1Doq8e/PXFLc+M1haIorF+/Hk9PTwYNGmQYOrVUqVJSwIUQQry0Qt8TT0jVA5CSrjfK/u/evcvYsWPZtm0bdnZ2TJ06FTc3N6O8lxBCiMKl0Bfxh7qM4m2dy0OuKorCxo0bGTduHNHR0Xh6erJkyRIqVKiQq+8jhBCi8Cr0RTz10eXpFYvl/gxmW7ZsITk5mZkzZ/Lee+/JoXMhhBC5qtAX8XvJ6QDolNw5nB4SEkLt2rXRaDTMnTuXuLg4KlasmCv7FkIIIR5X6LuGP1/OmMGsiI1VjvZz7949Bg4ciK+vLzt37gSgRIkSUsCFEEIYTaHvicc8zOiJl8nBXOJbtmxh7Nix3Lt3j3r16lGpUqXciieEEEI8U6Ev4smPrkrvUbPcC782OjqacePGsWHDBmxsbPj888/54IMPMDfPP7OhCSGEKLgKfRGPSEgDwN76xScb+eGHH9iwYQN16tRh6dKluLu753Y8IYQQ4pkKfRG3NNOQ9mjAl+yIiYnBwcEBS0tLhg4dSvHixenZs6f0voUQQuS5Qn1hm16vkKZX8KpYKlvb79ixA09PTxYuXAiApaUlffr0kQIuhBBCFYW6iCelZVzUZmP5/AMSMTExfPDBB/Tp04eYmBjs7OzyIp4QQgjxXIX6cHp8Ssb58Mi45Gdu89tvvzF69GgiIyN56623CAwMpGrVqnkVUQghhHimQl3E/x0vvcarRZ/6/JkzZ+jVqxeWlpZMnjyZESNGYGFRqJtMCCFEPlKoK1JMUirw5LjpaWlpWFpaUr16dT799FNat27N66+/rkZEIYQQ4pkKdRG/dj8egKRUHQCxsbF8+umnxMfHs2rVKjQaDWPGjFEzohBCCPFMRr2w7cCBA7Rs2RJfX1+CgoKeeD41NZUPP/wQX19funXrRkREhDHjPOFKVBwA1Vyc+f333/H09GTt2rXcvHmT+Pj4PM0ihBBCvCijFXGdTse0adNYsWIFwcHBbNu2jatXr2ba5pdffsHJyYndu3fzzjvvMHfuXGPFeSor84wff9/WDXTv3p2oqCjGjx/P7t27cXJyytMsQgghxIsyWhEPDQ2lXLlyuLm5YWVlRdu2bdmzZ0+mbfbu3UunTp0AaNmyJUeOHEFRsj/wSk4lpGScEz+yZyceHh7s2bOHcePGYWn54qO3CSGEEHnNaOfEtVotpUuXNiy7uLgQGhr6xDaurq4ZQSwscHR0JCYmhmLFij1332fPns2VjBYJsVigp2vzxrzfqwupqamEhITkyr4LI2m7nJM2zDlpw5yTNsy5vGpDk7ywzcPDA2tr6xzvp3ZtaFHuBPXrDsiFVIXbv/Ooi5cnbZhz0oY5J22Yc7nZhikpKc/tuBrtcLqLiwuRkZGGZa1Wi4uLyxPb3LlzB4D09HTi4+MpWvTp92wbi4WZJk/fTwghhMgtRivi1atXJywsjPDwcFJTUwkODsbHxyfTNj4+PmzcuBHIGBmtQYMGaDRSVIUQQojsMNrhdAsLCwICAhg8eDA6nY4uXbrg7u7OokWL8PDwoHnz5nTt2pWPP/4YX19fnJ2dWbBggbHiCCGEEAWOUc+Je3t74+3tnWndqFGjDI+tra1ZvHixMSMIIYQQBVahnsVMCCGEMGVSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFEmNXb6vzOcpaam5up+U1JScnV/hZW0Y85JG+actGHOSRvmXG614b/17lkzfGqUvJz7M4fi4+O5fPmy2jGEEEKIPFW5cmUcHR2fWG9SRVyv15OYmIilpaWMsS6EEKLAUxSFtLQ07O3tMTN78gy4SRVxIYQQQvyPXNgmhBBCmCgp4kIIIYSJkiIuhBBCmCgp4kIIIYSJKjRF/MCBA7Rs2RJfX1+CgoKeeD41NZUPP/wQX19funXrRkREhAop87es2nDlypW0adOG9u3bM2DAAG7duqVCyvwtqzb812+//UaVKlU4c+ZMHqYzHdlpx+3bt9OmTRvatm3LmDFj8jhh/pdVG96+fZt+/frRsWNH2rdvz/79+1VImX9NmDCBhg0b0q5du6c+rygK06dPx9fXl/bt23Pu3DnjBFEKgfT0dKV58+bKP//8o6SkpCjt27dXrly5kmmbH374QZk8ebKiKIqybds2ZdSoUSokzb+y04ZHjhxRkpKSFEVRlDVr1kgb/kd22lBRFCU+Pl7p3bu30q1bNyU0NFSFpPlbdtrxxo0bip+fn/LgwQNFURTl3r17akTNt7LThpMmTVLWrFmjKIqiXLlyRWnWrJkaUfOt48ePK2fPnlXatm371Of37dunDBo0SNHr9cqpU6eUrl27GiVHoeiJh4aGUq5cOdzc3LCysqJt27bs2bMn0zZ79+6lU6dOALRs2ZIjR448c4Scwig7bdigQQNsbW0BqFmzJpGRkWpEzbey04YAixYt4r333sPa2lqFlPlfdtrx559/pk+fPjg7OwNQvHhxNaLmW9lpQ41GQ0JCApAx0FapUqXUiJpv1a1b1/D5epo9e/bQsWNHNBoNNWvWJC4ujrt37+Z6jkJRxLVaLaVLlzYsu7i4oNVqn9jG1dUVAAsLCxwdHYmJicnTnPlZdtrwcevXr8fLyysvopmM7LThuXPniIyMpGnTpnmcznRkpx3DwsK4ceMGPXv2pHv37hw4cCCvY+Zr2WlDf39/tm7dipeXF0OGDGHSpEl5HdOk/beNS5cu/dy/mS+rUBRxkbc2b97M2bNnGTx4sNpRTIper2fWrFl88sknakcxeTqdjps3b7J69WrmzZvH5MmTiYuLUzuWSQkODqZTp04cOHCAoKAgxo0bh16vVzuW+I9CUcRdXFwyHdrVarW4uLg8sc2dO3cASE9PJz4+nqJFi+ZpzvwsO20IcPjwYZYvX86yZcuwsrLKy4j5XlZtmJiYyOXLl+nfvz8+Pj78/fffDB06VC5u+4/s/j77+PhgaWmJm5sb5cuXJywsLI+T5l/ZacP169fTunVrAGrVqkVKSoocnXwB/23jyMjIp/7NzKlCUcSrV69OWFgY4eHhpKamEhwcjI+PT6ZtfHx82LhxI5BxZXCDBg1kfPbHZKcNz58/T0BAAMuWLZNzkE+RVRs6Ojpy7Ngx9u7dy969e6lZsybLli2jevXqKqbOf7LzWWzRogXHjx8HIDo6mrCwMNzc3NSImy9lpw1dXV05cuQIANeuXSMlJYVixYqpEdck+fj4sGnTJhRF4e+//8bR0dEo1xWY1FSkL8vCwoKAgAAGDx6MTqejS5cuuLu7s2jRIjw8PGjevDldu3bl448/xtfXF2dnZxYsWKB27HwlO204e/ZskpKSGDVqFJDxR2D58uUqJ88/stOGImvZaccmTZrw559/0qZNG8zNzRk3bpwcWXtMdtpw/PjxTJo0iVWrVqHRaJg1a5Z0bB7z0Ucfcfz4cWJiYvDy8mLEiBGkp6cD0KtXL7y9vdm/fz++vr7Y2toyY8YMo+SQCVCEEEIIE1UoDqcLIYQQBZEUcSGEEMJESREXQgghTJQUcSGEEMJESREXQgghTJQUcSHygI+PD61atcLPzw8/P78sbzfx8fHh8uXLeZQu+7RaLf369TMsL1myhNTUVMPyokWL2L59uxrRiIiI4KefflLlvYVQi9xiJkQe8PHxYfny5VSuXNko26ulSpUqnDx5Ent7e6O/V3p6OhYWzx7a4tixY3z55Zds2LDB6FmEyC8KxWAvQuRHW7du5fvvvyctLQ2ATz75hIYNGz6xXWBgINu2bcPa2hqNRsP333+Pk5MTp0+fZu7cuSQmJgIwcuTIp06cMn78eCwsLLh69SoxMTHUrVuXgIAArKysuHfvHp999hn//PMPAIMGDaJjx47o9XqmTZvG0aNHsbKyws7OjnXr1hEREUGXLl04duwYU6dOBaBnz56YmZmxevVqZsyYgYeHB126dKFp06bs2LHDMMrXl19+ib29Pf7+/i+U3dzcnBs3bpCYmMjmzZsZM2YMN27cIC0tjbJlyzJjxgycnZ2ZNm0aERER+Pn5Ua5cORYvXsz169eZMWMGMTExpKWlMWDAALp06ZLjfzsh8g2jTHAqhMikWbNmSsuWLZUOHTooHTp0UA4cOKBER0crer1eURRFuXbtmtKkSZNM21+6dEmJiYlRateurSQnJyuKkjHXeFpamhIbG6v4+fkpWq1WURRF0Wq1SpMmTZTY2Ngn3vuTTz5R2rVrpyQkJChpaWnKu+++q6xevVpRFEUZNWqUsmDBAsM+GjVqpFy6dEk5d+6c0qpVK0Wn0ymKohjm5Q4PD1fq1atn2HflypWVhISETO/1774nTpyofPfdd4qiKEpaWprSqFEjJTw8/IWzd+rUSUlMTDSsu3//vuHx/PnzlTlz5iiKoihHjx5VOnXqZHguLS1N6dSpk3L16lVD27399tuGZSEKAumJC5FHFi9enOnweGhoKGPGjEGr1WJhYcG9e/eIioqiZMmShm0cHR0pW7Ys48aNo3HjxjRt2hQHBwdOnTpFREQE7733nmFbjUbDzZs3nzrWeps2bQyHvDt27MiuXbvo27cvR44cYfz48QCUKlUKb29vjh07RseOHUlPT+fTTz+lfv36NGvW7IV/3k6dOvHFF1/Qv39/Dhw4QMWKFSlTpgz79+9/oeytWrXCzs7OsLx582a2bt1KWloaSUlJlC9f/qnvHxYWxrVr1/joo48M69LS0rh+/TqvvfbaC/88QuRHUsSFUMlHH33E+PHjadGiBXq9nho1apCSkpJpG3Nzc37++WdOnjzJ0aNH6dy5MytWrEBRFKpUqcKaNWuMks3R0ZHg4GCOHTvG4cOHmTt3rmGCoOyqU6cOiYmJXLp0iY0bN9K5c2eAF87+eAE/ceIEP/74I+vWraNYsWJs3bqVn3/++amvUxSFokWLsnnz5hfKLYQpkavThVBJfHw8ZcqUAeDXX3/NdJX3vxISEoiOjqZevXqMHDmSypUrc+XKFWrVqsXNmzc5evSoYdvQ0FCUZ1ynunPnTpKSkkhPT2fz5s00aNAAgIYNGxqKYFRUFPv376dBgwZER0eTnJxMkyZNGDt2LI6OjoSHhz+xX3t7exISEp75M3bs2JGVK1fy119/0bJlS4AXzv64uLg4HBwcKFKkCKmpqfz666+G5xwcHDJlqVChAjY2NmzatMmw7tq1a8/NK4SpkZ64ECqZMGECw4YNw9nZmSZNmlCkSJEntklISGDEiBE8fPgQRVF4/fXXefvtt7G2tuarr75izpw5zJgxg7S0NNzc3Fi+fPlTZ5qqXr06AwcONHwh6N69OwCTJk0iICCA9u3bAzB27Fjc3d05d+4ckydPJj09HZ1Oh5eXFzVr1uT27duZ9jtw4ED69++PjY0Nq1evfuJ9O3bsSPPmzencuTO2trYAODs7v1D2xzVp0oQtW7bQsmVLihYtSp06dQzzrVepUoUKFSrQrl07KlasyOLFi1m+fDkzZszgm2++Qa/XU7x4cRYuXJjlv40QpkJuMROigBs/fjweHh707dtX7ShCiFwmh9OFEEIIEyU9cSGEEMJESU9cCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggTJUVcCCGEMFFSxIUQQggT9f8L+cItbrnZJwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 6 | AUC: 0.8589874116604624\n",
      "Fold 6 | Avergae_precesion: 0.36463997205677456\n",
      "Fold 6 | recall: 0.17664670658682635\n",
      "Fold 6 | F1: 0.2673716012084592\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.99      0.97     13998\n",
      "           1       0.55      0.18      0.27      1002\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.75      0.58      0.62     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13853   145]\n",
      " [  825   177]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 7 | AUC: 0.8590422740745114\n",
      "Fold 7 | Avergae_precesion: 0.406825545484374\n",
      "Fold 7 | recall: 0.1966067864271457\n",
      "Fold 7 | F1: 0.29579579579579585\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.99      0.97     13998\n",
      "           1       0.60      0.20      0.30      1002\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.77      0.59      0.63     15000\n",
      "weighted avg       0.92      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13865   133]\n",
      " [  805   197]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 8 | AUC: 0.8739358687967685\n",
      "Fold 8 | Avergae_precesion: 0.44231066779313616\n",
      "Fold 8 | recall: 0.1936127744510978\n",
      "Fold 8 | F1: 0.2980030721966206\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.99      0.97     13998\n",
      "           1       0.65      0.19      0.30      1002\n",
      "\n",
      "    accuracy                           0.94     15000\n",
      "   macro avg       0.80      0.59      0.63     15000\n",
      "weighted avg       0.93      0.94      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13892   106]\n",
      " [  808   194]]\n",
      "[LightGBM] [Warning] min_data_in_leaf is set with min_samples_leaf=47, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=47\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 9 | AUC: 0.8638826789912103\n",
      "Fold 9 | Avergae_precesion: 0.3724137507426916\n",
      "Fold 9 | recall: 0.16766467065868262\n",
      "Fold 9 | F1: 0.2547384382107657\n",
      "Classification report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.99      0.97     13998\n",
      "           1       0.53      0.17      0.25      1002\n",
      "\n",
      "    accuracy                           0.93     15000\n",
      "   macro avg       0.74      0.58      0.61     15000\n",
      "weighted avg       0.92      0.93      0.92     15000\n",
      "\n",
      "Confusion_matrix:\n",
      " [[13849   149]\n",
      " [  834   168]]\n",
      "--------------------------------------------------------------------------------\n",
      "\n",
      "Overall mean AUC score : 0.8652970875741328\n",
      "\n",
      "Overall mean F1 score : 0.2891585202564346\n",
      "\n",
      "Overall mean average precision score : 0.401173467129701\n",
      "\n",
      "Overall mean recall score : 0.1912995544305208\n",
      "CPU times: user 1h 30min 52s, sys: 20.7 s, total: 1h 31min 13s\n",
      "Wall time: 23min 43s\n"
     ]
    }
   ],
   "source": [
    "%%time \n",
    "# evaluate each strategy on the dataset\n",
    "results = list()\n",
    "# Setting a 10-fold stratified cross-validation (note: shuffle=True)\n",
    "SEED = 42\n",
    "FOLDS = 10\n",
    "skf = StratifiedKFold(n_splits=FOLDS, shuffle=True, random_state=SEED)\n",
    "# CV interations\n",
    "# Create arrays for the features and the response variable\n",
    "roc_auc = list()\n",
    "average_precision = list()\n",
    "#X1=X.iloc[0:2000,:].copy()\n",
    "#y1=y[0:2000].copy()\n",
    "oof = np.empty((X.shape[0],))\n",
    "#oof_bin = np.empty((X.shape[0],))\n",
    "predictions=[]\n",
    "mean_auc = 0\n",
    "mean_ap=0\n",
    "F1 = list()\n",
    "Recall=list()\n",
    "best_iteration = list()\n",
    "for fold, (train_idx, test_idx) in enumerate(skf.split(X, y)):    \n",
    "    X_train, y_train = X.iloc[list(train_idx), :], y.iloc[list(train_idx)]\n",
    "    X_test, y_test = X.iloc[list(test_idx), :],y.iloc[list(test_idx)]\n",
    "    pipe1_model_lgbm_woe_powertransformer2.fit(X_train,y_train,\n",
    "                                                  #classifier__verbose=False\n",
    "                                                  #,verbose=False\n",
    "                                                 )\n",
    "    preds = pipe1_model_lgbm_woe_powertransformer2.predict_proba(X_test)[:,1]\n",
    "    oof[test_idx] = preds\n",
    "    auc_score= roc_auc_score(y_true=y_test, y_score=preds)\n",
    "    roc_auc.append(auc_score)\n",
    "    ap=average_precision_score(y_true=y_test, y_score=preds)\n",
    "    plot_roc(y_test,preds)\n",
    "    mean_auc += auc_score / FOLDS\n",
    "    mean_ap+=ap/FOLDS\n",
    "    predictions.append(pipe1_model_lgbm_woe_powertransformer2.predict_proba(X_test_final)[:,1]) \n",
    "    y_predicted = pipe1_model_lgbm_woe_powertransformer2.predict(X_test)\n",
    "    recall = recall_score(y_test, y_predicted)\n",
    "    f1= f1_score(y_test, y_predicted)\n",
    "    print(f\"Fold {fold} | AUC: {auc_score}\")\n",
    "    print(f\"Fold {fold} | Avergae_precesion: {ap}\")\n",
    "    print(f\"Fold {fold} | recall: {recall}\")\n",
    "    print(f\"Fold {fold} | F1: {f1}\")\n",
    "    print('Classification report:\\n',classification_report(y_test,y_predicted))\n",
    "    print('Confusion_matrix:\\n',confusion_matrix(y_test,y_predicted))\n",
    "    roc_auc.append(auc_score)\n",
    "    F1.append(f1)\n",
    "    Recall.append(recall)\n",
    "    #Mean of the predictions\n",
    "print('-' * 80)  \n",
    "print(f\"\\nOverall mean AUC score : {mean_auc}\")\n",
    "print(f\"\\nOverall mean F1 score : {np.mean(F1)}\")\n",
    "print(f\"\\nOverall mean average precision score : {mean_ap}\")\n",
    "print(f\"\\nOverall mean recall score : {np.mean(Recall)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e399842c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:11.280512Z",
     "iopub.status.busy": "2021-12-28T12:30:11.279916Z",
     "iopub.status.idle": "2021-12-28T12:30:11.348557Z",
     "shell.execute_reply": "2021-12-28T12:30:11.348035Z"
    },
    "papermill": {
     "duration": 0.133267,
     "end_time": "2021-12-28T12:30:11.348699",
     "exception": false,
     "start_time": "2021-12-28T12:30:11.215432",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Modified Stlying DataFrame:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_a9e83_row0_col0, #T_a9e83_row2_col1, #T_a9e83_row2_col3, #T_a9e83_row3_col2 {\n",
       "  background-color: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_a9e83_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >model</th>\n",
       "      <th class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th class=\"col_heading level0 col2\" >f1_score</th>\n",
       "      <th class=\"col_heading level0 col3\" >average_precision_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_a9e83_row0_col0\" class=\"data row0 col0\" >m1_hgbc</td>\n",
       "      <td id=\"T_a9e83_row0_col1\" class=\"data row0 col1\" >0.865434</td>\n",
       "      <td id=\"T_a9e83_row0_col2\" class=\"data row0 col2\" >0.279313</td>\n",
       "      <td id=\"T_a9e83_row0_col3\" class=\"data row0 col3\" >0.403261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_a9e83_row1_col0\" class=\"data row1 col0\" >m2_catboost</td>\n",
       "      <td id=\"T_a9e83_row1_col1\" class=\"data row1 col1\" >0.866293</td>\n",
       "      <td id=\"T_a9e83_row1_col2\" class=\"data row1 col2\" >0.289173</td>\n",
       "      <td id=\"T_a9e83_row1_col3\" class=\"data row1 col3\" >0.400756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_a9e83_row2_col0\" class=\"data row2 col0\" >m3_xgbr</td>\n",
       "      <td id=\"T_a9e83_row2_col1\" class=\"data row2 col1\" >0.855243</td>\n",
       "      <td id=\"T_a9e83_row2_col2\" class=\"data row2 col2\" >0.294449</td>\n",
       "      <td id=\"T_a9e83_row2_col3\" class=\"data row2 col3\" >0.372796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_a9e83_row3_col0\" class=\"data row3 col0\" >m4_LR</td>\n",
       "      <td id=\"T_a9e83_row3_col1\" class=\"data row3 col1\" >0.862013</td>\n",
       "      <td id=\"T_a9e83_row3_col2\" class=\"data row3 col2\" >0.259481</td>\n",
       "      <td id=\"T_a9e83_row3_col3\" class=\"data row3 col3\" >0.381870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_a9e83_row4_col0\" class=\"data row4 col0\" >m5_RF</td>\n",
       "      <td id=\"T_a9e83_row4_col1\" class=\"data row4 col1\" >0.863872</td>\n",
       "      <td id=\"T_a9e83_row4_col2\" class=\"data row4 col2\" >0.276701</td>\n",
       "      <td id=\"T_a9e83_row4_col3\" class=\"data row4 col3\" >0.396000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_a9e83_row5_col0\" class=\"data row5 col0\" >m6_lgbm1</td>\n",
       "      <td id=\"T_a9e83_row5_col1\" class=\"data row5 col1\" >0.864935</td>\n",
       "      <td id=\"T_a9e83_row5_col2\" class=\"data row5 col2\" >0.292513</td>\n",
       "      <td id=\"T_a9e83_row5_col3\" class=\"data row5 col3\" >0.400006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a9e83_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_a9e83_row6_col0\" class=\"data row6 col0\" >m7_lgbm2</td>\n",
       "      <td id=\"T_a9e83_row6_col1\" class=\"data row6 col1\" >0.864804</td>\n",
       "      <td id=\"T_a9e83_row6_col2\" class=\"data row6 col2\" >0.294483</td>\n",
       "      <td id=\"T_a9e83_row6_col3\" class=\"data row6 col3\" >0.400505</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fa8e3f00e10>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_results={'model':['m1_hgbc','m2_catboost',\n",
    "               'm3_xgbr','m4_LR',\n",
    "               'm5_RF','m6_lgbm1',\n",
    "              'm7_lgbm2'],\n",
    "'auc' :[ 0.865434044881213,0.8662925048383615, \n",
    "        0.8552429800216585,0.8620132487013733,\n",
    "        0.8638723261807801,0.8649350128689158,\n",
    "       0.8648040512360318],\n",
    "\n",
    "'f1_score' : [0.27931324050967715,0.28917295975957824,\n",
    "              0.2944488401724624,0.25948063132387106,\n",
    "              0.2767006882442805,0.2925125077182382,\n",
    "             0.2944832224347166],\n",
    "\n",
    "'average_precision_score' :[0.4032613815557755,0.4007558117925276,\n",
    "                            0.3727962618723695,0.3818703865043032,\n",
    "                            0.3960003199342976,0.40000554322071746,\n",
    "                           0.40050455934231144]    \n",
    "     }\n",
    "\n",
    "df_results= pd.DataFrame.from_dict(data=data_results\n",
    "                         #,columns=['model','auc','f1_score', 'average_precision_score']\n",
    "                       )\n",
    "# Highlight the Min values in each column\n",
    "print(\"\\nModified Stlying DataFrame:\")\n",
    "df_results.style.highlight_min(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "93082ed5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:11.545490Z",
     "iopub.status.busy": "2021-12-28T12:30:11.544945Z",
     "iopub.status.idle": "2021-12-28T12:30:11.556307Z",
     "shell.execute_reply": "2021-12-28T12:30:11.555832Z"
    },
    "papermill": {
     "duration": 0.074859,
     "end_time": "2021-12-28T12:30:11.556439",
     "exception": false,
     "start_time": "2021-12-28T12:30:11.481580",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_9e782_row0_col3, #T_9e782_row1_col1, #T_9e782_row6_col0, #T_9e782_row6_col2 {\n",
       "  background-color: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_9e782_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >model</th>\n",
       "      <th class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th class=\"col_heading level0 col2\" >f1_score</th>\n",
       "      <th class=\"col_heading level0 col3\" >average_precision_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_9e782_row0_col0\" class=\"data row0 col0\" >m1_hgbc</td>\n",
       "      <td id=\"T_9e782_row0_col1\" class=\"data row0 col1\" >0.865434</td>\n",
       "      <td id=\"T_9e782_row0_col2\" class=\"data row0 col2\" >0.279313</td>\n",
       "      <td id=\"T_9e782_row0_col3\" class=\"data row0 col3\" >0.403261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_9e782_row1_col0\" class=\"data row1 col0\" >m2_catboost</td>\n",
       "      <td id=\"T_9e782_row1_col1\" class=\"data row1 col1\" >0.866293</td>\n",
       "      <td id=\"T_9e782_row1_col2\" class=\"data row1 col2\" >0.289173</td>\n",
       "      <td id=\"T_9e782_row1_col3\" class=\"data row1 col3\" >0.400756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_9e782_row2_col0\" class=\"data row2 col0\" >m3_xgbr</td>\n",
       "      <td id=\"T_9e782_row2_col1\" class=\"data row2 col1\" >0.855243</td>\n",
       "      <td id=\"T_9e782_row2_col2\" class=\"data row2 col2\" >0.294449</td>\n",
       "      <td id=\"T_9e782_row2_col3\" class=\"data row2 col3\" >0.372796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_9e782_row3_col0\" class=\"data row3 col0\" >m4_LR</td>\n",
       "      <td id=\"T_9e782_row3_col1\" class=\"data row3 col1\" >0.862013</td>\n",
       "      <td id=\"T_9e782_row3_col2\" class=\"data row3 col2\" >0.259481</td>\n",
       "      <td id=\"T_9e782_row3_col3\" class=\"data row3 col3\" >0.381870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_9e782_row4_col0\" class=\"data row4 col0\" >m5_RF</td>\n",
       "      <td id=\"T_9e782_row4_col1\" class=\"data row4 col1\" >0.863872</td>\n",
       "      <td id=\"T_9e782_row4_col2\" class=\"data row4 col2\" >0.276701</td>\n",
       "      <td id=\"T_9e782_row4_col3\" class=\"data row4 col3\" >0.396000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_9e782_row5_col0\" class=\"data row5 col0\" >m6_lgbm1</td>\n",
       "      <td id=\"T_9e782_row5_col1\" class=\"data row5 col1\" >0.864935</td>\n",
       "      <td id=\"T_9e782_row5_col2\" class=\"data row5 col2\" >0.292513</td>\n",
       "      <td id=\"T_9e782_row5_col3\" class=\"data row5 col3\" >0.400006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9e782_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_9e782_row6_col0\" class=\"data row6 col0\" >m7_lgbm2</td>\n",
       "      <td id=\"T_9e782_row6_col1\" class=\"data row6 col1\" >0.864804</td>\n",
       "      <td id=\"T_9e782_row6_col2\" class=\"data row6 col2\" >0.294483</td>\n",
       "      <td id=\"T_9e782_row6_col3\" class=\"data row6 col3\" >0.400505</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fa8e5545190>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_results.style.highlight_max(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "39459ce0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:11.680212Z",
     "iopub.status.busy": "2021-12-28T12:30:11.677583Z",
     "iopub.status.idle": "2021-12-28T12:30:11.699109Z",
     "shell.execute_reply": "2021-12-28T12:30:11.698674Z"
    },
    "papermill": {
     "duration": 0.083641,
     "end_time": "2021-12-28T12:30:11.699251",
     "exception": false,
     "start_time": "2021-12-28T12:30:11.615610",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_ba199_row0_col1, #T_ba199_row1_col3, #T_ba199_row3_col2 {\n",
       "  background-color: #0000ff;\n",
       "  color: #f1f1f1;\n",
       "  background-color: red;\n",
       "}\n",
       "#T_ba199_row0_col2 {\n",
       "  background-color: #2424fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row0_col3 {\n",
       "  background-color: #1414fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row1_col1 {\n",
       "  background-color: #1212fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row1_col2 {\n",
       "  background-color: #6868fa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row2_col1 {\n",
       "  background-color: #1d1dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row2_col2 {\n",
       "  background-color: #0d0dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row2_col3 {\n",
       "  background-color: #1919fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row3_col0 {\n",
       "  background-color: red;\n",
       "}\n",
       "#T_ba199_row3_col1 {\n",
       "  background-color: #2020fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row3_col3 {\n",
       "  background-color: #1616fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row4_col1 {\n",
       "  background-color: #3535fc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row4_col2 {\n",
       "  background-color: #7b7bf9;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row4_col3 {\n",
       "  background-color: #3a3afc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row5_col1 {\n",
       "  background-color: #5d5dfa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_ba199_row5_col2, #T_ba199_row6_col1, #T_ba199_row6_col3 {\n",
       "  background-color: #f0f0f3;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_ba199_row5_col3 {\n",
       "  background-color: #a9a9f6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_ba199_row6_col2 {\n",
       "  background-color: #0000ff;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_ba199_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >model</th>\n",
       "      <th class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th class=\"col_heading level0 col2\" >f1_score</th>\n",
       "      <th class=\"col_heading level0 col3\" >average_precision_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row0\" class=\"row_heading level0 row0\" >1</th>\n",
       "      <td id=\"T_ba199_row0_col0\" class=\"data row0 col0\" >m2_catboost</td>\n",
       "      <td id=\"T_ba199_row0_col1\" class=\"data row0 col1\" >0.866293</td>\n",
       "      <td id=\"T_ba199_row0_col2\" class=\"data row0 col2\" >0.289173</td>\n",
       "      <td id=\"T_ba199_row0_col3\" class=\"data row0 col3\" >0.400756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row1\" class=\"row_heading level0 row1\" >0</th>\n",
       "      <td id=\"T_ba199_row1_col0\" class=\"data row1 col0\" >m1_hgbc</td>\n",
       "      <td id=\"T_ba199_row1_col1\" class=\"data row1 col1\" >0.865434</td>\n",
       "      <td id=\"T_ba199_row1_col2\" class=\"data row1 col2\" >0.279313</td>\n",
       "      <td id=\"T_ba199_row1_col3\" class=\"data row1 col3\" >0.403261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row2\" class=\"row_heading level0 row2\" >5</th>\n",
       "      <td id=\"T_ba199_row2_col0\" class=\"data row2 col0\" >m6_lgbm1</td>\n",
       "      <td id=\"T_ba199_row2_col1\" class=\"data row2 col1\" >0.864935</td>\n",
       "      <td id=\"T_ba199_row2_col2\" class=\"data row2 col2\" >0.292513</td>\n",
       "      <td id=\"T_ba199_row2_col3\" class=\"data row2 col3\" >0.400006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row3\" class=\"row_heading level0 row3\" >6</th>\n",
       "      <td id=\"T_ba199_row3_col0\" class=\"data row3 col0\" >m7_lgbm2</td>\n",
       "      <td id=\"T_ba199_row3_col1\" class=\"data row3 col1\" >0.864804</td>\n",
       "      <td id=\"T_ba199_row3_col2\" class=\"data row3 col2\" >0.294483</td>\n",
       "      <td id=\"T_ba199_row3_col3\" class=\"data row3 col3\" >0.400505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_ba199_row4_col0\" class=\"data row4 col0\" >m5_RF</td>\n",
       "      <td id=\"T_ba199_row4_col1\" class=\"data row4 col1\" >0.863872</td>\n",
       "      <td id=\"T_ba199_row4_col2\" class=\"data row4 col2\" >0.276701</td>\n",
       "      <td id=\"T_ba199_row4_col3\" class=\"data row4 col3\" >0.396000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row5\" class=\"row_heading level0 row5\" >3</th>\n",
       "      <td id=\"T_ba199_row5_col0\" class=\"data row5 col0\" >m4_LR</td>\n",
       "      <td id=\"T_ba199_row5_col1\" class=\"data row5 col1\" >0.862013</td>\n",
       "      <td id=\"T_ba199_row5_col2\" class=\"data row5 col2\" >0.259481</td>\n",
       "      <td id=\"T_ba199_row5_col3\" class=\"data row5 col3\" >0.381870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ba199_level0_row6\" class=\"row_heading level0 row6\" >2</th>\n",
       "      <td id=\"T_ba199_row6_col0\" class=\"data row6 col0\" >m3_xgbr</td>\n",
       "      <td id=\"T_ba199_row6_col1\" class=\"data row6 col1\" >0.855243</td>\n",
       "      <td id=\"T_ba199_row6_col2\" class=\"data row6 col2\" >0.294449</td>\n",
       "      <td id=\"T_ba199_row6_col3\" class=\"data row6 col3\" >0.372796</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fa8e5e18590>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cm = sns.light_palette(\"blue\", as_cmap=True)\n",
    "\n",
    "df_results.sort_values(by='auc',ascending=False).style.background_gradient(cmap=cm).highlight_max(axis=0, color='red')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "401af87c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:11.823753Z",
     "iopub.status.busy": "2021-12-28T12:30:11.823101Z",
     "iopub.status.idle": "2021-12-28T12:30:11.843994Z",
     "shell.execute_reply": "2021-12-28T12:30:11.843396Z"
    },
    "papermill": {
     "duration": 0.084767,
     "end_time": "2021-12-28T12:30:11.844160",
     "exception": false,
     "start_time": "2021-12-28T12:30:11.759393",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_e3833_row0_col0 {\n",
       "  background-color: red;\n",
       "}\n",
       "#T_e3833_row0_col1 {\n",
       "  background-color: #2020fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row0_col2, #T_e3833_row3_col1, #T_e3833_row4_col3 {\n",
       "  background-color: #0000ff;\n",
       "  color: #f1f1f1;\n",
       "  background-color: red;\n",
       "}\n",
       "#T_e3833_row0_col3 {\n",
       "  background-color: #1616fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row1_col1, #T_e3833_row1_col3, #T_e3833_row6_col2 {\n",
       "  background-color: #f0f0f3;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e3833_row1_col2 {\n",
       "  background-color: #0000ff;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row2_col1 {\n",
       "  background-color: #1d1dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row2_col2 {\n",
       "  background-color: #0d0dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row2_col3 {\n",
       "  background-color: #1919fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row3_col2 {\n",
       "  background-color: #2424fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row3_col3 {\n",
       "  background-color: #1414fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row4_col1 {\n",
       "  background-color: #1212fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row4_col2 {\n",
       "  background-color: #6868fa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row5_col1 {\n",
       "  background-color: #3535fc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row5_col2 {\n",
       "  background-color: #7b7bf9;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row5_col3 {\n",
       "  background-color: #3a3afc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row6_col1 {\n",
       "  background-color: #5d5dfa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e3833_row6_col3 {\n",
       "  background-color: #a9a9f6;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_e3833_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >model</th>\n",
       "      <th class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th class=\"col_heading level0 col2\" >f1_score</th>\n",
       "      <th class=\"col_heading level0 col3\" >average_precision_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row0\" class=\"row_heading level0 row0\" >6</th>\n",
       "      <td id=\"T_e3833_row0_col0\" class=\"data row0 col0\" >m7_lgbm2</td>\n",
       "      <td id=\"T_e3833_row0_col1\" class=\"data row0 col1\" >0.864804</td>\n",
       "      <td id=\"T_e3833_row0_col2\" class=\"data row0 col2\" >0.294483</td>\n",
       "      <td id=\"T_e3833_row0_col3\" class=\"data row0 col3\" >0.400505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row1\" class=\"row_heading level0 row1\" >2</th>\n",
       "      <td id=\"T_e3833_row1_col0\" class=\"data row1 col0\" >m3_xgbr</td>\n",
       "      <td id=\"T_e3833_row1_col1\" class=\"data row1 col1\" >0.855243</td>\n",
       "      <td id=\"T_e3833_row1_col2\" class=\"data row1 col2\" >0.294449</td>\n",
       "      <td id=\"T_e3833_row1_col3\" class=\"data row1 col3\" >0.372796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row2\" class=\"row_heading level0 row2\" >5</th>\n",
       "      <td id=\"T_e3833_row2_col0\" class=\"data row2 col0\" >m6_lgbm1</td>\n",
       "      <td id=\"T_e3833_row2_col1\" class=\"data row2 col1\" >0.864935</td>\n",
       "      <td id=\"T_e3833_row2_col2\" class=\"data row2 col2\" >0.292513</td>\n",
       "      <td id=\"T_e3833_row2_col3\" class=\"data row2 col3\" >0.400006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row3\" class=\"row_heading level0 row3\" >1</th>\n",
       "      <td id=\"T_e3833_row3_col0\" class=\"data row3 col0\" >m2_catboost</td>\n",
       "      <td id=\"T_e3833_row3_col1\" class=\"data row3 col1\" >0.866293</td>\n",
       "      <td id=\"T_e3833_row3_col2\" class=\"data row3 col2\" >0.289173</td>\n",
       "      <td id=\"T_e3833_row3_col3\" class=\"data row3 col3\" >0.400756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row4\" class=\"row_heading level0 row4\" >0</th>\n",
       "      <td id=\"T_e3833_row4_col0\" class=\"data row4 col0\" >m1_hgbc</td>\n",
       "      <td id=\"T_e3833_row4_col1\" class=\"data row4 col1\" >0.865434</td>\n",
       "      <td id=\"T_e3833_row4_col2\" class=\"data row4 col2\" >0.279313</td>\n",
       "      <td id=\"T_e3833_row4_col3\" class=\"data row4 col3\" >0.403261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row5\" class=\"row_heading level0 row5\" >4</th>\n",
       "      <td id=\"T_e3833_row5_col0\" class=\"data row5 col0\" >m5_RF</td>\n",
       "      <td id=\"T_e3833_row5_col1\" class=\"data row5 col1\" >0.863872</td>\n",
       "      <td id=\"T_e3833_row5_col2\" class=\"data row5 col2\" >0.276701</td>\n",
       "      <td id=\"T_e3833_row5_col3\" class=\"data row5 col3\" >0.396000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3833_level0_row6\" class=\"row_heading level0 row6\" >3</th>\n",
       "      <td id=\"T_e3833_row6_col0\" class=\"data row6 col0\" >m4_LR</td>\n",
       "      <td id=\"T_e3833_row6_col1\" class=\"data row6 col1\" >0.862013</td>\n",
       "      <td id=\"T_e3833_row6_col2\" class=\"data row6 col2\" >0.259481</td>\n",
       "      <td id=\"T_e3833_row6_col3\" class=\"data row6 col3\" >0.381870</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fa8e4083890>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_results.sort_values(by='f1_score',ascending=False).style.background_gradient(cmap=cm).highlight_max(axis=0, color='red')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b6d01f96",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:11.983710Z",
     "iopub.status.busy": "2021-12-28T12:30:11.983060Z",
     "iopub.status.idle": "2021-12-28T12:30:11.985796Z",
     "shell.execute_reply": "2021-12-28T12:30:11.986228Z"
    },
    "papermill": {
     "duration": 0.080561,
     "end_time": "2021-12-28T12:30:11.986390",
     "exception": false,
     "start_time": "2021-12-28T12:30:11.905829",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_80dab_row0_col1 {\n",
       "  background-color: #1212fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row0_col2 {\n",
       "  background-color: #6868fa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row0_col3, #T_80dab_row1_col1, #T_80dab_row2_col2, #T_80dab_row6_col2 {\n",
       "  background-color: #0000ff;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row1_col2 {\n",
       "  background-color: #2424fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row1_col3 {\n",
       "  background-color: #1414fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row2_col1 {\n",
       "  background-color: #2020fd;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row2_col3 {\n",
       "  background-color: #1616fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row3_col1 {\n",
       "  background-color: #1d1dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row3_col2 {\n",
       "  background-color: #0d0dfe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row3_col3 {\n",
       "  background-color: #1919fe;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row4_col1 {\n",
       "  background-color: #3535fc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row4_col2 {\n",
       "  background-color: #7b7bf9;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row4_col3 {\n",
       "  background-color: #3a3afc;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row5_col1 {\n",
       "  background-color: #5d5dfa;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_80dab_row5_col2, #T_80dab_row6_col1, #T_80dab_row6_col3 {\n",
       "  background-color: #f0f0f3;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_80dab_row5_col3 {\n",
       "  background-color: #a9a9f6;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_80dab_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >model</th>\n",
       "      <th class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th class=\"col_heading level0 col2\" >f1_score</th>\n",
       "      <th class=\"col_heading level0 col3\" >average_precision_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_80dab_row0_col0\" class=\"data row0 col0\" >m1_hgbc</td>\n",
       "      <td id=\"T_80dab_row0_col1\" class=\"data row0 col1\" >0.865434</td>\n",
       "      <td id=\"T_80dab_row0_col2\" class=\"data row0 col2\" >0.279313</td>\n",
       "      <td id=\"T_80dab_row0_col3\" class=\"data row0 col3\" >0.403261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_80dab_row1_col0\" class=\"data row1 col0\" >m2_catboost</td>\n",
       "      <td id=\"T_80dab_row1_col1\" class=\"data row1 col1\" >0.866293</td>\n",
       "      <td id=\"T_80dab_row1_col2\" class=\"data row1 col2\" >0.289173</td>\n",
       "      <td id=\"T_80dab_row1_col3\" class=\"data row1 col3\" >0.400756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row2\" class=\"row_heading level0 row2\" >6</th>\n",
       "      <td id=\"T_80dab_row2_col0\" class=\"data row2 col0\" >m7_lgbm2</td>\n",
       "      <td id=\"T_80dab_row2_col1\" class=\"data row2 col1\" >0.864804</td>\n",
       "      <td id=\"T_80dab_row2_col2\" class=\"data row2 col2\" >0.294483</td>\n",
       "      <td id=\"T_80dab_row2_col3\" class=\"data row2 col3\" >0.400505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row3\" class=\"row_heading level0 row3\" >5</th>\n",
       "      <td id=\"T_80dab_row3_col0\" class=\"data row3 col0\" >m6_lgbm1</td>\n",
       "      <td id=\"T_80dab_row3_col1\" class=\"data row3 col1\" >0.864935</td>\n",
       "      <td id=\"T_80dab_row3_col2\" class=\"data row3 col2\" >0.292513</td>\n",
       "      <td id=\"T_80dab_row3_col3\" class=\"data row3 col3\" >0.400006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_80dab_row4_col0\" class=\"data row4 col0\" >m5_RF</td>\n",
       "      <td id=\"T_80dab_row4_col1\" class=\"data row4 col1\" >0.863872</td>\n",
       "      <td id=\"T_80dab_row4_col2\" class=\"data row4 col2\" >0.276701</td>\n",
       "      <td id=\"T_80dab_row4_col3\" class=\"data row4 col3\" >0.396000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row5\" class=\"row_heading level0 row5\" >3</th>\n",
       "      <td id=\"T_80dab_row5_col0\" class=\"data row5 col0\" >m4_LR</td>\n",
       "      <td id=\"T_80dab_row5_col1\" class=\"data row5 col1\" >0.862013</td>\n",
       "      <td id=\"T_80dab_row5_col2\" class=\"data row5 col2\" >0.259481</td>\n",
       "      <td id=\"T_80dab_row5_col3\" class=\"data row5 col3\" >0.381870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_80dab_level0_row6\" class=\"row_heading level0 row6\" >2</th>\n",
       "      <td id=\"T_80dab_row6_col0\" class=\"data row6 col0\" >m3_xgbr</td>\n",
       "      <td id=\"T_80dab_row6_col1\" class=\"data row6 col1\" >0.855243</td>\n",
       "      <td id=\"T_80dab_row6_col2\" class=\"data row6 col2\" >0.294449</td>\n",
       "      <td id=\"T_80dab_row6_col3\" class=\"data row6 col3\" >0.372796</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fa8e5545690>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_results.sort_values(by='average_precision_score',ascending=False).style.background_gradient(cmap=cm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "dde9dde0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:12.117065Z",
     "iopub.status.busy": "2021-12-28T12:30:12.116373Z",
     "iopub.status.idle": "2021-12-28T12:30:12.118984Z",
     "shell.execute_reply": "2021-12-28T12:30:12.118512Z"
    },
    "papermill": {
     "duration": 0.070448,
     "end_time": "2021-12-28T12:30:12.119121",
     "exception": false,
     "start_time": "2021-12-28T12:30:12.048673",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "#pipe_lgbm_woe_powertransformer.fit(X,y)\n",
    "#predictions =pipe_lgbm_woe_powertransformer.predict_proba(X_test_final)[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "d15c49f0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:12.247257Z",
     "iopub.status.busy": "2021-12-28T12:30:12.246526Z",
     "iopub.status.idle": "2021-12-28T12:30:12.257697Z",
     "shell.execute_reply": "2021-12-28T12:30:12.257092Z"
    },
    "papermill": {
     "duration": 0.077777,
     "end_time": "2021-12-28T12:30:12.257833",
     "exception": false,
     "start_time": "2021-12-28T12:30:12.180056",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "predictions = np.mean(np.column_stack(predictions), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "2fd7ea5c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-28T12:30:12.386593Z",
     "iopub.status.busy": "2021-12-28T12:30:12.385821Z",
     "iopub.status.idle": "2021-12-28T12:30:12.703285Z",
     "shell.execute_reply": "2021-12-28T12:30:12.702782Z"
    },
    "papermill": {
     "duration": 0.384185,
     "end_time": "2021-12-28T12:30:12.703416",
     "exception": false,
     "start_time": "2021-12-28T12:30:12.319231",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>Probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.059357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.043084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.015515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.064689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.105483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101498</th>\n",
       "      <td>101499</td>\n",
       "      <td>0.030125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101499</th>\n",
       "      <td>101500</td>\n",
       "      <td>0.398874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101500</th>\n",
       "      <td>101501</td>\n",
       "      <td>0.010927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101501</th>\n",
       "      <td>101502</td>\n",
       "      <td>0.069618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101502</th>\n",
       "      <td>101503</td>\n",
       "      <td>0.032488</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101503 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Id  Probability\n",
       "0            1     0.059357\n",
       "1            2     0.043084\n",
       "2            3     0.015515\n",
       "3            4     0.064689\n",
       "4            5     0.105483\n",
       "...        ...          ...\n",
       "101498  101499     0.030125\n",
       "101499  101500     0.398874\n",
       "101500  101501     0.010927\n",
       "101501  101502     0.069618\n",
       "101502  101503     0.032488\n",
       "\n",
       "[101503 rows x 2 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Save the predictions to a CSV file\n",
    "sub = pd.read_csv('../input/GiveMeSomeCredit/sampleEntry.csv')\n",
    "sub['Probability']=predictions\n",
    "sub.to_csv('pipe_lgbm_woe_powertransformer_estimator_XGB_CV_onlynum_fe_V3.csv', index=False)\n",
    "sub"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "41e89d93",
   "metadata": {
    "papermill": {
     "duration": 0.061438,
     "end_time": "2021-12-28T12:30:12.826527",
     "exception": false,
     "start_time": "2021-12-28T12:30:12.765089",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Conclusion: \n",
    "\n",
    "1. A model that meets the requirements and constraints of project stakeholders.\n",
    "1. A model that is sufficiently skillful given the time and resources available.\n",
    "1. A model that is skillful as compared to naive models.\n",
    "1. A model that is skillful relative to other tested models.\n",
    "1. A model that is skillful relative to the state-of-the-art."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 3983.221563,
   "end_time": "2021-12-28T12:30:13.804746",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2021-12-28T11:23:50.583183",
   "version": "2.3.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}