Feature selection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"></ul></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.io import *\n",
    "from fastai.structured import *\n",
    "from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier\n",
    "from pandas_summary import DataFrameSummary\n",
    "from IPython.display import display\n",
    "from sklearn import metrics\n",
    "import feather\n",
    "import lightgbm as lgb\n",
    "from sklearn.model_selection import StratifiedKFold, KFold\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import warnings\n",
    "import gc\n",
    "from sklearn.linear_model import Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlxtend.feature_selection import SequentialFeatureSelector as SFS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/pyarrow/pandas_compat.py:752: FutureWarning: .labels was deprecated in version 0.24.0. Use .codes instead.\n",
      "  labels, = index.labels\n"
     ]
    }
   ],
   "source": [
    "train_df2 = feather.read_dataframe('train_df_alpha')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(201917, 386)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_included = ['feature_1','feature_2','feature_3','transactions_count','subsector_id_nunique','merchant_id_nunique','merchant_category_id_nunique','purchase_Month_mean','purchase_Month_min','purchase_Month_max','purchase_Week_nunique','purchase_Week_mean','purchase_Week_min','purchase_Week_max','purchase_Dayofweek_mean','purchase_Dayofweek_min','purchase_Dayofweek_max','purchase_Day_nunique','purchase_Day_mean','purchase_Day_min','purchase_Day_max','purchase_Hour_nunique','purchase_Hour_mean','purchase_Hour_min','purchase_Hour_max','purchase_amount_sum','purchase_amount_max','purchase_amount_min','purchase_amount_mean','purchase_amount_var','purchase_amount_skew','installments_sum','installments_max','installments_mean','installments_var','installments_skew','month_lag_max','month_lag_min','month_lag_mean','month_lag_var','month_lag_skew','month_diff_mean','month_diff_var','month_diff_skew','purchased_on_weekend_mean','category_1_mean','category_2_mean','category_3_mean','card_id_count','price_mean','price_max','price_min','price_var','Christmas_Day_2017_mean','Children_day_2017_mean','Black_Friday_2017_mean','Mothers_Day_2018_mean','duration_mean','duration_min','duration_max','duration_var','duration_skew','amount_month_ratio_mean','amount_month_ratio_min','amount_month_ratio_max','amount_month_ratio_var','amount_month_ratio_skew','category_2_mean_mean','category_3_mean_mean','purchase_date_diff','purchase_date_average','purchase_date_uptonow','purchase_date_uptomin','transactions_count_old','subsector_id_nunique_old','merchant_id_nunique_old','merchant_category_id_nunique_old','purchase_Month_nunique','purchase_Month_mean_old','purchase_Month_min_old','purchase_Month_max_old','purchase_Week_nunique_old','purchase_Week_mean_old','purchase_Week_min_old','purchase_Week_max_old','purchase_Dayofweek_mean_old','purchase_Day_nunique_old','purchase_Day_mean_old','purchase_Day_min_old','purchase_Hour_nunique_old','purchase_Hour_mean_old','purchase_Hour_min_old','purchase_Hour_max_old','purchase_amount_sum_old','purchase_amount_max_old','purchase_amount_min_old','purchase_amount_mean_old','purchase_amount_var_old','purchase_amount_skew_old','installments_sum_old','installments_max_old','installments_mean_old','installments_var_old','installments_skew_old','month_lag_max_old','month_lag_min_old','month_lag_mean_old','month_lag_var_old','month_lag_skew_old','month_diff_max','month_diff_min','month_diff_mean_old','month_diff_var_old','month_diff_skew_old','authorized_flag_mean','purchased_on_weekend_mean_old','category_1_mean_old','category_2_mean_old','category_3_mean_old','card_id_count_old','price_sum','price_mean_old','price_max_old','price_min_old','price_var_old','Christmas_Day_2017_mean_old','Mothers_Day_2017_mean','fathers_day_2017_mean','Children_day_2017_mean_old','Valentine_Day_2017_mean','Black_Friday_2017_mean_old','Mothers_Day_2018_mean_old','duration_mean_old','duration_min_old','duration_max_old','duration_var_old','duration_skew_old','amount_month_ratio_mean_old','amount_month_ratio_min_old','amount_month_ratio_max_old','amount_month_ratio_var_old','amount_month_ratio_skew_old','category_2_mean_mean_old','category_3_mean_mean_old','purchase_date_diff_old','purchase_date_average_old','purchase_date_uptonow_old','purchase_date_uptomin_old','quarter','observed_elapsed_time','days_feature1','days_feature2','days_feature3','days_feature1_ratio','days_feature2_ratio','days_feature3_ratio','feature_sum','feature_mean','feature_max','feature_min','feature_var','card_id_total','card_id_count_total','card_id_count_ratio','purchase_amount_total','purchase_amount_total_mean','purchase_amount_total_max','purchase_amount_total_min','purchase_amount_sum_ratio','hist_first_buy','new_first_buy','hist_last_buy','new_last_buy','month_diff_ratio','installments_total','installments_ratio','price_total','CLV','CLV_old','CLV_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_cols = ['feature_1','feature_2','feature_3','transactions_count','subsector_id_nunique','merchant_id_nunique','merchant_category_id_nunique','purchase_Month_mean','purchase_Month_min','purchase_Month_max','purchase_Week_nunique','purchase_Week_mean','purchase_Week_min','purchase_Week_max','purchase_Dayofweek_mean','purchase_Dayofweek_min','purchase_Dayofweek_max','purchase_Day_nunique','purchase_Day_mean','purchase_Day_min','purchase_Day_max','purchase_Hour_nunique','purchase_Hour_mean','purchase_Hour_min','purchase_Hour_max','purchase_amount_sum','purchase_amount_max','purchase_amount_min','purchase_amount_mean','purchase_amount_var','purchase_amount_skew','installments_sum','installments_max','installments_mean','installments_var','installments_skew','month_lag_max','month_lag_min','month_lag_mean','month_lag_var','month_lag_skew','month_diff_mean','month_diff_var','month_diff_skew','purchased_on_weekend_mean','category_1_mean','category_2_mean','category_3_mean','card_id_count','price_mean','price_max','price_min','price_var','Christmas_Day_2017_mean','Children_day_2017_mean','Black_Friday_2017_mean','Mothers_Day_2018_mean','duration_mean','duration_min','duration_max','duration_var','duration_skew','amount_month_ratio_mean','amount_month_ratio_min','amount_month_ratio_max','amount_month_ratio_var','amount_month_ratio_skew','category_2_mean_mean','category_3_mean_mean','purchase_date_diff','purchase_date_average','purchase_date_uptonow','purchase_date_uptomin','transactions_count_old','subsector_id_nunique_old','merchant_id_nunique_old','merchant_category_id_nunique_old','purchase_Month_nunique','purchase_Month_mean_old','purchase_Month_min_old','purchase_Month_max_old','purchase_Week_nunique_old','purchase_Week_mean_old','purchase_Week_min_old','purchase_Week_max_old','purchase_Dayofweek_mean_old','purchase_Day_nunique_old','purchase_Day_mean_old','purchase_Day_min_old','purchase_Hour_nunique_old','purchase_Hour_mean_old','purchase_Hour_min_old','purchase_Hour_max_old','purchase_amount_sum_old','purchase_amount_max_old','purchase_amount_min_old','purchase_amount_mean_old','purchase_amount_var_old','purchase_amount_skew_old','installments_sum_old','installments_max_old','installments_mean_old','installments_var_old','installments_skew_old','month_lag_max_old','month_lag_min_old','month_lag_mean_old','month_lag_var_old','month_lag_skew_old','month_diff_max','month_diff_min','month_diff_mean_old','month_diff_var_old','month_diff_skew_old','authorized_flag_mean','purchased_on_weekend_mean_old','category_1_mean_old','category_2_mean_old','category_3_mean_old','card_id_count_old','price_sum','price_mean_old','price_max_old','price_min_old','price_var_old','Christmas_Day_2017_mean_old','Mothers_Day_2017_mean','fathers_day_2017_mean','Children_day_2017_mean_old','Valentine_Day_2017_mean','Black_Friday_2017_mean_old','Mothers_Day_2018_mean_old','duration_mean_old','duration_min_old','duration_max_old','duration_var_old','duration_skew_old','amount_month_ratio_mean_old','amount_month_ratio_min_old','amount_month_ratio_max_old','amount_month_ratio_var_old','amount_month_ratio_skew_old','category_2_mean_mean_old','category_3_mean_mean_old','purchase_date_diff_old','purchase_date_average_old','purchase_date_uptonow_old','purchase_date_uptomin_old','quarter','observed_elapsed_time','days_feature1','days_feature2','days_feature3','days_feature1_ratio','days_feature2_ratio','days_feature3_ratio','feature_sum','feature_mean','feature_max','feature_min','feature_var','card_id_total','card_id_count_total','card_id_count_ratio','purchase_amount_total','purchase_amount_total_mean','purchase_amount_total_max','purchase_amount_total_min','purchase_amount_sum_ratio','hist_first_buy','new_first_buy','hist_last_buy','new_last_buy','month_diff_ratio','installments_total','installments_ratio','price_total','CLV','CLV_old','CLV_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "for df in [train_df2]:\n",
    "    for f in ['purchase_date_max','purchase_date_min','purchase_date_max_old',\\\n",
    "                     'purchase_date_min_old', 'observation_date_old']:\n",
    "        df[f] = df[f].astype(np.int64) * 1e-9\n",
    "        \n",
    "cols_excluded = ['purchase_date_max', 'purchase_date_max_old', 'card_id', 'first_active_month',\n",
    "                 'target','outliers','card_id_size', 'card_id_size_old', \n",
    "                 'purchase_date_min', 'purchase_date_min_old','first_active_monthYear',\n",
    "                 'first_active_monthMonth',\n",
    "                 'first_active_monthWeek',\n",
    "                 'first_active_monthDay',\n",
    "                 'first_active_monthDayofweek',\n",
    "                 'first_active_monthDayofyear',\n",
    "                 'first_active_monthIs_month_end',\n",
    "                 'first_active_monthIs_month_start',\n",
    "                 'first_active_monthIs_quarter_end',\n",
    "                 'first_active_monthIs_quarter_start',\n",
    "                 'first_active_monthIs_year_end',\n",
    "                 'Black_Friday_2017_mean',\n",
    "                 'amount_month_ratio_max',\n",
    "                 'purchase_Month_mean_old',\n",
    "                 'purchase_amount_total_max',\n",
    "                 'first_active_monthIs_year_start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(180, 180)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols_included), len(base_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = [c for c in train_df2.columns.values if c not in cols_excluded]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "x, x_t, y, y_t = train_test_split(train_df2, train_df2['target'], \n",
    "                                 stratify=train_df2['outliers'], test_size=0.05, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((191821, 386), (10096, 386))"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.shape, x_t.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "x, x_t, y, y_t = train_test_split(x, y, \n",
    "                                 stratify=x['outliers'], test_size=.1, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((172638, 386), (19183, 386))"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.shape, x_t.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1887, 210)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(x['outliers']==1).sum(), (x_t['outliers']==1).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((172638,), (19183,))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.shape, y_t.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.85499487413145"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.shape[0]/train_df2.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_cols_sorte = ['purchase_amount_max', 'avg_trans_per_merchant_old', 'duration_min_old', 'purchase_date_uptonow',\n",
    "       'category_1_mean', 'authorized_flag_mean', 'purchase_date_uptonow_old', 'purchase_Week_mean',\n",
    "       'purchase_amount_mean', 'purchase_Month_nunique', 'month_diff_mean_old', 'category_1_mean_old',\n",
    "       'card_id_count_ratio', 'purchase_date_diff', 'duration_mean_old', 'month_lag_mean_old',\n",
    "       'purchase_date_average_old', 'purchase_date_diff_old', 'fathers_day_2017_mean', 'purchase_Month_mean',\n",
    "       'purchase_Day_mean', 'amount_month_ratio_min_old', 'installments_sum_old', 'month_lag_mean',\n",
    "       'purchase_date_average', 'days_feature1', 'purchase_Week_max', 'Children_day_2017_mean_old',\n",
    "       'Valentine_Day_2017_mean', 'days_feature2', 'month_lag_var_old', 'hist_last_buy',\n",
    "       'merchant_id_nunique_old', 'purchase_Hour_mean_old', 'purchase_date_uptomin_old',\n",
    "       'month_lag_skew_old', 'purchase_amount_mean_old', 'Mothers_Day_2018_mean', 'duration_var',\n",
    "       'installments_ratio', 'price_var_old', 'purchase_date_uptomin', 'month_lag_skew',\n",
    "       'amount_month_ratio_mean_old', 'purchase_Hour_mean', 'CLV', 'duration_min', 'amount_month_ratio_var',\n",
    "       'month_diff_skew_old', 'purchase_Week_nunique_old', 'price_max', 'month_diff_mean',\n",
    "       'Mothers_Day_2017_mean', 'purchase_Dayofweek_mean_old', 'purchase_amount_var', 'month_diff_ratio',\n",
    "       'purchase_Day_mean_old', 'purchase_amount_min', 'hist_first_buy', 'days_feature1_ratio',\n",
    "       'price_mean_old', 'merchant_category_id_nunique_old', 'purchase_amount_min_old', 'purchase_Week_min',\n",
    "       'duration_skew_old', 'CLV_ratio', 'purchase_amount_sum', 'purchase_Day_min',\n",
    "       'purchased_on_weekend_mean_old', 'Christmas_Day_2017_mean', 'category_3_mean_mean_old',\n",
    "       'new_last_buy', 'duration_max_old', 'new_first_buy', 'month_lag_max', 'month_lag_var',\n",
    "       'purchase_Day_max', 'price_var', 'installments_var_old', 'installments_total', 'days_feature2_ratio',\n",
    "       'purchase_amount_skew_old', 'price_min', 'days_feature3', 'category_3_mean_old', 'duration_mean',\n",
    "       'month_diff_var_old', 'Black_Friday_2017_mean_old', 'amount_month_ratio_skew_old',\n",
    "       'Black_Friday_2017_mean', 'amount_month_ratio_min', 'installments_mean', 'purchase_Month_mean_old',\n",
    "       'purchase_Week_mean_old', 'price_max_old', 'purchase_amount_max_old', 'price_sum',\n",
    "       'purchase_Hour_min', 'amount_month_ratio_max_old', 'purchase_amount_sum_ratio', 'duration_var_old',\n",
    "       'installments_mean_old', 'purchase_amount_total_min', 'category_2_mean_old', 'Children_day_2017_mean',\n",
    "       'installments_skew_old', 'Christmas_Day_2017_mean_old', 'purchase_amount_sum_mean',\n",
    "       'purchase_Week_min_old', 'purchase_amount_sum_old', 'amount_month_ratio_var_old', 'price_mean',\n",
    "       'purchase_Week_max_old', 'CLV_old', 'purchase_Dayofweek_mean', 'category_2_5.0_purchase_amount_mean',\n",
    "       'category_3_mean_mean', 'month_diff_skew', 'purchase_amount_skew', 'category_2_mean_mean_old',\n",
    "       'transactions_count_old', 'subsector_id_nunique_old', 'duration_skew', 'merchant_category_id_nunique',\n",
    "       'purchase_amount_var_old', 'month_diff_var', 'price_total', 'price_min_old', 'duration_max',\n",
    "       'elapsed_time', 'installments_var', 'purchase_Hour_nunique_old', 'amount_month_ratio_mean',\n",
    "       'card_id_total', 'purchase_amount_total', 'Mothers_Day_2018_mean_old', 'purchase_Day_nunique_old',\n",
    "       'purchase_amount_total_mean', 'days_feature3_ratio', 'month_diff_max', 'amount_month_ratio_skew',\n",
    "       'amount_month_ratio_max', 'purchase_Month_max', 'purchase_Hour_min_old', 'month_lag_min',\n",
    "       'installments_max', 'installments_sum', 'purchase_amount_total_max', 'feature_1',\n",
    "       'transactions_count', 'category_2_mean_mean', 'month_lag_max_old', 'subsector_id_nunique',\n",
    "       'installments_skew', 'month_lag_min_old', 'card_id_count_old', 'feature_var', 'purchase_Week_nunique',\n",
    "       'purchased_on_weekend_mean', 'purchase_Hour_max', 'merchant_id_nunique', 'category_3_mean',\n",
    "       'feature_sum', 'purchase_Dayofweek_min', 'purchase_Month_min', 'installments_max_old',\n",
    "       'category_2_mean', 'card_id_count_total', 'month_diff_min', 'purchase_Hour_nunique',\n",
    "       'purchase_Day_nunique', 'purchase_Month_min_old', 'purchase_Day_min_old', 'quarter', 'feature_2',\n",
    "       'card_id_count', 'purchase_Hour_max_old', 'purchase_Month_max_old', 'feature_max', 'feature_mean',\n",
    "       'purchase_Dayofweek_max', 'feature_3', 'feature_min']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_cols_sorted = ['purchase_amount_max', 'duration_min_old', 'purchase_date_uptonow', 'category_1_mean',\n",
    "       'purchase_Week_mean', 'purchase_date_uptonow_old', 'authorized_flag_mean',\n",
    "       'category_1_1_purchase_amount_sum', 'month_diff_mean_old', 'repurchase_merchant_rate_old',\n",
    "       'purchase_Month_nunique', 'purchase_date_diff', 'card_id_count_ratio', 'month_lag_mean_old',\n",
    "       'purchase_Month_mean', 'purchase_Day_mean', 'purchase_date_diff_old', 'fathers_day_2017_mean',\n",
    "       'month_lag_var_old', 'category_3_1_purchase_amount_mean', 'month_lag_mean',\n",
    "       'days_since_last_transaction', 'category_3_1_purchase_amount_sum', 'amount_month_ratio_min_old',\n",
    "       'days_feature1', 'purchase_amount_pct_75', 'category_1_1_purchase_amount_mean',\n",
    "       'Valentine_Day_2017_mean', 'days_feature2', 'category_3_0_purchase_amount_mean',\n",
    "       'category_3_2_purchase_amount_sum', 'category_1_mean_old', 'Children_day_2017_mean_old',\n",
    "       'purchase_Week_max', 'purchase_date_uptomin', 'duration_min', 'duration_var',\n",
    "       'days_since_last_transaction_old', 'purchase_date_uptomin_old', 'category_3_2_purchase_amount_mean',\n",
    "       'month_lag_skew', 'category_2_1.0_purchase_amount_sum', 'purchase_amount_mean', 'hist_last_buy',\n",
    "       'avg_spend_per_merchant_old', 'Mothers_Day_2018_mean', 'purchase_date_average_old',\n",
    "       'purchase_date_successive_diff_mean', 'amount_month_ratio_var', 'purchase_Hour_mean_old',\n",
    "       'purchase_amount_median_old', 'purchase_amount_var', 'month_diff_mean', 'installments_ratio',\n",
    "       'month_lag_skew_old', 'purchase_Week_min', 'installments_sum_old', 'duration_mean_old',\n",
    "       'hist_first_buy', 'authorized_flag_1_purchase_amount_mean', 'purchase_Hour_mean', 'price_max', 'CLV',\n",
    "       'purchase_Dayofweek_mean_old', 'Mothers_Day_2017_mean', 'purchase_date_successive_diff_mean_old',\n",
    "       'avg_trans_per_merchant_old', 'category_2_1.0_purchase_amount_mean', 'purchase_date_average',\n",
    "       'month_diff_ratio', 'price_var_old', 'Christmas_Day_2017_mean', 'days_feature1_ratio',\n",
    "       'merchant_id_nunique_old', 'purchase_amount_median', 'purchase_amount_pct_25_old',\n",
    "       'purchase_Day_mean_old', 'purchased_on_weekend_mean_old', 'month_diff_skew_old', 'month_lag_max',\n",
    "       'new_last_buy', 'purchase_Week_nunique_old', 'purchase_amount_min', 'purchase_Day_min',\n",
    "       'purchase_amount_pct_75_old', 'merchant_category_repurchase_old', 'Black_Friday_2017_mean',\n",
    "       'days_feature3', 'month_lag_var', 'new_first_buy', 'category_1_0_purchase_amount_mean', 'CLV_ratio',\n",
    "       'avg_spend_per_merchant', 'category_3_0_purchase_amount_sum', 'duration_mean', 'days_feature2_ratio',\n",
    "       'purchase_amount_sum', 'price_var', 'category_1_0_purchase_amount_sum', 'purchase_Day_max',\n",
    "       'month_diff_var_old', 'purchase_Hour_min', 'purchase_Week_mean_old', 'purchase_amount_pct_25',\n",
    "       'installments_mean', 'installments_total', 'merchant_category_id_nunique_old', 'duration_skew_old',\n",
    "       'purchase_amount_skew_old', 'amount_month_ratio_min', 'Black_Friday_2017_mean_old',\n",
    "       'category_3_mean_old', 'installments_var_old', 'purchase_Month_mean_old', 'purchase_amount_max_ptp',\n",
    "       'purchase_amount_sum_ptp', 'price_min', 'purchase_amount_min_old', 'purchase_Week_min_old',\n",
    "       'purchase_amount_max_mean', 'amount_month_ratio_skew_old', 'purchase_amount_sum_ratio',\n",
    "       'Christmas_Day_2017_mean_old', 'Children_day_2017_mean', 'purchase_amount_mean_ptp',\n",
    "       'purchase_amount_max_sum', 'purchase_amount_sum_max', 'category_3_mean_mean_old',\n",
    "       'purchase_amount_total_min', 'purchase_amount_count_ptp', 'amount_month_ratio_mean_old',\n",
    "       'category_2_mean_old', 'month_diff_skew', 'purchase_Dayofweek_mean', 'installments_mean_old',\n",
    "       'price_mean', 'inverse_avg_transactions_per_day_old', 'category_3_mean_mean',\n",
    "       'merchant_category_id_nunique', 'purchase_amount_skew', 'price_mean_old', 'duration_skew',\n",
    "       'price_max_old', 'price_sum', 'authorized_flag_1_purchase_amount_sum', 'duration_max_old',\n",
    "       'subsector_id_nunique_old', 'installments_skew_old', 'installments_var', 'purchase_Week_max_old',\n",
    "       'category_2_5.0_purchase_amount_sum', 'inverse_avg_transactions_per_day', 'month_diff_var',\n",
    "       'Mothers_Day_2018_mean_old', 'month_diff_max', 'elapsed_time', 'amount_month_ratio_skew',\n",
    "       'days_feature3_ratio', 'category_2_3.0_purchase_amount_mean', 'category_2_mean_mean_old',\n",
    "       'duration_max', 'duration_var_old', 'purchase_Hour_nunique_old', 'amount_month_ratio_max_old',\n",
    "       'amount_month_ratio_mean', 'category_2_5.0_purchase_amount_mean', 'installments_sum',\n",
    "       'purchase_amount_mean_sum', 'category_2_4.0_purchase_amount_sum', 'purchase_Month_max',\n",
    "       'purchase_amount_count_max', 'month_lag_min', 'price_min_old', 'amount_month_ratio_var_old',\n",
    "       'purchase_amount_mean_mean', 'avg_spend_per_transaction', 'price_total',\n",
    "       'category_2_3.0_purchase_amount_sum', 'installments_max', 'transactions_count', 'month_lag_min_old',\n",
    "       'purchase_Day_nunique_old', 'purchase_amount_max_old', 'CLV_old',\n",
    "       'category_2_4.0_purchase_amount_mean', 'category_2_2.0_purchase_amount_mean',\n",
    "       'purchase_amount_total_mean', 'amount_month_ratio_max', 'purchase_amount_mean_max',\n",
    "       'purchase_amount_var_old', 'purchase_Hour_min_old', 'feature_1', 'subsector_id_nunique',\n",
    "       'transactions_count_old', 'purchase_amount_sum_mean', 'month_lag_max_old', 'card_id_total',\n",
    "       'installments_skew', 'purchase_amount_sum_old', 'purchase_Week_nunique', 'purchase_Hour_max',\n",
    "       'purchase_amount_total_max', 'purchase_amount_count_mean', 'merchant_id_nunique',\n",
    "       'purchase_amount_mean_old', 'category_2_2.0_purchase_amount_sum', 'category_2_mean_mean',\n",
    "       'purchase_amount_total', 'purchase_Month_min', 'purchased_on_weekend_mean', 'category_3_mean',\n",
    "       'month_diff_min', 'purchase_Dayofweek_min', 'category_2_mean', 'feature_sum', 'feature_var',\n",
    "       'purchase_Hour_nunique', 'purchase_Day_nunique', 'installments_max_old', 'card_id_count_old',\n",
    "       'purchase_amount_max_max', 'purchase_Month_min_old', 'quarter', 'avg_spend_per_transaction_old',\n",
    "       'purchase_amount_sum_sum', 'feature_2', 'card_id_count_total', 'card_id_count',\n",
    "       'merchant_category_repurchase', 'feature_mean', 'purchase_Day_min_old', 'purchase_Month_max_old',\n",
    "       'purchase_Hour_max_old', 'purchase_Dayofweek_max', 'feature_max', 'purchase_amount_count_sum',\n",
    "       'purchased_before_issue', 'repurchase_merchant_rate', 'avg_trans_per_merchant', 'feature_3',\n",
    "       'feature_min']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_291_sorted = ['purchase_date_uptonow', 'duration_min_old', 'purchase_Month_nunique', 'purchase_amount_max',\n",
    "       'repurchase_merchant_rate_old', 'purchase_date_uptonow_old', 'authorized_flag_mean',\n",
    "       'month_diff_mean_old', 'purchase_Week_mean', 'authorized_flag_0_purchase_amount_sum',\n",
    "       'category_1_1_purchase_amount_sum', 'category_1_1_purchase_amount_sum_old', 'purchase_date_diff',\n",
    "       'purchase_date_diff_old', 'month_lag_mean_old', 'card_id_count_ratio', 'category_1_mean',\n",
    "       'purchase_Month_mean', 'purchase_Day_mean', 'fathers_day_2017_mean', 'month_lag_var_old',\n",
    "       'category_3_1_purchase_amount_sum', 'category_3_0_purchase_amount_mean_old',\n",
    "       'category_3_1_purchase_amount_mean', 'month_lag_mean', 'days_feature1', 'amount_month_ratio_min_old',\n",
    "       'category_1_1_purchase_amount_mean', 'purchase_Week_max', 'days_feature2', 'purchase_date_uptomin',\n",
    "       'days_since_last_transaction_old', 'purchase_date_average_old', 'month_lag_skew',\n",
    "       'days_since_last_transaction', 'category_1_mean_old', 'category_2_1.0_purchase_amount_mean_old',\n",
    "       'Children_day_2017_mean_old', 'Mothers_Day_2018_mean', 'purchase_Hour_mean_old',\n",
    "       'Valentine_Day_2017_mean', 'purchase_amount_max_mean_old', 'category_3_0_purchase_amount_mean',\n",
    "       'authorized_flag_1_purchase_amount_mean', 'category_3_2_purchase_amount_sum', 'duration_min',\n",
    "       'authorized_flag_0_purchase_amount_mean', 'purchase_date_successive_diff_mean_old',\n",
    "       'avg_spend_per_merchant_old', 'installments_sum_old', 'duration_var', 'duration_mean_old',\n",
    "       'purchase_date_successive_diff_mean', 'month_diff_mean', 'purchase_amount_median_old',\n",
    "       'purchase_date_uptomin_old', 'category_2_1.0_purchase_amount_sum', 'hist_last_buy',\n",
    "       'month_lag_skew_old', 'month_lag_max', 'category_3_0_purchase_amount_sum_old',\n",
    "       'avg_trans_per_merchant_old', 'category_3_2_purchase_amount_mean', 'installments_ratio',\n",
    "       'purchase_Hour_mean', 'purchase_Week_min', 'Mothers_Day_2017_mean', 'purchase_date_average',\n",
    "       'Christmas_Day_2017_mean', 'purchase_Dayofweek_mean_old', 'month_diff_ratio',\n",
    "       'category_2_1.0_purchase_amount_mean', 'category_1_0_purchase_amount_mean_old',\n",
    "       'purchase_amount_pct_25_old', 'purchase_Day_min', 'merchant_id_nunique_old',\n",
    "       'category_1_1_purchase_amount_mean_old', 'amount_month_ratio_var', 'month_lag_var',\n",
    "       'purchase_amount_pct_75_old', 'price_max', 'purchase_Week_nunique_old',\n",
    "       'category_1_0_purchase_amount_mean', 'category_1_0_purchase_amount_sum_old', 'month_diff_skew_old',\n",
    "       'new_last_buy', 'hist_first_buy', 'purchase_amount_max_max_old', 'price_var_old',\n",
    "       'purchase_amount_var', 'days_feature1_ratio', 'purchase_Day_max', 'merchant_category_repurchase_old',\n",
    "       'days_feature3', 'purchased_on_weekend_mean_old', 'duration_skew_old',\n",
    "       'category_2_1.0_purchase_amount_sum_old', 'category_3_1_purchase_amount_mean_old',\n",
    "       'merchant_category_id_nunique_old', 'purchase_Day_mean_old', 'days_feature2_ratio', 'new_first_buy',\n",
    "       'CLV_ratio', 'purchase_amount_skew_old', 'purchase_amount_max_sum_old', 'Black_Friday_2017_mean',\n",
    "       'category_3_0_purchase_amount_sum', 'purchase_amount_pct_75', 'category_1_0_purchase_amount_sum',\n",
    "       'Black_Friday_2017_mean_old', 'observed_elapsed_time', 'amount_month_ratio_min',\n",
    "       'amount_month_ratio_skew_old', 'purchase_amount_max_sum', 'purchase_amount_sum_ptp', 'CLV',\n",
    "       'purchase_amount_max_mean', 'purchase_amount_mean_max_old', 'purchase_amount_min_old',\n",
    "       'duration_mean', 'Children_day_2017_mean', 'installments_var_old', 'purchase_Month_mean_old',\n",
    "       'month_diff_var_old', 'inverse_avg_transactions_per_day_old', 'category_3_1_purchase_amount_sum_old',\n",
    "       'installments_total', 'purchase_Week_mean_old', 'purchase_amount_sum_max', 'purchase_Week_min_old',\n",
    "       'purchase_amount_sum_ratio', 'price_var', 'purchase_amount_sum_max_old', 'category_3_mean_old',\n",
    "       'purchase_amount_max_ptp', 'installments_mean', 'purchase_Hour_min', 'price_mean_old',\n",
    "       'Christmas_Day_2017_mean_old', 'category_3_mean_mean_old', 'subsector_id_nunique_old',\n",
    "       'Mothers_Day_2018_mean_old', 'installments_mean_old', 'purchase_amount_mean_ptp',\n",
    "       'purchase_amount_median', 'category_3_2_purchase_amount_mean_old', 'installments_skew_old',\n",
    "       'purchase_amount_min', 'authorized_flag_1_purchase_amount_sum', 'purchase_Dayofweek_mean',\n",
    "       'month_diff_max', 'purchase_Week_max_old', 'price_sum', 'inverse_avg_transactions_per_day',\n",
    "       'category_2_5.0_purchase_amount_sum_old', 'price_max_old', 'purchase_amount_count_ptp',\n",
    "       'amount_month_ratio_mean_old', 'month_diff_skew', 'purchase_amount_skew', 'price_min',\n",
    "       'category_3_mean_mean', 'purchase_amount_mean_sum_old', 'purchase_amount_mean',\n",
    "       'purchase_amount_pct_25', 'duration_max_old', 'category_2_mean_old', 'purchase_amount_total_min',\n",
    "       'installments_var', 'purchase_amount_max_ptp_old', 'category_3_2_purchase_amount_sum_old',\n",
    "       'days_feature3_ratio', 'duration_skew', 'category_2_5.0_purchase_amount_sum',\n",
    "       'purchase_amount_max_old', 'merchant_category_id_nunique', 'purchase_amount_sum_mean_old',\n",
    "       'amount_month_ratio_max_old', 'duration_max', 'purchase_amount_sum', 'purchase_Month_max',\n",
    "       'elapsed_time', 'month_diff_var', 'month_lag_min', 'purchase_amount_mean_sum',\n",
    "       'category_2_5.0_purchase_amount_mean', 'category_2_mean_mean_old', 'amount_month_ratio_skew',\n",
    "       'amount_month_ratio_mean', 'duration_var_old', 'category_2_5.0_purchase_amount_mean_old',\n",
    "       'purchase_Day_nunique_old', 'avg_spend_per_merchant', 'category_2_3.0_purchase_amount_mean',\n",
    "       'purchase_amount_count_max', 'amount_month_ratio_var_old', 'purchase_amount_mean_mean', 'price_mean',\n",
    "       'purchase_Hour_nunique_old', 'avg_spend_per_transaction', 'month_lag_max_old',\n",
    "       'purchase_amount_mean_mean_old', 'purchase_amount_sum_ptp_old', 'purchase_amount_mean_ptp_old',\n",
    "       'feature_1', 'installments_sum', 'month_lag_min_old', 'category_2_3.0_purchase_amount_sum',\n",
    "       'category_2_4.0_purchase_amount_sum', 'purchase_amount_count_max_old',\n",
    "       'category_2_3.0_purchase_amount_mean_old', 'price_min_old', 'purchase_amount_count_mean_old',\n",
    "       'price_total', 'purchase_amount_mean_max', 'purchase_amount_var_old', 'purchase_Hour_min_old',\n",
    "       'transactions_count_old', 'amount_month_ratio_max', 'category_2_2.0_purchase_amount_mean',\n",
    "       'purchase_Week_nunique', 'CLV_old', 'purchase_amount_total_mean',\n",
    "       'category_2_4.0_purchase_amount_mean', 'installments_max', 'purchase_amount_total',\n",
    "       'purchase_Month_min', 'purchase_amount_total_max', 'purchase_amount_count_mean',\n",
    "       'purchase_amount_sum_mean', 'purchase_Hour_max', 'card_id_total', 'transactions_count',\n",
    "       'purchase_amount_mean_old', 'category_2_2.0_purchase_amount_sum', 'month_diff_min',\n",
    "       'subsector_id_nunique', 'purchase_amount_sum_old', 'installments_skew', 'feature_sum',\n",
    "       'purchased_on_weekend_mean', 'feature_var', 'purchase_amount_max_max', 'category_2_mean_mean',\n",
    "       'purchase_Dayofweek_min', 'card_id_count_old', 'authorized_flag_1_purchase_amount_mean_old',\n",
    "       'purchase_Hour_nunique', 'purchase_amount_sum_sum_old', 'purchase_Month_min_old',\n",
    "       'installments_max_old', 'category_2_4.0_purchase_amount_sum_old', 'observation_date_old',\n",
    "       'category_2_3.0_purchase_amount_sum_old', 'feature_2', 'category_3_mean', 'quarter',\n",
    "       'purchase_Day_nunique', 'merchant_id_nunique', 'avg_spend_per_transaction_old', 'category_2_mean',\n",
    "       'merchant_category_repurchase', 'category_2_2.0_purchase_amount_mean_old',\n",
    "       'category_2_4.0_purchase_amount_mean_old', 'purchase_Day_min_old',\n",
    "       'category_2_2.0_purchase_amount_sum_old', 'purchase_Month_max_old', 'purchase_amount_count_ptp_old',\n",
    "       'authorized_flag_1_purchase_amount_sum_old', 'card_id_count_total', 'purchase_amount_sum_sum',\n",
    "       'feature_mean', 'card_id_count', 'purchase_Hour_max_old', 'purchase_Dayofweek_max',\n",
    "       'purchase_amount_count_sum', 'feature_max', 'purchased_before_issue', 'purchase_amount_count_sum_old',\n",
    "       'repurchase_merchant_rate', 'avg_trans_per_merchant', 'feature_3', 'feature_min']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 3.641 on CV\n",
    "cols_ful_352 = ['purchase_date_uptonow', 'month_diff_mean_old', 'purchase_amount_sum_ratio',\n",
    "       'repurchase_merchant_rate_old', 'purchase_Month_nunique', 'purchase_Week_mean',\n",
    "       'authorized_flag_mean', 'purchase_Month_mean', 'category_1_0_purchase_amount_sum_ratio', 'CLV',\n",
    "       'purchase_Week_max', 'purchase_amount_before_after_ratio', 'month_lag_mean_old',\n",
    "       'purchase_date_diff_old', 'fathers_day_2017_mean', 'authorized_flag_0_purchase_amount_sum',\n",
    "       'month_diff_mean_total', 'category_1_mean_old', 'CLV_ratio', 'category_1_1_purchase_amount_mean',\n",
    "       'Mothers_Day_2018_mean', 'category_1_mean_total', 'month_lag_var_old', 'purchase_amount_mean_ratio',\n",
    "       'Children_day_2017_mean_old', 'purchase_Week_min', 'Valentine_Day_2017_mean',\n",
    "       'purchase_date_uptonow_old', 'category_3_0_purchase_amount_sum_old', 'month_lag_mean_total',\n",
    "       'hist_first_buy', 'Mothers_Day_2017_mean', 'price_skew', 'month_lag_mean', 'purchase_Day_mean',\n",
    "       'month_diff_max', 'hist_last_buy', 'category_1_1_purchase_amount_sum',\n",
    "       'authorized_flag_0_purchase_amount_mean', 'category_1_0_purchase_amount_mean_ratio',\n",
    "       'purchase_date_uptomin', 'avg_trans_per_merchant_old', 'Christmas_Day_2017_mean',\n",
    "       'purchase_Hour_mean_old', 'category_2_1.0_purchase_amount_sum_ratio', 'installments_sum_old',\n",
    "       'days_feature2', 'purchase_Week_nunique_old', 'purchase_Dayofweek_mean_old',\n",
    "       'category_3_1_purchase_amount_sum', 'category_3_2_purchase_amount_mean', 'purchase_Hour_mean',\n",
    "       'month_diff_mean', 'month_lag_skew_old', 'category_3_2_purchase_amount_sum', 'month_diff_skew_old',\n",
    "       'Black_Friday_2017_mean', 'purchase_date_diff', 'Mothers_Day_2018_mean_old',\n",
    "       'category_1_0_purchase_amount_sum', 'purchased_on_weekend_mean_old',\n",
    "       'category_3_1_purchase_amount_mean', 'purchase_amount_pct_75_ratio',\n",
    "       'category_1_0_purchase_amount_sum_old', 'category_1_mean', 'purchase_Day_mean_old', 'days_feature3',\n",
    "       'purchase_date_uptomin_old', 'days_feature1', 'purchase_date_average',\n",
    "       'merchant_category_repurchase_old', 'purchase_date_average_old', 'card_id_count_ratio',\n",
    "       'duration_var', 'purchase_date_successive_diff_mean', 'purchase_date_successive_diff_mean_old',\n",
    "       'merchant_id_nunique_old', 'purchase_amount_sum', 'category_1_1_purchase_amount_sum_total',\n",
    "       'category_2_1.0_purchase_amount_sum', 'category_3_1_purchase_amount_sum_ratio',\n",
    "       'category_3_0_purchase_amount_sum_ratio', 'month_diff_min', 'month_lag_var',\n",
    "       'purchase_amount_median_ratio', 'category_2_1.0_purchase_amount_mean', 'Children_day_2017_mean',\n",
    "       'category_2_1.0_purchase_amount_mean_ratio', 'purchase_amount_fi_last_new_total', 'price_var_old',\n",
    "       'category_2_1.0_purchase_amount_sum_old', 'amount_month_ratio_min_old', 'days_feature1_ratio',\n",
    "       'category_3_0_purchase_amount_mean', 'purchase_Week_min_old', 'purchase_Hour_min',\n",
    "       'Black_Friday_2017_mean_old', 'purchase_Week_max_old', 'month_diff_ratio', 'purchase_amount_skew_old',\n",
    "       'authorized_flag_1_purchase_amount_sum', 'Christmas_Day_2017_mean_old',\n",
    "       'category_3_0_purchase_amount_sum', 'purchase_Day_max', 'days_feature2_ratio',\n",
    "       'purchase_amount_tail_sum_old', 'new_first_buy', 'purchase_Month_max', 'avg_spend_per_merchant_old',\n",
    "       'purchase_amount_tail_max_old', 'purchase_amount_first_last_diff', 'purchase_Month_mean_old',\n",
    "       'purchase_Week_mean_old', 'purchase_amount_var', 'purchase_amount_inter_quartile_ratio',\n",
    "       'duration_skew_old', 'transactions_count_old', 'duration_min_old',\n",
    "       'category_1_1_purchase_amount_mean_ratio', 'purchase_amount_pct_25_old',\n",
    "       'category_3_1_purchase_amount_mean_total', 'duration_max', 'installments_skew_old',\n",
    "       'purchase_Day_min', 'amount_month_ratio_max', 'CLV_old', 'purchase_amount_head_sum_old',\n",
    "       'amount_month_ratio_skew_old', 'installments_var_old', 'installments_ratio', 'month_diff_var_old',\n",
    "       'category_1_1_purchase_amount_sum_old', 'category_1_1_purchase_amount_sum_ratio',\n",
    "       'purchase_amount_median_old', 'category_3_1_purchase_amount_sum_old',\n",
    "       'category_3_0_purchase_amount_mean_ratio', 'installments_total', 'feature_1', 'month_lag_skew',\n",
    "       'merchant_category_id_nunique_old', 'new_last_buy', 'purchase_amount_head_max_old',\n",
    "       'observed_elapsed_time', 'category_1_1_purchase_amount_mean_old', 'amount_month_ratio_var',\n",
    "       'purchase_amount_pct_25_ratio', 'category_1_0_purchase_amount_mean',\n",
    "       'category_3_1_purchase_amount_mean_ratio', 'category_2_1.0_purchase_amount_mean_old',\n",
    "       'category_1_1_purchase_amount_mean_total', 'category_3_1_purchase_amount_mean_old',\n",
    "       'purchase_Dayofweek_mean', 'category_2_5.0_purchase_amount_sum',\n",
    "       'category_1_0_purchase_amount_sum_total', 'category_3_0_purchase_amount_mean_old', 'price_min',\n",
    "       'installments_mean_old', 'purchase_amount_skew', 'category_3_mean_old', 'purchase_amount_skew_ratio',\n",
    "       'price_min_old', 'category_3_2_purchase_amount_sum_ratio', 'category_1_0_purchase_amount_mean_old',\n",
    "       'category_3_1_purchase_amount_sum_total', 'amount_month_ratio_skew', 'purchase_amount_max',\n",
    "       'category_2_1.0_purchase_amount_sum_total', 'category_3_mean_mean_old', 'price_total',\n",
    "       'installments_pct_75_old', 'amount_month_ratio_min', 'purchase_Hour_nunique_old',\n",
    "       'category_2_1.0_purchase_amount_mean_total', 'installments_mean_total',\n",
    "       'purchase_amount_inter_quartile_old', 'category_1_0_purchase_amount_mean_total',\n",
    "       'purchase_amount_min_old', 'card_id_total', 'subsector_id_nunique_old', 'price_mean', 'duration_min',\n",
    "       'days_feature3_ratio', 'price_max_old', 'duration_skew', 'category_3_2_purchase_amount_sum_old',\n",
    "       'price_max', 'category_2_mean_old', 'category_3_0_purchase_amount_sum_total',\n",
    "       'category_3_2_purchase_amount_mean_old', 'category_3_0_purchase_amount_mean_total',\n",
    "       'installments_head_sum_old', 'duration_min_total', 'inverse_avg_transactions_per_day_old',\n",
    "       'month_diff_var', 'inverse_avg_transactions_per_day', 'purchase_amount_min',\n",
    "       'amount_month_ratio_min_total', 'category_2_mean_mean_old', 'price_mean_old',\n",
    "       'authorized_flag_1_purchase_amount_mean', 'purchase_amount_inter_quartile',\n",
    "       'installments_tail_sum_old', 'category_2_5.0_purchase_amount_mean_old',\n",
    "       'category_2_5.0_purchase_amount_mean', 'merchant_category_id_nunique', 'purchase_amount_total_min',\n",
    "       'purchase_amount_pct_25', 'category_3_2_purchase_amount_sum_total',\n",
    "       'category_3_2_purchase_amount_mean_total', 'authorized_flag_1_purchase_amount_sum_old',\n",
    "       'installments_var', 'purchase_Day_nunique_old', 'category_3_2_purchase_amount_mean_ratio',\n",
    "       'price_var', 'purchase_amount_pct_75_old', 'category_2_3.0_purchase_amount_mean',\n",
    "       'category_2_5.0_purchase_amount_sum_ratio', 'month_lag_min_old', 'amount_month_ratio_max_old',\n",
    "       'purchase_Month_min', 'purchase_amount_sum_old', 'price_sum', 'duration_max_old',\n",
    "       'amount_month_ratio_mean_old', 'category_2_3.0_purchase_amount_sum', 'amount_month_ratio_var_old',\n",
    "       'category_2_3.0_purchase_amount_mean_old', 'purchase_amount_head_sum', 'month_lag_min_total',\n",
    "       'amount_month_ratio_mean', 'month_lag_max_old', 'month_diff_skew',\n",
    "       'category_2_3.0_purchase_amount_sum_ratio', 'duration_var_old', 'purchase_Hour_min_old',\n",
    "       'duration_mean', 'purchase_amount_head_max', 'category_3_mean_mean',\n",
    "       'category_2_2.0_purchase_amount_mean', 'installments_mean', 'purchase_amount_median',\n",
    "       'category_2_3.0_purchase_amount_mean_ratio', 'purchase_amount_tail_sum', 'duration_max_total',\n",
    "       'purchase_amount_tail_max', 'purchase_Month_max_old', 'purchase_amount_max_old',\n",
    "       'category_2_5.0_purchase_amount_sum_old', 'card_id_count_old', 'feature_2',\n",
    "       'amount_month_ratio_mean_total', 'category_2_4.0_purchase_amount_sum', 'amount_month_ratio_max_total',\n",
    "       'purchase_amount_pct_75', 'month_lag_max_total', 'feature_var', 'duration_mean_old',\n",
    "       'category_2_2.0_purchase_amount_sum', 'category_2_4.0_purchase_amount_sum_ratio',\n",
    "       'purchase_amount_total', 'purchase_amount_var_old', 'purchase_amount_total_max',\n",
    "       'category_2_4.0_purchase_amount_mean', 'category_2_5.0_purchase_amount_mean_ratio',\n",
    "       'purchase_Hour_max', 'purchase_Dayofweek_min', 'purchased_on_weekend_mean', 'subsector_id_nunique',\n",
    "       'feature_sum', 'installments_skew', 'purchase_amount_fi_last_old_total', 'purchase_Month_min_old',\n",
    "       'purchase_amount_mean', 'avg_spend_per_transaction', 'installments_pct_75',\n",
    "       'purchase_amount_mean_old', 'merchant_id_nunique', 'duration_mean_total', 'transactions_count',\n",
    "       'purchase_amount_total_mean', 'installments_max_total', 'category_2_5.0_purchase_amount_mean_total',\n",
    "       'installments_max_old', 'card_id_count_total', 'category_2_5.0_purchase_amount_sum_total',\n",
    "       'category_2_3.0_purchase_amount_sum_total', 'category_2_3.0_purchase_amount_mean_total',\n",
    "       'installments_tail_sum', 'category_2_2.0_purchase_amount_sum_old', 'quarter', 'installments_sum',\n",
    "       'purchase_Week_nunique', 'purchase_amount_min_total', 'category_2_mean_mean',\n",
    "       'category_2_4.0_purchase_amount_mean_ratio', 'month_lag_max', 'purchase_Day_min_old',\n",
    "       'avg_spend_per_merchant', 'category_2_3.0_purchase_amount_sum_old', 'installments_tail_max_old',\n",
    "       'purchase_amount_sum_total', 'installments_median', 'purchase_Hour_nunique',\n",
    "       'category_2_2.0_purchase_amount_sum_ratio', 'category_2_2.0_purchase_amount_mean_total',\n",
    "       'installments_pct_25', 'purchase_amount_max_total', 'purchase_Day_nunique',\n",
    "       'category_2_2.0_purchase_amount_mean_old', 'feature_mean',\n",
    "       'category_2_4.0_purchase_amount_mean_total', 'category_2_4.0_purchase_amount_sum_total',\n",
    "       'installments_min', 'installments_head_max_old', 'category_2_mean', 'observation_date_old',\n",
    "       'category_2_4.0_purchase_amount_mean_old', 'category_3_mean',\n",
    "       'category_2_4.0_purchase_amount_sum_old', 'card_id_count', 'purchase_Hour_max_old',\n",
    "       'installments_max', 'avg_spend_per_transaction_old', 'installments_head_sum', 'installments_tail_max',\n",
    "       'category_2_2.0_purchase_amount_mean_ratio', 'purchase_amount_mean_total', 'installments_head_max',\n",
    "       'purchase_Dayofweek_max', 'merchant_category_repurchase',\n",
    "       'authorized_flag_1_purchase_amount_mean_old', 'installments_pct_25_old', 'feature_max',\n",
    "       'installments_median_old', 'purchased_before_issue', 'month_lag_min', 'repurchase_merchant_rate',\n",
    "       'category_2_2.0_purchase_amount_sum_total', 'feature_3', 'avg_trans_per_merchant', 'feature_min',\n",
    "       'installments_min_old']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['merchant_address_id_nunique',\n",
       " 'merchant_rating_nunique',\n",
       " 'merchant_rating_mean',\n",
       " 'merchant_rating_min',\n",
       " 'merchant_rating_max',\n",
       " 'purchase_amount_gmean',\n",
       " 'purchase_amount_hmean',\n",
       " 'category_4_mean',\n",
       " 'numerical_range_mean',\n",
       " 'numerical_range_min',\n",
       " 'numerical_range_max',\n",
       " 'numerical_range_skew',\n",
       " 'merchant_address_id_nunique_old',\n",
       " 'merchant_rating_nunique_old',\n",
       " 'merchant_rating_mean_old',\n",
       " 'merchant_rating_min_old',\n",
       " 'merchant_rating_max_old',\n",
       " 'purchase_amount_gmean_old',\n",
       " 'purchase_amount_hmean_old',\n",
       " 'category_4_mean_old',\n",
       " 'numerical_range_mean_old',\n",
       " 'numerical_range_min_old',\n",
       " 'numerical_range_max_old',\n",
       " 'numerical_range_skew_old',\n",
       " 'category_4_mean_total',\n",
       " 'category_4_mean_ratio',\n",
       " 'category_1_mean_ratio',\n",
       " 'numerical_range_mean_total',\n",
       " 'numerical_range_mean_ratio',\n",
       " 'merchant_rating_mean_ratio',\n",
       " 'purchase_amount_gmean_total',\n",
       " 'purchase_amount_hmean_total',\n",
       " 'purchase_amount_gmean_ratio',\n",
       " 'purchase_amount_hmean_ratio']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[c for c in df_train_columns if c not in cols_ful_352]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = base_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "376"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['feature_1', 'feature_2', 'feature_3']"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_columns[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# x = x[df_train_columns]\n",
    "# x_t = x_t[df_train_columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "lgb_param = {\n",
    "        'task': 'train',\n",
    "        'boosting': 'goss',\n",
    "        'objective': 'regression',\n",
    "        'metric': 'rmse',\n",
    "        'learning_rate': 0.01,\n",
    "        'subsample': 0.9855232997390695,\n",
    "        'max_depth': 7,\n",
    "        'top_rate': 0.9064148448434349,\n",
    "        'num_leaves': 123,\n",
    "        'min_child_weight': 41.9612869171337,\n",
    "        'other_rate': 0.0721768246018207,\n",
    "        'reg_alpha': 9.677537745007898,\n",
    "        'colsample_bytree': 0.5665320670155495,\n",
    "        'min_split_gain': 9.820197773625843,\n",
    "        'reg_lambda': 8.2532317400459,\n",
    "        'min_data_in_leaf': 21,\n",
    "        'verbose': -1,\n",
    "        'seed':int(2**5),\n",
    "        'bagging_seed':int(2**5),\n",
    "        'drop_seed':int(2**5)\n",
    "        }\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "trn_data = lgb.Dataset(x[df_train_columns], label=y)\n",
    "val_data = lgb.Dataset(x_t[df_train_columns], label=y_t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2min 37s, sys: 435 ms, total: 2min 37s\n",
      "Wall time: 26.5 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "3.645373200131201"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_round = 10000\n",
    "%time clf = lgb.train(lgb_param, trn_data, num_round, valid_sets = [trn_data, val_data], verbose_eval=0, early_stopping_rounds = 200)\n",
    "oof = clf.predict(x_t[df_train_columns], num_iteration=clf.best_iteration)\n",
    "np.sqrt(mean_squared_error(oof, y_t))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "339"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns[:-3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_train_columns = df_train_columns[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_add = ['merchant_address_id_nunique',\n",
    " 'merchant_rating_nunique',\n",
    " 'merchant_rating_mean',\n",
    " 'merchant_rating_min',\n",
    " 'merchant_rating_max',\n",
    " 'purchase_amount_gmean',\n",
    " 'purchase_amount_hmean',\n",
    " 'category_4_mean',\n",
    " 'numerical_range_mean',\n",
    " 'numerical_range_min',\n",
    " 'numerical_range_max',\n",
    " 'numerical_range_skew',\n",
    " 'merchant_address_id_nunique_old',\n",
    " 'merchant_rating_nunique_old',\n",
    " 'merchant_rating_mean_old',\n",
    " 'merchant_rating_min_old',\n",
    " 'merchant_rating_max_old',\n",
    " 'purchase_amount_gmean_old',\n",
    " 'purchase_amount_hmean_old',\n",
    " 'category_4_mean_old',\n",
    " 'numerical_range_mean_old',\n",
    " 'numerical_range_min_old',\n",
    " 'numerical_range_max_old',\n",
    " 'numerical_range_skew_old',\n",
    " 'category_4_mean_total',\n",
    " 'category_4_mean_ratio',\n",
    " 'category_1_mean_ratio',\n",
    " 'numerical_range_mean_total',\n",
    " 'numerical_range_mean_ratio',\n",
    " 'merchant_rating_mean_ratio',\n",
    " 'purchase_amount_gmean_total',\n",
    " 'purchase_amount_hmean_total',\n",
    " 'purchase_amount_gmean_ratio',\n",
    " 'purchase_amount_hmean_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cols_to_add = df_train_columns[:-3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cols_to_add = [c for c in df_train_columns if c not in all_cols_sorted[:3]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cols_to_add = [c for c in df_train_columns if c not in base_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_excluded = ['purchase_date_max', 'purchase_date_max_old', 'card_id', 'first_active_month',\n",
    "                 'target','outliers','card_id_size', 'card_id_size_old', \n",
    "                 'purchase_date_min', 'purchase_date_min_old','first_active_monthYear',\n",
    "                 'first_active_monthMonth',\n",
    "                 'first_active_monthWeek',\n",
    "                 'first_active_monthDay',\n",
    "                 'first_active_monthDayofweek',\n",
    "                 'first_active_monthDayofyear',\n",
    "                 'first_active_monthIs_month_end',\n",
    "                 'first_active_monthIs_month_start',\n",
    "                 'first_active_monthIs_quarter_end',\n",
    "                 'first_active_monthIs_quarter_start',\n",
    "                 'first_active_monthIs_year_end',\n",
    "                 'Black_Friday_2017_mean',\n",
    "                 'amount_month_ratio_max',\n",
    "                 'purchase_Month_mean_old',\n",
    "                 'purchase_amount_total_max',\n",
    "                 'first_active_monthIs_year_start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_included = ['feature_1','feature_2','feature_3','transactions_count','subsector_id_nunique','merchant_id_nunique','merchant_category_id_nunique','purchase_Month_mean','purchase_Month_min','purchase_Month_max','purchase_Week_nunique','purchase_Week_mean','purchase_Week_min','purchase_Week_max','purchase_Dayofweek_mean','purchase_Dayofweek_min','purchase_Dayofweek_max','purchase_Day_nunique','purchase_Day_mean','purchase_Day_min','purchase_Day_max','purchase_Hour_nunique','purchase_Hour_mean','purchase_Hour_min','purchase_Hour_max','purchase_amount_sum','purchase_amount_max','purchase_amount_min','purchase_amount_mean','purchase_amount_var','purchase_amount_skew','installments_sum','installments_max','installments_mean','installments_var','installments_skew','month_lag_max','month_lag_min','month_lag_mean','month_lag_var','month_lag_skew','month_diff_mean','month_diff_var','month_diff_skew','purchased_on_weekend_mean','category_1_mean','category_2_mean','category_3_mean','card_id_count','price_mean','price_max','price_min','price_var','Christmas_Day_2017_mean','Children_day_2017_mean','Black_Friday_2017_mean','Mothers_Day_2018_mean','duration_mean','duration_min','duration_max','duration_var','duration_skew','amount_month_ratio_mean','amount_month_ratio_min','amount_month_ratio_max','amount_month_ratio_var','amount_month_ratio_skew','category_2_mean_mean','category_3_mean_mean','purchase_date_diff','purchase_date_average','purchase_date_uptonow','purchase_date_uptomin','transactions_count_old','subsector_id_nunique_old','merchant_id_nunique_old','merchant_category_id_nunique_old','purchase_Month_nunique','purchase_Month_mean_old','purchase_Month_min_old','purchase_Month_max_old','purchase_Week_nunique_old','purchase_Week_mean_old','purchase_Week_min_old','purchase_Week_max_old','purchase_Dayofweek_mean_old','purchase_Day_nunique_old','purchase_Day_mean_old','purchase_Day_min_old','purchase_Hour_nunique_old','purchase_Hour_mean_old','purchase_Hour_min_old','purchase_Hour_max_old','purchase_amount_sum_old','purchase_amount_max_old','purchase_amount_min_old','purchase_amount_mean_old','purchase_amount_var_old','purchase_amount_skew_old','installments_sum_old','installments_max_old','installments_mean_old','installments_var_old','installments_skew_old','month_lag_max_old','month_lag_min_old','month_lag_mean_old','month_lag_var_old','month_lag_skew_old','month_diff_max','month_diff_min','month_diff_mean_old','month_diff_var_old','month_diff_skew_old','authorized_flag_mean','purchased_on_weekend_mean_old','category_1_mean_old','category_2_mean_old','category_3_mean_old','card_id_count_old','price_sum','price_mean_old','price_max_old','price_min_old','price_var_old','Christmas_Day_2017_mean_old','Mothers_Day_2017_mean','fathers_day_2017_mean','Children_day_2017_mean_old','Valentine_Day_2017_mean','Black_Friday_2017_mean_old','Mothers_Day_2018_mean_old','duration_mean_old','duration_min_old','duration_max_old','duration_var_old','duration_skew_old','amount_month_ratio_mean_old','amount_month_ratio_min_old','amount_month_ratio_max_old','amount_month_ratio_var_old','amount_month_ratio_skew_old','category_2_mean_mean_old','category_3_mean_mean_old','purchase_date_diff_old','purchase_date_average_old','purchase_date_uptonow_old','purchase_date_uptomin_old','quarter','observed_elapsed_time','days_feature1','days_feature2','days_feature3','days_feature1_ratio','days_feature2_ratio','days_feature3_ratio','feature_sum','feature_mean','feature_max','feature_min','feature_var','card_id_total','card_id_count_total','card_id_count_ratio','purchase_amount_total','purchase_amount_total_mean','purchase_amount_total_max','purchase_amount_total_min','purchase_amount_sum_ratio','hist_first_buy','new_first_buy','hist_last_buy','new_last_buy','month_diff_ratio','installments_total','installments_ratio','price_total','CLV','CLV_old','CLV_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "180"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols_included)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = [c for c in cols_included if c not in cols_excluded]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "exp_cols = ['merchant_address_id_nunique', 'merchant_rating_nunique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = df_train_columns + exp_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "178"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_tested = ['merchant_rating_mean',\n",
    "'merchant_rating_min',\n",
    "'merchant_rating_max',\n",
    "'purchase_amount_pct_75',\n",
    "'purchase_amount_pct_25',\n",
    "'purchase_amount_median',\n",
    "'purchase_amount_head_sum',\n",
    "'purchase_amount_head_max',\n",
    "'purchase_amount_tail_sum',\n",
    "'purchase_amount_tail_max',\n",
    "'purchase_amount_gmean',\n",
    "'purchase_amount_hmean',\n",
    "'installments_pct_75',\n",
    "'installments_pct_25',\n",
    "'installments_median',\n",
    "'installments_min',\n",
    "'installments_head_sum',\n",
    "'installments_head_max',\n",
    "'installments_tail_sum',\n",
    "'installments_tail_max',\n",
    "'category_4_mean',\n",
    "'numerical_range_mean',\n",
    "'numerical_range_min',\n",
    "'numerical_range_max',\n",
    "'numerical_range_skew',\n",
    "'purchase_date_successive_diff_mean',\n",
    "'inverse_avg_transactions_per_day',\n",
    "'repurchase_merchant_rate',\n",
    "'merchant_category_repurchase',\n",
    "'avg_spend_per_merchant',\n",
    "'avg_trans_per_merchant',\n",
    "'avg_spend_per_transaction',\n",
    "'merchant_address_id_nunique_old',\n",
    "'merchant_rating_nunique_old',\n",
    "'merchant_rating_mean_old',\n",
    "'merchant_rating_min_old',\n",
    "'merchant_rating_max_old',\n",
    "'purchase_amount_pct_75_old',\n",
    "'purchase_amount_pct_25_old',\n",
    "'purchase_amount_median_old',\n",
    "'purchase_amount_head_sum_old',\n",
    "'purchase_amount_head_max_old',\n",
    "'purchase_amount_tail_sum_old',\n",
    "'purchase_amount_tail_max_old',\n",
    "'purchase_amount_gmean_old',\n",
    "'purchase_amount_hmean_old',\n",
    "'installments_pct_75_old',\n",
    "'installments_pct_25_old',\n",
    "'installments_median_old',\n",
    "'installments_min_old',\n",
    "'installments_head_sum_old',\n",
    "'installments_head_max_old',\n",
    "'installments_tail_sum_old',\n",
    "'installments_tail_max_old',\n",
    "'category_4_mean_old',\n",
    "'numerical_range_mean_old',\n",
    "'numerical_range_min_old',\n",
    "'numerical_range_max_old',\n",
    "'numerical_range_skew_old',\n",
    "'price_skew',\n",
    "'purchase_date_successive_diff_mean_old',\n",
    "'inverse_avg_transactions_per_day_old',\n",
    "'repurchase_merchant_rate_old',\n",
    "'merchant_category_repurchase_old',\n",
    "'avg_spend_per_merchant_old',\n",
    "'avg_trans_per_merchant_old',\n",
    "'avg_spend_per_transaction_old',\n",
    "'category_1_0_purchase_amount_mean',\n",
    "'category_1_1_purchase_amount_mean',\n",
    "'category_1_0_purchase_amount_sum',\n",
    "'category_1_1_purchase_amount_sum',\n",
    "'category_2_1.0_purchase_amount_mean',\n",
    "'category_2_2.0_purchase_amount_mean',\n",
    "'category_2_3.0_purchase_amount_mean',\n",
    "'category_2_4.0_purchase_amount_mean',\n",
    "'category_2_5.0_purchase_amount_mean',\n",
    "'category_2_1.0_purchase_amount_sum',\n",
    "'category_2_2.0_purchase_amount_sum',\n",
    "'category_2_3.0_purchase_amount_sum',\n",
    "'category_2_4.0_purchase_amount_sum',\n",
    "'category_2_5.0_purchase_amount_sum',\n",
    "'category_3_0_purchase_amount_mean',\n",
    "'category_3_1_purchase_amount_mean',\n",
    "'category_3_2_purchase_amount_mean',\n",
    "'category_3_0_purchase_amount_sum',\n",
    "'category_3_1_purchase_amount_sum',\n",
    "'category_3_2_purchase_amount_sum',\n",
    "'authorized_flag_0_purchase_amount_mean',\n",
    "'authorized_flag_1_purchase_amount_mean',\n",
    "'authorized_flag_0_purchase_amount_sum',\n",
    "'authorized_flag_1_purchase_amount_sum',\n",
    "'category_1_0_purchase_amount_mean_old',\n",
    "'category_1_1_purchase_amount_mean_old',\n",
    "'category_1_0_purchase_amount_sum_old',\n",
    "'category_1_1_purchase_amount_sum_old',\n",
    "'category_2_1.0_purchase_amount_mean_old',\n",
    "'category_2_2.0_purchase_amount_mean_old',\n",
    "'category_2_3.0_purchase_amount_mean_old',\n",
    "'category_2_4.0_purchase_amount_mean_old',\n",
    "'category_2_5.0_purchase_amount_mean_old',\n",
    "'category_2_1.0_purchase_amount_sum_old',\n",
    "'category_2_2.0_purchase_amount_sum_old',\n",
    "'category_2_3.0_purchase_amount_sum_old',\n",
    "'category_2_4.0_purchase_amount_sum_old',\n",
    "'category_2_5.0_purchase_amount_sum_old',\n",
    "'category_3_0_purchase_amount_mean_old',\n",
    "'category_3_1_purchase_amount_mean_old',\n",
    "'category_3_2_purchase_amount_mean_old',\n",
    "'category_3_0_purchase_amount_sum_old',\n",
    "'category_3_1_purchase_amount_sum_old',\n",
    "'category_3_2_purchase_amount_sum_old',\n",
    "'authorized_flag_1_purchase_amount_mean_old',\n",
    "'authorized_flag_1_purchase_amount_sum_old',\n",
    "'observation_date_old',\n",
    "'purchased_before_issue',\n",
    "'month_diff_mean_total',\n",
    "'month_lag_mean_total',\n",
    "'month_lag_max_total',\n",
    "'month_lag_min_total',\n",
    "'category_1_mean_total',\n",
    "'category_4_mean_total',\n",
    "'category_4_mean_ratio',\n",
    "'category_1_mean_ratio',\n",
    "'numerical_range_mean_total',\n",
    "'numerical_range_mean_ratio',\n",
    "'merchant_rating_mean_ratio',\n",
    "'installments_mean_total',\n",
    "'installments_max_total',\n",
    "'duration_mean_total',\n",
    "'duration_min_total',\n",
    "'duration_max_total',\n",
    "'amount_month_ratio_mean_total',\n",
    "'amount_month_ratio_min_total',\n",
    "'amount_month_ratio_max_total',\n",
    "'category_1_0_purchase_amount_mean_total',\n",
    "'category_1_1_purchase_amount_mean_total',\n",
    "'category_1_0_purchase_amount_sum_total',\n",
    "'category_1_1_purchase_amount_sum_total',\n",
    "'category_2_1.0_purchase_amount_mean_total']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_add = [c for c in train_df2.columns.values if c not in df_train_columns and c not in cols_excluded and c not in cols_tested]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols_to_add)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.646777128761576 CV score\n"
     ]
    }
   ],
   "source": [
    "base_score = lgb_train_fn(x, y, df_train_columns, 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_score = 3.646777128761576"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "# base_score2 = 3.708340770058834"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "# base_score = 3.645373200131201"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "selected_cols = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_deltas_df = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols = df_train_columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "178"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(final_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_d = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cols_to_add = cols_to_add[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# final_cols = cols_ful_352[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# final_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "target = x['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((172638, 386), (172638,), (172638,))"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.shape, target.shape, y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def lgb_train_fn(train_df, target, trn_cols,  n_fold):\n",
    "    folds = StratifiedKFold(n_splits=n_fold, shuffle=True, random_state=4590)\n",
    "    oof = np.zeros(len(train_df))\n",
    "\n",
    "    for fold_, (trn_idx, val_idx) in enumerate(folds.split(train_df,train_df['outliers'].values)):\n",
    "        trn_data = lgb.Dataset(train_df.iloc[trn_idx][trn_cols], label=target.iloc[trn_idx])\n",
    "        val_data = lgb.Dataset(train_df.iloc[val_idx][trn_cols], label=target.iloc[val_idx])\n",
    "\n",
    "        num_round = 10000\n",
    "        clf = lgb.train(lgb_param, trn_data, num_round, valid_sets = [trn_data, val_data], verbose_eval=0, early_stopping_rounds = 200)\n",
    "        oof[val_idx] = clf.predict(train_df.iloc[val_idx][trn_cols], num_iteration=clf.best_iteration)\n",
    "\n",
    "    print(np.sqrt(mean_squared_error(oof, target)), 'CV score')\n",
    "    return np.sqrt(mean_squared_error(oof, target))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "178"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.646777128761576 CV score\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "3.646777128761576"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgb_train_fn(x, y, df_train_columns, 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "# base_score = 3.648290144094502"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(55, 178)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols_to_add), len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "179 lg_cols category_2_2.0_purchase_amount_mean_total\n",
      "3.6514452224208203 CV score\n",
      "179 lg_cols category_2_3.0_purchase_amount_mean_total\n",
      "3.660536446287273 CV score\n",
      "179 lg_cols category_2_4.0_purchase_amount_mean_total\n",
      "3.6602225789441944 CV score\n",
      "179 lg_cols category_2_5.0_purchase_amount_mean_total\n",
      "3.6597247082757463 CV score\n",
      "179 lg_cols category_2_1.0_purchase_amount_sum_total\n",
      "3.6603143895479304 CV score\n",
      "179 lg_cols category_2_2.0_purchase_amount_sum_total\n",
      "3.6600706825676332 CV score\n",
      "179 lg_cols category_2_3.0_purchase_amount_sum_total\n",
      "3.6476854944448283 CV score\n",
      "179 lg_cols category_2_4.0_purchase_amount_sum_total\n",
      "3.6603555892429696 CV score\n",
      "179 lg_cols category_2_5.0_purchase_amount_sum_total\n",
      "3.659777829445138 CV score\n",
      "179 lg_cols category_3_0_purchase_amount_mean_total\n",
      "3.659913962669658 CV score\n",
      "179 lg_cols category_3_1_purchase_amount_mean_total\n",
      "3.6597285298056192 CV score\n",
      "179 lg_cols category_3_2_purchase_amount_mean_total\n",
      "3.659882319087217 CV score\n",
      "179 lg_cols category_3_0_purchase_amount_sum_total\n",
      "3.6479348449317257 CV score\n",
      "179 lg_cols category_3_1_purchase_amount_sum_total\n",
      "3.660387299133759 CV score\n",
      "179 lg_cols category_3_2_purchase_amount_sum_total\n",
      "3.659941589344204 CV score\n",
      "179 lg_cols category_1_0_purchase_amount_mean_ratio\n",
      "3.65942836702963 CV score\n",
      "179 lg_cols category_1_1_purchase_amount_mean_ratio\n",
      "3.6604459228887323 CV score\n",
      "179 lg_cols category_1_0_purchase_amount_sum_ratio\n",
      "3.6478243192360633 CV score\n",
      "179 lg_cols category_1_1_purchase_amount_sum_ratio\n",
      "3.6539354985720753 CV score\n",
      "179 lg_cols category_2_1.0_purchase_amount_mean_ratio\n",
      "3.6605938422687077 CV score\n",
      "179 lg_cols category_2_2.0_purchase_amount_mean_ratio\n",
      "3.6596152798820314 CV score\n",
      "179 lg_cols category_2_3.0_purchase_amount_mean_ratio\n",
      "3.6601586813081175 CV score\n",
      "179 lg_cols category_2_4.0_purchase_amount_mean_ratio\n",
      "3.660169510176692 CV score\n",
      "179 lg_cols category_2_5.0_purchase_amount_mean_ratio\n",
      "3.660119585584449 CV score\n",
      "179 lg_cols category_2_1.0_purchase_amount_sum_ratio\n",
      "3.660755486742916 CV score\n",
      "179 lg_cols category_2_2.0_purchase_amount_sum_ratio\n",
      "3.6602730816511415 CV score\n",
      "179 lg_cols category_2_3.0_purchase_amount_sum_ratio\n",
      "3.660483530882945 CV score\n",
      "179 lg_cols category_2_4.0_purchase_amount_sum_ratio\n",
      "3.660291233602776 CV score\n",
      "179 lg_cols category_2_5.0_purchase_amount_sum_ratio\n",
      "3.6601505078271943 CV score\n",
      "179 lg_cols category_3_0_purchase_amount_mean_ratio\n",
      "3.6478891896294416 CV score\n",
      "179 lg_cols category_3_1_purchase_amount_mean_ratio\n",
      "3.6481657951484343 CV score\n",
      "179 lg_cols category_3_2_purchase_amount_mean_ratio\n",
      "3.6600092777161533 CV score\n",
      "179 lg_cols category_3_0_purchase_amount_sum_ratio\n",
      "3.6593589433756124 CV score\n",
      "179 lg_cols category_3_1_purchase_amount_sum_ratio\n",
      "3.659414622720865 CV score\n",
      "179 lg_cols category_3_2_purchase_amount_sum_ratio\n",
      "3.6600417203500903 CV score\n",
      "179 lg_cols purchase_amount_sum_total\n",
      "3.6605671810942786 CV score\n",
      "179 lg_cols purchase_amount_pct_75_ratio\n",
      "3.6595316066739323 CV score\n",
      "179 lg_cols purchase_amount_pct_25_ratio\n",
      "3.647808305962036 CV score\n",
      "179 lg_cols purchase_amount_inter_quartile\n",
      "3.6600297527574424 CV score\n",
      "179 lg_cols purchase_amount_inter_quartile_old\n",
      "3.6477438510850204 CV score\n",
      "179 lg_cols purchase_amount_inter_quartile_ratio\n",
      "3.6601784984265966 CV score\n",
      "179 lg_cols purchase_amount_mean_total\n",
      "3.654887912824904 CV score\n",
      "179 lg_cols purchase_amount_mean_ratio\n",
      "3.6604096609137975 CV score\n",
      "179 lg_cols purchase_amount_median_ratio\n",
      "3.660453524554481 CV score\n",
      "179 lg_cols purchase_amount_max_total\n",
      "3.6599772451477124 CV score\n",
      "179 lg_cols purchase_amount_min_total\n",
      "3.6599877387396034 CV score\n",
      "179 lg_cols purchase_amount_skew_ratio\n",
      "3.6603753569758317 CV score\n",
      "179 lg_cols purchase_amount_before_after_ratio\n",
      "3.659649189269222 CV score\n",
      "179 lg_cols purchase_amount_first_last_diff\n",
      "3.654994570709901 CV score\n",
      "179 lg_cols purchase_amount_fi_last_old_total\n",
      "3.660043478649737 CV score\n",
      "179 lg_cols purchase_amount_fi_last_new_total\n",
      "3.6604308942263404 CV score\n",
      "179 lg_cols purchase_amount_gmean_total\n",
      "3.6598804131807863 CV score\n",
      "179 lg_cols purchase_amount_hmean_total\n",
      "3.66010183912888 CV score\n",
      "179 lg_cols purchase_amount_gmean_ratio\n",
      "3.647285649353983 CV score\n",
      "179 lg_cols purchase_amount_hmean_ratio\n",
      "3.6595669246407527 CV score\n",
      "CPU times: user 7h 21min 4s, sys: 3min 1s, total: 7h 24min 5s\n",
      "Wall time: 1h 19min 49s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "for c in cols_to_add:\n",
    "    lgb_cols = final_cols + [c]\n",
    "    print(len(lgb_cols), 'lg_cols', c)\n",
    "    score = lgb_train_fn(x, y, lgb_cols, 5)\n",
    "#     trn_data = lgb.Dataset(x[lgb_cols], label=y)\n",
    "#     val_data = lgb.Dataset(x_t[lgb_cols], label=y_t)\n",
    "#     clf = lgb.train(lgb_param, trn_data, num_round, \n",
    "#                     valid_sets = [trn_data, val_data], verbose_eval=0, early_stopping_rounds = 200)\n",
    "#     oof = clf.predict(x_t[lgb_cols], num_iteration=clf.best_iteration)\n",
    "#     score = np.sqrt(mean_squared_error(y_t, oof))\n",
    "    delta = base_score - score\n",
    "    fe_d[c] = delta\n",
    "    if delta > 0:\n",
    "        base_score = score\n",
    "        selected_cols.append(c)\n",
    "        print('Selected cols', c)\n",
    "        print('Selected col delta', delta)\n",
    "        print(' score with col', score)\n",
    "        np.save('selecte_cols_extra', selected_cols)\n",
    "        final_cols = final_cols + [c]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "651"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(selected_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['merchant_address_id_nunique', 'merchant_rating_nunique']"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_colsected_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "__init__() got an unexpected keyword argument 'max_rows'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m~/anaconda3/envs/fastai/lib/python3.6/site-packages/IPython/core/formatters.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    343\u001b[0m             \u001b[0mmethod\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_real_method\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprint_method\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    344\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mmethod\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 345\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    346\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    347\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/fastai/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m_repr_html_\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    694\u001b[0m         \u001b[0mSee\u001b[0m \u001b[0mAlso\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    695\u001b[0m         \u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 696\u001b[0;31m         \u001b[0mto_html\u001b[0m \u001b[0;34m:\u001b[0m \u001b[0mConvert\u001b[0m \u001b[0mDataFrame\u001b[0m \u001b[0mto\u001b[0m \u001b[0mHTML\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    697\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    698\u001b[0m         \u001b[0mExamples\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/fastai/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36mto_html\u001b[0;34m(self, buf, columns, col_space, header, index, na_rep, formatters, float_format, sparsify, index_names, justify, bold_rows, classes, escape, max_rows, max_cols, show_dimensions, notebook, decimal, border, table_id)\u001b[0m\n\u001b[1;32m   2035\u001b[0m             \u001b[0mDictionary\u001b[0m \u001b[0mmapping\u001b[0m \u001b[0mcolumns\u001b[0m \u001b[0mcontaining\u001b[0m \u001b[0mdatetime\u001b[0m \u001b[0mtypes\u001b[0m \u001b[0mto\u001b[0m \u001b[0mstata\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2036\u001b[0m             \u001b[0minternal\u001b[0m \u001b[0mformat\u001b[0m \u001b[0mto\u001b[0m \u001b[0muse\u001b[0m \u001b[0mwhen\u001b[0m \u001b[0mwriting\u001b[0m \u001b[0mthe\u001b[0m \u001b[0mdates\u001b[0m\u001b[0;34m.\u001b[0m \u001b[0mOptions\u001b[0m \u001b[0mare\u001b[0m \u001b[0;34m'tc'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2037\u001b[0;31m             \u001b[0;34m'td'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'tm'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'tw'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'th'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'tq'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'ty'\u001b[0m\u001b[0;34m.\u001b[0m \u001b[0mColumn\u001b[0m \u001b[0mcan\u001b[0m \u001b[0mbe\u001b[0m \u001b[0meither\u001b[0m \u001b[0man\u001b[0m \u001b[0minteger\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2038\u001b[0m             \u001b[0;32mor\u001b[0m \u001b[0ma\u001b[0m \u001b[0mname\u001b[0m\u001b[0;34m.\u001b[0m \u001b[0mDatetime\u001b[0m \u001b[0mcolumns\u001b[0m \u001b[0mthat\u001b[0m \u001b[0mdo\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mhave\u001b[0m \u001b[0ma\u001b[0m \u001b[0mconversion\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2039\u001b[0m             \u001b[0mspecified\u001b[0m \u001b[0mwill\u001b[0m \u001b[0mbe\u001b[0m \u001b[0mconverted\u001b[0m \u001b[0mto\u001b[0m \u001b[0;34m'tc'\u001b[0m\u001b[0;34m.\u001b[0m \u001b[0mRaises\u001b[0m \u001b[0mNotImplementedError\u001b[0m \u001b[0;32mif\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/fastai/lib/python3.6/site-packages/pandas/io/formats/format.py\u001b[0m in \u001b[0;36mto_html\u001b[0;34m(self, classes, notebook, border)\u001b[0m\n\u001b[1;32m    751\u001b[0m             \u001b[0mneed_leadsp\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfmt_columns\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mis_numeric_dtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtypes\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    752\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 753\u001b[0;31m             \u001b[0;32mdef\u001b[0m \u001b[0mspace_format\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    754\u001b[0m                 if (y not in self.formatters and\n\u001b[1;32m    755\u001b[0m                         need_leadsp[x] and not restrict_formatting):\n",
      "\u001b[0;31mTypeError\u001b[0m: __init__() got an unexpected keyword argument 'max_rows'"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                                        0\n",
       "merchant_address_id_nunique      0.000127\n",
       "merchant_rating_nunique          0.000143\n",
       "merchant_rating_mean            -0.012485\n",
       "merchant_rating_min             -0.012664\n",
       "merchant_rating_max             -0.012596\n",
       "purchase_amount_gmean           -0.012499\n",
       "purchase_amount_hmean           -0.012374\n",
       "category_4_mean                 -0.012043\n",
       "numerical_range_mean            -0.012250\n",
       "numerical_range_min             -0.012239\n",
       "numerical_range_max             -0.012532\n",
       "numerical_range_skew            -0.012169\n",
       "merchant_address_id_nunique_old -0.012472\n",
       "merchant_rating_nunique_old     -0.012318\n",
       "merchant_rating_mean_old        -0.012137\n",
       "merchant_rating_min_old         -0.012592\n",
       "merchant_rating_max_old         -0.012672\n",
       "purchase_amount_gmean_old       -0.011832\n",
       "purchase_amount_hmean_old       -0.011543\n",
       "category_4_mean_old             -0.012462\n",
       "numerical_range_mean_old        -0.012551\n",
       "numerical_range_min_old         -0.012424\n",
       "numerical_range_max_old         -0.012606\n",
       "numerical_range_skew_old        -0.012983\n",
       "category_4_mean_total           -0.011930\n",
       "category_4_mean_ratio           -0.012365\n",
       "category_1_mean_ratio           -0.012232\n",
       "numerical_range_mean_total      -0.012636\n",
       "numerical_range_mean_ratio      -0.012441\n",
       "merchant_rating_mean_ratio      -0.012515\n",
       "purchase_amount_gmean_total     -0.012707\n",
       "purchase_amount_hmean_total     -0.012714\n",
       "purchase_amount_gmean_ratio     -0.012887\n",
       "purchase_amount_hmean_ratio     -0.012894"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame.from_dict(fe_d, orient='index')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['feature_1',\n",
       " 'feature_2',\n",
       " 'feature_3',\n",
       " 'transactions_count',\n",
       " 'subsector_id_nunique',\n",
       " 'merchant_id_nunique',\n",
       " 'merchant_category_id_nunique',\n",
       " 'purchase_Month_mean',\n",
       " 'purchase_Month_min',\n",
       " 'purchase_Month_max',\n",
       " 'purchase_Week_nunique',\n",
       " 'purchase_Week_mean',\n",
       " 'purchase_Week_min',\n",
       " 'purchase_Week_max',\n",
       " 'purchase_Dayofweek_mean',\n",
       " 'purchase_Dayofweek_min',\n",
       " 'purchase_Dayofweek_max',\n",
       " 'purchase_Day_nunique',\n",
       " 'purchase_Day_mean',\n",
       " 'purchase_Day_min',\n",
       " 'purchase_Day_max',\n",
       " 'purchase_Hour_nunique',\n",
       " 'purchase_Hour_mean',\n",
       " 'purchase_Hour_min',\n",
       " 'purchase_Hour_max',\n",
       " 'purchase_amount_sum',\n",
       " 'purchase_amount_max',\n",
       " 'purchase_amount_min',\n",
       " 'purchase_amount_mean',\n",
       " 'purchase_amount_var',\n",
       " 'purchase_amount_skew',\n",
       " 'installments_sum',\n",
       " 'installments_max',\n",
       " 'installments_mean',\n",
       " 'installments_var',\n",
       " 'installments_skew',\n",
       " 'month_lag_max',\n",
       " 'month_lag_min',\n",
       " 'month_lag_mean',\n",
       " 'month_lag_var',\n",
       " 'month_lag_skew',\n",
       " 'month_diff_mean',\n",
       " 'month_diff_var',\n",
       " 'month_diff_skew',\n",
       " 'purchased_on_weekend_mean',\n",
       " 'category_1_mean',\n",
       " 'category_2_mean',\n",
       " 'category_3_mean',\n",
       " 'card_id_count',\n",
       " 'price_mean',\n",
       " 'price_max',\n",
       " 'price_min',\n",
       " 'price_var',\n",
       " 'Christmas_Day_2017_mean',\n",
       " 'Children_day_2017_mean',\n",
       " 'Black_Friday_2017_mean',\n",
       " 'Mothers_Day_2018_mean',\n",
       " 'duration_mean',\n",
       " 'duration_min',\n",
       " 'duration_max',\n",
       " 'duration_var',\n",
       " 'duration_skew',\n",
       " 'amount_month_ratio_mean',\n",
       " 'amount_month_ratio_min',\n",
       " 'amount_month_ratio_max',\n",
       " 'amount_month_ratio_var',\n",
       " 'amount_month_ratio_skew',\n",
       " 'category_2_mean_mean',\n",
       " 'category_3_mean_mean',\n",
       " 'purchase_date_diff',\n",
       " 'purchase_date_average',\n",
       " 'purchase_date_uptonow',\n",
       " 'purchase_date_uptomin',\n",
       " 'transactions_count_old',\n",
       " 'subsector_id_nunique_old',\n",
       " 'merchant_id_nunique_old',\n",
       " 'merchant_category_id_nunique_old',\n",
       " 'purchase_Month_nunique',\n",
       " 'purchase_Month_mean_old',\n",
       " 'purchase_Month_min_old',\n",
       " 'purchase_Month_max_old',\n",
       " 'purchase_Week_nunique_old',\n",
       " 'purchase_Week_mean_old',\n",
       " 'purchase_Week_min_old',\n",
       " 'purchase_Week_max_old',\n",
       " 'purchase_Dayofweek_mean_old',\n",
       " 'purchase_Day_nunique_old',\n",
       " 'purchase_Day_mean_old',\n",
       " 'purchase_Day_min_old',\n",
       " 'purchase_Hour_nunique_old',\n",
       " 'purchase_Hour_mean_old',\n",
       " 'purchase_Hour_min_old',\n",
       " 'purchase_Hour_max_old',\n",
       " 'purchase_amount_sum_old',\n",
       " 'purchase_amount_max_old',\n",
       " 'purchase_amount_min_old',\n",
       " 'purchase_amount_mean_old',\n",
       " 'purchase_amount_var_old',\n",
       " 'purchase_amount_skew_old',\n",
       " 'installments_sum_old',\n",
       " 'installments_max_old',\n",
       " 'installments_mean_old',\n",
       " 'installments_var_old',\n",
       " 'installments_skew_old',\n",
       " 'month_lag_max_old',\n",
       " 'month_lag_min_old',\n",
       " 'month_lag_mean_old',\n",
       " 'month_lag_var_old',\n",
       " 'month_lag_skew_old',\n",
       " 'month_diff_max',\n",
       " 'month_diff_min',\n",
       " 'month_diff_mean_old',\n",
       " 'month_diff_var_old',\n",
       " 'month_diff_skew_old',\n",
       " 'authorized_flag_mean',\n",
       " 'purchased_on_weekend_mean_old',\n",
       " 'category_1_mean_old',\n",
       " 'category_2_mean_old',\n",
       " 'category_3_mean_old',\n",
       " 'card_id_count_old',\n",
       " 'price_sum',\n",
       " 'price_mean_old',\n",
       " 'price_max_old',\n",
       " 'price_min_old',\n",
       " 'price_var_old',\n",
       " 'Christmas_Day_2017_mean_old',\n",
       " 'Mothers_Day_2017_mean',\n",
       " 'fathers_day_2017_mean',\n",
       " 'Children_day_2017_mean_old',\n",
       " 'Valentine_Day_2017_mean',\n",
       " 'Black_Friday_2017_mean_old',\n",
       " 'Mothers_Day_2018_mean_old',\n",
       " 'duration_mean_old',\n",
       " 'duration_min_old',\n",
       " 'duration_max_old',\n",
       " 'duration_var_old',\n",
       " 'duration_skew_old',\n",
       " 'amount_month_ratio_mean_old',\n",
       " 'amount_month_ratio_min_old',\n",
       " 'amount_month_ratio_max_old',\n",
       " 'amount_month_ratio_var_old',\n",
       " 'amount_month_ratio_skew_old',\n",
       " 'category_2_mean_mean_old',\n",
       " 'category_3_mean_mean_old',\n",
       " 'purchase_date_diff_old',\n",
       " 'purchase_date_average_old',\n",
       " 'purchase_date_uptonow_old',\n",
       " 'purchase_date_uptomin_old',\n",
       " 'quarter',\n",
       " 'observed_elapsed_time',\n",
       " 'days_feature1',\n",
       " 'days_feature2',\n",
       " 'days_feature3',\n",
       " 'days_feature1_ratio',\n",
       " 'days_feature2_ratio',\n",
       " 'days_feature3_ratio',\n",
       " 'feature_sum',\n",
       " 'feature_mean',\n",
       " 'feature_max',\n",
       " 'feature_min',\n",
       " 'feature_var',\n",
       " 'card_id_total',\n",
       " 'card_id_count_total',\n",
       " 'card_id_count_ratio',\n",
       " 'purchase_amount_total',\n",
       " 'purchase_amount_total_mean',\n",
       " 'purchase_amount_total_max',\n",
       " 'purchase_amount_total_min',\n",
       " 'purchase_amount_sum_ratio',\n",
       " 'hist_first_buy',\n",
       " 'new_first_buy',\n",
       " 'hist_last_buy',\n",
       " 'new_last_buy',\n",
       " 'month_diff_ratio',\n",
       " 'installments_total',\n",
       " 'installments_ratio',\n",
       " 'price_total',\n",
       " 'CLV',\n",
       " 'CLV_old',\n",
       " 'CLV_ratio']"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.648599078789843"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('final_cols_352.txt', 'wb') as fp:\n",
    "    pickle.dump(final_cols, fp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_remove = ['duration_var',\n",
    "                 'duration_skew',\n",
    "                 'amount_month_ratio_mean',\n",
    "                 'amount_month_ratio_min',\n",
    "                 'amount_month_ratio_max',\n",
    "                 'amount_month_ratio_var',\n",
    "                 'amount_month_ratio_skew',\n",
    "                 'category_2_mean_mean',\n",
    "                 'category_3_mean_mean',\n",
    "                 'purchase_date_diff',\n",
    "                 'purchase_date_average',\n",
    "                 'purchase_date_uptonow',\n",
    "                 'purchase_date_uptomin',\n",
    "                 'transactions_count_old',\n",
    "                 'subsector_id_nunique_old',\n",
    "                 'merchant_id_nunique_old',\n",
    "                 'merchant_category_id_nunique_old',\n",
    "                 'purchase_Month_nunique',\n",
    "                 'purchase_Month_mean_old',\n",
    "                 'purchase_Month_min_old',\n",
    "                 'purchase_Month_max_old',\n",
    "                 'purchase_Week_nunique_old',\n",
    "                 'purchase_Week_mean_old',\n",
    "                 'purchase_Week_min_old',\n",
    "                 'purchase_Week_max_old',\n",
    "                 'purchase_Dayofweek_mean_old',\n",
    "                 'purchase_Day_nunique_old',\n",
    "                 'purchase_Day_mean_old',\n",
    "                 'purchase_Day_min_old',\n",
    "                 'purchase_Hour_nunique_old',\n",
    "                 'purchase_Hour_mean_old',\n",
    "                 'purchase_Hour_min_old',\n",
    "                 'purchase_Hour_max_old',\n",
    "                 'purchase_amount_sum_old',\n",
    "                 'purchase_amount_max_old',\n",
    "                 'purchase_amount_min_old',\n",
    "                 'purchase_amount_mean_old',\n",
    "                 'purchase_amount_var_old',\n",
    "                 'purchase_amount_skew_old',\n",
    "                 'installments_sum_old',\n",
    "                 'installments_max_old',\n",
    "                 'installments_mean_old',\n",
    "                 'installments_var_old',\n",
    "                 'installments_skew_old',\n",
    "                 'month_lag_max_old',\n",
    "                 'month_lag_min_old',\n",
    "                 'month_lag_mean_old',\n",
    "                 'month_lag_var_old',\n",
    "                 'month_lag_skew_old',\n",
    "                 'month_diff_max',\n",
    "                 'month_diff_min',\n",
    "                 'month_diff_mean_old',\n",
    "                 'month_diff_var_old',\n",
    "                 'month_diff_skew_old',\n",
    "                 'authorized_flag_mean',\n",
    "                 'purchased_on_weekend_mean_old',\n",
    "                 'category_1_mean_old',\n",
    "                 'category_2_mean_old',\n",
    "                 'category_3_mean_old',\n",
    "                 'card_id_count_old',\n",
    "                 'price_sum',\n",
    "                 'price_mean_old',\n",
    "                 'price_max_old',\n",
    "                 'price_min_old',\n",
    "                 'price_var_old',\n",
    "                 'Christmas_Day_2017_mean_old',\n",
    "                 'Mothers_Day_2017_mean',\n",
    "                 'fathers_day_2017_mean',\n",
    "                 'Children_day_2017_mean_old',\n",
    "                 'Valentine_Day_2017_mean',\n",
    "                 'Black_Friday_2017_mean_old',\n",
    "                 'Mothers_Day_2018_mean_old',\n",
    "                 'duration_mean_old',\n",
    "                 'duration_min_old',\n",
    "                 'duration_max_old',\n",
    "                 'duration_var_old',\n",
    "                 'duration_skew_old',\n",
    "                 'amount_month_ratio_mean_old',\n",
    "                 'amount_month_ratio_min_old',\n",
    "                 'amount_month_ratio_max_old',\n",
    "                 'amount_month_ratio_var_old',\n",
    "                 'amount_month_ratio_skew_old',\n",
    "                 'category_2_mean_mean_old',\n",
    "                 'category_3_mean_mean_old',\n",
    "                 'purchase_date_diff_old',\n",
    "                 'purchase_date_average_old',\n",
    "                 'purchase_date_uptonow_old',\n",
    "                 'purchase_date_uptomin_old',\n",
    "                 'quarter',\n",
    "                 'observed_elapsed_time',\n",
    "                 'days_feature1',\n",
    "                 'days_feature2',\n",
    "                 'days_feature3',\n",
    "                 'days_feature1_ratio',\n",
    "                 'days_feature2_ratio',\n",
    "                 'days_feature3_ratio',\n",
    "                 'feature_sum',\n",
    "                 'feature_mean',\n",
    "                 'feature_max',\n",
    "                 'feature_min',\n",
    "                 'feature_var',\n",
    "                 'card_id_total',\n",
    "                 'card_id_count_total',\n",
    "                 'card_id_count_ratio',\n",
    "                 'purchase_amount_total',\n",
    "                 'purchase_amount_total_mean',\n",
    "                 'purchase_amount_total_max',\n",
    "                 'purchase_amount_total_min',\n",
    "                 'purchase_amount_sum_ratio',\n",
    "                 'hist_first_buy',\n",
    "                 'new_first_buy',\n",
    "                 'hist_last_buy',\n",
    "                 'new_last_buy',\n",
    "                 'month_diff_ratio',\n",
    "                 'installments_total',\n",
    "                 'installments_ratio',\n",
    "                 'price_total',\n",
    "                 'CLV',\n",
    "                 'CLV_old',\n",
    "                 'CLV_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols = cols_to_remove + ['merchant_address_id_nunique', 'merchant_rating_nunique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "122"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(final_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "121"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len([l for l in final_cols if l!= 'merchant_address_id_nunique'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_score = 3.647901137421428"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = cols_included"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "180"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols = df_train_columns + ['merchant_address_id_nunique', 'merchant_rating_nunique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "182"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(final_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.6480832619147843"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.646777128761576"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "base_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['amount_month_ratio_max',\n",
       " 'purchase_Month_mean_old',\n",
       " 'purchase_amount_total_max']"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols = [c for c in df_train_columns if c not in selected_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "177"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(final_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols.remove('Black_Friday_2017_mean')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "59"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_cols.index('duration_var')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_remove = ['purchase_Month_min_old', 'amount_month_ratio_var', 'purchase_amount_total_min']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "175 lg_cols purchase_Month_min_old\n",
      "3.660311490783336 CV score\n",
      "175 lg_cols amount_month_ratio_var\n",
      "3.6591214030047143 CV score\n",
      "175 lg_cols purchase_amount_total_min\n",
      "3.6603760515089054 CV score\n",
      "CPU times: user 20min 39s, sys: 4.5 s, total: 20min 44s\n",
      "Wall time: 3min 29s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "for c in cols_to_remove:\n",
    "    lgb_cols = [l for l in final_cols if l!= c]\n",
    "    print(len(lgb_cols), 'lg_cols', c)\n",
    "    score = lgb_train_fn(x, y, lgb_cols, 5)\n",
    "#     trn_data = lgb.Dataset(x[lgb_cols], label=y)\n",
    "#     val_data = lgb.Dataset(x_t[lgb_cols], label=y_t)\n",
    "#     clf = lgb.train(lgb_param, trn_data, num_round, \n",
    "#                     valid_sets = [trn_data, val_data], verbose_eval=0, early_stopping_rounds = 200)\n",
    "#     oof = clf.predict(x_t[lgb_cols], num_iteration=clf.best_iteration)\n",
    "#     score = np.sqrt(mean_squared_error(y_t, oof))\n",
    "    delta = base_score - score\n",
    "    fe_d[c] = delta\n",
    "    if delta > 0:\n",
    "        base_score = score\n",
    "        selected_cols.append(c)\n",
    "        print('Selected cols', c)\n",
    "        print('Selected col delta', delta)\n",
    "        print(' score with col', score)\n",
    "        np.save('selecte_cold_prune', selected_cols)\n",
    "        final_cols.remove(c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['amount_month_ratio_max',\n",
       " 'purchase_Month_mean_old',\n",
       " 'purchase_amount_total_max']"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "339"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols_to_add)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['purchase_amount_max',\n",
       " 'avg_trans_per_merchant_old',\n",
       " 'duration_min_old',\n",
       " 'purchase_date_uptonow',\n",
       " 'category_1_mean',\n",
       " 'authorized_flag_mean',\n",
       " 'purchase_date_uptonow_old',\n",
       " 'purchase_Week_mean',\n",
       " 'purchase_amount_mean',\n",
       " 'purchase_Month_nunique',\n",
       " 'month_diff_mean_old',\n",
       " 'category_1_mean_old',\n",
       " 'duration_mean_old',\n",
       " 'month_lag_mean_old',\n",
       " 'purchase_date_average_old',\n",
       " 'fathers_day_2017_mean',\n",
       " 'purchase_Month_mean',\n",
       " 'installments_sum_old',\n",
       " 'month_lag_mean',\n",
       " 'days_feature1',\n",
       " 'purchase_Week_max',\n",
       " 'hist_last_buy',\n",
       " 'month_lag_skew',\n",
       " 'purchase_Week_nunique_old',\n",
       " 'Mothers_Day_2017_mean',\n",
       " 'purchase_Dayofweek_mean_old',\n",
       " 'merchant_category_id_nunique_old',\n",
       " 'new_last_buy',\n",
       " 'duration_max_old',\n",
       " 'new_first_buy',\n",
       " 'month_lag_max',\n",
       " 'price_var',\n",
       " 'installments_var_old',\n",
       " 'price_min',\n",
       " 'purchase_amount_sum_old',\n",
       " 'merchant_id_nunique',\n",
       " 'purchase_Hour_nunique',\n",
       " 'purchase_Day_nunique']"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_deltas_df = pd.DataFrame.from_dict(fe_d, orient='index')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "scrolled": true
   },
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>month_diff_mean_old</th>\n",
       "      <td>0.001600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>category_1_mean_old</th>\n",
       "      <td>0.009618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_ratio</th>\n",
       "      <td>-0.048055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_date_diff</th>\n",
       "      <td>-0.046296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration_mean_old</th>\n",
       "      <td>0.003146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_lag_mean_old</th>\n",
       "      <td>0.001767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_date_average_old</th>\n",
       "      <td>0.001088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_date_diff_old</th>\n",
       "      <td>-0.000456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fathers_day_2017_mean</th>\n",
       "      <td>0.000744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_mean</th>\n",
       "      <td>0.001524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_mean</th>\n",
       "      <td>-0.003207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>amount_month_ratio_min_old</th>\n",
       "      <td>-0.000794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_sum_old</th>\n",
       "      <td>0.002143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_lag_mean</th>\n",
       "      <td>0.000705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_date_average</th>\n",
       "      <td>-0.046009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature1</th>\n",
       "      <td>0.001636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_max</th>\n",
       "      <td>0.000332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Children_day_2017_mean_old</th>\n",
       "      <td>-0.000818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Valentine_Day_2017_mean</th>\n",
       "      <td>-0.002160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature2</th>\n",
       "      <td>-0.001431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_lag_var_old</th>\n",
       "      <td>-0.002556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hist_last_buy</th>\n",
       "      <td>0.000305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_id_nunique_old</th>\n",
       "      <td>-0.000595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_mean_old</th>\n",
       "      <td>-0.003817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_date_uptomin_old</th>\n",
       "      <td>-0.001833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_lag_skew_old</th>\n",
       "      <td>-0.002105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean_old</th>\n",
       "      <td>-0.002234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mothers_Day_2018_mean</th>\n",
       "      <td>-0.001466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration_var</th>\n",
       "      <td>-0.001066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_ratio</th>\n",
       "      <td>-0.001433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_skew</th>\n",
       "      <td>-0.003667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_lag_min_old</th>\n",
       "      <td>-0.001736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_old</th>\n",
       "      <td>-0.002559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_var</th>\n",
       "      <td>-0.001424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_nunique</th>\n",
       "      <td>-0.000757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchased_on_weekend_mean</th>\n",
       "      <td>-0.003562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_max</th>\n",
       "      <td>-0.002944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_id_nunique</th>\n",
       "      <td>0.000458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>category_3_mean</th>\n",
       "      <td>-0.001374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_sum</th>\n",
       "      <td>-0.001540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Dayofweek_min</th>\n",
       "      <td>-0.000571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_min</th>\n",
       "      <td>-0.001584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_max_old</th>\n",
       "      <td>-0.000841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>category_2_mean</th>\n",
       "      <td>-0.000791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_total</th>\n",
       "      <td>-0.000183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_diff_min</th>\n",
       "      <td>-0.000776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_nunique</th>\n",
       "      <td>0.000081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_nunique</th>\n",
       "      <td>0.000723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_min_old</th>\n",
       "      <td>-0.002328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_min_old</th>\n",
       "      <td>-0.000934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>quarter</th>\n",
       "      <td>-0.000695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_2</th>\n",
       "      <td>-0.001158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count</th>\n",
       "      <td>-0.001396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_max_old</th>\n",
       "      <td>-0.000579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_max_old</th>\n",
       "      <td>-0.000937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_max</th>\n",
       "      <td>-0.000260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_mean</th>\n",
       "      <td>-0.001213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Dayofweek_max</th>\n",
       "      <td>-0.000566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_3</th>\n",
       "      <td>-0.000582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_min</th>\n",
       "      <td>-0.000582</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>173 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   0\n",
       "month_diff_mean_old         0.001600\n",
       "category_1_mean_old         0.009618\n",
       "card_id_count_ratio        -0.048055\n",
       "purchase_date_diff         -0.046296\n",
       "duration_mean_old           0.003146\n",
       "month_lag_mean_old          0.001767\n",
       "purchase_date_average_old   0.001088\n",
       "purchase_date_diff_old     -0.000456\n",
       "fathers_day_2017_mean       0.000744\n",
       "purchase_Month_mean         0.001524\n",
       "purchase_Day_mean          -0.003207\n",
       "amount_month_ratio_min_old -0.000794\n",
       "installments_sum_old        0.002143\n",
       "month_lag_mean              0.000705\n",
       "purchase_date_average      -0.046009\n",
       "days_feature1               0.001636\n",
       "purchase_Week_max           0.000332\n",
       "Children_day_2017_mean_old -0.000818\n",
       "Valentine_Day_2017_mean    -0.002160\n",
       "days_feature2              -0.001431\n",
       "month_lag_var_old          -0.002556\n",
       "hist_last_buy               0.000305\n",
       "merchant_id_nunique_old    -0.000595\n",
       "purchase_Hour_mean_old     -0.003817\n",
       "purchase_date_uptomin_old  -0.001833\n",
       "month_lag_skew_old         -0.002105\n",
       "purchase_amount_mean_old   -0.002234\n",
       "Mothers_Day_2018_mean      -0.001466\n",
       "duration_var               -0.001066\n",
       "installments_ratio         -0.001433\n",
       "...                              ...\n",
       "installments_skew          -0.003667\n",
       "month_lag_min_old          -0.001736\n",
       "card_id_count_old          -0.002559\n",
       "feature_var                -0.001424\n",
       "purchase_Week_nunique      -0.000757\n",
       "purchased_on_weekend_mean  -0.003562\n",
       "purchase_Hour_max          -0.002944\n",
       "merchant_id_nunique         0.000458\n",
       "category_3_mean            -0.001374\n",
       "feature_sum                -0.001540\n",
       "purchase_Dayofweek_min     -0.000571\n",
       "purchase_Month_min         -0.001584\n",
       "installments_max_old       -0.000841\n",
       "category_2_mean            -0.000791\n",
       "card_id_count_total        -0.000183\n",
       "month_diff_min             -0.000776\n",
       "purchase_Hour_nunique       0.000081\n",
       "purchase_Day_nunique        0.000723\n",
       "purchase_Month_min_old     -0.002328\n",
       "purchase_Day_min_old       -0.000934\n",
       "quarter                    -0.000695\n",
       "feature_2                  -0.001158\n",
       "card_id_count              -0.001396\n",
       "purchase_Hour_max_old      -0.000579\n",
       "purchase_Month_max_old     -0.000937\n",
       "feature_max                -0.000260\n",
       "feature_mean               -0.001213\n",
       "purchase_Dayofweek_max     -0.000566\n",
       "feature_3                  -0.000582\n",
       "feature_min                -0.000582\n",
       "\n",
       "[173 rows x 1 columns]"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fe_deltas_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "collapsed": true
   },
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>transactions_count</th>\n",
       "      <td>-0.150052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subsector_id_nunique</th>\n",
       "      <td>-0.148867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_id_nunique</th>\n",
       "      <td>-0.150136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_category_id_nunique</th>\n",
       "      <td>-0.149662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_mean</th>\n",
       "      <td>-0.117247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_min</th>\n",
       "      <td>-0.118750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_max</th>\n",
       "      <td>-0.116200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_nunique</th>\n",
       "      <td>-0.149430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_mean</th>\n",
       "      <td>-0.106378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_min</th>\n",
       "      <td>-0.114731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_max</th>\n",
       "      <td>-0.110754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Dayofweek_mean</th>\n",
       "      <td>-0.151258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Dayofweek_min</th>\n",
       "      <td>-0.150686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Dayofweek_max</th>\n",
       "      <td>-0.147374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_nunique</th>\n",
       "      <td>-0.148126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_mean</th>\n",
       "      <td>-0.146837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_min</th>\n",
       "      <td>-0.145331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_max</th>\n",
       "      <td>-0.150433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_nunique</th>\n",
       "      <td>-0.150162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_mean</th>\n",
       "      <td>-0.149936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_min</th>\n",
       "      <td>-0.147686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Hour_max</th>\n",
       "      <td>-0.149023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum</th>\n",
       "      <td>-0.142631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_pct_75</th>\n",
       "      <td>-0.143030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_pct_25</th>\n",
       "      <td>-0.147739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean</th>\n",
       "      <td>-0.143986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_median</th>\n",
       "      <td>-0.144876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max</th>\n",
       "      <td>-0.139022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_min</th>\n",
       "      <td>-0.147256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_var</th>\n",
       "      <td>-0.144690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature2</th>\n",
       "      <td>-0.146333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature3</th>\n",
       "      <td>-0.153122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature1_ratio</th>\n",
       "      <td>-0.147171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature2_ratio</th>\n",
       "      <td>-0.143542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature3_ratio</th>\n",
       "      <td>-0.153021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_sum</th>\n",
       "      <td>-0.154089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_mean</th>\n",
       "      <td>-0.154089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_max</th>\n",
       "      <td>-0.156205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_min</th>\n",
       "      <td>-0.156205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_var</th>\n",
       "      <td>-0.156205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_total</th>\n",
       "      <td>-0.149681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_total</th>\n",
       "      <td>-0.149681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_ratio</th>\n",
       "      <td>-0.150981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total</th>\n",
       "      <td>-0.153602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total_mean</th>\n",
       "      <td>-0.149227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total_max</th>\n",
       "      <td>-0.146412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total_min</th>\n",
       "      <td>-0.150095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum_ratio</th>\n",
       "      <td>-0.152210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hist_first_buy</th>\n",
       "      <td>-0.149081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>new_first_buy</th>\n",
       "      <td>-0.145978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hist_last_buy</th>\n",
       "      <td>-0.144760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>new_last_buy</th>\n",
       "      <td>-0.141976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchased_before_issue</th>\n",
       "      <td>-0.156205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_diff_ratio</th>\n",
       "      <td>-0.143926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_total</th>\n",
       "      <td>-0.150340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_ratio</th>\n",
       "      <td>-0.156992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>price_total</th>\n",
       "      <td>-0.151305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLV</th>\n",
       "      <td>-0.142472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLV_old</th>\n",
       "      <td>-0.154081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLV_ratio</th>\n",
       "      <td>-0.148115</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>240 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     0\n",
       "transactions_count           -0.150052\n",
       "subsector_id_nunique         -0.148867\n",
       "merchant_id_nunique          -0.150136\n",
       "merchant_category_id_nunique -0.149662\n",
       "purchase_Month_mean          -0.117247\n",
       "purchase_Month_min           -0.118750\n",
       "purchase_Month_max           -0.116200\n",
       "purchase_Week_nunique        -0.149430\n",
       "purchase_Week_mean           -0.106378\n",
       "purchase_Week_min            -0.114731\n",
       "purchase_Week_max            -0.110754\n",
       "purchase_Dayofweek_mean      -0.151258\n",
       "purchase_Dayofweek_min       -0.150686\n",
       "purchase_Dayofweek_max       -0.147374\n",
       "purchase_Day_nunique         -0.148126\n",
       "purchase_Day_mean            -0.146837\n",
       "purchase_Day_min             -0.145331\n",
       "purchase_Day_max             -0.150433\n",
       "purchase_Hour_nunique        -0.150162\n",
       "purchase_Hour_mean           -0.149936\n",
       "purchase_Hour_min            -0.147686\n",
       "purchase_Hour_max            -0.149023\n",
       "purchase_amount_sum          -0.142631\n",
       "purchase_amount_pct_75       -0.143030\n",
       "purchase_amount_pct_25       -0.147739\n",
       "purchase_amount_mean         -0.143986\n",
       "purchase_amount_median       -0.144876\n",
       "purchase_amount_max          -0.139022\n",
       "purchase_amount_min          -0.147256\n",
       "purchase_amount_var          -0.144690\n",
       "...                                ...\n",
       "days_feature2                -0.146333\n",
       "days_feature3                -0.153122\n",
       "days_feature1_ratio          -0.147171\n",
       "days_feature2_ratio          -0.143542\n",
       "days_feature3_ratio          -0.153021\n",
       "feature_sum                  -0.154089\n",
       "feature_mean                 -0.154089\n",
       "feature_max                  -0.156205\n",
       "feature_min                  -0.156205\n",
       "feature_var                  -0.156205\n",
       "card_id_total                -0.149681\n",
       "card_id_count_total          -0.149681\n",
       "card_id_count_ratio          -0.150981\n",
       "purchase_amount_total        -0.153602\n",
       "purchase_amount_total_mean   -0.149227\n",
       "purchase_amount_total_max    -0.146412\n",
       "purchase_amount_total_min    -0.150095\n",
       "purchase_amount_sum_ratio    -0.152210\n",
       "hist_first_buy               -0.149081\n",
       "new_first_buy                -0.145978\n",
       "hist_last_buy                -0.144760\n",
       "new_last_buy                 -0.141976\n",
       "purchased_before_issue       -0.156205\n",
       "month_diff_ratio             -0.143926\n",
       "installments_total           -0.150340\n",
       "installments_ratio           -0.156992\n",
       "price_total                  -0.151305\n",
       "CLV                          -0.142472\n",
       "CLV_old                      -0.154081\n",
       "CLV_ratio                    -0.148115\n",
       "\n",
       "[240 rows x 1 columns]"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fe_deltas_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_df = fe_deltas_df.reset_index()\n",
    "fe_df.columns = ['feature', 'delta']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": true
   },
   "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>feature</th>\n",
       "      <th>delta</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>category_1_mean_old</td>\n",
       "      <td>0.009618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>month_lag_max</td>\n",
       "      <td>0.006409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>duration_mean_old</td>\n",
       "      <td>0.003146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>purchase_Dayofweek_mean_old</td>\n",
       "      <td>0.002901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>installments_sum_old</td>\n",
       "      <td>0.002143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>month_lag_skew</td>\n",
       "      <td>0.001768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>month_lag_mean_old</td>\n",
       "      <td>0.001767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>days_feature1</td>\n",
       "      <td>0.001636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>month_diff_mean_old</td>\n",
       "      <td>0.001600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>purchase_Month_mean</td>\n",
       "      <td>0.001524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>purchase_date_average_old</td>\n",
       "      <td>0.001088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>price_min</td>\n",
       "      <td>0.001069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>installments_var_old</td>\n",
       "      <td>0.000960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>duration_max_old</td>\n",
       "      <td>0.000808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>fathers_day_2017_mean</td>\n",
       "      <td>0.000744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>purchase_amount_sum_old</td>\n",
       "      <td>0.000725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>purchase_Day_nunique</td>\n",
       "      <td>0.000723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>month_lag_mean</td>\n",
       "      <td>0.000705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>purchase_Week_nunique_old</td>\n",
       "      <td>0.000625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>new_last_buy</td>\n",
       "      <td>0.000622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>price_var</td>\n",
       "      <td>0.000581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>merchant_id_nunique</td>\n",
       "      <td>0.000458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>purchase_Week_max</td>\n",
       "      <td>0.000332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>new_first_buy</td>\n",
       "      <td>0.000316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>hist_last_buy</td>\n",
       "      <td>0.000305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>Mothers_Day_2017_mean</td>\n",
       "      <td>0.000187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>merchant_category_id_nunique_old</td>\n",
       "      <td>0.000165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>purchase_Hour_nunique</td>\n",
       "      <td>0.000081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>purchase_amount_sum</td>\n",
       "      <td>-0.000047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>purchase_amount_skew_old</td>\n",
       "      <td>-0.000088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>price_total</td>\n",
       "      <td>-0.002657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>Children_day_2017_mean</td>\n",
       "      <td>-0.002756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>purchase_amount_total_min</td>\n",
       "      <td>-0.002769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>hist_first_buy</td>\n",
       "      <td>-0.002825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>elapsed_time</td>\n",
       "      <td>-0.002899</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>purchase_Hour_max</td>\n",
       "      <td>-0.002944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>month_diff_skew_old</td>\n",
       "      <td>-0.002970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>category_3_mean_mean</td>\n",
       "      <td>-0.002976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>price_mean</td>\n",
       "      <td>-0.003100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>Mothers_Day_2018_mean_old</td>\n",
       "      <td>-0.003107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>purchase_Dayofweek_mean</td>\n",
       "      <td>-0.003118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>amount_month_ratio_mean</td>\n",
       "      <td>-0.003175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>purchase_date_uptomin</td>\n",
       "      <td>-0.003181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>purchase_Day_mean</td>\n",
       "      <td>-0.003207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>price_mean_old</td>\n",
       "      <td>-0.003228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>purchase_amount_sum_ratio</td>\n",
       "      <td>-0.003252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>purchase_amount_var</td>\n",
       "      <td>-0.003354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>purchase_amount_total</td>\n",
       "      <td>-0.003512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>amount_month_ratio_var</td>\n",
       "      <td>-0.003514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>purchased_on_weekend_mean</td>\n",
       "      <td>-0.003562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>purchase_Day_nunique_old</td>\n",
       "      <td>-0.003631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>installments_skew</td>\n",
       "      <td>-0.003667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>category_2_mean_mean_old</td>\n",
       "      <td>-0.003671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>purchase_Hour_mean_old</td>\n",
       "      <td>-0.003817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>duration_max</td>\n",
       "      <td>-0.003918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>month_diff_skew</td>\n",
       "      <td>-0.003948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>purchase_Hour_mean</td>\n",
       "      <td>-0.004309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>purchase_date_average</td>\n",
       "      <td>-0.046009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>purchase_date_diff</td>\n",
       "      <td>-0.046296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>card_id_count_ratio</td>\n",
       "      <td>-0.048055</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>173 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                              feature     delta\n",
       "1                 category_1_mean_old  0.009618\n",
       "64                      month_lag_max  0.006409\n",
       "4                   duration_mean_old  0.003146\n",
       "43        purchase_Dayofweek_mean_old  0.002901\n",
       "12               installments_sum_old  0.002143\n",
       "32                     month_lag_skew  0.001768\n",
       "5                  month_lag_mean_old  0.001767\n",
       "15                      days_feature1  0.001636\n",
       "0                 month_diff_mean_old  0.001600\n",
       "9                 purchase_Month_mean  0.001524\n",
       "6           purchase_date_average_old  0.001088\n",
       "72                          price_min  0.001069\n",
       "68               installments_var_old  0.000960\n",
       "62                   duration_max_old  0.000808\n",
       "8               fathers_day_2017_mean  0.000744\n",
       "99            purchase_amount_sum_old  0.000725\n",
       "160              purchase_Day_nunique  0.000723\n",
       "13                     month_lag_mean  0.000705\n",
       "39          purchase_Week_nunique_old  0.000625\n",
       "61                       new_last_buy  0.000622\n",
       "67                          price_var  0.000581\n",
       "150               merchant_id_nunique  0.000458\n",
       "16                  purchase_Week_max  0.000332\n",
       "63                      new_first_buy  0.000316\n",
       "21                      hist_last_buy  0.000305\n",
       "42              Mothers_Day_2017_mean  0.000187\n",
       "51   merchant_category_id_nunique_old  0.000165\n",
       "159             purchase_Hour_nunique  0.000081\n",
       "56                purchase_amount_sum -0.000047\n",
       "71           purchase_amount_skew_old -0.000088\n",
       "..                                ...       ...\n",
       "116                       price_total -0.002657\n",
       "94             Children_day_2017_mean -0.002756\n",
       "92          purchase_amount_total_min -0.002769\n",
       "48                     hist_first_buy -0.002825\n",
       "119                      elapsed_time -0.002899\n",
       "149                 purchase_Hour_max -0.002944\n",
       "38                month_diff_skew_old -0.002970\n",
       "106              category_3_mean_mean -0.002976\n",
       "101                        price_mean -0.003100\n",
       "125         Mothers_Day_2018_mean_old -0.003107\n",
       "104           purchase_Dayofweek_mean -0.003118\n",
       "122           amount_month_ratio_mean -0.003175\n",
       "31              purchase_date_uptomin -0.003181\n",
       "10                  purchase_Day_mean -0.003207\n",
       "50                     price_mean_old -0.003228\n",
       "89          purchase_amount_sum_ratio -0.003252\n",
       "44                purchase_amount_var -0.003354\n",
       "124             purchase_amount_total -0.003512\n",
       "37             amount_month_ratio_var -0.003514\n",
       "148         purchased_on_weekend_mean -0.003562\n",
       "126          purchase_Day_nunique_old -0.003631\n",
       "143                 installments_skew -0.003667\n",
       "109          category_2_mean_mean_old -0.003671\n",
       "23             purchase_Hour_mean_old -0.003817\n",
       "118                      duration_max -0.003918\n",
       "107                   month_diff_skew -0.003948\n",
       "34                 purchase_Hour_mean -0.004309\n",
       "14              purchase_date_average -0.046009\n",
       "3                  purchase_date_diff -0.046296\n",
       "2                 card_id_count_ratio -0.048055\n",
       "\n",
       "[173 rows x 2 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fe_df.sort_values('delta', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "iter1_df = fe_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "re_iter_cols = fe_df.sort_values('delta', ascending=False)[:42]['feature'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": true
   },
   "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>feature</th>\n",
       "      <th>delta</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>avg_trans_per_merchant_old</td>\n",
       "      <td>0.002242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>purchase_amount_sum_mean</td>\n",
       "      <td>0.001915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>category_2_5.0_purchase_amount_mean</td>\n",
       "      <td>0.000357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>repurchase_merchant_rate_old</td>\n",
       "      <td>-0.000773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>category_1_0_purchase_amount_mean</td>\n",
       "      <td>-0.000885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>purchase_amount_pct_75</td>\n",
       "      <td>-0.000888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>purchase_amount_mean_max</td>\n",
       "      <td>-0.001402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>purchase_amount_pct_25</td>\n",
       "      <td>-0.001859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>purchase_amount_max_mean</td>\n",
       "      <td>-0.001862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>category_1_1_purchase_amount_sum</td>\n",
       "      <td>-0.002105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>category_2_4.0_purchase_amount_mean</td>\n",
       "      <td>-0.002119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>purchase_amount_mean_sum</td>\n",
       "      <td>-0.002155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>category_3_1_purchase_amount_sum</td>\n",
       "      <td>-0.002226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>purchase_date_successive_diff_mean_old</td>\n",
       "      <td>-0.002320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>avg_spend_per_merchant</td>\n",
       "      <td>-0.002344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>merchant_category_repurchase_old</td>\n",
       "      <td>-0.002618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>category_3_2_purchase_amount_mean</td>\n",
       "      <td>-0.002776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>purchase_amount_sum_sum</td>\n",
       "      <td>-0.002830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>category_3_0_purchase_amount_mean</td>\n",
       "      <td>-0.002930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>purchase_amount_sum_max</td>\n",
       "      <td>-0.002981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>avg_spend_per_merchant_old</td>\n",
       "      <td>-0.003635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>authorized_flag_0_purchase_amount_sum</td>\n",
       "      <td>-0.003639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>avg_spend_per_transaction</td>\n",
       "      <td>-0.003643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>category_3_0_purchase_amount_sum</td>\n",
       "      <td>-0.003715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>purchase_amount_count_sum</td>\n",
       "      <td>-0.003861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>purchase_amount_mean_ptp</td>\n",
       "      <td>-0.003875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>purchase_amount_count_max</td>\n",
       "      <td>-0.003880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>repurchase_merchant_rate</td>\n",
       "      <td>-0.003900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>avg_trans_per_merchant</td>\n",
       "      <td>-0.003900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>days_since_last_transaction</td>\n",
       "      <td>-0.003912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>purchase_amount_max_max</td>\n",
       "      <td>-0.003938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>purchase_amount_pct_25_old</td>\n",
       "      <td>-0.003973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>days_since_last_transaction_old</td>\n",
       "      <td>-0.004030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>category_2_1.0_purchase_amount_mean</td>\n",
       "      <td>-0.004325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>inverse_avg_transactions_per_day_old</td>\n",
       "      <td>-0.004325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>purchase_amount_sum_ptp</td>\n",
       "      <td>-0.004411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>category_2_3.0_purchase_amount_mean</td>\n",
       "      <td>-0.004500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>category_2_2.0_purchase_amount_sum</td>\n",
       "      <td>-0.004611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>avg_spend_per_transaction_old</td>\n",
       "      <td>-0.004752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>category_2_2.0_purchase_amount_mean</td>\n",
       "      <td>-0.004847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>purchase_date_successive_diff_mean</td>\n",
       "      <td>-0.004847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>merchant_category_repurchase</td>\n",
       "      <td>-0.005066</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   feature     delta\n",
       "0               avg_trans_per_merchant_old  0.002242\n",
       "2                 purchase_amount_sum_mean  0.001915\n",
       "13     category_2_5.0_purchase_amount_mean  0.000357\n",
       "1             repurchase_merchant_rate_old -0.000773\n",
       "10       category_1_0_purchase_amount_mean -0.000885\n",
       "6                   purchase_amount_pct_75 -0.000888\n",
       "9                 purchase_amount_mean_max -0.001402\n",
       "7                   purchase_amount_pct_25 -0.001859\n",
       "12                purchase_amount_max_mean -0.001862\n",
       "11        category_1_1_purchase_amount_sum -0.002105\n",
       "41     category_2_4.0_purchase_amount_mean -0.002119\n",
       "4                 purchase_amount_mean_sum -0.002155\n",
       "5         category_3_1_purchase_amount_sum -0.002226\n",
       "15  purchase_date_successive_diff_mean_old -0.002320\n",
       "8                   avg_spend_per_merchant -0.002344\n",
       "23        merchant_category_repurchase_old -0.002618\n",
       "14       category_3_2_purchase_amount_mean -0.002776\n",
       "20                 purchase_amount_sum_sum -0.002830\n",
       "3        category_3_0_purchase_amount_mean -0.002930\n",
       "31                 purchase_amount_sum_max -0.002981\n",
       "29              avg_spend_per_merchant_old -0.003635\n",
       "37   authorized_flag_0_purchase_amount_sum -0.003639\n",
       "33               avg_spend_per_transaction -0.003643\n",
       "22        category_3_0_purchase_amount_sum -0.003715\n",
       "28               purchase_amount_count_sum -0.003861\n",
       "17                purchase_amount_mean_ptp -0.003875\n",
       "25               purchase_amount_count_max -0.003880\n",
       "26                repurchase_merchant_rate -0.003900\n",
       "27                  avg_trans_per_merchant -0.003900\n",
       "38             days_since_last_transaction -0.003912\n",
       "30                 purchase_amount_max_max -0.003938\n",
       "19              purchase_amount_pct_25_old -0.003973\n",
       "40         days_since_last_transaction_old -0.004030\n",
       "35     category_2_1.0_purchase_amount_mean -0.004325\n",
       "32    inverse_avg_transactions_per_day_old -0.004325\n",
       "39                 purchase_amount_sum_ptp -0.004411\n",
       "16     category_2_3.0_purchase_amount_mean -0.004500\n",
       "34      category_2_2.0_purchase_amount_sum -0.004611\n",
       "36           avg_spend_per_transaction_old -0.004752\n",
       "18     category_2_2.0_purchase_amount_mean -0.004847\n",
       "21      purchase_date_successive_diff_mean -0.004847\n",
       "24            merchant_category_repurchase -0.005066"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fe_df.sort_values('delta', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics.scorer import make_scorer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse(actual, predict):\n",
    "    return mean_squared_error(actual, predict)\n",
    "\n",
    "rmse_score = make_scorer(rmse, greater_is_better = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = Ridge(alpha=1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import make_pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "sfs = SFS(clf, \n",
    "          k_features=160, \n",
    "          forward=True, \n",
    "          floating=False, \n",
    "          scoring='r2',\n",
    "          cv=5,\n",
    "          n_jobs=-1)\n",
    "pipe = make_pipeline(sfs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['explained_variance', 'r2', 'neg_median_absolute_error', 'neg_mean_absolute_error', 'neg_mean_squared_error', 'neg_mean_squared_log_error', 'accuracy', 'roc_auc', 'balanced_accuracy', 'average_precision', 'neg_log_loss', 'brier_score_loss', 'adjusted_rand_score', 'homogeneity_score', 'completeness_score', 'v_measure_score', 'mutual_info_score', 'adjusted_mutual_info_score', 'normalized_mutual_info_score', 'fowlkes_mallows_score', 'precision', 'precision_macro', 'precision_micro', 'precision_samples', 'precision_weighted', 'recall', 'recall_macro', 'recall_micro', 'recall_samples', 'recall_weighted', 'f1', 'f1_macro', 'f1_micro', 'f1_samples', 'f1_weighted'])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sklearn.metrics.SCORERS.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "180"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "x.fillna(0, inplace=True)\n",
    "x.replace(-np.inf, 0, inplace=True)\n",
    "x.replace(np.inf, 0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5806"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n",
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/sklearn/externals/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3min 55s, sys: 27.6 s, total: 4min 23s\n",
      "Wall time: 53min 40s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('sequentialfeatureselector', SequentialFeatureSelector(clone_estimator=True, cv=5,\n",
       "             estimator=Ridge(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001),\n",
       "             floating=False, forward=True, k_features=160, n_jobs=-1,\n",
       "             pre_dispatch='2*n_jobs', scoring='r2', verbose=0))])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time pipe.fit(x[df_train_columns].values, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.046826811091055645"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sfs.k_score_ #160 72% dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.04270958032826948"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sfs.k_score_ #40 72% dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.045418292964324404"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sfs.k_score_ #40 10% dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.040013368847060214"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sfs.k_score_ #160"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlxtend.plotting import plot_sequential_feature_selection as plot_sfs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best combination (ACC: 0.047): (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 58, 60, 61, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 95, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 121, 123, 125, 126, 127, 128, 129, 130, 131, 132, 133, 136, 138, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 179)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print('best combination (ACC: %.3f): %s\\n' % (sfs.k_score_, sfs.k_feature_idx_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best combination (ACC: -14.237): (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 30, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 82, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 161, 163, 165, 168, 169)\n",
      "\n",
      "all subsets:\n",
      " {1: {'feature_idx': (108,), 'cv_scores': array([-13.0142 , -16.20725, -14.21346, -15.21542]), 'avg_score': -14.662580626967827, 'feature_names': ('108',)}, 2: {'feature_idx': (77, 108), 'cv_scores': array([-12.91242, -16.08779, -14.02797, -15.15898]), 'avg_score': -14.54678860569546, 'feature_names': ('77', '108')}, 3: {'feature_idx': (26, 77, 108), 'cv_scores': array([-12.88092, -16.03325, -13.96852, -15.09211]), 'avg_score': -14.493701100309321, 'feature_names': ('26', '77', '108')}, 4: {'feature_idx': (26, 38, 77, 108), 'cv_scores': array([-12.87792, -15.97805, -13.93796, -15.06413]), 'avg_score': -14.464513770491575, 'feature_names': ('26', '38', '77', '108')}, 5: {'feature_idx': (15, 26, 38, 77, 108), 'cv_scores': array([-12.85053, -15.96103, -13.9042 , -15.03483]), 'avg_score': -14.437646206057769, 'feature_names': ('15', '26', '38', '77', '108')}, 6: {'feature_idx': (15, 26, 38, 45, 77, 108), 'cv_scores': array([-12.8569 , -15.89513, -13.86484, -15.03666]), 'avg_score': -14.413384360842043, 'feature_names': ('15', '26', '38', '45', '77', '108')}, 7: {'feature_idx': (15, 18, 26, 38, 45, 77, 108), 'cv_scores': array([-12.84666, -15.84775, -13.86794, -15.01833]), 'avg_score': -14.395170502481777, 'feature_names': ('15', '18', '26', '38', '45', '77', '108')}, 8: {'feature_idx': (15, 18, 26, 38, 45, 53, 77, 108), 'cv_scores': array([-12.7957 , -15.792  , -13.8546 , -15.07393]), 'avg_score': -14.379058620894034, 'feature_names': ('15', '18', '26', '38', '45', '53', '77', '108')}, 9: {'feature_idx': (15, 18, 26, 38, 45, 53, 77, 103, 108), 'cv_scores': array([-12.77647, -15.74626, -13.81276, -15.1081 ]), 'avg_score': -14.36089866265675, 'feature_names': ('15', '18', '26', '38', '45', '53', '77', '103', '108')}, 10: {'feature_idx': (15, 18, 26, 38, 45, 53, 77, 103, 108, 111), 'cv_scores': array([-12.75824, -15.73965, -13.79494, -15.08157]), 'avg_score': -14.343598346238322, 'feature_names': ('15', '18', '26', '38', '45', '53', '77', '103', '108', '111')}, 11: {'feature_idx': (15, 18, 26, 38, 39, 45, 53, 77, 103, 108, 111), 'cv_scores': array([-12.72236, -15.74664, -13.79699, -15.05373]), 'avg_score': -14.32993121640631, 'feature_names': ('15', '18', '26', '38', '39', '45', '53', '77', '103', '108', '111')}, 12: {'feature_idx': (4, 15, 18, 26, 38, 39, 45, 53, 77, 103, 108, 111), 'cv_scores': array([-12.7195 , -15.74314, -13.78879, -15.04444]), 'avg_score': -14.323968250775362, 'feature_names': ('4', '15', '18', '26', '38', '39', '45', '53', '77', '103', '108', '111')}, 13: {'feature_idx': (4, 15, 18, 26, 38, 39, 45, 53, 65, 77, 103, 108, 111), 'cv_scores': array([-12.71152, -15.73345, -13.78459, -15.03876]), 'avg_score': -14.317081754893344, 'feature_names': ('4', '15', '18', '26', '38', '39', '45', '53', '65', '77', '103', '108', '111')}, 14: {'feature_idx': (4, 15, 18, 26, 38, 39, 40, 45, 53, 65, 77, 103, 108, 111), 'cv_scores': array([-12.70539, -15.7206 , -13.78044, -15.04294]), 'avg_score': -14.31234062884421, 'feature_names': ('4', '15', '18', '26', '38', '39', '40', '45', '53', '65', '77', '103', '108', '111')}, 15: {'feature_idx': (4, 15, 18, 26, 38, 39, 40, 45, 53, 65, 77, 86, 103, 108, 111), 'cv_scores': array([-12.7012 , -15.72287, -13.77219, -15.03698]), 'avg_score': -14.308311494572207, 'feature_names': ('4', '15', '18', '26', '38', '39', '40', '45', '53', '65', '77', '86', '103', '108', '111')}, 16: {'feature_idx': (4, 15, 18, 26, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 108, 111), 'cv_scores': array([-12.67059, -15.73703, -13.76609, -15.01268]), 'avg_score': -14.296597835544574, 'feature_names': ('4', '15', '18', '26', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '108', '111')}, 17: {'feature_idx': (4, 15, 18, 26, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 108, 111, 124), 'cv_scores': array([-12.67051, -15.73601, -13.75841, -15.00676]), 'avg_score': -14.292921975358123, 'feature_names': ('4', '15', '18', '26', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '108', '111', '124')}, 18: {'feature_idx': (4, 15, 18, 26, 36, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 108, 111, 124), 'cv_scores': array([-12.67474, -15.72183, -13.75714, -14.99762]), 'avg_score': -14.287832301354667, 'feature_names': ('4', '15', '18', '26', '36', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '108', '111', '124')}, 19: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 108, 111, 124), 'cv_scores': array([-12.672  , -15.72424, -13.74743, -14.99219]), 'avg_score': -14.283965778075537, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '108', '111', '124')}, 20: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 107, 108, 111, 124), 'cv_scores': array([-12.65964, -15.72222, -13.74249, -14.99626]), 'avg_score': -14.28015105910031, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '107', '108', '111', '124')}, 21: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 107, 108, 111, 124, 139), 'cv_scores': array([-12.65614, -15.71689, -13.74362, -14.98677]), 'avg_score': -14.27585425350352, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '107', '108', '111', '124', '139')}, 22: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 86, 103, 107, 108, 111, 119, 124, 139), 'cv_scores': array([-12.64205, -15.70115, -13.75274, -14.99528]), 'avg_score': -14.27280538778122, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '86', '103', '107', '108', '111', '119', '124', '139')}, 23: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 86, 103, 107, 108, 111, 119, 124, 139), 'cv_scores': array([-12.64218, -15.6948 , -13.75413, -14.99164]), 'avg_score': -14.270690174445862, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '86', '103', '107', '108', '111', '119', '124', '139')}, 24: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 86, 103, 107, 108, 111, 119, 124, 131, 139), 'cv_scores': array([-12.64159, -15.6909 , -13.75297, -14.9885 ]), 'avg_score': -14.268491045372057, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '86', '103', '107', '108', '111', '119', '124', '131', '139')}, 25: {'feature_idx': (4, 15, 18, 26, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.64152, -15.68741, -13.75072, -14.98654]), 'avg_score': -14.266547656748429, 'feature_names': ('4', '15', '18', '26', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 26: {'feature_idx': (4, 15, 18, 26, 35, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63652, -15.68288, -13.75152, -14.98685]), 'avg_score': -14.264441022989793, 'feature_names': ('4', '15', '18', '26', '35', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 27: {'feature_idx': (4, 15, 18, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63315, -15.68082, -13.75378, -14.98049]), 'avg_score': -14.262062510838954, 'feature_names': ('4', '15', '18', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 28: {'feature_idx': (4, 15, 18, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 83, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.62917, -15.68132, -13.75077, -14.98146]), 'avg_score': -14.260679277218316, 'feature_names': ('4', '15', '18', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '83', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 29: {'feature_idx': (4, 15, 17, 18, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 83, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63191, -15.67738, -13.74624, -14.98132]), 'avg_score': -14.25921158521572, 'feature_names': ('4', '15', '17', '18', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '83', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 30: {'feature_idx': (4, 15, 17, 18, 19, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 73, 77, 80, 83, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63492, -15.67587, -13.7423 , -14.97839]), 'avg_score': -14.257868089049934, 'feature_names': ('4', '15', '17', '18', '19', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '73', '77', '80', '83', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 31: {'feature_idx': (4, 15, 17, 18, 19, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 66, 73, 77, 80, 83, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63617, -15.67518, -13.73776, -14.97545]), 'avg_score': -14.256138692755332, 'feature_names': ('4', '15', '17', '18', '19', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '66', '73', '77', '80', '83', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 32: {'feature_idx': (4, 15, 17, 18, 19, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 66, 67, 73, 77, 80, 83, 86, 103, 107, 108, 111, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63563, -15.67176, -13.7401 , -14.97043]), 'avg_score': -14.254480092273543, 'feature_names': ('4', '15', '17', '18', '19', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '66', '67', '73', '77', '80', '83', '86', '103', '107', '108', '111', '119', '124', '131', '139', '156')}, 33: {'feature_idx': (4, 15, 17, 18, 19, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 66, 67, 73, 77, 80, 83, 86, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63916, -15.66698, -13.73562, -14.97203]), 'avg_score': -14.25344634341284, 'feature_names': ('4', '15', '17', '18', '19', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '66', '67', '73', '77', '80', '83', '86', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 34: {'feature_idx': (4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 53, 65, 66, 67, 73, 77, 80, 83, 86, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.633  , -15.68207, -13.72831, -14.96642]), 'avg_score': -14.252448347576326, 'feature_names': ('4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '53', '65', '66', '67', '73', '77', '80', '83', '86', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 35: {'feature_idx': (4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 48, 53, 65, 66, 67, 73, 77, 80, 83, 86, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.6331 , -15.67874, -13.72546, -14.96769]), 'avg_score': -14.251247989235262, 'feature_names': ('4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '48', '53', '65', '66', '67', '73', '77', '80', '83', '86', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 36: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 48, 53, 65, 66, 67, 73, 77, 80, 83, 86, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63228, -15.67744, -13.7252 , -14.96638]), 'avg_score': -14.250322117472617, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '48', '53', '65', '66', '67', '73', '77', '80', '83', '86', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 37: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 48, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.63024, -15.67804, -13.72242, -14.96728]), 'avg_score': -14.249493403259613, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '48', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 38: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 48, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.6262 , -15.68164, -13.72917, -14.95823]), 'avg_score': -14.248808111829204, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '48', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 39: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.62421, -15.68015, -13.73056, -14.9583 ]), 'avg_score': -14.248303731910395, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 40: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156), 'cv_scores': array([-12.61941, -15.69064, -13.72688, -14.95409]), 'avg_score': -14.247754275733572, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156')}, 41: {'feature_idx': (1, 4, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156, 168), 'cv_scores': array([-12.61875, -15.6877 , -13.73061, -14.95223]), 'avg_score': -14.247323551336361, 'feature_names': ('1', '4', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156', '168')}, 42: {'feature_idx': (1, 4, 6, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156, 168), 'cv_scores': array([-12.61661, -15.6885 , -13.72749, -14.95247]), 'avg_score': -14.246267478334014, 'feature_names': ('1', '4', '6', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156', '168')}, 43: {'feature_idx': (1, 4, 6, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 131, 139, 156, 163, 168), 'cv_scores': array([-12.61526, -15.68968, -13.72591, -14.95265]), 'avg_score': -14.245872423859504, 'feature_names': ('1', '4', '6', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '131', '139', '156', '163', '168')}, 44: {'feature_idx': (1, 4, 6, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.61507, -15.68991, -13.72534, -14.95165]), 'avg_score': -14.245493188756553, 'feature_names': ('1', '4', '6', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '128', '131', '139', '156', '163', '168')}, 45: {'feature_idx': (1, 4, 6, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.61659, -15.68856, -13.72351, -14.95189]), 'avg_score': -14.245135336305808, 'feature_names': ('1', '4', '6', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '128', '131', '139', '156', '163', '168')}, 46: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.61132, -15.6868 , -13.72421, -14.95392]), 'avg_score': -14.244060668506563, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '119', '124', '128', '131', '139', '156', '163', '168')}, 47: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 100, 102, 103, 107, 108, 111, 112, 113, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.60859, -15.6851 , -13.72911, -14.9519 ]), 'avg_score': -14.243673491839168, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '100', '102', '103', '107', '108', '111', '112', '113', '119', '124', '128', '131', '139', '156', '163', '168')}, 48: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 107, 108, 111, 112, 113, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.60583, -15.68399, -13.73159, -14.95179]), 'avg_score': -14.243299280783758, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '107', '108', '111', '112', '113', '119', '124', '128', '131', '139', '156', '163', '168')}, 49: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 107, 108, 111, 112, 113, 119, 124, 128, 131, 139, 156, 163, 168), 'cv_scores': array([-12.60685, -15.69174, -13.72696, -14.94617]), 'avg_score': -14.242928166526621, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '107', '108', '111', '112', '113', '119', '124', '128', '131', '139', '156', '163', '168')}, 50: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 107, 108, 111, 112, 113, 119, 124, 128, 131, 139, 156, 163, 165, 168), 'cv_scores': array([-12.60375, -15.69076, -13.72611, -14.95026]), 'avg_score': -14.242718960533416, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '107', '108', '111', '112', '113', '119', '124', '128', '131', '139', '156', '163', '165', '168')}, 51: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 139, 156, 163, 165, 168), 'cv_scores': array([-12.60388, -15.69048, -13.72582, -14.95047]), 'avg_score': -14.242661585306845, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '139', '156', '163', '165', '168')}, 52: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 139, 156, 163, 165, 168), 'cv_scores': array([-12.60206, -15.69123, -13.72854, -14.94859]), 'avg_score': -14.242603862291057, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '139', '156', '163', '165', '168')}, 53: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 139, 153, 156, 163, 165, 168), 'cv_scores': array([-12.60076, -15.68894, -13.72915, -14.95141]), 'avg_score': -14.242564860340295, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '139', '153', '156', '163', '165', '168')}, 54: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.58602, -15.68411, -13.73804, -14.94774]), 'avg_score': -14.23897812116543, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '139', '150', '153', '156', '163', '165', '168')}, 55: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.60753, -15.67569, -13.73075, -14.93285]), 'avg_score': -14.236701445519335, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 56: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 63, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.60612, -15.67792, -13.73115, -14.93135]), 'avg_score': -14.236632543129993, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '63', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 57: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 63, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.62003, -15.67484, -13.72394, -14.92554]), 'avg_score': -14.23608939260324, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '63', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 58: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 63, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61594, -15.67473, -13.72497, -14.92691]), 'avg_score': -14.235635141372693, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '63', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 59: {'feature_idx': (1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61288, -15.67149, -13.72385, -14.92012]), 'avg_score': -14.232085307417446, 'feature_names': ('1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 60: {'feature_idx': (0, 1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61284, -15.67145, -13.7238 , -14.92008]), 'avg_score': -14.232044794835419, 'feature_names': ('0', '1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 61: {'feature_idx': (0, 1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61279, -15.6715 , -13.72374, -14.92006]), 'avg_score': -14.232022424763754, 'feature_names': ('0', '1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '150', '153', '156', '163', '165', '168')}, 62: {'feature_idx': (0, 1, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 148, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61278, -15.67149, -13.72374, -14.92005]), 'avg_score': -14.23201416732762, 'feature_names': ('0', '1', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '148', '150', '153', '156', '163', '165', '168')}, 63: {'feature_idx': (0, 1, 2, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 148, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61278, -15.67149, -13.72374, -14.92005]), 'avg_score': -14.232014030947475, 'feature_names': ('0', '1', '2', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '148', '150', '153', '156', '163', '165', '168')}, 64: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 148, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61278, -15.67149, -13.72374, -14.92005]), 'avg_score': -14.232014119105088, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '148', '150', '153', '156', '163', '165', '168')}, 65: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 148, 149, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61277, -15.67149, -13.72375, -14.92005]), 'avg_score': -14.232015237913389, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '148', '149', '150', '153', '156', '163', '165', '168')}, 66: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 146, 148, 149, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61278, -15.67149, -13.72375, -14.92006]), 'avg_score': -14.23201900804353, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '146', '148', '149', '150', '153', '156', '163', '165', '168')}, 67: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 156, 163, 165, 168), 'cv_scores': array([-12.61281, -15.67158, -13.72379, -14.92005]), 'avg_score': -14.232057497818296, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '156', '163', '165', '168')}, 68: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61442, -15.67131, -13.72307, -14.91964]), 'avg_score': -14.232110216203328, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '156', '161', '163', '165', '168')}, 69: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61487, -15.67168, -13.72288, -14.91946]), 'avg_score': -14.232223443354492, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '156', '161', '163', '165', '168')}, 70: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 154, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61566, -15.6643 , -13.73428, -14.91327]), 'avg_score': -14.2318758123211, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '154', '156', '161', '163', '165', '168')}, 71: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 154, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61566, -15.6643 , -13.73428, -14.91327]), 'avg_score': -14.231875812628074, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '154', '156', '161', '163', '165', '168')}, 72: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 131, 136, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61566, -15.6643 , -13.73428, -14.91327]), 'avg_score': -14.23187581297891, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '131', '136', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 73: {'feature_idx': (0, 1, 2, 3, 4, 6, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 131, 136, 138, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62369, -15.66267, -13.73052, -14.91101]), 'avg_score': -14.231971638340166, 'feature_names': ('0', '1', '2', '3', '4', '6', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '131', '136', '138', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 74: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 131, 136, 138, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62436, -15.66276, -13.73032, -14.91101]), 'avg_score': -14.232110970420631, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '131', '136', '138', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 75: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 131, 136, 138, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.6244 , -15.66282, -13.73082, -14.91109]), 'avg_score': -14.232281607044996, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '131', '136', '138', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 76: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 130, 131, 136, 138, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62421, -15.66281, -13.73105, -14.91152]), 'avg_score': -14.232397861300214, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '130', '131', '136', '138', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 77: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 91, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 130, 131, 136, 138, 139, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62613, -15.66063, -13.72718, -14.90648]), 'avg_score': -14.230101708893699, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '91', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '130', '131', '136', '138', '139', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 78: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 91, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62617, -15.66193, -13.72702, -14.90639]), 'avg_score': -14.230377123327822, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '91', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 79: {'feature_idx': (0, 1, 2, 3, 4, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 91, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62607, -15.66298, -13.72671, -14.90676]), 'avg_score': -14.230629325493421, 'feature_names': ('0', '1', '2', '3', '4', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '91', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 80: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 83, 86, 89, 91, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62595, -15.66298, -13.72699, -14.90778]), 'avg_score': -14.230926300300531, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '83', '86', '89', '91', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 81: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 81, 83, 86, 89, 91, 100, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62586, -15.66635, -13.72617, -14.90687]), 'avg_score': -14.231312666538528, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '81', '83', '86', '89', '91', '100', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 82: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 81, 83, 86, 89, 91, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62441, -15.66919, -13.72778, -14.90539]), 'avg_score': -14.231691723954661, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '81', '83', '86', '89', '91', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 83: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 81, 83, 86, 87, 89, 91, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62465, -15.66921, -13.72764, -14.90695]), 'avg_score': -14.23211232879219, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '81', '83', '86', '87', '89', '91', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 84: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 73, 77, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.624  , -15.67465, -13.72623, -14.9051 ]), 'avg_score': -14.232493986687619, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '73', '77', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 85: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62494, -15.67637, -13.726  , -14.90447]), 'avg_score': -14.232944625661972, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 86: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.62951, -15.67465, -13.72537, -14.90399]), 'avg_score': -14.23337953217725, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 87: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.6313 , -15.67447, -13.72422, -14.90541]), 'avg_score': -14.233847706003965, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 88: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63313, -15.67618, -13.72306, -14.90503]), 'avg_score': -14.234350236487892, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 89: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63403, -15.67498, -13.72336, -14.90718]), 'avg_score': -14.23488638952231, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 90: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63303, -15.67718, -13.72324, -14.90841]), 'avg_score': -14.235462913654406, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 91: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63618, -15.67818, -13.72265, -14.9072 ]), 'avg_score': -14.236053460997397, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 92: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 136, 138, 139, 141, 146, 147, 148, 149, 150, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63597, -15.678  , -13.72512, -14.90755]), 'avg_score': -14.236658025413018, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '136', '138', '139', '141', '146', '147', '148', '149', '150', '153', '154', '155', '156', '161', '163', '165', '168')}, 93: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 136, 138, 139, 141, 146, 147, 148, 149, 150, 151, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.63601, -15.67529, -13.73324, -14.90468]), 'avg_score': -14.237305185878284, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '136', '138', '139', '141', '146', '147', '148', '149', '150', '151', '153', '154', '155', '156', '161', '163', '165', '168')}, 94: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61704, -15.69423, -13.72383, -14.90099]), 'avg_score': -14.234021641912147, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '153', '154', '155', '156', '161', '163', '165', '168')}, 95: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61561, -15.69944, -13.72205, -14.90093]), 'avg_score': -14.234509669325035, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '153', '154', '155', '156', '161', '163', '165', '168')}, 96: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61586, -15.69937, -13.72323, -14.90124]), 'avg_score': -14.234923603959968, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '135', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '153', '154', '155', '156', '161', '163', '165', '168')}, 97: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.61364, -15.6966 , -13.72095, -14.91024]), 'avg_score': -14.235358153140506, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '135', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '152', '153', '154', '155', '156', '161', '163', '165', '168')}, 98: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 82, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 161, 163, 165, 168), 'cv_scores': array([-12.6128 , -15.70145, -13.7202 , -14.90946]), 'avg_score': -14.235977111789339, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '82', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '135', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '152', '153', '154', '155', '156', '161', '163', '165', '168')}, 99: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 82, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 161, 163, 165, 168, 169), 'cv_scores': array([-12.61279, -15.70212, -13.72158, -14.90981]), 'avg_score': -14.23657370836351, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '82', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '135', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '152', '153', '154', '155', '156', '161', '163', '165', '168', '169')}, 100: {'feature_idx': (0, 1, 2, 3, 4, 5, 6, 10, 14, 15, 17, 18, 19, 21, 22, 25, 26, 27, 28, 30, 31, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 53, 56, 62, 63, 64, 65, 66, 67, 70, 71, 73, 77, 79, 80, 81, 82, 83, 86, 87, 89, 91, 96, 100, 101, 102, 103, 104, 106, 107, 108, 111, 112, 113, 116, 118, 119, 122, 124, 128, 129, 130, 131, 134, 135, 136, 138, 139, 141, 143, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 161, 163, 165, 168, 169), 'cv_scores': array([-12.6128 , -15.704  , -13.72271, -14.90972]), 'avg_score': -14.237308120241192, 'feature_names': ('0', '1', '2', '3', '4', '5', '6', '10', '14', '15', '17', '18', '19', '21', '22', '25', '26', '27', '28', '30', '31', '34', '35', '36', '37', '38', '39', '40', '42', '43', '44', '45', '46', '48', '49', '53', '56', '62', '63', '64', '65', '66', '67', '70', '71', '73', '77', '79', '80', '81', '82', '83', '86', '87', '89', '91', '96', '100', '101', '102', '103', '104', '106', '107', '108', '111', '112', '113', '116', '118', '119', '122', '124', '128', '129', '130', '131', '134', '135', '136', '138', '139', '141', '143', '146', '147', '148', '149', '150', '151', '152', '153', '154', '155', '156', '161', '163', '165', '168', '169')}}\n"
     ]
    }
   ],
   "source": [
    "print('best combination (ACC: %.3f): %s\\n' % (sfs.k_score_, sfs.k_feature_idx_))\n",
    "print('all subsets:\\n', sfs.subsets_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_sfs(sfs.get_metric_dict(), kind='std_err');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1 = plot_sfs(sfs.get_metric_dict(), kind='std_err')\n",
    "plt.title('Sequential Forward Selection (w. StdDev)')\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1 = plot_sfs(sfs.get_metric_dict(), kind='std_err')\n",
    "plt.title('Sequential Forward Selection (w. StdDev)')\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  0,   1,   2,   3,   4,   5,   6,   7,   8,   9,  10,  11,  12,  13,  14,  15,  16,  17,  18,  19,\n",
       "        20,  21,  22,  23,  24,  25,  26,  27,  28,  29,  30,  31,  32,  33,  34,  35,  36,  37,  38,  39,\n",
       "        40,  41,  42,  43,  44,  45,  46,  47,  48,  49,  50,  51,  52,  53,  54,  55,  56,  57,  58,  59,\n",
       "        60,  61,  62,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,  73,  74,  75,  76,  77,  78,  79,\n",
       "        80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,  91,  92,  93,  94,  95,  96,  97,  98,  99,\n",
       "       100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119,\n",
       "       120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139,\n",
       "       140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.arange(160)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1 = plot_sfs(sfs.get_metric_dict(), kind='std_err')\n",
    "plt.title('Sequential Forward Selection (w. StdDev)')\n",
    "pos = np.arange(160)\n",
    "ticks = plt.xticks(pos[::10], pos[::10], rotation=90)\n",
    "plt.grid()\n",
    "plt.show() #full 72% dataset\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "feats_se_100 = [0, 1, 2, 3, 4, 5, 6, 9, 10, 13, 14, 15, 17, 19, 21, 23, 25, 26, 27, 28, 30, 31, 32, 33, 34, 35, 36, 37, 40, 41, 43, 44, 45, 46, 47, 48, 49, 50, 52, 56, 59, 63, 66, 67, 68, 70, 71, 73, 75, 76, 80, 83, 84, 85, 89, 91, 92, 94, 95, 99, 100, 103, 104, 107, 109, 110, 111, 114, 115, 119, 120, 123, 127, 130, 131, 133, 135, 137, 138, 143, 145, 146, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 168, 172, 173, 174, 175, 176, 177, 179]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_fe_100 = x.columns[feats_se_100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "feats_se_80 = [0, 1, 2, 3, 4, 6, 9, 13, 14, 15, 19, 21, 23, 26, 27, 28, 31, 32, 34, 35, 36, 37, 40, 41, 43, 44, 45, 46, 47, 48, 50, 52, 56, 59, 63, 66, 67, 68, 70, 71, 75, 76, 80, 83, 89, 92, 94, 99, 100, 104, 107, 109, 110, 111, 114, 115, 120, 123, 127, 130, 131, 135, 137, 138, 143, 146, 153, 154, 155, 156, 157, 159, 160, 161, 162, 168, 173, 174, 177, 179]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['feature_1', 'feature_2', 'feature_3', 'transactions_count',\n",
       "       'subsector_id_nunique', 'merchant_category_id_nunique',\n",
       "       'purchase_Month_max', 'purchase_Week_max', 'purchase_Dayofweek_mean',\n",
       "       'purchase_Dayofweek_min', 'purchase_Day_min', 'purchase_Hour_nunique',\n",
       "       'purchase_Hour_min', 'purchase_amount_max', 'purchase_amount_min',\n",
       "       'purchase_amount_mean', 'installments_sum', 'installments_max',\n",
       "       'installments_var', 'installments_skew', 'month_lag_max',\n",
       "       'month_lag_min', 'month_lag_skew', 'month_diff_mean', 'month_diff_skew',\n",
       "       'purchased_on_weekend_mean', 'category_1_mean', 'category_2_mean',\n",
       "       'category_3_mean', 'card_id_count', 'price_max', 'price_var',\n",
       "       'Mothers_Day_2018_mean', 'duration_max', 'amount_month_ratio_min',\n",
       "       'amount_month_ratio_skew', 'category_2_mean_mean',\n",
       "       'category_3_mean_mean', 'purchase_date_average',\n",
       "       'purchase_date_uptonow', 'merchant_id_nunique_old',\n",
       "       'merchant_category_id_nunique_old', 'purchase_Month_max_old',\n",
       "       'purchase_Week_min_old', 'purchase_Hour_nunique_old',\n",
       "       'purchase_Hour_max_old', 'purchase_amount_max_old',\n",
       "       'installments_sum_old', 'installments_max_old', 'month_lag_max_old',\n",
       "       'month_lag_var_old', 'month_diff_max', 'month_diff_min',\n",
       "       'month_diff_mean_old', 'authorized_flag_mean',\n",
       "       'purchased_on_weekend_mean_old', 'price_sum', 'price_min_old',\n",
       "       'fathers_day_2017_mean', 'Black_Friday_2017_mean_old',\n",
       "       'Mothers_Day_2018_mean_old', 'duration_var_old',\n",
       "       'amount_month_ratio_mean_old', 'amount_month_ratio_min_old',\n",
       "       'category_3_mean_mean_old', 'purchase_date_uptonow_old',\n",
       "       'days_feature1_ratio', 'days_feature2_ratio', 'days_feature3_ratio',\n",
       "       'feature_sum', 'feature_mean', 'feature_min', 'feature_var',\n",
       "       'card_id_total', 'card_id_count_total', 'purchase_amount_sum_ratio',\n",
       "       'month_diff_ratio', 'installments_total', 'CLV', 'CLV_ratio'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.columns[feats_se_80]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_fe_80 = x.columns[feats_se_80]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_100 = [0, 1, 2, 3, 4, 5, 6, 9, 10, 13, 14, 15, 17, 19, 21, 23, 25, 26, 27, 28, 30, 31, 32, 33, 34, 35, 36, 37, 40, 41, 43, 44, 45, 46, 47, 48, 49, 50, 52, 56, 59, 63, 66, 67, 68, 70, 71, 73, 75, 76, 80, 83, 84, 85, 89, 91, 92, 94, 95, 99, 100, 103, 104, 107, 109, 110, 111, 114, 115, 119, 120, 123, 127, 130, 131, 133, 135, 137, 138, 143, 145, 146, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 168, 172, 173, 174, 175, 176, 177, 179]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_fe_100 = x.columns[fe_100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['feature_1', 'feature_2', 'feature_3', 'transactions_count',\n",
       "       'subsector_id_nunique', 'merchant_id_nunique',\n",
       "       'merchant_category_id_nunique', 'purchase_Month_max',\n",
       "       'purchase_Week_nunique', 'purchase_Week_max', 'purchase_Dayofweek_mean',\n",
       "       'purchase_Dayofweek_min', 'purchase_Day_nunique', 'purchase_Day_min',\n",
       "       'purchase_Hour_nunique', 'purchase_Hour_min', 'purchase_amount_sum',\n",
       "       'purchase_amount_max', 'purchase_amount_min', 'purchase_amount_mean',\n",
       "       'purchase_amount_skew', 'installments_sum', 'installments_max',\n",
       "       'installments_mean', 'installments_var', 'installments_skew',\n",
       "       'month_lag_max', 'month_lag_min', 'month_lag_skew', 'month_diff_mean',\n",
       "       'month_diff_skew', 'purchased_on_weekend_mean', 'category_1_mean',\n",
       "       'category_2_mean', 'category_3_mean', 'card_id_count', 'price_mean',\n",
       "       'price_max', 'price_var', 'Mothers_Day_2018_mean', 'duration_max',\n",
       "       'amount_month_ratio_min', 'amount_month_ratio_skew',\n",
       "       'category_2_mean_mean', 'category_3_mean_mean', 'purchase_date_average',\n",
       "       'purchase_date_uptonow', 'transactions_count_old',\n",
       "       'merchant_id_nunique_old', 'merchant_category_id_nunique_old',\n",
       "       'purchase_Month_max_old', 'purchase_Week_min_old',\n",
       "       'purchase_Week_max_old', 'purchase_Dayofweek_mean_old',\n",
       "       'purchase_Hour_nunique_old', 'purchase_Hour_min_old',\n",
       "       'purchase_Hour_max_old', 'purchase_amount_max_old',\n",
       "       'purchase_amount_min_old', 'installments_sum_old',\n",
       "       'installments_max_old', 'installments_skew_old', 'month_lag_max_old',\n",
       "       'month_lag_var_old', 'month_diff_max', 'month_diff_min',\n",
       "       'month_diff_mean_old', 'authorized_flag_mean',\n",
       "       'purchased_on_weekend_mean_old', 'card_id_count_old', 'price_sum',\n",
       "       'price_min_old', 'fathers_day_2017_mean', 'Black_Friday_2017_mean_old',\n",
       "       'Mothers_Day_2018_mean_old', 'duration_min_old', 'duration_var_old',\n",
       "       'amount_month_ratio_mean_old', 'amount_month_ratio_min_old',\n",
       "       'category_3_mean_mean_old', 'purchase_date_average_old',\n",
       "       'purchase_date_uptonow_old', 'days_feature1_ratio',\n",
       "       'days_feature2_ratio', 'days_feature3_ratio', 'feature_sum',\n",
       "       'feature_mean', 'feature_max', 'feature_min', 'feature_var',\n",
       "       'card_id_total', 'card_id_count_total', 'purchase_amount_sum_ratio',\n",
       "       'new_last_buy', 'month_diff_ratio', 'installments_total',\n",
       "       'installments_ratio', 'price_total', 'CLV', 'CLV_ratio'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols_fe_100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "fe_160 = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 55, 56, 58, 59, 61, 63, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 95, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 123, 125, 126, 127, 128, 129, 130, 131, 132, 133, 135, 136, 137, 138, 139, 141, 142, 143, 144, 145, 146, 147, 148, 150, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 167, 168, 170, 172, 173, 174, 175, 176, 177, 178, 179]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_fe_160 = x.columns[fe_160]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['feature_1', 'feature_2', 'feature_3', 'transactions_count', 'subsector_id_nunique',\n",
       "       'merchant_id_nunique', 'merchant_category_id_nunique', 'purchase_Month_mean', 'purchase_Month_min',\n",
       "       'purchase_Month_max', 'purchase_Week_nunique', 'purchase_Week_mean', 'purchase_Week_min',\n",
       "       'purchase_Week_max', 'purchase_Dayofweek_mean', 'purchase_Dayofweek_min', 'purchase_Dayofweek_max',\n",
       "       'purchase_Day_nunique', 'purchase_Day_mean', 'purchase_Day_min', 'purchase_Day_max',\n",
       "       'purchase_Hour_nunique', 'purchase_Hour_mean', 'purchase_Hour_min', 'purchase_Hour_max',\n",
       "       'purchase_amount_sum', 'purchase_amount_max', 'purchase_amount_min', 'purchase_amount_mean',\n",
       "       'purchase_amount_skew', 'installments_sum', 'installments_max', 'installments_mean',\n",
       "       'installments_var', 'installments_skew', 'month_lag_max', 'month_lag_min', 'month_lag_mean',\n",
       "       'month_lag_var', 'month_lag_skew', 'month_diff_mean', 'month_diff_var', 'month_diff_skew',\n",
       "       'purchased_on_weekend_mean', 'category_1_mean', 'category_2_mean', 'category_3_mean', 'card_id_count',\n",
       "       'price_mean', 'price_max', 'price_min', 'price_var', 'Christmas_Day_2017_mean',\n",
       "       'Black_Friday_2017_mean', 'Mothers_Day_2018_mean', 'duration_min', 'duration_max', 'duration_skew',\n",
       "       'amount_month_ratio_min', 'amount_month_ratio_var', 'amount_month_ratio_skew', 'category_2_mean_mean',\n",
       "       'category_3_mean_mean', 'purchase_date_diff', 'purchase_date_average', 'purchase_date_uptonow',\n",
       "       'purchase_date_uptomin', 'transactions_count_old', 'subsector_id_nunique_old',\n",
       "       'merchant_id_nunique_old', 'merchant_category_id_nunique_old', 'purchase_Month_nunique',\n",
       "       'purchase_Month_mean_old', 'purchase_Month_min_old', 'purchase_Month_max_old',\n",
       "       'purchase_Week_nunique_old', 'purchase_Week_mean_old', 'purchase_Week_min_old',\n",
       "       'purchase_Week_max_old', 'purchase_Dayofweek_mean_old', 'purchase_Day_nunique_old',\n",
       "       'purchase_Day_mean_old', 'purchase_Day_min_old', 'purchase_Hour_nunique_old',\n",
       "       'purchase_Hour_mean_old', 'purchase_Hour_min_old', 'purchase_Hour_max_old', 'purchase_amount_min_old',\n",
       "       'purchase_amount_var_old', 'purchase_amount_skew_old', 'installments_sum_old', 'installments_max_old',\n",
       "       'installments_mean_old', 'installments_var_old', 'installments_skew_old', 'month_lag_max_old',\n",
       "       'month_lag_min_old', 'month_lag_mean_old', 'month_lag_var_old', 'month_lag_skew_old',\n",
       "       'month_diff_max', 'month_diff_min', 'month_diff_mean_old', 'month_diff_var_old',\n",
       "       'month_diff_skew_old', 'authorized_flag_mean', 'purchased_on_weekend_mean_old', 'category_1_mean_old',\n",
       "       'category_2_mean_old', 'category_3_mean_old', 'card_id_count_old', 'price_sum', 'price_mean_old',\n",
       "       'price_min_old', 'Christmas_Day_2017_mean_old', 'Mothers_Day_2017_mean', 'fathers_day_2017_mean',\n",
       "       'Children_day_2017_mean_old', 'Valentine_Day_2017_mean', 'Black_Friday_2017_mean_old',\n",
       "       'Mothers_Day_2018_mean_old', 'duration_mean_old', 'duration_min_old', 'duration_var_old',\n",
       "       'duration_skew_old', 'amount_month_ratio_mean_old', 'amount_month_ratio_min_old',\n",
       "       'amount_month_ratio_max_old', 'amount_month_ratio_skew_old', 'category_2_mean_mean_old',\n",
       "       'category_3_mean_mean_old', 'purchase_date_diff_old', 'purchase_date_average_old',\n",
       "       'purchase_date_uptonow_old', 'purchase_date_uptomin_old', 'quarter', 'days_feature1', 'days_feature3',\n",
       "       'days_feature1_ratio', 'days_feature2_ratio', 'days_feature3_ratio', 'feature_sum', 'feature_mean',\n",
       "       'feature_max', 'feature_min', 'feature_var', 'card_id_total', 'card_id_count_total',\n",
       "       'card_id_count_ratio', 'purchase_amount_total_min', 'purchase_amount_sum_ratio', 'new_first_buy',\n",
       "       'new_last_buy', 'month_diff_ratio', 'installments_total', 'installments_ratio', 'price_total', 'CLV',\n",
       "       'CLV_old', 'CLV_ratio'], dtype=object)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols_fe_160.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "160"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(fe_160)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['subsector_id_nunique', 'purchase_Month_max', 'purchase_Week_max',\n",
       "       'purchase_Dayofweek_min', 'purchase_Day_min', 'purchase_Hour_min',\n",
       "       'purchase_amount_max', 'installments_sum', 'installments_mean',\n",
       "       'installments_var', 'month_lag_max', 'month_lag_min', 'month_lag_skew',\n",
       "       'month_diff_mean', 'month_diff_skew', 'purchased_on_weekend_mean',\n",
       "       'category_2_mean', 'price_max', 'Mothers_Day_2018_mean',\n",
       "       'amount_month_ratio_skew', 'purchase_date_uptonow',\n",
       "       'merchant_category_id_nunique_old', 'purchase_Month_max_old',\n",
       "       'purchase_Week_min_old', 'purchase_Hour_nunique_old',\n",
       "       'installments_sum_old', 'installments_max_old', 'month_diff_min',\n",
       "       'month_diff_mean_old', 'authorized_flag_mean',\n",
       "       'purchased_on_weekend_mean_old', 'category_1_mean_old',\n",
       "       'fathers_day_2017_mean', 'Mothers_Day_2018_mean_old',\n",
       "       'amount_month_ratio_min_old', 'purchase_date_uptonow_old',\n",
       "       'feature_mean', 'feature_min', 'feature_var', 'month_diff_ratio'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fe_40 = [4, 9, 13, 15, 19, 23, 26, 31, 33, 34, 36, 37, 40, 41, 43, 44, 46, 50, 56, 66, 71, 76, 80, 83, 89, 99, 100, 110, 111, 114, 115, 116, 127, 131, 138, 146, 157, 159, 160, 173]\n",
    "cols_fe_40 = x.columns[fe_40]\n",
    "cols_fe_40"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['feature_1', 'feature_2', 'feature_3', 'transactions_count', 'subsector_id_nunique',\n",
       "       'merchant_id_nunique', 'merchant_category_id_nunique', 'purchase_Month_mean', 'purchase_Month_min',\n",
       "       'purchase_Month_max', 'purchase_Week_nunique', 'purchase_Week_mean', 'purchase_Week_min',\n",
       "       'purchase_Week_max', 'purchase_Dayofweek_mean', 'purchase_Dayofweek_min', 'purchase_Dayofweek_max',\n",
       "       'purchase_Day_nunique', 'purchase_Day_mean', 'purchase_Day_min', 'purchase_Day_max',\n",
       "       'purchase_Hour_nunique', 'purchase_Hour_mean', 'purchase_Hour_min', 'purchase_Hour_max',\n",
       "       'purchase_amount_sum', 'purchase_amount_max', 'purchase_amount_min', 'purchase_amount_mean',\n",
       "       'purchase_amount_var', 'purchase_amount_skew', 'installments_sum', 'installments_max',\n",
       "       'installments_mean', 'installments_var', 'installments_skew', 'month_lag_max', 'month_lag_min',\n",
       "       'month_lag_mean', 'month_lag_var', 'month_lag_skew', 'month_diff_mean', 'month_diff_var',\n",
       "       'month_diff_skew', 'purchased_on_weekend_mean', 'category_1_mean', 'category_2_mean',\n",
       "       'category_3_mean', 'card_id_count', 'price_mean', 'price_max', 'price_min', 'Christmas_Day_2017_mean',\n",
       "       'Children_day_2017_mean', 'Black_Friday_2017_mean', 'Mothers_Day_2018_mean', 'duration_min',\n",
       "       'duration_var', 'duration_skew', 'amount_month_ratio_min', 'amount_month_ratio_max',\n",
       "       'amount_month_ratio_var', 'amount_month_ratio_skew', 'category_2_mean_mean', 'category_3_mean_mean',\n",
       "       'purchase_date_diff', 'purchase_date_average', 'purchase_date_uptonow', 'purchase_date_uptomin',\n",
       "       'transactions_count_old', 'subsector_id_nunique_old', 'merchant_id_nunique_old',\n",
       "       'merchant_category_id_nunique_old', 'purchase_Month_nunique', 'purchase_Month_mean_old',\n",
       "       'purchase_Month_min_old', 'purchase_Month_max_old', 'purchase_Week_nunique_old',\n",
       "       'purchase_Week_mean_old', 'purchase_Week_min_old', 'purchase_Week_max_old',\n",
       "       'purchase_Dayofweek_mean_old', 'purchase_Day_nunique_old', 'purchase_Day_mean_old',\n",
       "       'purchase_Day_min_old', 'purchase_Hour_nunique_old', 'purchase_Hour_mean_old',\n",
       "       'purchase_Hour_min_old', 'purchase_Hour_max_old', 'purchase_amount_min_old',\n",
       "       'purchase_amount_skew_old', 'installments_sum_old', 'installments_max_old', 'installments_mean_old',\n",
       "       'installments_var_old', 'installments_skew_old', 'month_lag_max_old', 'month_lag_min_old',\n",
       "       'month_lag_mean_old', 'month_lag_var_old', 'month_lag_skew_old', 'month_diff_max', 'month_diff_min',\n",
       "       'month_diff_mean_old', 'month_diff_var_old', 'month_diff_skew_old', 'authorized_flag_mean',\n",
       "       'purchased_on_weekend_mean_old', 'category_1_mean_old', 'category_2_mean_old', 'category_3_mean_old',\n",
       "       'card_id_count_old', 'price_mean_old', 'price_min_old', 'Christmas_Day_2017_mean_old',\n",
       "       'Mothers_Day_2017_mean', 'fathers_day_2017_mean', 'Children_day_2017_mean_old',\n",
       "       'Valentine_Day_2017_mean', 'Black_Friday_2017_mean_old', 'Mothers_Day_2018_mean_old',\n",
       "       'duration_mean_old', 'duration_min_old', 'duration_skew_old', 'amount_month_ratio_min_old',\n",
       "       'amount_month_ratio_skew_old', 'category_2_mean_mean_old', 'category_3_mean_mean_old',\n",
       "       'purchase_date_diff_old', 'purchase_date_average_old', 'purchase_date_uptonow_old',\n",
       "       'purchase_date_uptomin_old', 'quarter', 'elapsed_time', 'days_feature1', 'days_feature2',\n",
       "       'days_feature3', 'days_feature1_ratio', 'days_feature2_ratio', 'days_feature3_ratio', 'feature_sum',\n",
       "       'feature_mean', 'feature_max', 'feature_min', 'feature_var', 'card_id_total', 'card_id_count_total',\n",
       "       'card_id_count_ratio', 'purchase_amount_total_min', 'purchase_amount_sum_ratio', 'hist_first_buy',\n",
       "       'new_first_buy', 'hist_last_buy', 'new_last_buy', 'month_diff_ratio', 'installments_total',\n",
       "       'installments_ratio', 'price_total', 'CLV', 'CLV_ratio'], dtype=object)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_160 = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 58, 60, 61, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 95, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 121, 123, 125, 126, 127, 128, 129, 130, 131, 132, 133, 136, 138, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 179]\n",
    "cols_ful_160 = x.columns[full_160]\n",
    "cols_ful_160.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_fe_160 = ['feature_1', 'feature_2', 'feature_3', 'transactions_count', 'subsector_id_nunique',\n",
    "       'merchant_id_nunique', 'merchant_category_id_nunique', 'purchase_Month_mean', 'purchase_Month_min',\n",
    "       'purchase_Month_max', 'purchase_Week_nunique', 'purchase_Week_mean', 'purchase_Week_min',\n",
    "       'purchase_Week_max', 'purchase_Dayofweek_mean', 'purchase_Dayofweek_min', 'purchase_Dayofweek_max',\n",
    "       'purchase_Day_nunique', 'purchase_Day_mean', 'purchase_Day_min', 'purchase_Day_max',\n",
    "       'purchase_Hour_nunique', 'purchase_Hour_mean', 'purchase_Hour_min', 'purchase_Hour_max',\n",
    "       'purchase_amount_sum', 'purchase_amount_max', 'purchase_amount_min', 'purchase_amount_mean',\n",
    "       'purchase_amount_skew', 'installments_sum', 'installments_max', 'installments_mean',\n",
    "       'installments_var', 'installments_skew', 'month_lag_max', 'month_lag_min', 'month_lag_mean',\n",
    "       'month_lag_var', 'month_lag_skew', 'month_diff_mean', 'month_diff_var', 'month_diff_skew',\n",
    "       'purchased_on_weekend_mean', 'category_1_mean', 'category_2_mean', 'category_3_mean', 'card_id_count',\n",
    "       'price_mean', 'price_max', 'price_min', 'price_var', 'Christmas_Day_2017_mean',\n",
    "       'Black_Friday_2017_mean', 'Mothers_Day_2018_mean', 'duration_min', 'duration_max', 'duration_skew',\n",
    "       'amount_month_ratio_min', 'amount_month_ratio_var', 'amount_month_ratio_skew', 'category_2_mean_mean',\n",
    "       'category_3_mean_mean', 'purchase_date_diff', 'purchase_date_average', 'purchase_date_uptonow',\n",
    "       'purchase_date_uptomin', 'transactions_count_old', 'subsector_id_nunique_old',\n",
    "       'merchant_id_nunique_old', 'merchant_category_id_nunique_old', 'purchase_Month_nunique',\n",
    "       'purchase_Month_mean_old', 'purchase_Month_min_old', 'purchase_Month_max_old',\n",
    "       'purchase_Week_nunique_old', 'purchase_Week_mean_old', 'purchase_Week_min_old',\n",
    "       'purchase_Week_max_old', 'purchase_Dayofweek_mean_old', 'purchase_Day_nunique_old',\n",
    "       'purchase_Day_mean_old', 'purchase_Day_min_old', 'purchase_Hour_nunique_old',\n",
    "       'purchase_Hour_mean_old', 'purchase_Hour_min_old', 'purchase_Hour_max_old', 'purchase_amount_min_old',\n",
    "       'purchase_amount_var_old', 'purchase_amount_skew_old', 'installments_sum_old', 'installments_max_old',\n",
    "       'installments_mean_old', 'installments_var_old', 'installments_skew_old', 'month_lag_max_old',\n",
    "       'month_lag_min_old', 'month_lag_mean_old', 'month_lag_var_old', 'month_lag_skew_old',\n",
    "       'month_diff_max', 'month_diff_min', 'month_diff_mean_old', 'month_diff_var_old',\n",
    "       'month_diff_skew_old', 'authorized_flag_mean', 'purchased_on_weekend_mean_old', 'category_1_mean_old',\n",
    "       'category_2_mean_old', 'category_3_mean_old', 'card_id_count_old', 'price_sum', 'price_mean_old',\n",
    "       'price_min_old', 'Christmas_Day_2017_mean_old', 'Mothers_Day_2017_mean', 'fathers_day_2017_mean',\n",
    "       'Children_day_2017_mean_old', 'Valentine_Day_2017_mean', 'Black_Friday_2017_mean_old',\n",
    "       'Mothers_Day_2018_mean_old', 'duration_mean_old', 'duration_min_old', 'duration_var_old',\n",
    "       'duration_skew_old', 'amount_month_ratio_mean_old', 'amount_month_ratio_min_old',\n",
    "       'amount_month_ratio_max_old', 'amount_month_ratio_skew_old', 'category_2_mean_mean_old',\n",
    "       'category_3_mean_mean_old', 'purchase_date_diff_old', 'purchase_date_average_old',\n",
    "       'purchase_date_uptonow_old', 'purchase_date_uptomin_old', 'quarter', 'days_feature1', 'days_feature3',\n",
    "       'days_feature1_ratio', 'days_feature2_ratio', 'days_feature3_ratio', 'feature_sum', 'feature_mean',\n",
    "       'feature_max', 'feature_min', 'feature_var', 'card_id_total', 'card_id_count_total',\n",
    "       'card_id_count_ratio', 'purchase_amount_total_min', 'purchase_amount_sum_ratio', 'new_first_buy',\n",
    "       'new_last_buy', 'month_diff_ratio', 'installments_total', 'installments_ratio', 'price_total', 'CLV',\n",
    "       'CLV_old', 'CLV_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['feature_1', 'feature_2', 'feature_3', 'transactions_count',\n",
       "       'subsector_id_nunique', 'merchant_id_nunique',\n",
       "       'merchant_category_id_nunique', 'purchase_Month_mean',\n",
       "       'purchase_Month_min', 'purchase_Month_max',\n",
       "       ...\n",
       "       'hist_first_buy', 'new_first_buy', 'hist_last_buy', 'new_last_buy',\n",
       "       'month_diff_ratio', 'installments_total', 'installments_ratio',\n",
       "       'price_total', 'CLV', 'CLV_ratio'],\n",
       "      dtype='object', length=160)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols_ful_160"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['purchase_amount_var',\n",
       " 'Children_day_2017_mean',\n",
       " 'duration_var',\n",
       " 'amount_month_ratio_max',\n",
       " 'elapsed_time',\n",
       " 'days_feature2',\n",
       " 'hist_first_buy',\n",
       " 'hist_last_buy']"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[c for c in cols_ful_160 if c not in cols_fe_160]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['feature_1', 'feature_2', 'feature_3', 'transactions_count', 'subsector_id_nunique',\n",
       "       'merchant_id_nunique', 'merchant_category_id_nunique', 'purchase_Month_mean', 'purchase_Month_min',\n",
       "       'purchase_Month_max', 'purchase_Week_nunique', 'purchase_Week_mean', 'purchase_Week_min',\n",
       "       'purchase_Week_max', 'purchase_Dayofweek_mean', 'purchase_Dayofweek_min', 'purchase_Dayofweek_max',\n",
       "       'purchase_Day_nunique', 'purchase_Day_mean', 'purchase_Day_min', 'purchase_Day_max',\n",
       "       'purchase_Hour_nunique', 'purchase_Hour_mean', 'purchase_Hour_min', 'purchase_Hour_max',\n",
       "       'purchase_amount_sum', 'purchase_amount_max', 'purchase_amount_min', 'purchase_amount_mean',\n",
       "       'purchase_amount_var', 'purchase_amount_skew', 'installments_sum', 'installments_max',\n",
       "       'installments_mean', 'installments_var', 'installments_skew', 'month_lag_max', 'month_lag_min',\n",
       "       'month_lag_mean', 'month_lag_var', 'month_lag_skew', 'month_diff_mean', 'month_diff_var',\n",
       "       'month_diff_skew', 'purchased_on_weekend_mean', 'category_1_mean', 'category_2_mean',\n",
       "       'category_3_mean', 'card_id_count', 'price_mean', 'price_max', 'price_min', 'Christmas_Day_2017_mean',\n",
       "       'Children_day_2017_mean', 'Black_Friday_2017_mean', 'Mothers_Day_2018_mean', 'duration_min',\n",
       "       'duration_var', 'duration_skew', 'amount_month_ratio_min', 'amount_month_ratio_max',\n",
       "       'amount_month_ratio_var', 'amount_month_ratio_skew', 'category_2_mean_mean', 'category_3_mean_mean',\n",
       "       'purchase_date_diff', 'purchase_date_average', 'purchase_date_uptonow', 'purchase_date_uptomin',\n",
       "       'transactions_count_old', 'subsector_id_nunique_old', 'merchant_id_nunique_old',\n",
       "       'merchant_category_id_nunique_old', 'purchase_Month_nunique', 'purchase_Month_mean_old',\n",
       "       'purchase_Month_min_old', 'purchase_Month_max_old', 'purchase_Week_nunique_old',\n",
       "       'purchase_Week_mean_old', 'purchase_Week_min_old', 'purchase_Week_max_old',\n",
       "       'purchase_Dayofweek_mean_old', 'purchase_Day_nunique_old', 'purchase_Day_mean_old',\n",
       "       'purchase_Day_min_old', 'purchase_Hour_nunique_old', 'purchase_Hour_mean_old',\n",
       "       'purchase_Hour_min_old', 'purchase_Hour_max_old', 'purchase_amount_min_old',\n",
       "       'purchase_amount_skew_old', 'installments_sum_old', 'installments_max_old', 'installments_mean_old',\n",
       "       'installments_var_old', 'installments_skew_old', 'month_lag_max_old', 'month_lag_min_old',\n",
       "       'month_lag_mean_old', 'month_lag_var_old', 'month_lag_skew_old', 'month_diff_max', 'month_diff_min',\n",
       "       'month_diff_mean_old', 'month_diff_var_old', 'month_diff_skew_old', 'authorized_flag_mean',\n",
       "       'purchased_on_weekend_mean_old', 'category_1_mean_old', 'category_2_mean_old', 'category_3_mean_old',\n",
       "       'card_id_count_old', 'price_mean_old', 'price_min_old', 'Christmas_Day_2017_mean_old',\n",
       "       'Mothers_Day_2017_mean', 'fathers_day_2017_mean', 'Children_day_2017_mean_old',\n",
       "       'Valentine_Day_2017_mean', 'Black_Friday_2017_mean_old', 'Mothers_Day_2018_mean_old',\n",
       "       'duration_mean_old', 'duration_min_old', 'duration_skew_old', 'amount_month_ratio_min_old',\n",
       "       'amount_month_ratio_skew_old', 'category_2_mean_mean_old', 'category_3_mean_mean_old',\n",
       "       'purchase_date_diff_old', 'purchase_date_average_old', 'purchase_date_uptonow_old',\n",
       "       'purchase_date_uptomin_old', 'quarter', 'elapsed_time', 'days_feature1', 'days_feature2',\n",
       "       'days_feature3', 'days_feature1_ratio', 'days_feature2_ratio', 'days_feature3_ratio', 'feature_sum',\n",
       "       'feature_mean', 'feature_max', 'feature_min', 'feature_var', 'card_id_total', 'card_id_count_total',\n",
       "       'card_id_count_ratio', 'purchase_amount_total_min', 'purchase_amount_sum_ratio', 'hist_first_buy',\n",
       "       'new_first_buy', 'hist_last_buy', 'new_last_buy', 'month_diff_ratio', 'installments_total',\n",
       "       'installments_ratio', 'price_total', 'CLV', 'CLV_ratio'], dtype=object)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols_ful_160.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "229px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}