Elo - ML model - alpha.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\"><li><span><a href=\"#Load-imports-and-dfs\" data-toc-modified-id=\"Load-imports-and-dfs-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Load imports and dfs</a></span></li><li><span><a href=\"#Blend\" data-toc-modified-id=\"Blend-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Blend</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "IPython.notebook.set_autosave_interval(120000)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Autosaving every 120 seconds\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 4\n",
    "%autosave 120"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load imports and dfs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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')\n",
    "test_df2 = feather.read_dataframe('test_df_alpha')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "for df in [train_df2, test_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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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",
    "                 'first_active_monthIs_year_start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "code_folding": []
   },
   "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": [
    "df_train_columns = [c for c in cols_included if c not in cols_excluded]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = [c for c in train_df2.columns.values if c not in cols_excluded and c in test_df2.columns.values]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "target = train_df2['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((201917, 386), (123623, 384))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df2.shape, test_df2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df2['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "180"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "exp_cols = ['merchant_address_id_nunique', 'merchant_rating_nunique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = df_train_columns + exp_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "remove_cols = ['Black_Friday_2017_mean']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = [c for c in df_train_columns if c not in remove_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "181"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_train_columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cols =['purchase_date_uptonow',\n",
    " 'month_diff_mean_old',\n",
    " 'purchase_amount_sum_ratio',\n",
    " 'month_lag_min',\n",
    " 'purchased_before_issue',\n",
    " 'installments_median_old',\n",
    " 'feature_max',\n",
    " 'purchase_amount_sum_old',\n",
    " 'purchase_Month_min',\n",
    " 'month_lag_min_old',\n",
    " 'category_2_5.0_purchase_amount_sum_ratio',\n",
    " 'price_var',\n",
    " 'purchase_amount_pct_25',\n",
    " 'price_mean_old',\n",
    " 'amount_month_ratio_min_total',\n",
    " 'purchase_amount_min',\n",
    " 'duration_min_total',\n",
    " 'price_max_old',\n",
    " 'days_feature3_ratio',\n",
    " 'price_mean',\n",
    " 'subsector_id_nunique_old',\n",
    " 'purchase_Hour_nunique_old',\n",
    " 'amount_month_ratio_min',\n",
    " 'category_1_0_purchase_amount_mean_old',\n",
    " 'category_1_1_purchase_amount_mean_total',\n",
    " 'category_2_1.0_purchase_amount_mean_old',\n",
    " 'category_3_1_purchase_amount_mean_ratio',\n",
    " 'observed_elapsed_time',\n",
    " 'purchase_amount_head_sum_old',\n",
    " 'transactions_count_old',\n",
    " 'purchase_Week_mean_old',\n",
    " 'category_2_1.0_purchase_amount_mean_ratio',\n",
    " 'month_diff_min',\n",
    " 'merchant_id_nunique_old',\n",
    " 'purchase_date_successive_diff_mean_old',\n",
    " 'purchase_date_successive_diff_mean',\n",
    " 'duration_var',\n",
    " 'card_id_count_ratio',\n",
    " 'category_3_2_purchase_amount_mean',\n",
    " 'category_3_1_purchase_amount_sum',\n",
    " 'purchase_Dayofweek_mean_old',\n",
    " 'purchase_Week_nunique_old',\n",
    " 'days_feature2',\n",
    " 'category_1_1_purchase_amount_sum',\n",
    " 'hist_first_buy',\n",
    " 'purchase_date_uptonow_old',\n",
    " 'Valentine_Day_2017_mean',\n",
    " 'Children_day_2017_mean_old',\n",
    " 'Mothers_Day_2018_mean',\n",
    " 'authorized_flag_mean',\n",
    " 'purchase_Week_mean',\n",
    " 'purchase_Month_nunique',\n",
    " 'repurchase_merchant_rate_old']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_columns = final_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "181"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_train_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold 0\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttraining's rmse: 3.63751\tvalid_1's rmse: 3.70283\n",
      "[200]\ttraining's rmse: 3.5499\tvalid_1's rmse: 3.66297\n",
      "[300]\ttraining's rmse: 3.50003\tvalid_1's rmse: 3.64955\n",
      "[400]\ttraining's rmse: 3.46597\tvalid_1's rmse: 3.64469\n",
      "[500]\ttraining's rmse: 3.44051\tvalid_1's rmse: 3.64277\n",
      "[600]\ttraining's rmse: 3.41838\tvalid_1's rmse: 3.64165\n",
      "[700]\ttraining's rmse: 3.39872\tvalid_1's rmse: 3.64155\n",
      "[800]\ttraining's rmse: 3.37874\tvalid_1's rmse: 3.64171\n",
      "Early stopping, best iteration is:\n",
      "[657]\ttraining's rmse: 3.40722\tvalid_1's rmse: 3.64135\n",
      "fold 1\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttraining's rmse: 3.63836\tvalid_1's rmse: 3.6947\n",
      "[200]\ttraining's rmse: 3.54941\tvalid_1's rmse: 3.65822\n",
      "[300]\ttraining's rmse: 3.49881\tvalid_1's rmse: 3.64757\n",
      "[400]\ttraining's rmse: 3.46362\tvalid_1's rmse: 3.64443\n",
      "[500]\ttraining's rmse: 3.43779\tvalid_1's rmse: 3.64291\n",
      "[600]\ttraining's rmse: 3.41711\tvalid_1's rmse: 3.64223\n",
      "[700]\ttraining's rmse: 3.39745\tvalid_1's rmse: 3.64173\n",
      "[800]\ttraining's rmse: 3.37854\tvalid_1's rmse: 3.6413\n",
      "[900]\ttraining's rmse: 3.35986\tvalid_1's rmse: 3.64115\n",
      "[1000]\ttraining's rmse: 3.34277\tvalid_1's rmse: 3.64143\n",
      "[1100]\ttraining's rmse: 3.32626\tvalid_1's rmse: 3.64198\n",
      "Early stopping, best iteration is:\n",
      "[900]\ttraining's rmse: 3.35986\tvalid_1's rmse: 3.64115\n",
      "fold 2\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttraining's rmse: 3.63537\tvalid_1's rmse: 3.70505\n",
      "[200]\ttraining's rmse: 3.54545\tvalid_1's rmse: 3.6695\n",
      "[300]\ttraining's rmse: 3.49725\tvalid_1's rmse: 3.65909\n",
      "[400]\ttraining's rmse: 3.46236\tvalid_1's rmse: 3.65463\n",
      "[500]\ttraining's rmse: 3.43749\tvalid_1's rmse: 3.65228\n",
      "[600]\ttraining's rmse: 3.4163\tvalid_1's rmse: 3.65132\n",
      "[700]\ttraining's rmse: 3.39726\tvalid_1's rmse: 3.6505\n",
      "[800]\ttraining's rmse: 3.37906\tvalid_1's rmse: 3.65016\n",
      "[900]\ttraining's rmse: 3.36086\tvalid_1's rmse: 3.65005\n",
      "[1000]\ttraining's rmse: 3.3441\tvalid_1's rmse: 3.64991\n",
      "[1100]\ttraining's rmse: 3.32774\tvalid_1's rmse: 3.64996\n",
      "[1200]\ttraining's rmse: 3.31093\tvalid_1's rmse: 3.65003\n",
      "[1300]\ttraining's rmse: 3.29567\tvalid_1's rmse: 3.65033\n",
      "Early stopping, best iteration is:\n",
      "[1140]\ttraining's rmse: 3.3216\tvalid_1's rmse: 3.64972\n",
      "fold 3\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttraining's rmse: 3.63767\tvalid_1's rmse: 3.69675\n",
      "[200]\ttraining's rmse: 3.54858\tvalid_1's rmse: 3.66166\n",
      "[300]\ttraining's rmse: 3.49612\tvalid_1's rmse: 3.65176\n",
      "[400]\ttraining's rmse: 3.46258\tvalid_1's rmse: 3.64794\n",
      "[500]\ttraining's rmse: 3.43783\tvalid_1's rmse: 3.64621\n",
      "[600]\ttraining's rmse: 3.41779\tvalid_1's rmse: 3.64491\n",
      "[700]\ttraining's rmse: 3.40032\tvalid_1's rmse: 3.64437\n",
      "[800]\ttraining's rmse: 3.38385\tvalid_1's rmse: 3.64394\n",
      "[900]\ttraining's rmse: 3.36771\tvalid_1's rmse: 3.644\n",
      "[1000]\ttraining's rmse: 3.35041\tvalid_1's rmse: 3.644\n",
      "Early stopping, best iteration is:\n",
      "[822]\ttraining's rmse: 3.38019\tvalid_1's rmse: 3.64384\n",
      "fold 4\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttraining's rmse: 3.63789\tvalid_1's rmse: 3.68739\n",
      "[200]\ttraining's rmse: 3.54864\tvalid_1's rmse: 3.65692\n",
      "[300]\ttraining's rmse: 3.49539\tvalid_1's rmse: 3.648\n",
      "[400]\ttraining's rmse: 3.46339\tvalid_1's rmse: 3.64546\n",
      "[500]\ttraining's rmse: 3.44078\tvalid_1's rmse: 3.64455\n",
      "[600]\ttraining's rmse: 3.4215\tvalid_1's rmse: 3.64425\n",
      "[700]\ttraining's rmse: 3.4044\tvalid_1's rmse: 3.64442\n",
      "[800]\ttraining's rmse: 3.38799\tvalid_1's rmse: 3.64477\n",
      "Early stopping, best iteration is:\n",
      "[646]\ttraining's rmse: 3.41359\tvalid_1's rmse: 3.64415\n",
      "3.644042133990217 CV score\n",
      "CPU times: user 1h 7min 5s, sys: 4.44 s, total: 1h 7min 9s\n",
      "Wall time: 16min 13s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "n_fold =5\n",
    "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**n_fold),\n",
    "        'bagging_seed':int(2**n_fold),\n",
    "        'drop_seed':int(2**n_fold)\n",
    "        }\n",
    "folds = StratifiedKFold(n_splits=n_fold, shuffle=True, random_state=4590)\n",
    "oof = np.zeros(len(train_df2))\n",
    "predictions = np.zeros(len(test_df2))\n",
    "feature_importance_df = pd.DataFrame()\n",
    "\n",
    "for fold_, (trn_idx, val_idx) in enumerate(folds.split(train_df2,train_df2['outliers'].values)):\n",
    "    print(\"fold {}\".format(fold_))\n",
    "    trn_data = lgb.Dataset(train_df2.iloc[trn_idx][df_train_columns], label=target.iloc[trn_idx])\n",
    "    val_data = lgb.Dataset(train_df2.iloc[val_idx][df_train_columns], label=target.iloc[val_idx])\n",
    "\n",
    "    num_round = 10000\n",
    "    clf = lgb.train(param, trn_data, num_round, valid_sets = [trn_data, val_data], verbose_eval=100, early_stopping_rounds = 200)\n",
    "    oof[val_idx] = clf.predict(train_df2.iloc[val_idx][df_train_columns], num_iteration=clf.best_iteration)\n",
    "    \n",
    "    fold_importance_df = pd.DataFrame()\n",
    "    fold_importance_df[\"Feature\"] = df_train_columns\n",
    "    fold_importance_df[\"importance\"] = clf.feature_importance()\n",
    "    fold_importance_df[\"fold\"] = fold_ + 1\n",
    "    feature_importance_df = pd.concat([feature_importance_df, fold_importance_df], axis=0)\n",
    "    \n",
    "    predictions += clf.predict(test_df2[df_train_columns], num_iteration=clf.best_iteration) / folds.n_splits\n",
    "\n",
    "print(np.sqrt(mean_squared_error(oof, target)), 'CV score')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Base feats - f1,f2,f3 - 3.8421399267932315\n",
    "2. 3.6552368169417204, 3.692\n",
    "3. 3.654460362787254, 3.692\n",
    "4. 3.6538177315106624, 3.691\n",
    "5. 3.6552638224814196- after trimming purchase_amount\n",
    "6. 3.6544381493750677 - base_cols 180\n",
    "7. 3.653984525732515 - added purchase_amount_pct_25_old\n",
    "8. 3.653161955069063 - base_cols - 180 - 16th feb - 181 has first_active_monthElapsed score 3.653410894233743\n",
    "9. 3.6535590767314026 - added 'avg_trans_per_merchant_old','purchase_amount_sum_mean', 'category_2_5.0_purchase_amount_mean'\n",
    "10. 3.6573988258459713 - with 38 cols\n",
    "11. 3.655024925251293 - with 241 cols\n",
    "12. 3.6527864665829197 - with 283 cols - added missed out agg in train, test & added observed date, lb - 3.816\n",
    "13. 3.6535891905746243 - with 181 cols - base + observed_elapsed time, **3.691** better than 3.693 of 180 base cols\n",
    "14. 3.653084106006047 - 414 cols \n",
    "15. 3.648418646163904 - rescaled purchase_amount  - 3.690 on LB\n",
    "16. 3.645283649529094 - adjusted purchase_date_uptomin, purchase_date_uptonow to observed time - 3.690 on LB\n",
    "17. 3.641896345010267 - 342 cols, LB - 3.717 - maybe blend later (not working)\n",
    "18. 3.654142065931516 - 53 cols\n",
    "19. 3.644520662740318  - with 'merchant_address_id_nunique', 'merchant_rating_nunique' + 180 cols\n",
    "20. 3.644042133990217 -  with 'merchant_address_id_nunique', 'merchant_rating_nunique' + 180 cols - Black_Friday_2017_mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "cols = (feature_importance_df[[\"Feature\", \"importance\"]]\n",
    "        .groupby(\"Feature\")\n",
    "        .mean()\n",
    "        .sort_values(by=\"importance\", ascending=False)[:1000].index)\n",
    "\n",
    "best_features = feature_importance_df.loc[feature_importance_df.Feature.isin(cols)]\n",
    "\n",
    "plt.figure(figsize=(14,25))\n",
    "sns.barplot(x=\"importance\",\n",
    "            y=\"Feature\",\n",
    "            data=best_features.sort_values(by=\"importance\",\n",
    "                                           ascending=False))\n",
    "plt.title('LightGBM Features (avg over folds)')\n",
    "plt.tight_layout()\n",
    "plt.savefig('lgbm_importances.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_ft_df = best_features.groupby('Feature').agg({'importance': ['mean']})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_ft_df.columns = ['_'.join(c).strip() for c in best_ft_df.columns.values]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_ft_df = best_ft_df.reset_index().sort_values('importance_mean', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "fts_sorted_by_imp = best_ft_df.sort_values('importance_mean', ascending=False)['Feature'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_rows', 500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "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>index</th>\n",
       "      <th>Feature</th>\n",
       "      <th>importance_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>330</td>\n",
       "      <td>purchase_date_uptonow</td>\n",
       "      <td>1183.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>205</td>\n",
       "      <td>month_diff_mean_old</td>\n",
       "      <td>1077.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>310</td>\n",
       "      <td>purchase_amount_sum_ratio</td>\n",
       "      <td>927.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>337</td>\n",
       "      <td>repurchase_merchant_rate_old</td>\n",
       "      <td>901.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>266</td>\n",
       "      <td>purchase_Month_nunique</td>\n",
       "      <td>804.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>269</td>\n",
       "      <td>purchase_Week_mean</td>\n",
       "      <td>804.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>32</td>\n",
       "      <td>authorized_flag_mean</td>\n",
       "      <td>736.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>262</td>\n",
       "      <td>purchase_Month_mean</td>\n",
       "      <td>629.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>50</td>\n",
       "      <td>category_1_0_purchase_amount_sum_ratio</td>\n",
       "      <td>576.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2</td>\n",
       "      <td>CLV</td>\n",
       "      <td>552.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>267</td>\n",
       "      <td>purchase_Week_max</td>\n",
       "      <td>550.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>275</td>\n",
       "      <td>purchase_amount_before_after_ratio</td>\n",
       "      <td>543.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>217</td>\n",
       "      <td>month_lag_mean_old</td>\n",
       "      <td>531.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>325</td>\n",
       "      <td>purchase_date_diff_old</td>\n",
       "      <td>531.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>154</td>\n",
       "      <td>fathers_day_2017_mean</td>\n",
       "      <td>511.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>27</td>\n",
       "      <td>authorized_flag_0_purchase_amount_sum</td>\n",
       "      <td>503.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>206</td>\n",
       "      <td>month_diff_mean_total</td>\n",
       "      <td>498.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>61</td>\n",
       "      <td>category_1_mean_old</td>\n",
       "      <td>474.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>4</td>\n",
       "      <td>CLV_ratio</td>\n",
       "      <td>469.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>52</td>\n",
       "      <td>category_1_1_purchase_amount_mean</td>\n",
       "      <td>452.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>10</td>\n",
       "      <td>Mothers_Day_2018_mean</td>\n",
       "      <td>451.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>62</td>\n",
       "      <td>category_1_mean_total</td>\n",
       "      <td>443.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>225</td>\n",
       "      <td>month_lag_var_old</td>\n",
       "      <td>427.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>291</td>\n",
       "      <td>purchase_amount_mean_ratio</td>\n",
       "      <td>427.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>6</td>\n",
       "      <td>Children_day_2017_mean_old</td>\n",
       "      <td>427.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>271</td>\n",
       "      <td>purchase_Week_min</td>\n",
       "      <td>423.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>12</td>\n",
       "      <td>Valentine_Day_2017_mean</td>\n",
       "      <td>417.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>331</td>\n",
       "      <td>purchase_date_uptonow_old</td>\n",
       "      <td>413.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>112</td>\n",
       "      <td>category_3_0_purchase_amount_sum_old</td>\n",
       "      <td>405.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>218</td>\n",
       "      <td>month_lag_mean_total</td>\n",
       "      <td>402.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>163</td>\n",
       "      <td>hist_first_buy</td>\n",
       "      <td>390.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>9</td>\n",
       "      <td>Mothers_Day_2017_mean</td>\n",
       "      <td>389.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>236</td>\n",
       "      <td>price_skew</td>\n",
       "      <td>373.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>216</td>\n",
       "      <td>month_lag_mean</td>\n",
       "      <td>372.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>242</td>\n",
       "      <td>purchase_Day_mean</td>\n",
       "      <td>366.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>203</td>\n",
       "      <td>month_diff_max</td>\n",
       "      <td>364.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>164</td>\n",
       "      <td>hist_last_buy</td>\n",
       "      <td>361.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>56</td>\n",
       "      <td>category_1_1_purchase_amount_sum</td>\n",
       "      <td>358.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>26</td>\n",
       "      <td>authorized_flag_0_purchase_amount_mean</td>\n",
       "      <td>357.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>46</td>\n",
       "      <td>category_1_0_purchase_amount_mean_ratio</td>\n",
       "      <td>353.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>328</td>\n",
       "      <td>purchase_date_uptomin</td>\n",
       "      <td>350.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>38</td>\n",
       "      <td>avg_trans_per_merchant_old</td>\n",
       "      <td>343.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>7</td>\n",
       "      <td>Christmas_Day_2017_mean</td>\n",
       "      <td>326.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>255</td>\n",
       "      <td>purchase_Hour_mean_old</td>\n",
       "      <td>322.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>69</td>\n",
       "      <td>category_2_1.0_purchase_amount_sum_ratio</td>\n",
       "      <td>320.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>187</td>\n",
       "      <td>installments_sum_old</td>\n",
       "      <td>316.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>137</td>\n",
       "      <td>days_feature2</td>\n",
       "      <td>314.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>274</td>\n",
       "      <td>purchase_Week_nunique_old</td>\n",
       "      <td>309.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>250</td>\n",
       "      <td>purchase_Dayofweek_mean_old</td>\n",
       "      <td>308.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>119</td>\n",
       "      <td>category_3_1_purchase_amount_sum</td>\n",
       "      <td>300.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>123</td>\n",
       "      <td>category_3_2_purchase_amount_mean</td>\n",
       "      <td>297.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>254</td>\n",
       "      <td>purchase_Hour_mean</td>\n",
       "      <td>289.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>204</td>\n",
       "      <td>month_diff_mean</td>\n",
       "      <td>285.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>223</td>\n",
       "      <td>month_lag_skew_old</td>\n",
       "      <td>285.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>127</td>\n",
       "      <td>category_3_2_purchase_amount_sum</td>\n",
       "      <td>283.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>210</td>\n",
       "      <td>month_diff_skew_old</td>\n",
       "      <td>282.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>0</td>\n",
       "      <td>Black_Friday_2017_mean</td>\n",
       "      <td>281.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>324</td>\n",
       "      <td>purchase_date_diff</td>\n",
       "      <td>280.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>11</td>\n",
       "      <td>Mothers_Day_2018_mean_old</td>\n",
       "      <td>280.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>48</td>\n",
       "      <td>category_1_0_purchase_amount_sum</td>\n",
       "      <td>270.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>334</td>\n",
       "      <td>purchased_on_weekend_mean_old</td>\n",
       "      <td>266.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>115</td>\n",
       "      <td>category_3_1_purchase_amount_mean</td>\n",
       "      <td>264.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>304</td>\n",
       "      <td>purchase_amount_pct_75_ratio</td>\n",
       "      <td>262.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>49</td>\n",
       "      <td>category_1_0_purchase_amount_sum_old</td>\n",
       "      <td>256.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>60</td>\n",
       "      <td>category_1_mean</td>\n",
       "      <td>253.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>243</td>\n",
       "      <td>purchase_Day_mean_old</td>\n",
       "      <td>251.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>139</td>\n",
       "      <td>days_feature3</td>\n",
       "      <td>251.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>329</td>\n",
       "      <td>purchase_date_uptomin_old</td>\n",
       "      <td>250.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>135</td>\n",
       "      <td>days_feature1</td>\n",
       "      <td>247.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>322</td>\n",
       "      <td>purchase_date_average</td>\n",
       "      <td>245.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>200</td>\n",
       "      <td>merchant_category_repurchase_old</td>\n",
       "      <td>245.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>323</td>\n",
       "      <td>purchase_date_average_old</td>\n",
       "      <td>241.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>41</td>\n",
       "      <td>card_id_count_ratio</td>\n",
       "      <td>238.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>152</td>\n",
       "      <td>duration_var</td>\n",
       "      <td>237.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>326</td>\n",
       "      <td>purchase_date_successive_diff_mean</td>\n",
       "      <td>237.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>327</td>\n",
       "      <td>purchase_date_successive_diff_mean_old</td>\n",
       "      <td>236.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>202</td>\n",
       "      <td>merchant_id_nunique_old</td>\n",
       "      <td>235.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>308</td>\n",
       "      <td>purchase_amount_sum</td>\n",
       "      <td>235.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>59</td>\n",
       "      <td>category_1_1_purchase_amount_sum_total</td>\n",
       "      <td>233.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>67</td>\n",
       "      <td>category_2_1.0_purchase_amount_sum</td>\n",
       "      <td>232.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>121</td>\n",
       "      <td>category_3_1_purchase_amount_sum_ratio</td>\n",
       "      <td>231.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>113</td>\n",
       "      <td>category_3_0_purchase_amount_sum_ratio</td>\n",
       "      <td>229.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>207</td>\n",
       "      <td>month_diff_min</td>\n",
       "      <td>227.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>224</td>\n",
       "      <td>month_lag_var</td>\n",
       "      <td>226.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>295</td>\n",
       "      <td>purchase_amount_median_ratio</td>\n",
       "      <td>226.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>63</td>\n",
       "      <td>category_2_1.0_purchase_amount_mean</td>\n",
       "      <td>224.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>5</td>\n",
       "      <td>Children_day_2017_mean</td>\n",
       "      <td>222.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>65</td>\n",
       "      <td>category_2_1.0_purchase_amount_mean_ratio</td>\n",
       "      <td>218.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>276</td>\n",
       "      <td>purchase_amount_fi_last_new_total</td>\n",
       "      <td>218.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>240</td>\n",
       "      <td>price_var_old</td>\n",
       "      <td>218.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>68</td>\n",
       "      <td>category_2_1.0_purchase_amount_sum_old</td>\n",
       "      <td>218.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>20</td>\n",
       "      <td>amount_month_ratio_min_old</td>\n",
       "      <td>215.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>136</td>\n",
       "      <td>days_feature1_ratio</td>\n",
       "      <td>213.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>107</td>\n",
       "      <td>category_3_0_purchase_amount_mean</td>\n",
       "      <td>213.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>272</td>\n",
       "      <td>purchase_Week_min_old</td>\n",
       "      <td>211.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>256</td>\n",
       "      <td>purchase_Hour_min</td>\n",
       "      <td>203.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>1</td>\n",
       "      <td>Black_Friday_2017_mean_old</td>\n",
       "      <td>202.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>268</td>\n",
       "      <td>purchase_Week_max_old</td>\n",
       "      <td>198.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>208</td>\n",
       "      <td>month_diff_ratio</td>\n",
       "      <td>197.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>306</td>\n",
       "      <td>purchase_amount_skew_old</td>\n",
       "      <td>196.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>30</td>\n",
       "      <td>authorized_flag_1_purchase_amount_sum</td>\n",
       "      <td>193.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>8</td>\n",
       "      <td>Christmas_Day_2017_mean_old</td>\n",
       "      <td>193.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>111</td>\n",
       "      <td>category_3_0_purchase_amount_sum</td>\n",
       "      <td>193.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>241</td>\n",
       "      <td>purchase_Day_max</td>\n",
       "      <td>190.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>138</td>\n",
       "      <td>days_feature2_ratio</td>\n",
       "      <td>189.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>315</td>\n",
       "      <td>purchase_amount_tail_sum_old</td>\n",
       "      <td>188.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>226</td>\n",
       "      <td>new_first_buy</td>\n",
       "      <td>186.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>260</td>\n",
       "      <td>purchase_Month_max</td>\n",
       "      <td>186.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>34</td>\n",
       "      <td>avg_spend_per_merchant_old</td>\n",
       "      <td>185.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>313</td>\n",
       "      <td>purchase_amount_tail_max_old</td>\n",
       "      <td>184.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>278</td>\n",
       "      <td>purchase_amount_first_last_diff</td>\n",
       "      <td>183.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>263</td>\n",
       "      <td>purchase_Month_mean_old</td>\n",
       "      <td>180.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>270</td>\n",
       "      <td>purchase_Week_mean_old</td>\n",
       "      <td>180.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>320</td>\n",
       "      <td>purchase_amount_var</td>\n",
       "      <td>180.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>285</td>\n",
       "      <td>purchase_amount_inter_quartile_ratio</td>\n",
       "      <td>179.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>151</td>\n",
       "      <td>duration_skew_old</td>\n",
       "      <td>178.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>341</td>\n",
       "      <td>transactions_count_old</td>\n",
       "      <td>178.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>148</td>\n",
       "      <td>duration_min_old</td>\n",
       "      <td>177.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>54</td>\n",
       "      <td>category_1_1_purchase_amount_mean_ratio</td>\n",
       "      <td>175.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>300</td>\n",
       "      <td>purchase_amount_pct_25_old</td>\n",
       "      <td>175.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>118</td>\n",
       "      <td>category_3_1_purchase_amount_mean_total</td>\n",
       "      <td>175.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>141</td>\n",
       "      <td>duration_max</td>\n",
       "      <td>174.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>185</td>\n",
       "      <td>installments_skew_old</td>\n",
       "      <td>173.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>244</td>\n",
       "      <td>purchase_Day_min</td>\n",
       "      <td>173.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>13</td>\n",
       "      <td>amount_month_ratio_max</td>\n",
       "      <td>171.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>3</td>\n",
       "      <td>CLV_old</td>\n",
       "      <td>171.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>282</td>\n",
       "      <td>purchase_amount_head_sum_old</td>\n",
       "      <td>169.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>23</td>\n",
       "      <td>amount_month_ratio_skew_old</td>\n",
       "      <td>169.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>194</td>\n",
       "      <td>installments_var_old</td>\n",
       "      <td>169.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>183</td>\n",
       "      <td>installments_ratio</td>\n",
       "      <td>167.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>212</td>\n",
       "      <td>month_diff_var_old</td>\n",
       "      <td>165.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>57</td>\n",
       "      <td>category_1_1_purchase_amount_sum_old</td>\n",
       "      <td>163.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>58</td>\n",
       "      <td>category_1_1_purchase_amount_sum_ratio</td>\n",
       "      <td>163.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>294</td>\n",
       "      <td>purchase_amount_median_old</td>\n",
       "      <td>161.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>120</td>\n",
       "      <td>category_3_1_purchase_amount_sum_old</td>\n",
       "      <td>161.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>109</td>\n",
       "      <td>category_3_0_purchase_amount_mean_ratio</td>\n",
       "      <td>161.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>192</td>\n",
       "      <td>installments_total</td>\n",
       "      <td>159.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>155</td>\n",
       "      <td>feature_1</td>\n",
       "      <td>159.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>222</td>\n",
       "      <td>month_lag_skew</td>\n",
       "      <td>158.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>227</td>\n",
       "      <td>new_last_buy</td>\n",
       "      <td>158.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>198</td>\n",
       "      <td>merchant_category_id_nunique_old</td>\n",
       "      <td>158.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>280</td>\n",
       "      <td>purchase_amount_head_max_old</td>\n",
       "      <td>157.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>229</td>\n",
       "      <td>observed_elapsed_time</td>\n",
       "      <td>157.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>53</td>\n",
       "      <td>category_1_1_purchase_amount_mean_old</td>\n",
       "      <td>157.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>24</td>\n",
       "      <td>amount_month_ratio_var</td>\n",
       "      <td>156.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>301</td>\n",
       "      <td>purchase_amount_pct_25_ratio</td>\n",
       "      <td>155.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>44</td>\n",
       "      <td>category_1_0_purchase_amount_mean</td>\n",
       "      <td>154.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>117</td>\n",
       "      <td>category_3_1_purchase_amount_mean_ratio</td>\n",
       "      <td>153.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>64</td>\n",
       "      <td>category_2_1.0_purchase_amount_mean_old</td>\n",
       "      <td>153.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>55</td>\n",
       "      <td>category_1_1_purchase_amount_mean_total</td>\n",
       "      <td>151.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>116</td>\n",
       "      <td>category_3_1_purchase_amount_mean_old</td>\n",
       "      <td>147.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>249</td>\n",
       "      <td>purchase_Dayofweek_mean</td>\n",
       "      <td>147.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>99</td>\n",
       "      <td>category_2_5.0_purchase_amount_sum</td>\n",
       "      <td>146.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>51</td>\n",
       "      <td>category_1_0_purchase_amount_sum_total</td>\n",
       "      <td>146.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>108</td>\n",
       "      <td>category_3_0_purchase_amount_mean_old</td>\n",
       "      <td>144.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>234</td>\n",
       "      <td>price_min</td>\n",
       "      <td>142.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>173</td>\n",
       "      <td>installments_mean_old</td>\n",
       "      <td>141.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>305</td>\n",
       "      <td>purchase_amount_skew</td>\n",
       "      <td>141.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>134</td>\n",
       "      <td>category_3_mean_old</td>\n",
       "      <td>141.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>307</td>\n",
       "      <td>purchase_amount_skew_ratio</td>\n",
       "      <td>140.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>235</td>\n",
       "      <td>price_min_old</td>\n",
       "      <td>139.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>129</td>\n",
       "      <td>category_3_2_purchase_amount_sum_ratio</td>\n",
       "      <td>137.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>45</td>\n",
       "      <td>category_1_0_purchase_amount_mean_old</td>\n",
       "      <td>137.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163</th>\n",
       "      <td>122</td>\n",
       "      <td>category_3_1_purchase_amount_sum_total</td>\n",
       "      <td>133.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>164</th>\n",
       "      <td>22</td>\n",
       "      <td>amount_month_ratio_skew</td>\n",
       "      <td>131.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>286</td>\n",
       "      <td>purchase_amount_max</td>\n",
       "      <td>131.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>70</td>\n",
       "      <td>category_2_1.0_purchase_amount_sum_total</td>\n",
       "      <td>131.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167</th>\n",
       "      <td>133</td>\n",
       "      <td>category_3_mean_mean_old</td>\n",
       "      <td>130.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>238</td>\n",
       "      <td>price_total</td>\n",
       "      <td>127.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>182</td>\n",
       "      <td>installments_pct_75_old</td>\n",
       "      <td>127.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>19</td>\n",
       "      <td>amount_month_ratio_min</td>\n",
       "      <td>126.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>259</td>\n",
       "      <td>purchase_Hour_nunique_old</td>\n",
       "      <td>125.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>66</td>\n",
       "      <td>category_2_1.0_purchase_amount_mean_total</td>\n",
       "      <td>124.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>174</td>\n",
       "      <td>installments_mean_total</td>\n",
       "      <td>124.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>284</td>\n",
       "      <td>purchase_amount_inter_quartile_old</td>\n",
       "      <td>123.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>47</td>\n",
       "      <td>category_1_0_purchase_amount_mean_total</td>\n",
       "      <td>122.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>297</td>\n",
       "      <td>purchase_amount_min_old</td>\n",
       "      <td>121.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>43</td>\n",
       "      <td>card_id_total</td>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>339</td>\n",
       "      <td>subsector_id_nunique_old</td>\n",
       "      <td>119.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>232</td>\n",
       "      <td>price_mean</td>\n",
       "      <td>118.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>180</th>\n",
       "      <td>147</td>\n",
       "      <td>duration_min</td>\n",
       "      <td>118.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>181</th>\n",
       "      <td>140</td>\n",
       "      <td>days_feature3_ratio</td>\n",
       "      <td>117.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>231</td>\n",
       "      <td>price_max_old</td>\n",
       "      <td>116.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183</th>\n",
       "      <td>150</td>\n",
       "      <td>duration_skew</td>\n",
       "      <td>115.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>128</td>\n",
       "      <td>category_3_2_purchase_amount_sum_old</td>\n",
       "      <td>115.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>230</td>\n",
       "      <td>price_max</td>\n",
       "      <td>112.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>106</td>\n",
       "      <td>category_2_mean_old</td>\n",
       "      <td>111.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>187</th>\n",
       "      <td>114</td>\n",
       "      <td>category_3_0_purchase_amount_sum_total</td>\n",
       "      <td>111.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>124</td>\n",
       "      <td>category_3_2_purchase_amount_mean_old</td>\n",
       "      <td>111.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>110</td>\n",
       "      <td>category_3_0_purchase_amount_mean_total</td>\n",
       "      <td>110.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>190</th>\n",
       "      <td>168</td>\n",
       "      <td>installments_head_sum_old</td>\n",
       "      <td>109.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>191</th>\n",
       "      <td>149</td>\n",
       "      <td>duration_min_total</td>\n",
       "      <td>107.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>196</td>\n",
       "      <td>inverse_avg_transactions_per_day_old</td>\n",
       "      <td>106.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>193</th>\n",
       "      <td>211</td>\n",
       "      <td>month_diff_var</td>\n",
       "      <td>105.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>194</th>\n",
       "      <td>195</td>\n",
       "      <td>inverse_avg_transactions_per_day</td>\n",
       "      <td>104.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>296</td>\n",
       "      <td>purchase_amount_min</td>\n",
       "      <td>103.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>21</td>\n",
       "      <td>amount_month_ratio_min_total</td>\n",
       "      <td>102.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>105</td>\n",
       "      <td>category_2_mean_mean_old</td>\n",
       "      <td>101.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>233</td>\n",
       "      <td>price_mean_old</td>\n",
       "      <td>100.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>28</td>\n",
       "      <td>authorized_flag_1_purchase_amount_mean</td>\n",
       "      <td>99.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>283</td>\n",
       "      <td>purchase_amount_inter_quartile</td>\n",
       "      <td>98.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>191</td>\n",
       "      <td>installments_tail_sum_old</td>\n",
       "      <td>98.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>202</th>\n",
       "      <td>96</td>\n",
       "      <td>category_2_5.0_purchase_amount_mean_old</td>\n",
       "      <td>98.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>203</th>\n",
       "      <td>95</td>\n",
       "      <td>category_2_5.0_purchase_amount_mean</td>\n",
       "      <td>96.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204</th>\n",
       "      <td>197</td>\n",
       "      <td>merchant_category_id_nunique</td>\n",
       "      <td>96.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205</th>\n",
       "      <td>319</td>\n",
       "      <td>purchase_amount_total_min</td>\n",
       "      <td>96.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>299</td>\n",
       "      <td>purchase_amount_pct_25</td>\n",
       "      <td>95.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>207</th>\n",
       "      <td>130</td>\n",
       "      <td>category_3_2_purchase_amount_sum_total</td>\n",
       "      <td>95.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>126</td>\n",
       "      <td>category_3_2_purchase_amount_mean_total</td>\n",
       "      <td>95.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>209</th>\n",
       "      <td>31</td>\n",
       "      <td>authorized_flag_1_purchase_amount_sum_old</td>\n",
       "      <td>95.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>193</td>\n",
       "      <td>installments_var</td>\n",
       "      <td>94.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>247</td>\n",
       "      <td>purchase_Day_nunique_old</td>\n",
       "      <td>93.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212</th>\n",
       "      <td>125</td>\n",
       "      <td>category_3_2_purchase_amount_mean_ratio</td>\n",
       "      <td>93.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>239</td>\n",
       "      <td>price_var</td>\n",
       "      <td>92.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>214</th>\n",
       "      <td>303</td>\n",
       "      <td>purchase_amount_pct_75_old</td>\n",
       "      <td>91.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>215</th>\n",
       "      <td>79</td>\n",
       "      <td>category_2_3.0_purchase_amount_mean</td>\n",
       "      <td>89.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>216</th>\n",
       "      <td>101</td>\n",
       "      <td>category_2_5.0_purchase_amount_sum_ratio</td>\n",
       "      <td>88.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>220</td>\n",
       "      <td>month_lag_min_old</td>\n",
       "      <td>88.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>218</th>\n",
       "      <td>14</td>\n",
       "      <td>amount_month_ratio_max_old</td>\n",
       "      <td>86.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>219</th>\n",
       "      <td>264</td>\n",
       "      <td>purchase_Month_min</td>\n",
       "      <td>85.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220</th>\n",
       "      <td>309</td>\n",
       "      <td>purchase_amount_sum_old</td>\n",
       "      <td>83.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>221</th>\n",
       "      <td>237</td>\n",
       "      <td>price_sum</td>\n",
       "      <td>83.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>222</th>\n",
       "      <td>142</td>\n",
       "      <td>duration_max_old</td>\n",
       "      <td>83.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223</th>\n",
       "      <td>17</td>\n",
       "      <td>amount_month_ratio_mean_old</td>\n",
       "      <td>79.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>83</td>\n",
       "      <td>category_2_3.0_purchase_amount_sum</td>\n",
       "      <td>78.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>225</th>\n",
       "      <td>25</td>\n",
       "      <td>amount_month_ratio_var_old</td>\n",
       "      <td>78.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226</th>\n",
       "      <td>80</td>\n",
       "      <td>category_2_3.0_purchase_amount_mean_old</td>\n",
       "      <td>76.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>227</th>\n",
       "      <td>281</td>\n",
       "      <td>purchase_amount_head_sum</td>\n",
       "      <td>76.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>228</th>\n",
       "      <td>221</td>\n",
       "      <td>month_lag_min_total</td>\n",
       "      <td>76.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>229</th>\n",
       "      <td>16</td>\n",
       "      <td>amount_month_ratio_mean</td>\n",
       "      <td>76.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>230</th>\n",
       "      <td>214</td>\n",
       "      <td>month_lag_max_old</td>\n",
       "      <td>76.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>231</th>\n",
       "      <td>209</td>\n",
       "      <td>month_diff_skew</td>\n",
       "      <td>75.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>85</td>\n",
       "      <td>category_2_3.0_purchase_amount_sum_ratio</td>\n",
       "      <td>74.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>153</td>\n",
       "      <td>duration_var_old</td>\n",
       "      <td>74.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>234</th>\n",
       "      <td>257</td>\n",
       "      <td>purchase_Hour_min_old</td>\n",
       "      <td>73.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>235</th>\n",
       "      <td>279</td>\n",
       "      <td>purchase_amount_head_max</td>\n",
       "      <td>73.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>236</th>\n",
       "      <td>144</td>\n",
       "      <td>duration_mean</td>\n",
       "      <td>73.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>132</td>\n",
       "      <td>category_3_mean_mean</td>\n",
       "      <td>72.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>71</td>\n",
       "      <td>category_2_2.0_purchase_amount_mean</td>\n",
       "      <td>72.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>172</td>\n",
       "      <td>installments_mean</td>\n",
       "      <td>71.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>293</td>\n",
       "      <td>purchase_amount_median</td>\n",
       "      <td>70.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>81</td>\n",
       "      <td>category_2_3.0_purchase_amount_mean_ratio</td>\n",
       "      <td>70.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>314</td>\n",
       "      <td>purchase_amount_tail_sum</td>\n",
       "      <td>70.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>243</th>\n",
       "      <td>143</td>\n",
       "      <td>duration_max_total</td>\n",
       "      <td>69.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>244</th>\n",
       "      <td>312</td>\n",
       "      <td>purchase_amount_tail_max</td>\n",
       "      <td>69.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>245</th>\n",
       "      <td>261</td>\n",
       "      <td>purchase_Month_max_old</td>\n",
       "      <td>68.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>287</td>\n",
       "      <td>purchase_amount_max_old</td>\n",
       "      <td>68.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>100</td>\n",
       "      <td>category_2_5.0_purchase_amount_sum_old</td>\n",
       "      <td>67.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248</th>\n",
       "      <td>40</td>\n",
       "      <td>card_id_count_old</td>\n",
       "      <td>67.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>156</td>\n",
       "      <td>feature_2</td>\n",
       "      <td>66.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>18</td>\n",
       "      <td>amount_month_ratio_mean_total</td>\n",
       "      <td>66.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>91</td>\n",
       "      <td>category_2_4.0_purchase_amount_sum</td>\n",
       "      <td>65.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>15</td>\n",
       "      <td>amount_month_ratio_max_total</td>\n",
       "      <td>65.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>253</th>\n",
       "      <td>302</td>\n",
       "      <td>purchase_amount_pct_75</td>\n",
       "      <td>63.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>254</th>\n",
       "      <td>215</td>\n",
       "      <td>month_lag_max_total</td>\n",
       "      <td>62.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>255</th>\n",
       "      <td>162</td>\n",
       "      <td>feature_var</td>\n",
       "      <td>61.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>145</td>\n",
       "      <td>duration_mean_old</td>\n",
       "      <td>61.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>257</th>\n",
       "      <td>75</td>\n",
       "      <td>category_2_2.0_purchase_amount_sum</td>\n",
       "      <td>61.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>258</th>\n",
       "      <td>93</td>\n",
       "      <td>category_2_4.0_purchase_amount_sum_ratio</td>\n",
       "      <td>60.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>259</th>\n",
       "      <td>316</td>\n",
       "      <td>purchase_amount_total</td>\n",
       "      <td>60.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>260</th>\n",
       "      <td>321</td>\n",
       "      <td>purchase_amount_var_old</td>\n",
       "      <td>58.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>261</th>\n",
       "      <td>317</td>\n",
       "      <td>purchase_amount_total_max</td>\n",
       "      <td>58.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>262</th>\n",
       "      <td>87</td>\n",
       "      <td>category_2_4.0_purchase_amount_mean</td>\n",
       "      <td>58.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>263</th>\n",
       "      <td>97</td>\n",
       "      <td>category_2_5.0_purchase_amount_mean_ratio</td>\n",
       "      <td>57.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264</th>\n",
       "      <td>252</td>\n",
       "      <td>purchase_Hour_max</td>\n",
       "      <td>56.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>265</th>\n",
       "      <td>251</td>\n",
       "      <td>purchase_Dayofweek_min</td>\n",
       "      <td>55.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>266</th>\n",
       "      <td>333</td>\n",
       "      <td>purchased_on_weekend_mean</td>\n",
       "      <td>54.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>267</th>\n",
       "      <td>338</td>\n",
       "      <td>subsector_id_nunique</td>\n",
       "      <td>53.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>184</td>\n",
       "      <td>installments_skew</td>\n",
       "      <td>53.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>161</td>\n",
       "      <td>feature_sum</td>\n",
       "      <td>53.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>270</th>\n",
       "      <td>277</td>\n",
       "      <td>purchase_amount_fi_last_old_total</td>\n",
       "      <td>52.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>271</th>\n",
       "      <td>265</td>\n",
       "      <td>purchase_Month_min_old</td>\n",
       "      <td>52.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>272</th>\n",
       "      <td>289</td>\n",
       "      <td>purchase_amount_mean</td>\n",
       "      <td>50.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>273</th>\n",
       "      <td>35</td>\n",
       "      <td>avg_spend_per_transaction</td>\n",
       "      <td>50.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>274</th>\n",
       "      <td>181</td>\n",
       "      <td>installments_pct_75</td>\n",
       "      <td>48.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>275</th>\n",
       "      <td>290</td>\n",
       "      <td>purchase_amount_mean_old</td>\n",
       "      <td>48.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>276</th>\n",
       "      <td>201</td>\n",
       "      <td>merchant_id_nunique</td>\n",
       "      <td>46.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>277</th>\n",
       "      <td>146</td>\n",
       "      <td>duration_mean_total</td>\n",
       "      <td>46.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>278</th>\n",
       "      <td>340</td>\n",
       "      <td>transactions_count</td>\n",
       "      <td>45.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>318</td>\n",
       "      <td>purchase_amount_total_mean</td>\n",
       "      <td>44.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>280</th>\n",
       "      <td>171</td>\n",
       "      <td>installments_max_total</td>\n",
       "      <td>44.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>281</th>\n",
       "      <td>98</td>\n",
       "      <td>category_2_5.0_purchase_amount_mean_total</td>\n",
       "      <td>43.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>170</td>\n",
       "      <td>installments_max_old</td>\n",
       "      <td>41.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>283</th>\n",
       "      <td>42</td>\n",
       "      <td>card_id_count_total</td>\n",
       "      <td>41.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>284</th>\n",
       "      <td>102</td>\n",
       "      <td>category_2_5.0_purchase_amount_sum_total</td>\n",
       "      <td>41.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>285</th>\n",
       "      <td>86</td>\n",
       "      <td>category_2_3.0_purchase_amount_sum_total</td>\n",
       "      <td>40.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>286</th>\n",
       "      <td>82</td>\n",
       "      <td>category_2_3.0_purchase_amount_mean_total</td>\n",
       "      <td>38.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>287</th>\n",
       "      <td>190</td>\n",
       "      <td>installments_tail_sum</td>\n",
       "      <td>38.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>288</th>\n",
       "      <td>76</td>\n",
       "      <td>category_2_2.0_purchase_amount_sum_old</td>\n",
       "      <td>38.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289</th>\n",
       "      <td>335</td>\n",
       "      <td>quarter</td>\n",
       "      <td>37.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>290</th>\n",
       "      <td>186</td>\n",
       "      <td>installments_sum</td>\n",
       "      <td>37.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291</th>\n",
       "      <td>273</td>\n",
       "      <td>purchase_Week_nunique</td>\n",
       "      <td>37.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>292</th>\n",
       "      <td>298</td>\n",
       "      <td>purchase_amount_min_total</td>\n",
       "      <td>36.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>293</th>\n",
       "      <td>104</td>\n",
       "      <td>category_2_mean_mean</td>\n",
       "      <td>34.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>294</th>\n",
       "      <td>89</td>\n",
       "      <td>category_2_4.0_purchase_amount_mean_ratio</td>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>213</td>\n",
       "      <td>month_lag_max</td>\n",
       "      <td>33.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>245</td>\n",
       "      <td>purchase_Day_min_old</td>\n",
       "      <td>32.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>297</th>\n",
       "      <td>33</td>\n",
       "      <td>avg_spend_per_merchant</td>\n",
       "      <td>32.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>311</td>\n",
       "      <td>purchase_amount_sum_total</td>\n",
       "      <td>31.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>84</td>\n",
       "      <td>category_2_3.0_purchase_amount_sum_old</td>\n",
       "      <td>31.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300</th>\n",
       "      <td>189</td>\n",
       "      <td>installments_tail_max_old</td>\n",
       "      <td>31.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>301</th>\n",
       "      <td>175</td>\n",
       "      <td>installments_median</td>\n",
       "      <td>29.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>258</td>\n",
       "      <td>purchase_Hour_nunique</td>\n",
       "      <td>28.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>303</th>\n",
       "      <td>77</td>\n",
       "      <td>category_2_2.0_purchase_amount_sum_ratio</td>\n",
       "      <td>26.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>304</th>\n",
       "      <td>74</td>\n",
       "      <td>category_2_2.0_purchase_amount_mean_total</td>\n",
       "      <td>26.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305</th>\n",
       "      <td>179</td>\n",
       "      <td>installments_pct_25</td>\n",
       "      <td>26.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>288</td>\n",
       "      <td>purchase_amount_max_total</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>307</th>\n",
       "      <td>246</td>\n",
       "      <td>purchase_Day_nunique</td>\n",
       "      <td>25.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>308</th>\n",
       "      <td>72</td>\n",
       "      <td>category_2_2.0_purchase_amount_mean_old</td>\n",
       "      <td>25.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>309</th>\n",
       "      <td>159</td>\n",
       "      <td>feature_mean</td>\n",
       "      <td>24.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>310</th>\n",
       "      <td>90</td>\n",
       "      <td>category_2_4.0_purchase_amount_mean_total</td>\n",
       "      <td>24.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>94</td>\n",
       "      <td>category_2_4.0_purchase_amount_sum_total</td>\n",
       "      <td>23.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>177</td>\n",
       "      <td>installments_min</td>\n",
       "      <td>22.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>313</th>\n",
       "      <td>166</td>\n",
       "      <td>installments_head_max_old</td>\n",
       "      <td>22.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>314</th>\n",
       "      <td>103</td>\n",
       "      <td>category_2_mean</td>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>315</th>\n",
       "      <td>228</td>\n",
       "      <td>observation_date_old</td>\n",
       "      <td>21.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>88</td>\n",
       "      <td>category_2_4.0_purchase_amount_mean_old</td>\n",
       "      <td>21.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>317</th>\n",
       "      <td>131</td>\n",
       "      <td>category_3_mean</td>\n",
       "      <td>21.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>318</th>\n",
       "      <td>92</td>\n",
       "      <td>category_2_4.0_purchase_amount_sum_old</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319</th>\n",
       "      <td>39</td>\n",
       "      <td>card_id_count</td>\n",
       "      <td>20.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>320</th>\n",
       "      <td>253</td>\n",
       "      <td>purchase_Hour_max_old</td>\n",
       "      <td>20.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>321</th>\n",
       "      <td>169</td>\n",
       "      <td>installments_max</td>\n",
       "      <td>20.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>322</th>\n",
       "      <td>36</td>\n",
       "      <td>avg_spend_per_transaction_old</td>\n",
       "      <td>19.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>323</th>\n",
       "      <td>167</td>\n",
       "      <td>installments_head_sum</td>\n",
       "      <td>19.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>324</th>\n",
       "      <td>188</td>\n",
       "      <td>installments_tail_max</td>\n",
       "      <td>18.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>73</td>\n",
       "      <td>category_2_2.0_purchase_amount_mean_ratio</td>\n",
       "      <td>18.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>292</td>\n",
       "      <td>purchase_amount_mean_total</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>165</td>\n",
       "      <td>installments_head_max</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>248</td>\n",
       "      <td>purchase_Dayofweek_max</td>\n",
       "      <td>16.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>199</td>\n",
       "      <td>merchant_category_repurchase</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>330</th>\n",
       "      <td>29</td>\n",
       "      <td>authorized_flag_1_purchase_amount_mean_old</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>331</th>\n",
       "      <td>180</td>\n",
       "      <td>installments_pct_25_old</td>\n",
       "      <td>14.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>332</th>\n",
       "      <td>158</td>\n",
       "      <td>feature_max</td>\n",
       "      <td>14.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>333</th>\n",
       "      <td>176</td>\n",
       "      <td>installments_median_old</td>\n",
       "      <td>13.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334</th>\n",
       "      <td>332</td>\n",
       "      <td>purchased_before_issue</td>\n",
       "      <td>11.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>335</th>\n",
       "      <td>219</td>\n",
       "      <td>month_lag_min</td>\n",
       "      <td>8.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>336</th>\n",
       "      <td>336</td>\n",
       "      <td>repurchase_merchant_rate</td>\n",
       "      <td>7.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>78</td>\n",
       "      <td>category_2_2.0_purchase_amount_sum_total</td>\n",
       "      <td>7.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>338</th>\n",
       "      <td>157</td>\n",
       "      <td>feature_3</td>\n",
       "      <td>4.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>339</th>\n",
       "      <td>37</td>\n",
       "      <td>avg_trans_per_merchant</td>\n",
       "      <td>3.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>340</th>\n",
       "      <td>160</td>\n",
       "      <td>feature_min</td>\n",
       "      <td>1.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>178</td>\n",
       "      <td>installments_min_old</td>\n",
       "      <td>0.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     index                                     Feature  importance_mean\n",
       "0      330                       purchase_date_uptonow           1183.8\n",
       "1      205                         month_diff_mean_old           1077.0\n",
       "2      310                   purchase_amount_sum_ratio            927.6\n",
       "3      337                repurchase_merchant_rate_old            901.4\n",
       "4      266                      purchase_Month_nunique            804.8\n",
       "5      269                          purchase_Week_mean            804.2\n",
       "6       32                        authorized_flag_mean            736.6\n",
       "7      262                         purchase_Month_mean            629.0\n",
       "8       50      category_1_0_purchase_amount_sum_ratio            576.8\n",
       "9        2                                         CLV            552.0\n",
       "10     267                           purchase_Week_max            550.6\n",
       "11     275          purchase_amount_before_after_ratio            543.0\n",
       "12     217                          month_lag_mean_old            531.8\n",
       "13     325                      purchase_date_diff_old            531.2\n",
       "14     154                       fathers_day_2017_mean            511.8\n",
       "15      27       authorized_flag_0_purchase_amount_sum            503.4\n",
       "16     206                       month_diff_mean_total            498.4\n",
       "17      61                         category_1_mean_old            474.6\n",
       "18       4                                   CLV_ratio            469.6\n",
       "19      52           category_1_1_purchase_amount_mean            452.2\n",
       "20      10                       Mothers_Day_2018_mean            451.4\n",
       "21      62                       category_1_mean_total            443.0\n",
       "22     225                           month_lag_var_old            427.6\n",
       "23     291                  purchase_amount_mean_ratio            427.6\n",
       "24       6                  Children_day_2017_mean_old            427.2\n",
       "25     271                           purchase_Week_min            423.2\n",
       "26      12                     Valentine_Day_2017_mean            417.8\n",
       "27     331                   purchase_date_uptonow_old            413.0\n",
       "28     112        category_3_0_purchase_amount_sum_old            405.8\n",
       "29     218                        month_lag_mean_total            402.2\n",
       "30     163                              hist_first_buy            390.2\n",
       "31       9                       Mothers_Day_2017_mean            389.2\n",
       "32     236                                  price_skew            373.4\n",
       "33     216                              month_lag_mean            372.2\n",
       "34     242                           purchase_Day_mean            366.4\n",
       "35     203                              month_diff_max            364.2\n",
       "36     164                               hist_last_buy            361.8\n",
       "37      56            category_1_1_purchase_amount_sum            358.0\n",
       "38      26      authorized_flag_0_purchase_amount_mean            357.8\n",
       "39      46     category_1_0_purchase_amount_mean_ratio            353.0\n",
       "40     328                       purchase_date_uptomin            350.2\n",
       "41      38                  avg_trans_per_merchant_old            343.2\n",
       "42       7                     Christmas_Day_2017_mean            326.0\n",
       "43     255                      purchase_Hour_mean_old            322.0\n",
       "44      69    category_2_1.0_purchase_amount_sum_ratio            320.2\n",
       "45     187                        installments_sum_old            316.0\n",
       "46     137                               days_feature2            314.8\n",
       "47     274                   purchase_Week_nunique_old            309.0\n",
       "48     250                 purchase_Dayofweek_mean_old            308.0\n",
       "49     119            category_3_1_purchase_amount_sum            300.6\n",
       "50     123           category_3_2_purchase_amount_mean            297.0\n",
       "51     254                          purchase_Hour_mean            289.6\n",
       "52     204                             month_diff_mean            285.8\n",
       "53     223                          month_lag_skew_old            285.0\n",
       "54     127            category_3_2_purchase_amount_sum            283.2\n",
       "55     210                         month_diff_skew_old            282.8\n",
       "56       0                      Black_Friday_2017_mean            281.0\n",
       "57     324                          purchase_date_diff            280.8\n",
       "58      11                   Mothers_Day_2018_mean_old            280.4\n",
       "59      48            category_1_0_purchase_amount_sum            270.0\n",
       "60     334               purchased_on_weekend_mean_old            266.6\n",
       "61     115           category_3_1_purchase_amount_mean            264.0\n",
       "62     304                purchase_amount_pct_75_ratio            262.4\n",
       "63      49        category_1_0_purchase_amount_sum_old            256.8\n",
       "64      60                             category_1_mean            253.4\n",
       "65     243                       purchase_Day_mean_old            251.4\n",
       "66     139                               days_feature3            251.0\n",
       "67     329                   purchase_date_uptomin_old            250.2\n",
       "68     135                               days_feature1            247.2\n",
       "69     322                       purchase_date_average            245.8\n",
       "70     200            merchant_category_repurchase_old            245.6\n",
       "71     323                   purchase_date_average_old            241.0\n",
       "72      41                         card_id_count_ratio            238.2\n",
       "73     152                                duration_var            237.2\n",
       "74     326          purchase_date_successive_diff_mean            237.2\n",
       "75     327      purchase_date_successive_diff_mean_old            236.2\n",
       "76     202                     merchant_id_nunique_old            235.8\n",
       "77     308                         purchase_amount_sum            235.6\n",
       "78      59      category_1_1_purchase_amount_sum_total            233.6\n",
       "79      67          category_2_1.0_purchase_amount_sum            232.2\n",
       "80     121      category_3_1_purchase_amount_sum_ratio            231.0\n",
       "81     113      category_3_0_purchase_amount_sum_ratio            229.8\n",
       "82     207                              month_diff_min            227.6\n",
       "83     224                               month_lag_var            226.4\n",
       "84     295                purchase_amount_median_ratio            226.0\n",
       "85      63         category_2_1.0_purchase_amount_mean            224.2\n",
       "86       5                      Children_day_2017_mean            222.4\n",
       "87      65   category_2_1.0_purchase_amount_mean_ratio            218.8\n",
       "88     276           purchase_amount_fi_last_new_total            218.6\n",
       "89     240                               price_var_old            218.6\n",
       "90      68      category_2_1.0_purchase_amount_sum_old            218.0\n",
       "91      20                  amount_month_ratio_min_old            215.8\n",
       "92     136                         days_feature1_ratio            213.6\n",
       "93     107           category_3_0_purchase_amount_mean            213.6\n",
       "94     272                       purchase_Week_min_old            211.2\n",
       "95     256                           purchase_Hour_min            203.8\n",
       "96       1                  Black_Friday_2017_mean_old            202.8\n",
       "97     268                       purchase_Week_max_old            198.0\n",
       "98     208                            month_diff_ratio            197.2\n",
       "99     306                    purchase_amount_skew_old            196.6\n",
       "100     30       authorized_flag_1_purchase_amount_sum            193.6\n",
       "101      8                 Christmas_Day_2017_mean_old            193.4\n",
       "102    111            category_3_0_purchase_amount_sum            193.0\n",
       "103    241                            purchase_Day_max            190.6\n",
       "104    138                         days_feature2_ratio            189.4\n",
       "105    315                purchase_amount_tail_sum_old            188.4\n",
       "106    226                               new_first_buy            186.8\n",
       "107    260                          purchase_Month_max            186.4\n",
       "108     34                  avg_spend_per_merchant_old            185.2\n",
       "109    313                purchase_amount_tail_max_old            184.8\n",
       "110    278             purchase_amount_first_last_diff            183.6\n",
       "111    263                     purchase_Month_mean_old            180.8\n",
       "112    270                      purchase_Week_mean_old            180.4\n",
       "113    320                         purchase_amount_var            180.0\n",
       "114    285        purchase_amount_inter_quartile_ratio            179.2\n",
       "115    151                           duration_skew_old            178.6\n",
       "116    341                      transactions_count_old            178.0\n",
       "117    148                            duration_min_old            177.8\n",
       "118     54     category_1_1_purchase_amount_mean_ratio            175.4\n",
       "119    300                  purchase_amount_pct_25_old            175.2\n",
       "120    118     category_3_1_purchase_amount_mean_total            175.0\n",
       "121    141                                duration_max            174.0\n",
       "122    185                       installments_skew_old            173.8\n",
       "123    244                            purchase_Day_min            173.0\n",
       "124     13                      amount_month_ratio_max            171.4\n",
       "125      3                                     CLV_old            171.2\n",
       "126    282                purchase_amount_head_sum_old            169.6\n",
       "127     23                 amount_month_ratio_skew_old            169.4\n",
       "128    194                        installments_var_old            169.0\n",
       "129    183                          installments_ratio            167.6\n",
       "130    212                          month_diff_var_old            165.0\n",
       "131     57        category_1_1_purchase_amount_sum_old            163.6\n",
       "132     58      category_1_1_purchase_amount_sum_ratio            163.2\n",
       "133    294                  purchase_amount_median_old            161.8\n",
       "134    120        category_3_1_purchase_amount_sum_old            161.6\n",
       "135    109     category_3_0_purchase_amount_mean_ratio            161.2\n",
       "136    192                          installments_total            159.4\n",
       "137    155                                   feature_1            159.2\n",
       "138    222                              month_lag_skew            158.8\n",
       "139    227                                new_last_buy            158.4\n",
       "140    198            merchant_category_id_nunique_old            158.4\n",
       "141    280                purchase_amount_head_max_old            157.8\n",
       "142    229                       observed_elapsed_time            157.8\n",
       "143     53       category_1_1_purchase_amount_mean_old            157.2\n",
       "144     24                      amount_month_ratio_var            156.2\n",
       "145    301                purchase_amount_pct_25_ratio            155.2\n",
       "146     44           category_1_0_purchase_amount_mean            154.8\n",
       "147    117     category_3_1_purchase_amount_mean_ratio            153.8\n",
       "148     64     category_2_1.0_purchase_amount_mean_old            153.4\n",
       "149     55     category_1_1_purchase_amount_mean_total            151.6\n",
       "150    116       category_3_1_purchase_amount_mean_old            147.8\n",
       "151    249                     purchase_Dayofweek_mean            147.0\n",
       "152     99          category_2_5.0_purchase_amount_sum            146.8\n",
       "153     51      category_1_0_purchase_amount_sum_total            146.6\n",
       "154    108       category_3_0_purchase_amount_mean_old            144.0\n",
       "155    234                                   price_min            142.4\n",
       "156    173                       installments_mean_old            141.8\n",
       "157    305                        purchase_amount_skew            141.6\n",
       "158    134                         category_3_mean_old            141.0\n",
       "159    307                  purchase_amount_skew_ratio            140.4\n",
       "160    235                               price_min_old            139.2\n",
       "161    129      category_3_2_purchase_amount_sum_ratio            137.8\n",
       "162     45       category_1_0_purchase_amount_mean_old            137.4\n",
       "163    122      category_3_1_purchase_amount_sum_total            133.6\n",
       "164     22                     amount_month_ratio_skew            131.8\n",
       "165    286                         purchase_amount_max            131.8\n",
       "166     70    category_2_1.0_purchase_amount_sum_total            131.6\n",
       "167    133                    category_3_mean_mean_old            130.6\n",
       "168    238                                 price_total            127.0\n",
       "169    182                     installments_pct_75_old            127.0\n",
       "170     19                      amount_month_ratio_min            126.2\n",
       "171    259                   purchase_Hour_nunique_old            125.6\n",
       "172     66   category_2_1.0_purchase_amount_mean_total            124.6\n",
       "173    174                     installments_mean_total            124.4\n",
       "174    284          purchase_amount_inter_quartile_old            123.4\n",
       "175     47     category_1_0_purchase_amount_mean_total            122.8\n",
       "176    297                     purchase_amount_min_old            121.6\n",
       "177     43                               card_id_total            120.0\n",
       "178    339                    subsector_id_nunique_old            119.8\n",
       "179    232                                  price_mean            118.8\n",
       "180    147                                duration_min            118.6\n",
       "181    140                         days_feature3_ratio            117.6\n",
       "182    231                               price_max_old            116.2\n",
       "183    150                               duration_skew            115.8\n",
       "184    128        category_3_2_purchase_amount_sum_old            115.0\n",
       "185    230                                   price_max            112.6\n",
       "186    106                         category_2_mean_old            111.6\n",
       "187    114      category_3_0_purchase_amount_sum_total            111.6\n",
       "188    124       category_3_2_purchase_amount_mean_old            111.2\n",
       "189    110     category_3_0_purchase_amount_mean_total            110.4\n",
       "190    168                   installments_head_sum_old            109.2\n",
       "191    149                          duration_min_total            107.4\n",
       "192    196        inverse_avg_transactions_per_day_old            106.8\n",
       "193    211                              month_diff_var            105.4\n",
       "194    195            inverse_avg_transactions_per_day            104.2\n",
       "195    296                         purchase_amount_min            103.6\n",
       "196     21                amount_month_ratio_min_total            102.8\n",
       "197    105                    category_2_mean_mean_old            101.0\n",
       "198    233                              price_mean_old            100.4\n",
       "199     28      authorized_flag_1_purchase_amount_mean             99.2\n",
       "200    283              purchase_amount_inter_quartile             98.8\n",
       "201    191                   installments_tail_sum_old             98.6\n",
       "202     96     category_2_5.0_purchase_amount_mean_old             98.0\n",
       "203     95         category_2_5.0_purchase_amount_mean             96.8\n",
       "204    197                merchant_category_id_nunique             96.6\n",
       "205    319                   purchase_amount_total_min             96.0\n",
       "206    299                      purchase_amount_pct_25             95.8\n",
       "207    130      category_3_2_purchase_amount_sum_total             95.6\n",
       "208    126     category_3_2_purchase_amount_mean_total             95.2\n",
       "209     31   authorized_flag_1_purchase_amount_sum_old             95.2\n",
       "210    193                            installments_var             94.8\n",
       "211    247                    purchase_Day_nunique_old             93.8\n",
       "212    125     category_3_2_purchase_amount_mean_ratio             93.0\n",
       "213    239                                   price_var             92.8\n",
       "214    303                  purchase_amount_pct_75_old             91.0\n",
       "215     79         category_2_3.0_purchase_amount_mean             89.2\n",
       "216    101    category_2_5.0_purchase_amount_sum_ratio             88.6\n",
       "217    220                           month_lag_min_old             88.4\n",
       "218     14                  amount_month_ratio_max_old             86.4\n",
       "219    264                          purchase_Month_min             85.4\n",
       "220    309                     purchase_amount_sum_old             83.8\n",
       "221    237                                   price_sum             83.2\n",
       "222    142                            duration_max_old             83.2\n",
       "223     17                 amount_month_ratio_mean_old             79.8\n",
       "224     83          category_2_3.0_purchase_amount_sum             78.4\n",
       "225     25                  amount_month_ratio_var_old             78.0\n",
       "226     80     category_2_3.0_purchase_amount_mean_old             76.8\n",
       "227    281                    purchase_amount_head_sum             76.4\n",
       "228    221                         month_lag_min_total             76.4\n",
       "229     16                     amount_month_ratio_mean             76.2\n",
       "230    214                           month_lag_max_old             76.0\n",
       "231    209                             month_diff_skew             75.6\n",
       "232     85    category_2_3.0_purchase_amount_sum_ratio             74.6\n",
       "233    153                            duration_var_old             74.4\n",
       "234    257                       purchase_Hour_min_old             73.6\n",
       "235    279                    purchase_amount_head_max             73.2\n",
       "236    144                               duration_mean             73.2\n",
       "237    132                        category_3_mean_mean             72.2\n",
       "238     71         category_2_2.0_purchase_amount_mean             72.0\n",
       "239    172                           installments_mean             71.0\n",
       "240    293                      purchase_amount_median             70.2\n",
       "241     81   category_2_3.0_purchase_amount_mean_ratio             70.0\n",
       "242    314                    purchase_amount_tail_sum             70.0\n",
       "243    143                          duration_max_total             69.2\n",
       "244    312                    purchase_amount_tail_max             69.2\n",
       "245    261                      purchase_Month_max_old             68.8\n",
       "246    287                     purchase_amount_max_old             68.4\n",
       "247    100      category_2_5.0_purchase_amount_sum_old             67.4\n",
       "248     40                           card_id_count_old             67.2\n",
       "249    156                                   feature_2             66.4\n",
       "250     18               amount_month_ratio_mean_total             66.2\n",
       "251     91          category_2_4.0_purchase_amount_sum             65.6\n",
       "252     15                amount_month_ratio_max_total             65.0\n",
       "253    302                      purchase_amount_pct_75             63.8\n",
       "254    215                         month_lag_max_total             62.6\n",
       "255    162                                 feature_var             61.6\n",
       "256    145                           duration_mean_old             61.4\n",
       "257     75          category_2_2.0_purchase_amount_sum             61.2\n",
       "258     93    category_2_4.0_purchase_amount_sum_ratio             60.8\n",
       "259    316                       purchase_amount_total             60.2\n",
       "260    321                     purchase_amount_var_old             58.4\n",
       "261    317                   purchase_amount_total_max             58.4\n",
       "262     87         category_2_4.0_purchase_amount_mean             58.0\n",
       "263     97   category_2_5.0_purchase_amount_mean_ratio             57.8\n",
       "264    252                           purchase_Hour_max             56.6\n",
       "265    251                      purchase_Dayofweek_min             55.0\n",
       "266    333                   purchased_on_weekend_mean             54.6\n",
       "267    338                        subsector_id_nunique             53.8\n",
       "268    184                           installments_skew             53.0\n",
       "269    161                                 feature_sum             53.0\n",
       "270    277           purchase_amount_fi_last_old_total             52.8\n",
       "271    265                      purchase_Month_min_old             52.6\n",
       "272    289                        purchase_amount_mean             50.6\n",
       "273     35                   avg_spend_per_transaction             50.2\n",
       "274    181                         installments_pct_75             48.8\n",
       "275    290                    purchase_amount_mean_old             48.8\n",
       "276    201                         merchant_id_nunique             46.2\n",
       "277    146                         duration_mean_total             46.2\n",
       "278    340                          transactions_count             45.0\n",
       "279    318                  purchase_amount_total_mean             44.8\n",
       "280    171                      installments_max_total             44.8\n",
       "281     98   category_2_5.0_purchase_amount_mean_total             43.2\n",
       "282    170                        installments_max_old             41.8\n",
       "283     42                         card_id_count_total             41.4\n",
       "284    102    category_2_5.0_purchase_amount_sum_total             41.0\n",
       "285     86    category_2_3.0_purchase_amount_sum_total             40.2\n",
       "286     82   category_2_3.0_purchase_amount_mean_total             38.8\n",
       "287    190                       installments_tail_sum             38.2\n",
       "288     76      category_2_2.0_purchase_amount_sum_old             38.0\n",
       "289    335                                     quarter             37.8\n",
       "290    186                            installments_sum             37.6\n",
       "291    273                       purchase_Week_nunique             37.2\n",
       "292    298                   purchase_amount_min_total             36.8\n",
       "293    104                        category_2_mean_mean             34.4\n",
       "294     89   category_2_4.0_purchase_amount_mean_ratio             33.4\n",
       "295    213                               month_lag_max             33.0\n",
       "296    245                        purchase_Day_min_old             32.8\n",
       "297     33                      avg_spend_per_merchant             32.2\n",
       "298    311                   purchase_amount_sum_total             31.0\n",
       "299     84      category_2_3.0_purchase_amount_sum_old             31.0\n",
       "300    189                   installments_tail_max_old             31.0\n",
       "301    175                         installments_median             29.8\n",
       "302    258                       purchase_Hour_nunique             28.6\n",
       "303     77    category_2_2.0_purchase_amount_sum_ratio             26.8\n",
       "304     74   category_2_2.0_purchase_amount_mean_total             26.6\n",
       "305    179                         installments_pct_25             26.6\n",
       "306    288                   purchase_amount_max_total             26.0\n",
       "307    246                        purchase_Day_nunique             25.8\n",
       "308     72     category_2_2.0_purchase_amount_mean_old             25.0\n",
       "309    159                                feature_mean             24.8\n",
       "310     90   category_2_4.0_purchase_amount_mean_total             24.6\n",
       "311     94    category_2_4.0_purchase_amount_sum_total             23.6\n",
       "312    177                            installments_min             22.6\n",
       "313    166                   installments_head_max_old             22.6\n",
       "314    103                             category_2_mean             21.6\n",
       "315    228                        observation_date_old             21.4\n",
       "316     88     category_2_4.0_purchase_amount_mean_old             21.2\n",
       "317    131                             category_3_mean             21.2\n",
       "318     92      category_2_4.0_purchase_amount_sum_old             21.0\n",
       "319     39                               card_id_count             20.8\n",
       "320    253                       purchase_Hour_max_old             20.8\n",
       "321    169                            installments_max             20.2\n",
       "322     36               avg_spend_per_transaction_old             19.8\n",
       "323    167                       installments_head_sum             19.6\n",
       "324    188                       installments_tail_max             18.6\n",
       "325     73   category_2_2.0_purchase_amount_mean_ratio             18.4\n",
       "326    292                  purchase_amount_mean_total             18.0\n",
       "327    165                       installments_head_max             17.2\n",
       "328    248                      purchase_Dayofweek_max             16.4\n",
       "329    199                merchant_category_repurchase             16.0\n",
       "330     29  authorized_flag_1_purchase_amount_mean_old             16.0\n",
       "331    180                     installments_pct_25_old             14.8\n",
       "332    158                                 feature_max             14.2\n",
       "333    176                     installments_median_old             13.4\n",
       "334    332                      purchased_before_issue             11.8\n",
       "335    219                               month_lag_min              8.2\n",
       "336    336                    repurchase_merchant_rate              7.8\n",
       "337     78    category_2_2.0_purchase_amount_sum_total              7.2\n",
       "338    157                                   feature_3              4.6\n",
       "339     37                      avg_trans_per_merchant              3.6\n",
       "340    160                                 feature_min              1.8\n",
       "341    178                        installments_min_old              0.2"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_ft_df.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['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'], dtype=object)"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fts_sorted_by_imp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "picked_feats = best_features.sort_values(by='importance', ascending=False)[:800]['Feature'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse(y_true, y_pred):\n",
    "    return np.sqrt(mean_squared_error(y_true, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def kfold_lightgbm(train_df, test_df, num_folds, stratified = False, debug= False):\n",
    "    print(\"Starting LightGBM. Train shape: {}, test shape: {}\".format(train_df.shape, test_df.shape))\n",
    "\n",
    "    # Cross validation model\n",
    "    if stratified:\n",
    "        folds = StratifiedKFold(n_splits= num_folds, shuffle=False, random_state=326)\n",
    "    else:\n",
    "        folds = KFold(n_splits= num_folds, shuffle=True, random_state=326)\n",
    "\n",
    "    # Create arrays and dataframes to store results\n",
    "    oof_preds = np.zeros(train_df.shape[0])\n",
    "    sub_preds = np.zeros(test_df.shape[0])\n",
    "    feature_importance_df = pd.DataFrame()\n",
    "    feats = [c for c in cols_included if c not in cols_excluded]\n",
    "    print(\"Total feats picked is \", len(feats))\n",
    "    # k-fold\n",
    "    for n_fold, (train_idx, valid_idx) in enumerate(folds.split(train_df[feats], train_df['outliers'])):\n",
    "        train_x, train_y = train_df[feats].iloc[train_idx], train_df['target'].iloc[train_idx]\n",
    "        valid_x, valid_y = train_df[feats].iloc[valid_idx], train_df['target'].iloc[valid_idx]\n",
    "\n",
    "        # set data structure\n",
    "        lgb_train = lgb.Dataset(train_x,\n",
    "                                label=train_y,\n",
    "                                free_raw_data=False)\n",
    "        lgb_test = lgb.Dataset(valid_x,\n",
    "                               label=valid_y,\n",
    "                               free_raw_data=False)\n",
    "\n",
    "        # params optimized by optuna\n",
    "        params ={\n",
    "                'task': 'train',\n",
    "                'boosting': 'goss',\n",
    "                'objective': 'regression',\n",
    "                'metric': 'rmse',\n",
    "                'learning_rate': 0.01,\n",
    "                'subsample': 0.9855232997390695,\n",
    "                'max_depth': 7,\n",
    "                'top_rate': 0.9064148448434349,\n",
    "                'num_leaves': 63,\n",
    "                'min_child_weight': 41.9612869171337,\n",
    "                'other_rate': 0.0721768246018207,\n",
    "                'reg_alpha': 9.677537745007898,\n",
    "                'colsample_bytree': 0.5665320670155495,\n",
    "                'min_split_gain': 9.820197773625843,\n",
    "                'reg_lambda': 8.2532317400459,\n",
    "                'min_data_in_leaf': 21,\n",
    "                'verbose': -1,\n",
    "                'seed':int(2**n_fold),\n",
    "                'bagging_seed':int(2**n_fold),\n",
    "                'drop_seed':int(2**n_fold)\n",
    "                }\n",
    "\n",
    "        reg = lgb.train(\n",
    "                        params,\n",
    "                        lgb_train,\n",
    "                        valid_sets=[lgb_train, lgb_test],\n",
    "                        valid_names=['train', 'test'],\n",
    "                        num_boost_round=10000,\n",
    "                        early_stopping_rounds= 200,\n",
    "                        verbose_eval=100\n",
    "                        )\n",
    "\n",
    "        oof_preds[valid_idx] = reg.predict(valid_x, num_iteration=reg.best_iteration)\n",
    "        sub_preds += reg.predict(test_df[feats], num_iteration=reg.best_iteration) / folds.n_splits\n",
    "\n",
    "        fold_importance_df = pd.DataFrame()\n",
    "        fold_importance_df[\"feature\"] = feats\n",
    "        fold_importance_df[\"importance\"] = np.log1p(reg.feature_importance(importance_type='gain', iteration=reg.best_iteration))\n",
    "        fold_importance_df[\"fold\"] = n_fold + 1\n",
    "        feature_importance_df = pd.concat([feature_importance_df, fold_importance_df], axis=0)\n",
    "        print('Fold %2d RMSE : %.6f' % (n_fold + 1, rmse(valid_y, oof_preds[valid_idx])))\n",
    "        del reg, train_x, train_y, valid_x, valid_y\n",
    "        gc.collect()\n",
    "\n",
    "    # display importances\n",
    "    display_importances(feature_importance_df)\n",
    "\n",
    "    if not debug:\n",
    "        # save submission file\n",
    "        submission_name = pd.to_datetime(\"today\")\n",
    "        sub_df = pd.DataFrame({\"card_id\":test_df2[\"card_id\"].values})\n",
    "        sub_df[\"target\"] = sub_preds\n",
    "        sub_df.to_csv(f'submission_kfold-{submission_name}.csv', index=False)\n",
    "#         test_df.loc[:,'target'] = sub_preds\n",
    "#         test_df = test_df.reset_index()\n",
    "#         test_df[['card_id', 'target']].to_csv(submission_name, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Display/plot feature importance\n",
    "def display_importances(feature_importance_df_):\n",
    "    cols = feature_importance_df_[[\"feature\", \"importance\"]].groupby(\"feature\").mean().sort_values(by=\"importance\", ascending=False)[:40].index\n",
    "    best_features = feature_importance_df_.loc[feature_importance_df_.feature.isin(cols)]\n",
    "\n",
    "    plt.figure(figsize=(8, 10))\n",
    "    sns.barplot(x=\"importance\", y=\"feature\", data=best_features.sort_values(by=\"importance\", ascending=False))\n",
    "    plt.title('LightGBM Features (avg over folds)')\n",
    "    plt.tight_layout()\n",
    "    plt.savefig('lgbm_importances.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse(y_true, y_pred):\n",
    "    return np.sqrt(mean_squared_error(y_true, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "753"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting LightGBM. Train shape: (201917, 352), test shape: (123623, 350)\n",
      "Total feats picked is  180\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.64653\ttest's rmse: 3.7208\n",
      "[200]\ttrain's rmse: 3.56532\ttest's rmse: 3.68475\n",
      "[300]\ttrain's rmse: 3.51858\ttest's rmse: 3.67335\n",
      "[400]\ttrain's rmse: 3.48479\ttest's rmse: 3.66973\n",
      "[500]\ttrain's rmse: 3.45881\ttest's rmse: 3.66879\n",
      "[600]\ttrain's rmse: 3.43738\ttest's rmse: 3.66824\n",
      "[700]\ttrain's rmse: 3.419\ttest's rmse: 3.66831\n",
      "Early stopping, best iteration is:\n",
      "[587]\ttrain's rmse: 3.44063\ttest's rmse: 3.66803\n",
      "Fold  1 RMSE : 3.668031\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.63502\ttest's rmse: 3.80651\n",
      "[200]\ttrain's rmse: 3.55747\ttest's rmse: 3.75925\n",
      "[300]\ttrain's rmse: 3.51211\ttest's rmse: 3.7414\n",
      "[400]\ttrain's rmse: 3.48009\ttest's rmse: 3.73399\n",
      "[500]\ttrain's rmse: 3.45577\ttest's rmse: 3.73027\n",
      "[600]\ttrain's rmse: 3.43528\ttest's rmse: 3.72839\n",
      "[700]\ttrain's rmse: 3.41644\ttest's rmse: 3.7269\n",
      "[800]\ttrain's rmse: 3.39871\ttest's rmse: 3.72616\n",
      "[900]\ttrain's rmse: 3.38231\ttest's rmse: 3.72582\n",
      "[1000]\ttrain's rmse: 3.36576\ttest's rmse: 3.72544\n",
      "[1100]\ttrain's rmse: 3.34983\ttest's rmse: 3.72478\n",
      "[1200]\ttrain's rmse: 3.3352\ttest's rmse: 3.72425\n",
      "[1300]\ttrain's rmse: 3.32032\ttest's rmse: 3.72441\n",
      "Early stopping, best iteration is:\n",
      "[1146]\ttrain's rmse: 3.34304\ttest's rmse: 3.72411\n",
      "Fold  2 RMSE : 3.724107\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.67023\ttest's rmse: 3.60903\n",
      "[200]\ttrain's rmse: 211521\ttest's rmse: 213585\n",
      "[300]\ttrain's rmse: 1.29516e+10\ttest's rmse: 1.29298e+10\n",
      "Early stopping, best iteration is:\n",
      "[100]\ttrain's rmse: 3.67023\ttest's rmse: 3.60903\n",
      "Fold  3 RMSE : 3.609033\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.67269\ttest's rmse: 3.5863\n",
      "[200]\ttrain's rmse: 3.58936\ttest's rmse: 3.55447\n",
      "[300]\ttrain's rmse: 3.54219\ttest's rmse: 3.5461\n",
      "[400]\ttrain's rmse: 3.5118\ttest's rmse: 3.5424\n",
      "[500]\ttrain's rmse: 3.48754\ttest's rmse: 3.54119\n",
      "[600]\ttrain's rmse: 3.46802\ttest's rmse: 3.54081\n",
      "[700]\ttrain's rmse: 3.44929\ttest's rmse: 3.54103\n",
      "Early stopping, best iteration is:\n",
      "[585]\ttrain's rmse: 3.47086\ttest's rmse: 3.54075\n",
      "Fold  4 RMSE : 3.540754\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.64673\ttest's rmse: 3.7143\n",
      "[200]\ttrain's rmse: 3.56642\ttest's rmse: 3.6859\n",
      "[300]\ttrain's rmse: 3.51866\ttest's rmse: 3.67655\n",
      "[400]\ttrain's rmse: 3.48616\ttest's rmse: 3.67272\n",
      "[500]\ttrain's rmse: 3.46256\ttest's rmse: 3.671\n",
      "[600]\ttrain's rmse: 3.44302\ttest's rmse: 3.66941\n",
      "[700]\ttrain's rmse: 3.42437\ttest's rmse: 3.66872\n",
      "[800]\ttrain's rmse: 3.40589\ttest's rmse: 3.66817\n",
      "[900]\ttrain's rmse: 3.38931\ttest's rmse: 3.66783\n",
      "[1000]\ttrain's rmse: 3.37418\ttest's rmse: 3.66754\n",
      "[1100]\ttrain's rmse: 3.35885\ttest's rmse: 3.66765\n",
      "[1200]\ttrain's rmse: 3.3443\ttest's rmse: 3.66793\n",
      "Early stopping, best iteration is:\n",
      "[1077]\ttrain's rmse: 3.36263\ttest's rmse: 3.66744\n",
      "Fold  5 RMSE : 3.667440\n",
      "Training until validation scores don't improve for 200 rounds.\n",
      "[100]\ttrain's rmse: 3.63939\ttest's rmse: 3.76362\n",
      "[200]\ttrain's rmse: 3.5589\ttest's rmse: 3.7256\n",
      "[300]\ttrain's rmse: 3.51388\ttest's rmse: 3.71431\n",
      "[400]\ttrain's rmse: 3.48239\ttest's rmse: 3.70963\n",
      "[500]\ttrain's rmse: 3.46041\ttest's rmse: 3.70718\n",
      "[600]\ttrain's rmse: 3.43993\ttest's rmse: 3.7056\n",
      "[700]\ttrain's rmse: 3.42134\ttest's rmse: 3.70445\n",
      "[800]\ttrain's rmse: 3.40435\ttest's rmse: 3.70414\n",
      "[900]\ttrain's rmse: 3.38976\ttest's rmse: 3.70366\n",
      "[1000]\ttrain's rmse: 3.3736\ttest's rmse: 3.7034\n",
      "[1100]\ttrain's rmse: 3.36068\ttest's rmse: 3.70273\n",
      "[1200]\ttrain's rmse: 3.34739\ttest's rmse: 3.7021\n",
      "[1300]\ttrain's rmse: 3.33209\ttest's rmse: 3.70196\n",
      "[1400]\ttrain's rmse: 3.31797\ttest's rmse: 3.70196\n",
      "[1500]\ttrain's rmse: 3.30283\ttest's rmse: 3.70177\n",
      "[1600]\ttrain's rmse: 3.28893\ttest's rmse: 3.70193\n",
      "[1700]\ttrain's rmse: 3.27536\ttest's rmse: 3.70217\n",
      "Early stopping, best iteration is:\n",
      "[1544]\ttrain's rmse: 3.29686\ttest's rmse: 3.70154\n",
      "Fold  6 RMSE : 3.701545\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/chandrasekhar/anaconda3/envs/fastai/lib/python3.6/site-packages/scipy/stats/stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.\n",
      "  return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "kfold_lightgbm(train_df2, test_df2, num_folds=6, stratified=False, debug=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_name = pd.to_datetime(\"today\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub_df = pd.DataFrame({\"card_id\":test_df2[\"card_id\"].values})\n",
    "sub_df[\"target\"] = predictions\n",
    "sub_df.to_csv(f'submission-{submission_name}.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Blend with df outliers.ipynb'\r\n",
      "'Blending submissions.ipynb'\r\n",
      "'Combining model without outlier.ipynb'\r\n",
      "'Elo - ML model - 3.690.ipynb'\r\n",
      "'Elo - ML model - 3.693.ipynb'\r\n",
      "'Elo - ML model - 3.695 beta.ipynb'\r\n",
      "'Elo - ML model - 3.695.ipynb'\r\n",
      "'Elo - ML model - 3.696.ipynb'\r\n",
      "'Elo - ML model - alpha.ipynb'\r\n",
      "'Elo - ML model - cat aggregates.ipynb'\r\n",
      "'Elo - ML model 3.691 beta new.ipynb'\r\n",
      "'Elo - ML model 3.691.ipynb'\r\n",
      "'Elo - ML model clean.ipynb'\r\n",
      "'Elo - ML model proc_df.ipynb'\r\n",
      "'Elo - ML model-RF copy.ipynb'\r\n",
      "'Elo - ML model-backup.ipynb'\r\n",
      "'Elo - ML model-new.ipynb'\r\n",
      "'Elo RF- New.ipynb'\r\n",
      "'Elo RF.ipynb'\r\n",
      "'Elo end 2 end.ipynb'\r\n",
      "'Elo feature engineering - 3.691 - beta.ipynb'\r\n",
      "'Elo feature engineering - 3.691.ipynb'\r\n",
      "'Elo feature engineering - 3.696.ipynb'\r\n",
      "'Elo feature engineering 3.690.ipynb'\r\n",
      "'Elo feature engineering 3.693 .ipynb'\r\n",
      "'Elo feature engineering 3.695 beta.ipynb'\r\n",
      "'Elo feature engineering 3.695.ipynb'\r\n",
      "'Elo feature engineering alpha.ipynb'\r\n",
      "'Elo feature engineering using agg.ipynb'\r\n",
      "'Elo feature engineering-backup.ipynb'\r\n",
      "'Elo feature engineering-proc_df - new-Copy1.ipynb'\r\n",
      "'Elo feature engineering-proc_df - new.ipynb'\r\n",
      "'Elo feature engineering-proc_df-OLD.ipynb'\r\n",
      "'Elo feature engineering-proc_df.ipynb'\r\n",
      "'Elo merchant category reco - using RF.ipynb'\r\n",
      "'Elo outlier predictions & blending .ipynb'\r\n",
      "'Feature selection using Boruta.ipynb'\r\n",
      "'Feature selection.ipynb'\r\n",
      "'Null importances - beta.ipynb'\r\n",
      "'Outlier classification using Autoencoder.ipynb'\r\n",
      " README.md\r\n",
      "'Scrap pad.ipynb'\r\n",
      "'Scrap pad.py'\r\n",
      "'TS Fresh.ipynb'\r\n",
      "'TS Fresh.py'\r\n",
      "'Train & test clean.ipynb'\r\n",
      "'Using DL-Copy1.ipynb'\r\n",
      "'Using DL.ipynb'\r\n",
      "'Using Feature tools - old.ipynb'\r\n",
      "'Using feature tools-Copy1.ipynb'\r\n",
      "'Using feature tools.ipynb'\r\n",
      " actual_importances_ditribution_rf.csv\r\n",
      " adf\r\n",
      " alla.pkl\r\n",
      " authorised_trans_proc_agg\r\n",
      "'blend-2019-02-02 23:24:52.864854.csv'\r\n",
      " blend-{submission_name}\r\n",
      " blend_kfol_st.csv\r\n",
      " chandu_mean1.csv\r\n",
      " cols_keep.npy\r\n",
      " comb_15.csv\r\n",
      "'combined_submission-2019-02-15 15:07:09.944221.csv'\r\n",
      "'combined_submission-2019-02-15 15:09:20.708055.csv'\r\n",
      "'combined_submission-2019-02-20 00:21:17.597315.csv'\r\n",
      "'combined_submission-2019-02-20 00:25:02.521379.csv'\r\n",
      "'combined_submission-2019-02-20 00:26:39.776651.csv'\r\n",
      " combining_submission.csv\r\n",
      " corr_scores_null_importances.csv\r\n",
      " correlation\r\n",
      " data\r\n",
      " df_auth\r\n",
      " df_hot_dropped_all\r\n",
      " df_hot_ext\r\n",
      " df_hot_test\r\n",
      " df_hot_test_ext\r\n",
      " df_new\r\n",
      " df_non_auth\r\n",
      " df_proc_dfed\r\n",
      " df_test_n_o\r\n",
      " df_xo\r\n",
      " fastai\r\n",
      " fm_new_hist\r\n",
      " hist_trans\r\n",
      " hist_trans_3_696\r\n",
      " hist_trans_agg\r\n",
      " hist_trans_agg_3_696\r\n",
      " hist_trans_agg_beta\r\n",
      " hist_trans_agg_cat\r\n",
      " hist_trans_agg_month\r\n",
      " hist_trans_agg_n\r\n",
      " hist_trans_agg_proc\r\n",
      " hist_trans_beta\r\n",
      " hist_trans_purchase_aggs_alpha\r\n",
      "'kfold lightgbm.ipynb'\r\n",
      " latent_representation.png\r\n",
      " lgbm_importances.png\r\n",
      " lla\r\n",
      " new_hist_prc\r\n",
      " new_hist_trans\r\n",
      " new_hist_trans_3_696\r\n",
      " new_hist_trans_agg\r\n",
      " new_hist_trans_agg_3_696\r\n",
      " new_hist_trans_agg_beta\r\n",
      " new_hist_trans_agg_cat\r\n",
      " new_hist_trans_agg_month\r\n",
      " new_hist_trans_agg_n\r\n",
      " new_hist_trans_beta\r\n",
      " new_hist_trans_purchase_aggs_alpha\r\n",
      " new_trans_agg_proc\r\n",
      " new_trans_proc_agg\r\n",
      " non_authorised_trans_proc\r\n",
      " non_authorised_trans_proc_agg\r\n",
      " null_importances_distribution_rf.csv\r\n",
      " oof_lgb_3.npy\r\n",
      " oof_xgb_3.npy\r\n",
      " original.png\r\n",
      "'outlier_combining_submission-2019-02-19 02:12:16.662625.csv'\r\n",
      "'outlier_combining_submission-2019-02-19 02:28:13.811057.csv'\r\n",
      "'outlier_combining_submission-2019-02-19 02:48:44.830259.csv'\r\n",
      " predictions_lgb_3.npy\r\n",
      " predictions_xgb_3.npy\r\n",
      "'scrap pad 2.ipynb'\r\n",
      "'submission-2019-01-11 09:49:53.716164.csv'\r\n",
      "'submission-2019-01-11 10:18:09.909061.csv'\r\n",
      "'submission-2019-01-12 13:53:27.215115.csv'\r\n",
      "'submission-2019-01-12 16:27:31.239044.csv'\r\n",
      "'submission-2019-01-12 17:25:56.678553.csv'\r\n",
      "'submission-2019-01-12 17:33:29.475835.csv'\r\n",
      "'submission-2019-01-12 17:42:01.350278.csv'\r\n",
      "'submission-2019-01-12 22:12:59.947458.csv'\r\n",
      "'submission-2019-01-13 15:59:45.792347.csv'\r\n",
      "'submission-2019-01-13 18:24:51.237991.csv'\r\n",
      "'submission-2019-01-14 19:04:33.028609.csv'\r\n",
      "'submission-2019-01-14 19:56:09.906235.csv'\r\n",
      "'submission-2019-01-15 22:18:34.294443.csv'\r\n",
      "'submission-2019-01-15 22:24:38.071972.csv'\r\n",
      "'submission-2019-01-15 22:29:20.144417.csv'\r\n",
      "'submission-2019-01-16 00:35:46.490517.csv'\r\n",
      "'submission-2019-01-16 00:36:26.004443.csv'\r\n",
      "'submission-2019-01-16 12:20:00.957329.csv'\r\n",
      "'submission-2019-01-16 21:37:16.421927.csv'\r\n",
      "'submission-2019-01-17 10:20:06.724992.csv'\r\n",
      "'submission-2019-01-17 11:09:17.495216.csv'\r\n",
      "'submission-2019-01-17 23:23:57.743505.csv'\r\n",
      "'submission-2019-01-18 01:18:08.182115.csv'\r\n",
      "'submission-2019-01-18 01:42:20.587834.csv'\r\n",
      "'submission-2019-01-18 01:47:04.024727.csv'\r\n",
      "'submission-2019-01-18 16:16:45.475456.csv'\r\n",
      "'submission-2019-01-18 16:27:58.560480.csv'\r\n",
      "'submission-2019-01-18 17:22:46.188675.csv'\r\n",
      "'submission-2019-01-20 00:30:10.643280.csv'\r\n",
      "'submission-2019-01-20 01:28:12.077657.csv'\r\n",
      "'submission-2019-01-20 01:46:39.187580.csv'\r\n",
      "'submission-2019-01-20 02:32:16.439302.csv'\r\n",
      "'submission-2019-01-20 17:01:54.207459.csv'\r\n",
      "'submission-2019-01-20 17:27:46.797180.csv'\r\n",
      "'submission-2019-01-20 18:29:37.773532.csv'\r\n",
      "'submission-2019-01-20 19:11:58.593189.csv'\r\n",
      "'submission-2019-01-20 19:35:24.892043.csv'\r\n",
      "'submission-2019-01-20 20:47:32.743729.csv'\r\n",
      "'submission-2019-01-22 09:33:02.341491.csv'\r\n",
      "'submission-2019-01-22 10:35:57.465665.csv'\r\n",
      "'submission-2019-01-31 22:00:10.810901.csv'\r\n",
      "'submission-2019-01-31 22:04:41.815902.csv'\r\n",
      "'submission-2019-02-02 23:13:59.670630.csv'\r\n",
      "'submission-2019-02-02 23:24:52.864854.csv'\r\n",
      "'submission-2019-02-03 14:42:22.636262.csv'\r\n",
      "'submission-2019-02-03 14:43:29.797523.csv'\r\n",
      "'submission-2019-02-10 14:58:45.312994.csv'\r\n",
      "'submission-2019-02-10 17:36:17.986042.csv'\r\n",
      "'submission-2019-02-10 18:01:58.093819.csv'\r\n",
      "'submission-2019-02-12 22:57:39.862967.csv'\r\n",
      "'submission-2019-02-13 00:42:12.274968.csv'\r\n",
      "'submission-2019-02-13 01:51:25.440947.csv'\r\n",
      "'submission-2019-02-13 22:48:45.005871.csv'\r\n",
      "'submission-2019-02-15 14:22:06.881490.csv'\r\n",
      "'submission-2019-02-15 14:45:36.932530.csv'\r\n",
      "'submission-2019-02-15 18:08:33.904515.csv'\r\n",
      "'submission-2019-02-16 13:14:48.321120.csv'\r\n",
      "'submission-2019-02-16 13:22:43.779348.csv'\r\n",
      "'submission-2019-02-16 23:41:55.685750.csv'\r\n",
      "'submission-2019-02-16 23:52:03.148358.csv'\r\n",
      "'submission-2019-02-19 00:14:48.737899.csv'\r\n",
      "'submission-2019-02-19 00:42:33.872294.csv'\r\n",
      "'submission-2019-02-19 00:51:12.314011.csv'\r\n",
      "'submission-2019-02-19 00:51:48.972772.csv'\r\n",
      "'submission_kfold-2019-02-13 00:55:25.132146.csv'\r\n",
      "'submission_kfold-2019-02-13 01:29:28.856630.csv'\r\n",
      "'submission_kfold-2019-02-13 01:59:08.971995.csv'\r\n",
      "'submission_kfold-2019-02-13 23:21:24.051900.csv'\r\n",
      "'submission_kfold-2019-02-20 00:58:19.913926.csv'\r\n",
      " submissions\r\n",
      " target_all\r\n",
      " target_outliers_removed\r\n",
      " test_clean.csv\r\n",
      " test_clean.csv.zip\r\n",
      " test_df\r\n",
      " test_df_3.695\r\n",
      " test_df_3691\r\n",
      " test_df_3_696\r\n",
      " test_df_alpha\r\n",
      " test_df_beta\r\n",
      " test_df_proc\r\n",
      " test_df_proc_new\r\n",
      " test_df_proc_new_m\r\n",
      " train_clean.csv\r\n",
      " train_clean.csv.zip\r\n",
      " train_clean0\r\n",
      " train_df\r\n",
      " train_df_3691\r\n",
      " train_df_3_696\r\n",
      " train_df_alpha\r\n",
      " train_df_beta\r\n",
      " train_df_proc\r\n",
      " train_df_proc_new\r\n",
      " train_df_proc_new_m\r\n"
     ]
    }
   ],
   "source": [
    "!ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='submission-2019-02-19 00:14:48.737899.csv' target='_blank'>submission-2019-02-19 00:14:48.737899.csv</a><br>"
      ],
      "text/plain": [
       "/home/chandrasekhar/private/kaggle/submission-2019-02-19 00:14:48.737899.csv"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FileLink(f'submission-{submission_name}.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Blend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "p1 = pd.read_csv(f'submission-{submission_name}.csv')\n",
    "p2 = pd.read_csv('submission-2019-02-02 23:13:59.670630.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "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>card_id</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_ID_0ab67a22ab</td>\n",
       "      <td>-2.960024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C_ID_130fd0cbdd</td>\n",
       "      <td>-0.421305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C_ID_b709037bc5</td>\n",
       "      <td>-1.011111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C_ID_d27d835a9f</td>\n",
       "      <td>-0.141994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C_ID_2b5e3df5c2</td>\n",
       "      <td>-1.270823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           card_id    target\n",
       "0  C_ID_0ab67a22ab -2.960024\n",
       "1  C_ID_130fd0cbdd -0.421305\n",
       "2  C_ID_b709037bc5 -1.011111\n",
       "3  C_ID_d27d835a9f -0.141994\n",
       "4  C_ID_2b5e3df5c2 -1.270823"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "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>card_id</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_ID_0ab67a22ab</td>\n",
       "      <td>-3.072875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C_ID_130fd0cbdd</td>\n",
       "      <td>-0.348145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C_ID_b709037bc5</td>\n",
       "      <td>-0.949767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C_ID_d27d835a9f</td>\n",
       "      <td>-0.149493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C_ID_2b5e3df5c2</td>\n",
       "      <td>-1.212316</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           card_id    target\n",
       "0  C_ID_0ab67a22ab -3.072875\n",
       "1  C_ID_130fd0cbdd -0.348145\n",
       "2  C_ID_b709037bc5 -0.949767\n",
       "3  C_ID_d27d835a9f -0.149493\n",
       "4  C_ID_2b5e3df5c2 -1.212316"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "p3 = p1.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "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>card_id</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_ID_0ab67a22ab</td>\n",
       "      <td>-2.960024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C_ID_130fd0cbdd</td>\n",
       "      <td>-0.421305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C_ID_b709037bc5</td>\n",
       "      <td>-1.011111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C_ID_d27d835a9f</td>\n",
       "      <td>-0.141994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C_ID_2b5e3df5c2</td>\n",
       "      <td>-1.270823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           card_id    target\n",
       "0  C_ID_0ab67a22ab -2.960024\n",
       "1  C_ID_130fd0cbdd -0.421305\n",
       "2  C_ID_b709037bc5 -1.011111\n",
       "3  C_ID_d27d835a9f -0.141994\n",
       "4  C_ID_2b5e3df5c2 -1.270823"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "p3 = pd.DataFrame({\"card_id\":test_df2[\"card_id\"].values})\n",
    "p3[\"target\"] = p1['target']*.5 + p2['target']*.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "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>card_id</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_ID_0ab67a22ab</td>\n",
       "      <td>-3.016449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C_ID_130fd0cbdd</td>\n",
       "      <td>-0.384725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C_ID_b709037bc5</td>\n",
       "      <td>-0.980439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C_ID_d27d835a9f</td>\n",
       "      <td>-0.145743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C_ID_2b5e3df5c2</td>\n",
       "      <td>-1.241569</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           card_id    target\n",
       "0  C_ID_0ab67a22ab -3.016449\n",
       "1  C_ID_130fd0cbdd -0.384725\n",
       "2  C_ID_b709037bc5 -0.980439\n",
       "3  C_ID_d27d835a9f -0.145743\n",
       "4  C_ID_2b5e3df5c2 -1.241569"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "p3.to_csv('blend_kfol_st.csv',index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "message = 'using blend kfold + stratified, purchase amount trimmed'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "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>card_id</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_ID_0ab67a22ab</td>\n",
       "      <td>-3.016449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C_ID_130fd0cbdd</td>\n",
       "      <td>-0.384725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>C_ID_b709037bc5</td>\n",
       "      <td>-0.980439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C_ID_d27d835a9f</td>\n",
       "      <td>-0.145743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>C_ID_2b5e3df5c2</td>\n",
       "      <td>-1.241569</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           card_id    target\n",
       "0  C_ID_0ab67a22ab -3.016449\n",
       "1  C_ID_130fd0cbdd -0.384725\n",
       "2  C_ID_b709037bc5 -0.980439\n",
       "3  C_ID_d27d835a9f -0.145743\n",
       "4  C_ID_2b5e3df5c2 -1.241569"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = f'submission-{submission_name}.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'submission-2019-02-19 00:51:48.972772.csv'"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filename"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100%|#######################################| 4.24M/4.24M [00:22<00:00, 197kB/s]\n",
      "Successfully submitted to Elo Merchant Category Recommendation"
     ]
    }
   ],
   "source": [
    "!kaggle competitions submit elo-merchant-category-recommendation -f 'submission_kfold-2019-02-20 00:58:19.913926.csv' -m \"180 cols with kfold + uniformly distributed target\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = train_df2.copy()\n",
    "train['rounded_target'] = train['target'].round(0)\n",
    "train = train.sort_values('rounded_target').reset_index(drop=True)\n",
    "vc = train['rounded_target'].value_counts()\n",
    "vc = dict(sorted(vc.items()))\n",
    "df = pd.DataFrame()\n",
    "train['indexcol'],i = 0,1\n",
    "for k,v in vc.items():\n",
    "    step = train.shape[0]/v\n",
    "    indent = train.shape[0]/(v+1)\n",
    "    df2 = train[train['rounded_target'] == k].sample(v, random_state=120).reset_index(drop=True)\n",
    "    for j in range(0, v):\n",
    "        df2.at[j, 'indexcol'] = indent + j*step + 0.000001*i\n",
    "    df = pd.concat([df2,df])\n",
    "    i+=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{-33.0: 2207,\n",
       " -18.0: 1,\n",
       " -17.0: 3,\n",
       " -16.0: 2,\n",
       " -15.0: 6,\n",
       " -14.0: 8,\n",
       " -13.0: 6,\n",
       " -12.0: 5,\n",
       " -11.0: 14,\n",
       " -10.0: 31,\n",
       " -9.0: 56,\n",
       " -8.0: 141,\n",
       " -7.0: 284,\n",
       " -6.0: 704,\n",
       " -5.0: 1515,\n",
       " -4.0: 3067,\n",
       " -3.0: 6685,\n",
       " -2.0: 15652,\n",
       " -1.0: 38864,\n",
       " 0.0: 68531,\n",
       " 1.0: 37614,\n",
       " 2.0: 14648,\n",
       " 3.0: 6400,\n",
       " 4.0: 2907,\n",
       " 5.0: 1388,\n",
       " 6.0: 652,\n",
       " 7.0: 313,\n",
       " 8.0: 103,\n",
       " 9.0: 47,\n",
       " 10.0: 27,\n",
       " 11.0: 11,\n",
       " 12.0: 10,\n",
       " 13.0: 4,\n",
       " 14.0: 3,\n",
       " 15.0: 4,\n",
       " 16.0: 2,\n",
       " 18.0: 2}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df2 = df.sort_values('indexcol', ascending=True).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df2['indexcol'], train_df2['rounded_target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(201917, 352)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "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",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>first_active_month</th>\n",
       "      <td>2017-12-01 00:00:00</td>\n",
       "      <td>2016-12-01 00:00:00</td>\n",
       "      <td>2016-09-01 00:00:00</td>\n",
       "      <td>2017-11-01 00:00:00</td>\n",
       "      <td>2017-11-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id</th>\n",
       "      <td>C_ID_176ecc65bd</td>\n",
       "      <td>C_ID_6082929d3a</td>\n",
       "      <td>C_ID_a87bbd85a7</td>\n",
       "      <td>C_ID_92e4809531</td>\n",
       "      <td>C_ID_d2dd6bbf9a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_1</th>\n",
       "      <td>0.0131446</td>\n",
       "      <td>0.0107116</td>\n",
       "      <td>0.0104794</td>\n",
       "      <td>0.00805849</td>\n",
       "      <td>0.0107116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_2</th>\n",
       "      <td>0.0113848</td>\n",
       "      <td>0.0113848</td>\n",
       "      <td>0.0113848</td>\n",
       "      <td>0.0113848</td>\n",
       "      <td>0.0113848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feature_3</th>\n",
       "      <td>0.0114275</td>\n",
       "      <td>0.0102828</td>\n",
       "      <td>0.0114275</td>\n",
       "      <td>0.0102828</td>\n",
       "      <td>0.0102828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>target</th>\n",
       "      <td>-0.279039</td>\n",
       "      <td>-0.926039</td>\n",
       "      <td>0.475326</td>\n",
       "      <td>0.516026</td>\n",
       "      <td>-0.100927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthYear</th>\n",
       "      <td>2017</td>\n",
       "      <td>2016</td>\n",
       "      <td>2016</td>\n",
       "      <td>2017</td>\n",
       "      <td>2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthMonth</th>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthWeek</th>\n",
       "      <td>48</td>\n",
       "      <td>48</td>\n",
       "      <td>35</td>\n",
       "      <td>44</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthDay</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthDayofweek</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthDayofyear</th>\n",
       "      <td>335</td>\n",
       "      <td>336</td>\n",
       "      <td>245</td>\n",
       "      <td>305</td>\n",
       "      <td>305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_month_end</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_month_start</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_quarter_end</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_quarter_start</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_year_end</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthIs_year_start</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_active_monthElapsed</th>\n",
       "      <td>1512086400</td>\n",
       "      <td>1480550400</td>\n",
       "      <td>1472688000</td>\n",
       "      <td>1509494400</td>\n",
       "      <td>1509494400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>transactions_count</th>\n",
       "      <td>10</td>\n",
       "      <td>46</td>\n",
       "      <td>22</td>\n",
       "      <td>6</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>city_id_nunique</th>\n",
       "      <td>5</td>\n",
       "      <td>15</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_category_id_nunique</th>\n",
       "      <td>7</td>\n",
       "      <td>20</td>\n",
       "      <td>17</td>\n",
       "      <td>4</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>merchant_id_nunique</th>\n",
       "      <td>10</td>\n",
       "      <td>46</td>\n",
       "      <td>22</td>\n",
       "      <td>6</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>state_id_nunique</th>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subsector_id_nunique</th>\n",
       "      <td>7</td>\n",
       "      <td>13</td>\n",
       "      <td>11</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Year_nunique</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Month_nunique</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Week_nunique</th>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_Day_nunique</th>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "      <td>13</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum</th>\n",
       "      <td>-0.450675</td>\n",
       "      <td>-22.5638</td>\n",
       "      <td>-15.5147</td>\n",
       "      <td>-3.1597</td>\n",
       "      <td>-4.21238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum_sum</th>\n",
       "      <td>7.36008</td>\n",
       "      <td>-390.031</td>\n",
       "      <td>-322.03</td>\n",
       "      <td>-14.6139</td>\n",
       "      <td>-14.7675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum_ptp</th>\n",
       "      <td>23.4604</td>\n",
       "      <td>358.046</td>\n",
       "      <td>284.74</td>\n",
       "      <td>12.2301</td>\n",
       "      <td>1.36254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_sum_max</th>\n",
       "      <td>15.4103</td>\n",
       "      <td>-0.619333</td>\n",
       "      <td>-18.6449</td>\n",
       "      <td>-1.1919</td>\n",
       "      <td>-6.70246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max_mean</th>\n",
       "      <td>9.49366</td>\n",
       "      <td>2.61874</td>\n",
       "      <td>-0.588055</td>\n",
       "      <td>1.90307</td>\n",
       "      <td>3.17877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max_sum</th>\n",
       "      <td>18.9873</td>\n",
       "      <td>7.85621</td>\n",
       "      <td>-1.17611</td>\n",
       "      <td>3.80614</td>\n",
       "      <td>6.35754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max_ptp</th>\n",
       "      <td>14.2301</td>\n",
       "      <td>8.07011</td>\n",
       "      <td>0.129754</td>\n",
       "      <td>1.82123</td>\n",
       "      <td>4.16685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max_max</th>\n",
       "      <td>16.6087</td>\n",
       "      <td>7.45078</td>\n",
       "      <td>-0.523178</td>\n",
       "      <td>2.81368</td>\n",
       "      <td>5.26219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean_mean</th>\n",
       "      <td>-0.04237</td>\n",
       "      <td>-0.593192</td>\n",
       "      <td>-0.712977</td>\n",
       "      <td>-0.280704</td>\n",
       "      <td>-0.284128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean_sum</th>\n",
       "      <td>-0.08474</td>\n",
       "      <td>-1.77958</td>\n",
       "      <td>-1.42595</td>\n",
       "      <td>-0.561408</td>\n",
       "      <td>-0.568255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean_ptp</th>\n",
       "      <td>0.490273</td>\n",
       "      <td>0.0499709</td>\n",
       "      <td>0.00826695</td>\n",
       "      <td>0.164107</td>\n",
       "      <td>0.269552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_mean_max</th>\n",
       "      <td>0.202767</td>\n",
       "      <td>-0.569362</td>\n",
       "      <td>-0.708844</td>\n",
       "      <td>-0.19865</td>\n",
       "      <td>-0.149352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elapsed_time</th>\n",
       "      <td>426</td>\n",
       "      <td>791</td>\n",
       "      <td>882</td>\n",
       "      <td>456</td>\n",
       "      <td>456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_total</th>\n",
       "      <td>114</td>\n",
       "      <td>708</td>\n",
       "      <td>476</td>\n",
       "      <td>49</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_id_count_total</th>\n",
       "      <td>114</td>\n",
       "      <td>708</td>\n",
       "      <td>476</td>\n",
       "      <td>49</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total</th>\n",
       "      <td>6.9094</td>\n",
       "      <td>-412.594</td>\n",
       "      <td>-337.545</td>\n",
       "      <td>-17.7736</td>\n",
       "      <td>-18.9798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_total_mean</th>\n",
       "      <td>0.0257024</td>\n",
       "      <td>-1.07969</td>\n",
       "      <td>-1.41453</td>\n",
       "      <td>-0.866476</td>\n",
       "      <td>-0.474238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installments_total</th>\n",
       "      <td>124</td>\n",
       "      <td>742</td>\n",
       "      <td>0</td>\n",
       "      <td>74</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hist_first_buy</th>\n",
       "      <td>4</td>\n",
       "      <td>31</td>\n",
       "      <td>125</td>\n",
       "      <td>6</td>\n",
       "      <td>-62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>new_first_buy</th>\n",
       "      <td>90</td>\n",
       "      <td>465</td>\n",
       "      <td>551</td>\n",
       "      <td>121</td>\n",
       "      <td>140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hist_last_buy</th>\n",
       "      <td>89</td>\n",
       "      <td>454</td>\n",
       "      <td>540</td>\n",
       "      <td>117</td>\n",
       "      <td>116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>new_last_buy</th>\n",
       "      <td>132</td>\n",
       "      <td>515</td>\n",
       "      <td>598</td>\n",
       "      <td>172</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchase_amount_max_total</th>\n",
       "      <td>3.01714</td>\n",
       "      <td>0.574974</td>\n",
       "      <td>-1.31004</td>\n",
       "      <td>-0.932425</td>\n",
       "      <td>0.144826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purchased_before_issue</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>outliers</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days</th>\n",
       "      <td>62</td>\n",
       "      <td>427</td>\n",
       "      <td>518</td>\n",
       "      <td>92</td>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature1</th>\n",
       "      <td>310</td>\n",
       "      <td>1708</td>\n",
       "      <td>1554</td>\n",
       "      <td>92</td>\n",
       "      <td>368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature2</th>\n",
       "      <td>62</td>\n",
       "      <td>427</td>\n",
       "      <td>518</td>\n",
       "      <td>92</td>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_feature3</th>\n",
       "      <td>62</td>\n",
       "      <td>0</td>\n",
       "      <td>518</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rounded_target</th>\n",
       "      <td>-0</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>indexcol</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>170 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                      0                    1  \\\n",
       "first_active_month                  2017-12-01 00:00:00  2016-12-01 00:00:00   \n",
       "card_id                                 C_ID_176ecc65bd      C_ID_6082929d3a   \n",
       "feature_1                                     0.0131446            0.0107116   \n",
       "feature_2                                     0.0113848            0.0113848   \n",
       "feature_3                                     0.0114275            0.0102828   \n",
       "target                                        -0.279039            -0.926039   \n",
       "first_active_monthYear                             2017                 2016   \n",
       "first_active_monthMonth                              12                   12   \n",
       "first_active_monthWeek                               48                   48   \n",
       "first_active_monthDay                                 1                    1   \n",
       "first_active_monthDayofweek                           4                    3   \n",
       "first_active_monthDayofyear                         335                  336   \n",
       "first_active_monthIs_month_end                    False                False   \n",
       "first_active_monthIs_month_start                   True                 True   \n",
       "first_active_monthIs_quarter_end                  False                False   \n",
       "first_active_monthIs_quarter_start                False                False   \n",
       "first_active_monthIs_year_end                     False                False   \n",
       "first_active_monthIs_year_start                   False                False   \n",
       "first_active_monthElapsed                    1512086400           1480550400   \n",
       "transactions_count                                   10                   46   \n",
       "city_id_nunique                                       5                   15   \n",
       "merchant_category_id_nunique                          7                   20   \n",
       "merchant_id_nunique                                  10                   46   \n",
       "state_id_nunique                                      2                    8   \n",
       "subsector_id_nunique                                  7                   13   \n",
       "purchase_Year_nunique                                 1                    1   \n",
       "purchase_Month_nunique                                2                    2   \n",
       "purchase_Week_nunique                                 5                    9   \n",
       "purchase_Day_nunique                                  7                   17   \n",
       "purchase_amount_sum                           -0.450675             -22.5638   \n",
       "...                                                 ...                  ...   \n",
       "purchase_amount_sum_sum                         7.36008             -390.031   \n",
       "purchase_amount_sum_ptp                         23.4604              358.046   \n",
       "purchase_amount_sum_max                         15.4103            -0.619333   \n",
       "purchase_amount_max_mean                        9.49366              2.61874   \n",
       "purchase_amount_max_sum                         18.9873              7.85621   \n",
       "purchase_amount_max_ptp                         14.2301              8.07011   \n",
       "purchase_amount_max_max                         16.6087              7.45078   \n",
       "purchase_amount_mean_mean                      -0.04237            -0.593192   \n",
       "purchase_amount_mean_sum                       -0.08474             -1.77958   \n",
       "purchase_amount_mean_ptp                       0.490273            0.0499709   \n",
       "purchase_amount_mean_max                       0.202767            -0.569362   \n",
       "elapsed_time                                        426                  791   \n",
       "card_id_total                                       114                  708   \n",
       "card_id_count_total                                 114                  708   \n",
       "purchase_amount_total                            6.9094             -412.594   \n",
       "purchase_amount_total_mean                    0.0257024             -1.07969   \n",
       "installments_total                                  124                  742   \n",
       "hist_first_buy                                        4                   31   \n",
       "new_first_buy                                        90                  465   \n",
       "hist_last_buy                                        89                  454   \n",
       "new_last_buy                                        132                  515   \n",
       "purchase_amount_max_total                       3.01714             0.574974   \n",
       "purchased_before_issue                            False                False   \n",
       "outliers                                              0                    0   \n",
       "days                                                 62                  427   \n",
       "days_feature1                                       310                 1708   \n",
       "days_feature2                                        62                  427   \n",
       "days_feature3                                        62                    0   \n",
       "rounded_target                                       -0                   -1   \n",
       "indexcol                                              2                    5   \n",
       "\n",
       "                                                      2                    3  \\\n",
       "first_active_month                  2016-09-01 00:00:00  2017-11-01 00:00:00   \n",
       "card_id                                 C_ID_a87bbd85a7      C_ID_92e4809531   \n",
       "feature_1                                     0.0104794           0.00805849   \n",
       "feature_2                                     0.0113848            0.0113848   \n",
       "feature_3                                     0.0114275            0.0102828   \n",
       "target                                         0.475326             0.516026   \n",
       "first_active_monthYear                             2016                 2017   \n",
       "first_active_monthMonth                               9                   11   \n",
       "first_active_monthWeek                               35                   44   \n",
       "first_active_monthDay                                 1                    1   \n",
       "first_active_monthDayofweek                           3                    2   \n",
       "first_active_monthDayofyear                         245                  305   \n",
       "first_active_monthIs_month_end                    False                False   \n",
       "first_active_monthIs_month_start                   True                 True   \n",
       "first_active_monthIs_quarter_end                  False                False   \n",
       "first_active_monthIs_quarter_start                False                False   \n",
       "first_active_monthIs_year_end                     False                False   \n",
       "first_active_monthIs_year_start                   False                False   \n",
       "first_active_monthElapsed                    1472688000           1509494400   \n",
       "transactions_count                                   22                    6   \n",
       "city_id_nunique                                       3                    5   \n",
       "merchant_category_id_nunique                         17                    4   \n",
       "merchant_id_nunique                                  22                    6   \n",
       "state_id_nunique                                      1                    3   \n",
       "subsector_id_nunique                                 11                    4   \n",
       "purchase_Year_nunique                                 1                    1   \n",
       "purchase_Month_nunique                                2                    2   \n",
       "purchase_Week_nunique                                 5                    5   \n",
       "purchase_Day_nunique                                 13                    6   \n",
       "purchase_amount_sum                            -15.5147              -3.1597   \n",
       "...                                                 ...                  ...   \n",
       "purchase_amount_sum_sum                         -322.03             -14.6139   \n",
       "purchase_amount_sum_ptp                          284.74              12.2301   \n",
       "purchase_amount_sum_max                        -18.6449              -1.1919   \n",
       "purchase_amount_max_mean                      -0.588055              1.90307   \n",
       "purchase_amount_max_sum                        -1.17611              3.80614   \n",
       "purchase_amount_max_ptp                        0.129754              1.82123   \n",
       "purchase_amount_max_max                       -0.523178              2.81368   \n",
       "purchase_amount_mean_mean                     -0.712977            -0.280704   \n",
       "purchase_amount_mean_sum                       -1.42595            -0.561408   \n",
       "purchase_amount_mean_ptp                     0.00826695             0.164107   \n",
       "purchase_amount_mean_max                      -0.708844             -0.19865   \n",
       "elapsed_time                                        882                  456   \n",
       "card_id_total                                       476                   49   \n",
       "card_id_count_total                                 476                   49   \n",
       "purchase_amount_total                          -337.545             -17.7736   \n",
       "purchase_amount_total_mean                     -1.41453            -0.866476   \n",
       "installments_total                                    0                   74   \n",
       "hist_first_buy                                      125                    6   \n",
       "new_first_buy                                       551                  121   \n",
       "hist_last_buy                                       540                  117   \n",
       "new_last_buy                                        598                  172   \n",
       "purchase_amount_max_total                      -1.31004            -0.932425   \n",
       "purchased_before_issue                            False                False   \n",
       "outliers                                              0                    0   \n",
       "days                                                518                   92   \n",
       "days_feature1                                      1554                   92   \n",
       "days_feature2                                       518                   92   \n",
       "days_feature3                                       518                    0   \n",
       "rounded_target                                        0                    1   \n",
       "indexcol                                              5                    5   \n",
       "\n",
       "                                                      4  \n",
       "first_active_month                  2017-11-01 00:00:00  \n",
       "card_id                                 C_ID_d2dd6bbf9a  \n",
       "feature_1                                     0.0107116  \n",
       "feature_2                                     0.0113848  \n",
       "feature_3                                     0.0102828  \n",
       "target                                        -0.100927  \n",
       "first_active_monthYear                             2017  \n",
       "first_active_monthMonth                              11  \n",
       "first_active_monthWeek                               44  \n",
       "first_active_monthDay                                 1  \n",
       "first_active_monthDayofweek                           2  \n",
       "first_active_monthDayofyear                         305  \n",
       "first_active_monthIs_month_end                    False  \n",
       "first_active_monthIs_month_start                   True  \n",
       "first_active_monthIs_quarter_end                  False  \n",
       "first_active_monthIs_quarter_start                False  \n",
       "first_active_monthIs_year_end                     False  \n",
       "first_active_monthIs_year_start                   False  \n",
       "first_active_monthElapsed                    1509494400  \n",
       "transactions_count                                   16  \n",
       "city_id_nunique                                       6  \n",
       "merchant_category_id_nunique                         13  \n",
       "merchant_id_nunique                                  16  \n",
       "state_id_nunique                                      3  \n",
       "subsector_id_nunique                                  8  \n",
       "purchase_Year_nunique                                 1  \n",
       "purchase_Month_nunique                                2  \n",
       "purchase_Week_nunique                                 6  \n",
       "purchase_Day_nunique                                  9  \n",
       "purchase_amount_sum                            -4.21238  \n",
       "...                                                 ...  \n",
       "purchase_amount_sum_sum                        -14.7675  \n",
       "purchase_amount_sum_ptp                         1.36254  \n",
       "purchase_amount_sum_max                        -6.70246  \n",
       "purchase_amount_max_mean                        3.17877  \n",
       "purchase_amount_max_sum                         6.35754  \n",
       "purchase_amount_max_ptp                         4.16685  \n",
       "purchase_amount_max_max                         5.26219  \n",
       "purchase_amount_mean_mean                     -0.284128  \n",
       "purchase_amount_mean_sum                      -0.568255  \n",
       "purchase_amount_mean_ptp                       0.269552  \n",
       "purchase_amount_mean_max                      -0.149352  \n",
       "elapsed_time                                        456  \n",
       "card_id_total                                        86  \n",
       "card_id_count_total                                  86  \n",
       "purchase_amount_total                          -18.9798  \n",
       "purchase_amount_total_mean                    -0.474238  \n",
       "installments_total                                  180  \n",
       "hist_first_buy                                      -62  \n",
       "new_first_buy                                       140  \n",
       "hist_last_buy                                       116  \n",
       "new_last_buy                                        180  \n",
       "purchase_amount_max_total                      0.144826  \n",
       "purchased_before_issue                             True  \n",
       "outliers                                              0  \n",
       "days                                                 92  \n",
       "days_feature1                                       368  \n",
       "days_feature2                                        92  \n",
       "days_feature3                                         0  \n",
       "rounded_target                                       -0  \n",
       "indexcol                                              8  \n",
       "\n",
       "[170 rows x 5 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head().T"
   ]
  },
  {
   "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": "165px"
   },
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}