lightgbm_ver5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type='text/css'>\n",
       ".datatable table.frame { margin-bottom: 0; }\n",
       ".datatable table.frame thead { border-bottom: none; }\n",
       ".datatable table.frame tr.coltypes td {  color: #FFFFFF;  line-height: 6px;  padding: 0 0.5em;}\n",
       ".datatable .bool    { background: #DDDD99; }\n",
       ".datatable .object  { background: #565656; }\n",
       ".datatable .int     { background: #5D9E5D; }\n",
       ".datatable .float   { background: #4040CC; }\n",
       ".datatable .str     { background: #CC4040; }\n",
       ".datatable .time    { background: #40CC40; }\n",
       ".datatable .row_index {  background: var(--jp-border-color3);  border-right: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  font-size: 9px;}\n",
       ".datatable .frame tbody td { text-align: left; }\n",
       ".datatable .frame tr.coltypes .row_index {  background: var(--jp-border-color0);}\n",
       ".datatable th:nth-child(2) { padding-left: 12px; }\n",
       ".datatable .hellipsis {  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .vellipsis {  background: var(--jp-layout-color0);  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .na {  color: var(--jp-cell-editor-border-color);  font-size: 80%;}\n",
       ".datatable .sp {  opacity: 0.25;}\n",
       ".datatable .footer { font-size: 9px; }\n",
       ".datatable .frame_dimensions {  background: var(--jp-border-color3);  border-top: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  display: inline-block;  opacity: 0.6;  padding: 1px 10px 1px 5px;}\n",
       "</style>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "# 日本語フォントの指定\n",
    "mpl.rcParams['font.family'] = 'IPAexGothic' # ここにインストールしたフォント名を入れる\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import GroupKFold\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from nyaggle.feature.category_encoder import TargetEncoder\n",
    "\n",
    "from nyaggle.experiment import run_experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_ver1 = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/train_feature_ver1.csv')\n",
    "test_ver1 = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/test_feature_ver1.csv')\n",
    "\n",
    "train_ver2 = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/train_feature_ver2.csv')\n",
    "test_ver2 = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/test_feature_ver2.csv')\n",
    "\n",
    "era = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/era.csv')\n",
    "sample_submission = pd.read_csv('yukiCup2023Autumn-atmaCup_dataset/yukiCup 2023 Autumn x atmaCup/sample_submission.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.concat([train_ver1, train_ver2.iloc[:, 17:]], axis=1)\n",
    "test = pd.concat([test_ver1, test_ver2.iloc[:, 16:]], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['棟名', '種別2', '国', '時代', '都道府県 ※美工品は「所有者住所（所在都道府県）」', '所在地', '保管施設の名称', '所有者名', '管理団体又は責任者']\n",
      "['緯度', '経度', '開始', '終了', '緯度*経度', '終了-開始', 'label_棟名', 'label_種別2', 'label_国', 'label_時代', 'label_都道府県 ※美工品は「所有者住所（所在都道府県）」', 'label_所在地', 'label_保管施設の名称', 'label_所有者名', 'label_管理団体又は責任者', 'count_棟名', 'count_種別2', 'count_国', 'count_時代', 'count_都道府県 ※美工品は「所有者住所（所在都道府県）」', 'count_所在地', 'count_保管施設の名称', 'count_所有者名', 'count_管理団体又は責任者', '棟名_緯度_min', '棟名_緯度_max', '棟名_緯度_mean', '棟名_緯度_std', '棟名_経度_min', '棟名_経度_max', '棟名_経度_mean', '棟名_経度_std', '棟名_開始_min', '棟名_開始_max', '棟名_開始_mean', '棟名_開始_std', '棟名_終了_min', '棟名_終了_max', '棟名_終了_mean', '棟名_終了_std', '棟名_緯度*経度_min', '棟名_緯度*経度_max', '棟名_緯度*経度_mean', '棟名_緯度*経度_std', '棟名_終了-開始_min', '棟名_終了-開始_max', '棟名_終了-開始_mean', '棟名_終了-開始_std', '種別2_緯度_min', '種別2_緯度_max', '種別2_緯度_mean', '種別2_緯度_std', '種別2_経度_min', '種別2_経度_max', '種別2_経度_mean', '種別2_経度_std', '種別2_開始_min', '種別2_開始_max', '種別2_開始_mean', '種別2_開始_std', '種別2_終了_min', '種別2_終了_max', '種別2_終了_mean', '種別2_終了_std', '種別2_緯度*経度_min', '種別2_緯度*経度_max', '種別2_緯度*経度_mean', '種別2_緯度*経度_std', '種別2_終了-開始_min', '種別2_終了-開始_max', '種別2_終了-開始_mean', '種別2_終了-開始_std', '国_緯度_min', '国_緯度_max', '国_緯度_mean', '国_緯度_std', '国_経度_min', '国_経度_max', '国_経度_mean', '国_経度_std', '国_開始_min', '国_開始_max', '国_開始_mean', '国_開始_std', '国_終了_min', '国_終了_max', '国_終了_mean', '国_終了_std', '国_緯度*経度_min', '国_緯度*経度_max', '国_緯度*経度_mean', '国_緯度*経度_std', '国_終了-開始_min', '国_終了-開始_max', '国_終了-開始_mean', '国_終了-開始_std', '時代_緯度_min', '時代_緯度_max', '時代_緯度_mean', '時代_緯度_std', '時代_経度_min', '時代_経度_max', '時代_経度_mean', '時代_経度_std', '時代_開始_min', '時代_開始_max', '時代_開始_mean', '時代_開始_std', '時代_終了_min', '時代_終了_max', '時代_終了_mean', '時代_終了_std', '時代_緯度*経度_min', '時代_緯度*経度_max', '時代_緯度*経度_mean', '時代_緯度*経度_std', '時代_終了-開始_min', '時代_終了-開始_max', '時代_終了-開始_mean', '時代_終了-開始_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_経度_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_経度_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_経度_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_経度_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_開始_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_開始_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_開始_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_開始_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度*経度_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度*経度_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度*経度_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_緯度*経度_std', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了-開始_min', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了-開始_max', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了-開始_mean', '都道府県 ※美工品は「所有者住所（所在都道府県）」_終了-開始_std', '所在地_緯度_min', '所在地_緯度_max', '所在地_緯度_mean', '所在地_緯度_std', '所在地_経度_min', '所在地_経度_max', '所在地_経度_mean', '所在地_経度_std', '所在地_開始_min', '所在地_開始_max', '所在地_開始_mean', '所在地_開始_std', '所在地_終了_min', '所在地_終了_max', '所在地_終了_mean', '所在地_終了_std', '所在地_緯度*経度_min', '所在地_緯度*経度_max', '所在地_緯度*経度_mean', '所在地_緯度*経度_std', '所在地_終了-開始_min', '所在地_終了-開始_max', '所在地_終了-開始_mean', '所在地_終了-開始_std', '保管施設の名称_緯度_min', '保管施設の名称_緯度_max', '保管施設の名称_緯度_mean', '保管施設の名称_緯度_std', '保管施設の名称_経度_min', '保管施設の名称_経度_max', '保管施設の名称_経度_mean', '保管施設の名称_経度_std', '保管施設の名称_開始_min', '保管施設の名称_開始_max', '保管施設の名称_開始_mean', '保管施設の名称_開始_std', '保管施設の名称_終了_min', '保管施設の名称_終了_max', '保管施設の名称_終了_mean', '保管施設の名称_終了_std', '保管施設の名称_緯度*経度_min', '保管施設の名称_緯度*経度_max', '保管施設の名称_緯度*経度_mean', '保管施設の名称_緯度*経度_std', '保管施設の名称_終了-開始_min', '保管施設の名称_終了-開始_max', '保管施設の名称_終了-開始_mean', '保管施設の名称_終了-開始_std', '所有者名_緯度_min', '所有者名_緯度_max', '所有者名_緯度_mean', '所有者名_緯度_std', '所有者名_経度_min', '所有者名_経度_max', '所有者名_経度_mean', '所有者名_経度_std', '所有者名_開始_min', '所有者名_開始_max', '所有者名_開始_mean', '所有者名_開始_std', '所有者名_終了_min', '所有者名_終了_max', '所有者名_終了_mean', '所有者名_終了_std', '所有者名_緯度*経度_min', '所有者名_緯度*経度_max', '所有者名_緯度*経度_mean', '所有者名_緯度*経度_std', '所有者名_終了-開始_min', '所有者名_終了-開始_max', '所有者名_終了-開始_mean', '所有者名_終了-開始_std', '管理団体又は責任者_緯度_min', '管理団体又は責任者_緯度_max', '管理団体又は責任者_緯度_mean', '管理団体又は責任者_緯度_std', '管理団体又は責任者_経度_min', '管理団体又は責任者_経度_max', '管理団体又は責任者_経度_mean', '管理団体又は責任者_経度_std', '管理団体又は責任者_開始_min', '管理団体又は責任者_開始_max', '管理団体又は責任者_開始_mean', '管理団体又は責任者_開始_std', '管理団体又は責任者_終了_min', '管理団体又は責任者_終了_max', '管理団体又は責任者_終了_mean', '管理団体又は責任者_終了_std', '管理団体又は責任者_緯度*経度_min', '管理団体又は責任者_緯度*経度_max', '管理団体又は責任者_緯度*経度_mean', '管理団体又は責任者_緯度*経度_std', '管理団体又は責任者_終了-開始_min', '管理団体又は責任者_終了-開始_max', '管理団体又は責任者_終了-開始_mean', '管理団体又は責任者_終了-開始_std', '1', '2', '3', '4', '5', '6', '7', '8', '9', '10', '11', '12', '13', '14', '15', '16', '17', '18', '19', '20', '21', '22', '23', '24', '25', '26', '27', '28', '29', '30', '31', '32', '33', '34', '35', '36', '37', '38', '39', '40', '41', '42', '43', '44', '45', '46', '47', '48', '49', '50', '51', '52', '53', '54', '55', '56', '57', '58', '59', '60', '61', '62', '63', '64', '65', '66', '67', '68', '69', '70', '71', '72', '73', '74', '75', '76', '77', '78', '79', '80', '81', '82', '83', '84', '85', '86', '87', '88', '89', '90', '91', '92', '93', '94', '95', '96', '97', '98', '99', '100', '101', '102', '103', '104', '105', '106', '107', '108', '109', '110', '111', '112', '113', '114', '115', '116', '117', '118', '119', '120', '121', '122', '123', '124', '125', '126', '127', '128', '129', '130', '131', '132', '133', '134', '135', '136', '137', '138', '139', '140', '141', '142', '143', '144', '145', '146', '147', '148', '149', '150', '151', '152', '153', '154', '155', '156', '157', '158', '159', '160', '161', '162', '163', '164', '165', '166', '167', '168', '169', '170', '171', '172', '173', '174', '175', '176', '177', '178', '179', '180', '181', '182', '183', '184', '185', '186', '187', '188', '189', '190', '191', '192', '193', '194', '195', '196', '197', '198', '199', '200', '201', '202', '203', '204', '205', '206', '207', '208', '209', '210', '211', '212', '213', '214', '215', '216', '217', '218', '219', '220', '221', '222', '223', '224', '225', '226', '227', '228', '229', '230', '231', '232', '233', '234', '235', '236', '237', '238', '239', '240', '241', '242', '243', '244', '245', '246', '247', '248', '249', '250', '251', '252', '253', '254', '255', '256', '257', '258', '259', '260', '261', '262', '263', '264', '265', '266', '267', '268', '269', '270', '271', '272', '273', '274', '275', '276', '277', '278', '279', '280', '281', '282', '283', '284', '285', '286', '287', '288', '289', '290', '291', '292', '293', '294', '295', '296', '297', '298', '299', '300', '301', '302', '303', '304', '305', '306', '307', '308', '309', '310', '311', '312', '313', '314', '315', '316', '317', '318', '319', '320', '321', '322', '323', '324', '325', '326', '327', '328', '329', '330', '331', '332', '333', '334', '335', '336', '337', '338', '339', '340', '341', '342', '343', '344', '345', '346', '347', '348', '349', '350', '351', '352', '353', '354', '355', '356', '357', '358', '359', '360', '361', '362', '363', '364', '365', '366', '367', '368', '369', '370', '371', '372', '373', '374', '375', '376', '377', '378', '379', '380', '381', '382', '383', '384', '385', '386', '387', '388', '389', '390', '391', '392', '393', '394', '395', '396', '397', '398', '399', '400', '401', '402', '403', '404', '405', '406', '407', '408', '409', '410', '411', '412', '413', '414', '415', '416', '417', '418', '419', '420', '421', '422', '423', '424', '425', '426', '427', '428', '429', '430', '431', '432', '433', '434', '435', '436', '437', '438', '439', '440', '441', '442', '443', '444', '445', '446', '447', '448', '449', '450', '451', '452', '453', '454', '455', '456', '457', '458', '459', '460', '461', '462', '463', '464', '465', '466', '467', '468', '469', '470', '471', '472', '473', '474', '475', '476', '477', '478', '479', '480', '481', '482', '483', '484', '485', '486', '487', '488', '489', '490', '491', '492', '493', '494', '495', '496', '497', '498', '499', '500', '501', '502', '503', '504', '505', '506', '507', '508', '509', '510', '511']\n",
      "is_kokuhou\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:2: DeprecationWarning: `np.object` is a deprecated alias for the builtin `object`. To silence this warning, use `object` by itself. Doing this will not modify any behavior and is safe. \n",
      "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
      "  \n",
      "/opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:6: DeprecationWarning: `np.object` is a deprecated alias for the builtin `object`. To silence this warning, use `object` by itself. Doing this will not modify any behavior and is safe. \n",
      "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "#categorical変数抽出\n",
    "cat_cols = [c for c in test.columns if train[c].dtype == np.object]\n",
    "#cat_cols.remove('string') # これは取り除かないといけない\n",
    "\n",
    "# numerical変数抽出\n",
    "num_cols = [c for c in test.columns if train[c].dtype != np.object]\n",
    "\n",
    "#目的変数抽出\n",
    "target_col = 'is_kokuhou'\n",
    "\n",
    "print(cat_cols)\n",
    "print(num_cols)\n",
    "print(target_col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# target encoding\n",
    "\n",
    "# kf = KFold(4)\n",
    "# kf = KFold(4, groups=train.categorical)　#GroupKFold\n",
    "kf = StratifiedKFold(5, shuffle=True, random_state=42531) #StratifiedKFold\n",
    "\n",
    "# Target encoding with K-fold\n",
    "# te = TargetEncoder(kf.split(train))\n",
    "# te = TargetEncoder(kf.split(train)) #GroupKFold\n",
    "te = TargetEncoder(kf.split(train, train[target_col])) #StratifiedKFold\n",
    "\n",
    "# use fit/fit_transform to train data, then apply transform to test data\n",
    "train.loc[:, cat_cols] = te.fit_transform(train[cat_cols], train[target_col])\n",
    "test.loc[:, cat_cols] = te.transform(test[cat_cols])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "760\n"
     ]
    }
   ],
   "source": [
    "print(len(cat_cols) + len(num_cols))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2065: UserWarning: Using categorical_feature in Dataset.\n",
      "  _log_warning('Using categorical_feature in Dataset.')\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2068: UserWarning: categorical_feature in Dataset is overridden.\n",
      "New categorical_feature is []\n",
      "  _log_warning('categorical_feature in Dataset is overridden.\\n'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] boosting is set=dart, boosting_type=gbdt will be ignored. Current value: boosting=dart\n",
      "[1]\tvalid_0's binary_logloss: 0.259296\n",
      "[2]\tvalid_0's binary_logloss: 0.255975\n",
      "[3]\tvalid_0's binary_logloss: 0.253076\n",
      "[4]\tvalid_0's binary_logloss: 0.250717\n",
      "[5]\tvalid_0's binary_logloss: 0.248844\n",
      "[6]\tvalid_0's binary_logloss: 0.247017\n",
      "[7]\tvalid_0's binary_logloss: 0.246519\n",
      "[8]\tvalid_0's binary_logloss: 0.245276\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/callback.py:223: UserWarning: Early stopping is not available in dart mode\n",
      "  _log_warning('Early stopping is not available in dart mode')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[9]\tvalid_0's binary_logloss: 0.273209\n",
      "[10]\tvalid_0's binary_logloss: 0.270533\n",
      "[11]\tvalid_0's binary_logloss: 0.267628\n",
      "[12]\tvalid_0's binary_logloss: 0.267628\n",
      "[13]\tvalid_0's binary_logloss: 0.267748\n",
      "[14]\tvalid_0's binary_logloss: 0.264898\n",
      "[15]\tvalid_0's binary_logloss: 0.262232\n",
      "[16]\tvalid_0's binary_logloss: 0.259734\n",
      "[17]\tvalid_0's binary_logloss: 0.257811\n",
      "[18]\tvalid_0's binary_logloss: 0.255596\n",
      "[19]\tvalid_0's binary_logloss: 0.298936\n",
      "[20]\tvalid_0's binary_logloss: 0.294026\n",
      "[21]\tvalid_0's binary_logloss: 0.294945\n",
      "[22]\tvalid_0's binary_logloss: 0.289949\n",
      "[23]\tvalid_0's binary_logloss: 0.285357\n",
      "[24]\tvalid_0's binary_logloss: 0.285196\n",
      "[25]\tvalid_0's binary_logloss: 0.28094\n",
      "[26]\tvalid_0's binary_logloss: 0.277437\n",
      "[27]\tvalid_0's binary_logloss: 0.273812\n",
      "[28]\tvalid_0's binary_logloss: 0.270148\n",
      "[29]\tvalid_0's binary_logloss: 0.267143\n",
      "[30]\tvalid_0's binary_logloss: 0.264078\n",
      "[31]\tvalid_0's binary_logloss: 0.261156\n",
      "[32]\tvalid_0's binary_logloss: 0.260166\n",
      "[33]\tvalid_0's binary_logloss: 0.261018\n",
      "[34]\tvalid_0's binary_logloss: 0.260705\n",
      "[35]\tvalid_0's binary_logloss: 0.26133\n",
      "[36]\tvalid_0's binary_logloss: 0.258393\n",
      "[37]\tvalid_0's binary_logloss: 0.258064\n",
      "[38]\tvalid_0's binary_logloss: 0.255814\n",
      "[39]\tvalid_0's binary_logloss: 0.256629\n",
      "[40]\tvalid_0's binary_logloss: 0.257528\n",
      "[41]\tvalid_0's binary_logloss: 0.257911\n",
      "[42]\tvalid_0's binary_logloss: 0.271969\n",
      "[43]\tvalid_0's binary_logloss: 0.268728\n",
      "[44]\tvalid_0's binary_logloss: 0.265574\n",
      "[45]\tvalid_0's binary_logloss: 0.262798\n",
      "[46]\tvalid_0's binary_logloss: 0.260658\n",
      "[47]\tvalid_0's binary_logloss: 0.261571\n",
      "[48]\tvalid_0's binary_logloss: 0.261658\n",
      "[49]\tvalid_0's binary_logloss: 0.262994\n",
      "[50]\tvalid_0's binary_logloss: 0.263605\n",
      "[51]\tvalid_0's binary_logloss: 0.264771\n",
      "[52]\tvalid_0's binary_logloss: 0.266233\n",
      "[53]\tvalid_0's binary_logloss: 0.265801\n",
      "[54]\tvalid_0's binary_logloss: 0.262548\n",
      "[55]\tvalid_0's binary_logloss: 0.259975\n",
      "[56]\tvalid_0's binary_logloss: 0.2576\n",
      "[57]\tvalid_0's binary_logloss: 0.255336\n",
      "[58]\tvalid_0's binary_logloss: 0.253508\n",
      "[59]\tvalid_0's binary_logloss: 0.254326\n",
      "[60]\tvalid_0's binary_logloss: 0.255228\n",
      "[61]\tvalid_0's binary_logloss: 0.256185\n",
      "[62]\tvalid_0's binary_logloss: 0.256773\n",
      "[63]\tvalid_0's binary_logloss: 0.257134\n",
      "[64]\tvalid_0's binary_logloss: 0.25469\n",
      "[65]\tvalid_0's binary_logloss: 0.252503\n",
      "[66]\tvalid_0's binary_logloss: 0.250356\n",
      "[67]\tvalid_0's binary_logloss: 0.248731\n",
      "[68]\tvalid_0's binary_logloss: 0.247151\n",
      "[69]\tvalid_0's binary_logloss: 0.245927\n",
      "[70]\tvalid_0's binary_logloss: 0.246618\n",
      "[71]\tvalid_0's binary_logloss: 0.246942\n",
      "[72]\tvalid_0's binary_logloss: 0.247552\n",
      "[73]\tvalid_0's binary_logloss: 0.246042\n",
      "[74]\tvalid_0's binary_logloss: 0.244495\n",
      "[75]\tvalid_0's binary_logloss: 0.243017\n",
      "[76]\tvalid_0's binary_logloss: 0.243283\n",
      "[77]\tvalid_0's binary_logloss: 0.243977\n",
      "[78]\tvalid_0's binary_logloss: 0.24248\n",
      "[79]\tvalid_0's binary_logloss: 0.243052\n",
      "[80]\tvalid_0's binary_logloss: 0.242051\n",
      "[81]\tvalid_0's binary_logloss: 0.242718\n",
      "[82]\tvalid_0's binary_logloss: 0.243225\n",
      "[83]\tvalid_0's binary_logloss: 0.243752\n",
      "[84]\tvalid_0's binary_logloss: 0.242574\n",
      "[85]\tvalid_0's binary_logloss: 0.241282\n",
      "[86]\tvalid_0's binary_logloss: 0.240387\n",
      "[87]\tvalid_0's binary_logloss: 0.239313\n",
      "[88]\tvalid_0's binary_logloss: 0.238347\n",
      "[89]\tvalid_0's binary_logloss: 0.238847\n",
      "[90]\tvalid_0's binary_logloss: 0.239054\n",
      "[91]\tvalid_0's binary_logloss: 0.237766\n",
      "[92]\tvalid_0's binary_logloss: 0.237295\n",
      "[93]\tvalid_0's binary_logloss: 0.236368\n",
      "[94]\tvalid_0's binary_logloss: 0.238253\n",
      "[95]\tvalid_0's binary_logloss: 0.238882\n",
      "[96]\tvalid_0's binary_logloss: 0.237664\n",
      "[97]\tvalid_0's binary_logloss: 0.23964\n",
      "[98]\tvalid_0's binary_logloss: 0.23815\n",
      "[99]\tvalid_0's binary_logloss: 0.237568\n",
      "[100]\tvalid_0's binary_logloss: 0.23777\n",
      "[101]\tvalid_0's binary_logloss: 0.238219\n",
      "[102]\tvalid_0's binary_logloss: 0.23846\n",
      "[103]\tvalid_0's binary_logloss: 0.237333\n",
      "[104]\tvalid_0's binary_logloss: 0.237773\n",
      "[105]\tvalid_0's binary_logloss: 0.23674\n",
      "[106]\tvalid_0's binary_logloss: 0.235818\n",
      "[107]\tvalid_0's binary_logloss: 0.236196\n",
      "[108]\tvalid_0's binary_logloss: 0.23677\n",
      "[109]\tvalid_0's binary_logloss: 0.235756\n",
      "[110]\tvalid_0's binary_logloss: 0.236196\n",
      "[111]\tvalid_0's binary_logloss: 0.235451\n",
      "[112]\tvalid_0's binary_logloss: 0.234604\n",
      "[113]\tvalid_0's binary_logloss: 0.233951\n",
      "[114]\tvalid_0's binary_logloss: 0.233333\n",
      "[115]\tvalid_0's binary_logloss: 0.232942\n",
      "[116]\tvalid_0's binary_logloss: 0.232433\n",
      "[117]\tvalid_0's binary_logloss: 0.232039\n",
      "[118]\tvalid_0's binary_logloss: 0.232312\n",
      "[119]\tvalid_0's binary_logloss: 0.232563\n",
      "[120]\tvalid_0's binary_logloss: 0.232314\n",
      "[121]\tvalid_0's binary_logloss: 0.232481\n",
      "[122]\tvalid_0's binary_logloss: 0.231817\n",
      "[123]\tvalid_0's binary_logloss: 0.231743\n",
      "[124]\tvalid_0's binary_logloss: 0.231856\n",
      "[125]\tvalid_0's binary_logloss: 0.231914\n",
      "[126]\tvalid_0's binary_logloss: 0.231545\n",
      "[127]\tvalid_0's binary_logloss: 0.231239\n",
      "[128]\tvalid_0's binary_logloss: 0.23137\n",
      "[129]\tvalid_0's binary_logloss: 0.231611\n",
      "[130]\tvalid_0's binary_logloss: 0.231802\n",
      "[131]\tvalid_0's binary_logloss: 0.231626\n",
      "[132]\tvalid_0's binary_logloss: 0.23164\n",
      "[133]\tvalid_0's binary_logloss: 0.231347\n",
      "[134]\tvalid_0's binary_logloss: 0.231456\n",
      "[135]\tvalid_0's binary_logloss: 0.231521\n",
      "[136]\tvalid_0's binary_logloss: 0.231208\n",
      "[137]\tvalid_0's binary_logloss: 0.230882\n",
      "[138]\tvalid_0's binary_logloss: 0.230615\n",
      "[139]\tvalid_0's binary_logloss: 0.230364\n",
      "[140]\tvalid_0's binary_logloss: 0.230356\n",
      "[141]\tvalid_0's binary_logloss: 0.229917\n",
      "[142]\tvalid_0's binary_logloss: 0.229982\n",
      "[143]\tvalid_0's binary_logloss: 0.229432\n",
      "[144]\tvalid_0's binary_logloss: 0.228964\n",
      "[145]\tvalid_0's binary_logloss: 0.229008\n",
      "[146]\tvalid_0's binary_logloss: 0.228878\n",
      "[147]\tvalid_0's binary_logloss: 0.228798\n",
      "[148]\tvalid_0's binary_logloss: 0.229007\n",
      "[149]\tvalid_0's binary_logloss: 0.22913\n",
      "[150]\tvalid_0's binary_logloss: 0.229238\n",
      "[151]\tvalid_0's binary_logloss: 0.22924\n",
      "[152]\tvalid_0's binary_logloss: 0.229344\n",
      "[153]\tvalid_0's binary_logloss: 0.22943\n",
      "[154]\tvalid_0's binary_logloss: 0.229213\n",
      "[155]\tvalid_0's binary_logloss: 0.228872\n",
      "[156]\tvalid_0's binary_logloss: 0.228905\n",
      "[157]\tvalid_0's binary_logloss: 0.228949\n",
      "[158]\tvalid_0's binary_logloss: 0.229237\n",
      "[159]\tvalid_0's binary_logloss: 0.228991\n",
      "[160]\tvalid_0's binary_logloss: 0.228782\n",
      "[161]\tvalid_0's binary_logloss: 0.228643\n",
      "[162]\tvalid_0's binary_logloss: 0.228671\n",
      "[163]\tvalid_0's binary_logloss: 0.228642\n",
      "[164]\tvalid_0's binary_logloss: 0.228666\n",
      "[165]\tvalid_0's binary_logloss: 0.22851\n",
      "[166]\tvalid_0's binary_logloss: 0.228499\n",
      "[167]\tvalid_0's binary_logloss: 0.228517\n",
      "[168]\tvalid_0's binary_logloss: 0.228444\n",
      "[169]\tvalid_0's binary_logloss: 0.22807\n",
      "[170]\tvalid_0's binary_logloss: 0.228052\n",
      "[171]\tvalid_0's binary_logloss: 0.227701\n",
      "[172]\tvalid_0's binary_logloss: 0.227585\n",
      "[173]\tvalid_0's binary_logloss: 0.227522\n",
      "[174]\tvalid_0's binary_logloss: 0.227508\n",
      "[175]\tvalid_0's binary_logloss: 0.227236\n",
      "[176]\tvalid_0's binary_logloss: 0.227211\n",
      "[177]\tvalid_0's binary_logloss: 0.227184\n",
      "[178]\tvalid_0's binary_logloss: 0.227357\n",
      "[179]\tvalid_0's binary_logloss: 0.227578\n",
      "[180]\tvalid_0's binary_logloss: 0.227542\n",
      "[181]\tvalid_0's binary_logloss: 0.227123\n",
      "[182]\tvalid_0's binary_logloss: 0.226971\n",
      "[183]\tvalid_0's binary_logloss: 0.226915\n",
      "[184]\tvalid_0's binary_logloss: 0.226984\n",
      "[185]\tvalid_0's binary_logloss: 0.227219\n",
      "[186]\tvalid_0's binary_logloss: 0.22715\n",
      "[187]\tvalid_0's binary_logloss: 0.227218\n",
      "[188]\tvalid_0's binary_logloss: 0.22714\n",
      "[189]\tvalid_0's binary_logloss: 0.226997\n",
      "[190]\tvalid_0's binary_logloss: 0.227125\n",
      "[191]\tvalid_0's binary_logloss: 0.226979\n",
      "[192]\tvalid_0's binary_logloss: 0.227117\n",
      "[193]\tvalid_0's binary_logloss: 0.227351\n",
      "[194]\tvalid_0's binary_logloss: 0.227442\n",
      "[195]\tvalid_0's binary_logloss: 0.227387\n",
      "[196]\tvalid_0's binary_logloss: 0.22759\n",
      "[197]\tvalid_0's binary_logloss: 0.227932\n",
      "[198]\tvalid_0's binary_logloss: 0.227865\n",
      "[199]\tvalid_0's binary_logloss: 0.227956\n",
      "[200]\tvalid_0's binary_logloss: 0.227807\n",
      "[201]\tvalid_0's binary_logloss: 0.227982\n",
      "[202]\tvalid_0's binary_logloss: 0.228106\n",
      "[203]\tvalid_0's binary_logloss: 0.227894\n",
      "[204]\tvalid_0's binary_logloss: 0.227881\n",
      "[205]\tvalid_0's binary_logloss: 0.227751\n",
      "[206]\tvalid_0's binary_logloss: 0.227616\n",
      "[207]\tvalid_0's binary_logloss: 0.227701\n",
      "[208]\tvalid_0's binary_logloss: 0.227496\n",
      "[209]\tvalid_0's binary_logloss: 0.227717\n",
      "[210]\tvalid_0's binary_logloss: 0.227279\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2065: UserWarning: Using categorical_feature in Dataset.\n",
      "  _log_warning('Using categorical_feature in Dataset.')\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2068: UserWarning: categorical_feature in Dataset is overridden.\n",
      "New categorical_feature is []\n",
      "  _log_warning('categorical_feature in Dataset is overridden.\\n'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] boosting is set=dart, boosting_type=gbdt will be ignored. Current value: boosting=dart\n",
      "[1]\tvalid_0's binary_logloss: 0.259477\n",
      "[2]\tvalid_0's binary_logloss: 0.255951\n",
      "[3]\tvalid_0's binary_logloss: 0.251897\n",
      "[4]\tvalid_0's binary_logloss: 0.249026\n",
      "[5]\tvalid_0's binary_logloss: 0.246278\n",
      "[6]\tvalid_0's binary_logloss: 0.244507\n",
      "[7]\tvalid_0's binary_logloss: 0.244777\n",
      "[8]\tvalid_0's binary_logloss: 0.242741\n",
      "[9]\tvalid_0's binary_logloss: 0.273295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/callback.py:223: UserWarning: Early stopping is not available in dart mode\n",
      "  _log_warning('Early stopping is not available in dart mode')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10]\tvalid_0's binary_logloss: 0.270145\n",
      "[11]\tvalid_0's binary_logloss: 0.267005\n",
      "[12]\tvalid_0's binary_logloss: 0.267005\n",
      "[13]\tvalid_0's binary_logloss: 0.267233\n",
      "[14]\tvalid_0's binary_logloss: 0.264974\n",
      "[15]\tvalid_0's binary_logloss: 0.262391\n",
      "[16]\tvalid_0's binary_logloss: 0.259674\n",
      "[17]\tvalid_0's binary_logloss: 0.257125\n",
      "[18]\tvalid_0's binary_logloss: 0.255114\n",
      "[19]\tvalid_0's binary_logloss: 0.301665\n",
      "[20]\tvalid_0's binary_logloss: 0.296916\n",
      "[21]\tvalid_0's binary_logloss: 0.297934\n",
      "[22]\tvalid_0's binary_logloss: 0.293299\n",
      "[23]\tvalid_0's binary_logloss: 0.289066\n",
      "[24]\tvalid_0's binary_logloss: 0.288557\n",
      "[25]\tvalid_0's binary_logloss: 0.284537\n",
      "[26]\tvalid_0's binary_logloss: 0.280458\n",
      "[27]\tvalid_0's binary_logloss: 0.276733\n",
      "[28]\tvalid_0's binary_logloss: 0.273205\n",
      "[29]\tvalid_0's binary_logloss: 0.270087\n",
      "[30]\tvalid_0's binary_logloss: 0.267092\n",
      "[31]\tvalid_0's binary_logloss: 0.264466\n",
      "[32]\tvalid_0's binary_logloss: 0.263384\n",
      "[33]\tvalid_0's binary_logloss: 0.264298\n",
      "[34]\tvalid_0's binary_logloss: 0.264109\n",
      "[35]\tvalid_0's binary_logloss: 0.264734\n",
      "[36]\tvalid_0's binary_logloss: 0.262052\n",
      "[37]\tvalid_0's binary_logloss: 0.261785\n",
      "[38]\tvalid_0's binary_logloss: 0.259745\n",
      "[39]\tvalid_0's binary_logloss: 0.260472\n",
      "[40]\tvalid_0's binary_logloss: 0.261295\n",
      "[41]\tvalid_0's binary_logloss: 0.261681\n",
      "[42]\tvalid_0's binary_logloss: 0.276933\n",
      "[43]\tvalid_0's binary_logloss: 0.273663\n",
      "[44]\tvalid_0's binary_logloss: 0.270527\n",
      "[45]\tvalid_0's binary_logloss: 0.267377\n",
      "[46]\tvalid_0's binary_logloss: 0.264894\n",
      "[47]\tvalid_0's binary_logloss: 0.265835\n",
      "[48]\tvalid_0's binary_logloss: 0.266016\n",
      "[49]\tvalid_0's binary_logloss: 0.267333\n",
      "[50]\tvalid_0's binary_logloss: 0.268064\n",
      "[51]\tvalid_0's binary_logloss: 0.269285\n",
      "[52]\tvalid_0's binary_logloss: 0.270552\n",
      "[53]\tvalid_0's binary_logloss: 0.270114\n",
      "[54]\tvalid_0's binary_logloss: 0.26739\n",
      "[55]\tvalid_0's binary_logloss: 0.264362\n",
      "[56]\tvalid_0's binary_logloss: 0.261763\n",
      "[57]\tvalid_0's binary_logloss: 0.259525\n",
      "[58]\tvalid_0's binary_logloss: 0.257443\n",
      "[59]\tvalid_0's binary_logloss: 0.258278\n",
      "[60]\tvalid_0's binary_logloss: 0.2593\n",
      "[61]\tvalid_0's binary_logloss: 0.260308\n",
      "[62]\tvalid_0's binary_logloss: 0.261044\n",
      "[63]\tvalid_0's binary_logloss: 0.261311\n",
      "[64]\tvalid_0's binary_logloss: 0.258979\n",
      "[65]\tvalid_0's binary_logloss: 0.25691\n",
      "[66]\tvalid_0's binary_logloss: 0.255088\n",
      "[67]\tvalid_0's binary_logloss: 0.253159\n",
      "[68]\tvalid_0's binary_logloss: 0.251268\n",
      "[69]\tvalid_0's binary_logloss: 0.249638\n",
      "[70]\tvalid_0's binary_logloss: 0.250357\n",
      "[71]\tvalid_0's binary_logloss: 0.250711\n",
      "[72]\tvalid_0's binary_logloss: 0.251376\n",
      "[73]\tvalid_0's binary_logloss: 0.249894\n",
      "[74]\tvalid_0's binary_logloss: 0.248217\n",
      "[75]\tvalid_0's binary_logloss: 0.246881\n",
      "[76]\tvalid_0's binary_logloss: 0.247202\n",
      "[77]\tvalid_0's binary_logloss: 0.247894\n",
      "[78]\tvalid_0's binary_logloss: 0.246373\n",
      "[79]\tvalid_0's binary_logloss: 0.246971\n",
      "[80]\tvalid_0's binary_logloss: 0.245688\n",
      "[81]\tvalid_0's binary_logloss: 0.24635\n",
      "[82]\tvalid_0's binary_logloss: 0.246942\n",
      "[83]\tvalid_0's binary_logloss: 0.24754\n",
      "[84]\tvalid_0's binary_logloss: 0.245976\n",
      "[85]\tvalid_0's binary_logloss: 0.244763\n",
      "[86]\tvalid_0's binary_logloss: 0.243105\n",
      "[87]\tvalid_0's binary_logloss: 0.241688\n",
      "[88]\tvalid_0's binary_logloss: 0.240722\n",
      "[89]\tvalid_0's binary_logloss: 0.241294\n",
      "[90]\tvalid_0's binary_logloss: 0.241643\n",
      "[91]\tvalid_0's binary_logloss: 0.240626\n",
      "[92]\tvalid_0's binary_logloss: 0.239185\n",
      "[93]\tvalid_0's binary_logloss: 0.238415\n",
      "[94]\tvalid_0's binary_logloss: 0.240758\n",
      "[95]\tvalid_0's binary_logloss: 0.241319\n",
      "[96]\tvalid_0's binary_logloss: 0.240349\n",
      "[97]\tvalid_0's binary_logloss: 0.242685\n",
      "[98]\tvalid_0's binary_logloss: 0.24142\n",
      "[99]\tvalid_0's binary_logloss: 0.240219\n",
      "[100]\tvalid_0's binary_logloss: 0.240458\n",
      "[101]\tvalid_0's binary_logloss: 0.241087\n",
      "[102]\tvalid_0's binary_logloss: 0.241575\n",
      "[103]\tvalid_0's binary_logloss: 0.240597\n",
      "[104]\tvalid_0's binary_logloss: 0.241127\n",
      "[105]\tvalid_0's binary_logloss: 0.240304\n",
      "[106]\tvalid_0's binary_logloss: 0.239321\n",
      "[107]\tvalid_0's binary_logloss: 0.23983\n",
      "[108]\tvalid_0's binary_logloss: 0.240493\n",
      "[109]\tvalid_0's binary_logloss: 0.239746\n",
      "[110]\tvalid_0's binary_logloss: 0.240078\n",
      "[111]\tvalid_0's binary_logloss: 0.238977\n",
      "[112]\tvalid_0's binary_logloss: 0.238182\n",
      "[113]\tvalid_0's binary_logloss: 0.237459\n",
      "[114]\tvalid_0's binary_logloss: 0.23646\n",
      "[115]\tvalid_0's binary_logloss: 0.235795\n",
      "[116]\tvalid_0's binary_logloss: 0.235439\n",
      "[117]\tvalid_0's binary_logloss: 0.23499\n",
      "[118]\tvalid_0's binary_logloss: 0.235227\n",
      "[119]\tvalid_0's binary_logloss: 0.235434\n",
      "[120]\tvalid_0's binary_logloss: 0.234739\n",
      "[121]\tvalid_0's binary_logloss: 0.234929\n",
      "[122]\tvalid_0's binary_logloss: 0.234355\n",
      "[123]\tvalid_0's binary_logloss: 0.234179\n",
      "[124]\tvalid_0's binary_logloss: 0.23439\n",
      "[125]\tvalid_0's binary_logloss: 0.234564\n",
      "[126]\tvalid_0's binary_logloss: 0.234114\n",
      "[127]\tvalid_0's binary_logloss: 0.233881\n",
      "[128]\tvalid_0's binary_logloss: 0.233961\n",
      "[129]\tvalid_0's binary_logloss: 0.234161\n",
      "[130]\tvalid_0's binary_logloss: 0.234371\n",
      "[131]\tvalid_0's binary_logloss: 0.234152\n",
      "[132]\tvalid_0's binary_logloss: 0.233726\n",
      "[133]\tvalid_0's binary_logloss: 0.233207\n",
      "[134]\tvalid_0's binary_logloss: 0.233382\n",
      "[135]\tvalid_0's binary_logloss: 0.23355\n",
      "[136]\tvalid_0's binary_logloss: 0.233335\n",
      "[137]\tvalid_0's binary_logloss: 0.233139\n",
      "[138]\tvalid_0's binary_logloss: 0.232747\n",
      "[139]\tvalid_0's binary_logloss: 0.232739\n",
      "[140]\tvalid_0's binary_logloss: 0.232753\n",
      "[141]\tvalid_0's binary_logloss: 0.232622\n",
      "[142]\tvalid_0's binary_logloss: 0.232675\n",
      "[143]\tvalid_0's binary_logloss: 0.232627\n",
      "[144]\tvalid_0's binary_logloss: 0.232396\n",
      "[145]\tvalid_0's binary_logloss: 0.232425\n",
      "[146]\tvalid_0's binary_logloss: 0.232467\n",
      "[147]\tvalid_0's binary_logloss: 0.232349\n",
      "[148]\tvalid_0's binary_logloss: 0.232124\n",
      "[149]\tvalid_0's binary_logloss: 0.232523\n",
      "[150]\tvalid_0's binary_logloss: 0.232583\n",
      "[151]\tvalid_0's binary_logloss: 0.232613\n",
      "[152]\tvalid_0's binary_logloss: 0.232617\n",
      "[153]\tvalid_0's binary_logloss: 0.23276\n",
      "[154]\tvalid_0's binary_logloss: 0.2326\n",
      "[155]\tvalid_0's binary_logloss: 0.232834\n",
      "[156]\tvalid_0's binary_logloss: 0.232912\n",
      "[157]\tvalid_0's binary_logloss: 0.232972\n",
      "[158]\tvalid_0's binary_logloss: 0.232547\n",
      "[159]\tvalid_0's binary_logloss: 0.232361\n",
      "[160]\tvalid_0's binary_logloss: 0.232326\n",
      "[161]\tvalid_0's binary_logloss: 0.232161\n",
      "[162]\tvalid_0's binary_logloss: 0.232163\n",
      "[163]\tvalid_0's binary_logloss: 0.232224\n",
      "[164]\tvalid_0's binary_logloss: 0.23225\n",
      "[165]\tvalid_0's binary_logloss: 0.232053\n",
      "[166]\tvalid_0's binary_logloss: 0.232053\n",
      "[167]\tvalid_0's binary_logloss: 0.232077\n",
      "[168]\tvalid_0's binary_logloss: 0.231965\n",
      "[169]\tvalid_0's binary_logloss: 0.231722\n",
      "[170]\tvalid_0's binary_logloss: 0.23182\n",
      "[171]\tvalid_0's binary_logloss: 0.231658\n",
      "[172]\tvalid_0's binary_logloss: 0.231622\n",
      "[173]\tvalid_0's binary_logloss: 0.231549\n",
      "[174]\tvalid_0's binary_logloss: 0.231508\n",
      "[175]\tvalid_0's binary_logloss: 0.231344\n",
      "[176]\tvalid_0's binary_logloss: 0.231427\n",
      "[177]\tvalid_0's binary_logloss: 0.231371\n",
      "[178]\tvalid_0's binary_logloss: 0.231474\n",
      "[179]\tvalid_0's binary_logloss: 0.231208\n",
      "[180]\tvalid_0's binary_logloss: 0.231144\n",
      "[181]\tvalid_0's binary_logloss: 0.231155\n",
      "[182]\tvalid_0's binary_logloss: 0.231316\n",
      "[183]\tvalid_0's binary_logloss: 0.231285\n",
      "[184]\tvalid_0's binary_logloss: 0.231474\n",
      "[185]\tvalid_0's binary_logloss: 0.23142\n",
      "[186]\tvalid_0's binary_logloss: 0.231329\n",
      "[187]\tvalid_0's binary_logloss: 0.23142\n",
      "[188]\tvalid_0's binary_logloss: 0.231187\n",
      "[189]\tvalid_0's binary_logloss: 0.231251\n",
      "[190]\tvalid_0's binary_logloss: 0.231028\n",
      "[191]\tvalid_0's binary_logloss: 0.231291\n",
      "[192]\tvalid_0's binary_logloss: 0.231279\n",
      "[193]\tvalid_0's binary_logloss: 0.231261\n",
      "[194]\tvalid_0's binary_logloss: 0.231384\n",
      "[195]\tvalid_0's binary_logloss: 0.231344\n",
      "[196]\tvalid_0's binary_logloss: 0.231333\n",
      "[197]\tvalid_0's binary_logloss: 0.231443\n",
      "[198]\tvalid_0's binary_logloss: 0.231093\n",
      "[199]\tvalid_0's binary_logloss: 0.231332\n",
      "[200]\tvalid_0's binary_logloss: 0.230987\n",
      "[201]\tvalid_0's binary_logloss: 0.231366\n",
      "[202]\tvalid_0's binary_logloss: 0.231601\n",
      "[203]\tvalid_0's binary_logloss: 0.231481\n",
      "[204]\tvalid_0's binary_logloss: 0.23179\n",
      "[205]\tvalid_0's binary_logloss: 0.231623\n",
      "[206]\tvalid_0's binary_logloss: 0.231467\n",
      "[207]\tvalid_0's binary_logloss: 0.231893\n",
      "[208]\tvalid_0's binary_logloss: 0.231769\n",
      "[209]\tvalid_0's binary_logloss: 0.231746\n",
      "[210]\tvalid_0's binary_logloss: 0.231594\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2065: UserWarning: Using categorical_feature in Dataset.\n",
      "  _log_warning('Using categorical_feature in Dataset.')\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2068: UserWarning: categorical_feature in Dataset is overridden.\n",
      "New categorical_feature is []\n",
      "  _log_warning('categorical_feature in Dataset is overridden.\\n'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] boosting is set=dart, boosting_type=gbdt will be ignored. Current value: boosting=dart\n",
      "[1]\tvalid_0's binary_logloss: 0.259138\n",
      "[2]\tvalid_0's binary_logloss: 0.253862\n",
      "[3]\tvalid_0's binary_logloss: 0.2492\n",
      "[4]\tvalid_0's binary_logloss: 0.245896\n",
      "[5]\tvalid_0's binary_logloss: 0.242315\n",
      "[6]\tvalid_0's binary_logloss: 0.239222\n",
      "[7]\tvalid_0's binary_logloss: 0.239446\n",
      "[8]\tvalid_0's binary_logloss: 0.236605\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/callback.py:223: UserWarning: Early stopping is not available in dart mode\n",
      "  _log_warning('Early stopping is not available in dart mode')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[9]\tvalid_0's binary_logloss: 0.267867\n",
      "[10]\tvalid_0's binary_logloss: 0.264161\n",
      "[11]\tvalid_0's binary_logloss: 0.260502\n",
      "[12]\tvalid_0's binary_logloss: 0.260502\n",
      "[13]\tvalid_0's binary_logloss: 0.261075\n",
      "[14]\tvalid_0's binary_logloss: 0.257547\n",
      "[15]\tvalid_0's binary_logloss: 0.254404\n",
      "[16]\tvalid_0's binary_logloss: 0.251515\n",
      "[17]\tvalid_0's binary_logloss: 0.248342\n",
      "[18]\tvalid_0's binary_logloss: 0.245673\n",
      "[19]\tvalid_0's binary_logloss: 0.292892\n",
      "[20]\tvalid_0's binary_logloss: 0.287156\n",
      "[21]\tvalid_0's binary_logloss: 0.288394\n",
      "[22]\tvalid_0's binary_logloss: 0.282708\n",
      "[23]\tvalid_0's binary_logloss: 0.277663\n",
      "[24]\tvalid_0's binary_logloss: 0.277371\n",
      "[25]\tvalid_0's binary_logloss: 0.27281\n",
      "[26]\tvalid_0's binary_logloss: 0.26831\n",
      "[27]\tvalid_0's binary_logloss: 0.264017\n",
      "[28]\tvalid_0's binary_logloss: 0.260426\n",
      "[29]\tvalid_0's binary_logloss: 0.256901\n",
      "[30]\tvalid_0's binary_logloss: 0.253395\n",
      "[31]\tvalid_0's binary_logloss: 0.250237\n",
      "[32]\tvalid_0's binary_logloss: 0.249076\n",
      "[33]\tvalid_0's binary_logloss: 0.250347\n",
      "[34]\tvalid_0's binary_logloss: 0.250356\n",
      "[35]\tvalid_0's binary_logloss: 0.251436\n",
      "[36]\tvalid_0's binary_logloss: 0.248108\n",
      "[37]\tvalid_0's binary_logloss: 0.247638\n",
      "[38]\tvalid_0's binary_logloss: 0.244709\n",
      "[39]\tvalid_0's binary_logloss: 0.245922\n",
      "[40]\tvalid_0's binary_logloss: 0.247115\n",
      "[41]\tvalid_0's binary_logloss: 0.24809\n",
      "[42]\tvalid_0's binary_logloss: 0.263907\n",
      "[43]\tvalid_0's binary_logloss: 0.259932\n",
      "[44]\tvalid_0's binary_logloss: 0.256457\n",
      "[45]\tvalid_0's binary_logloss: 0.252892\n",
      "[46]\tvalid_0's binary_logloss: 0.249641\n",
      "[47]\tvalid_0's binary_logloss: 0.250952\n",
      "[48]\tvalid_0's binary_logloss: 0.251183\n",
      "[49]\tvalid_0's binary_logloss: 0.252851\n",
      "[50]\tvalid_0's binary_logloss: 0.253784\n",
      "[51]\tvalid_0's binary_logloss: 0.255296\n",
      "[52]\tvalid_0's binary_logloss: 0.257067\n",
      "[53]\tvalid_0's binary_logloss: 0.25642\n",
      "[54]\tvalid_0's binary_logloss: 0.252889\n",
      "[55]\tvalid_0's binary_logloss: 0.249757\n",
      "[56]\tvalid_0's binary_logloss: 0.246706\n",
      "[57]\tvalid_0's binary_logloss: 0.243637\n",
      "[58]\tvalid_0's binary_logloss: 0.241357\n",
      "[59]\tvalid_0's binary_logloss: 0.242455\n",
      "[60]\tvalid_0's binary_logloss: 0.243815\n",
      "[61]\tvalid_0's binary_logloss: 0.245059\n",
      "[62]\tvalid_0's binary_logloss: 0.245963\n",
      "[63]\tvalid_0's binary_logloss: 0.246424\n",
      "[64]\tvalid_0's binary_logloss: 0.243419\n",
      "[65]\tvalid_0's binary_logloss: 0.240818\n",
      "[66]\tvalid_0's binary_logloss: 0.238148\n",
      "[67]\tvalid_0's binary_logloss: 0.236017\n",
      "[68]\tvalid_0's binary_logloss: 0.234325\n",
      "[69]\tvalid_0's binary_logloss: 0.232742\n",
      "[70]\tvalid_0's binary_logloss: 0.233595\n",
      "[71]\tvalid_0's binary_logloss: 0.23417\n",
      "[72]\tvalid_0's binary_logloss: 0.235127\n",
      "[73]\tvalid_0's binary_logloss: 0.233288\n",
      "[74]\tvalid_0's binary_logloss: 0.23173\n",
      "[75]\tvalid_0's binary_logloss: 0.230117\n",
      "[76]\tvalid_0's binary_logloss: 0.230676\n",
      "[77]\tvalid_0's binary_logloss: 0.231854\n",
      "[78]\tvalid_0's binary_logloss: 0.230392\n",
      "[79]\tvalid_0's binary_logloss: 0.231133\n",
      "[80]\tvalid_0's binary_logloss: 0.229322\n",
      "[81]\tvalid_0's binary_logloss: 0.230259\n",
      "[82]\tvalid_0's binary_logloss: 0.230983\n",
      "[83]\tvalid_0's binary_logloss: 0.231834\n",
      "[84]\tvalid_0's binary_logloss: 0.230247\n",
      "[85]\tvalid_0's binary_logloss: 0.228829\n",
      "[86]\tvalid_0's binary_logloss: 0.227829\n",
      "[87]\tvalid_0's binary_logloss: 0.226597\n",
      "[88]\tvalid_0's binary_logloss: 0.225486\n",
      "[89]\tvalid_0's binary_logloss: 0.2262\n",
      "[90]\tvalid_0's binary_logloss: 0.226752\n",
      "[91]\tvalid_0's binary_logloss: 0.225442\n",
      "[92]\tvalid_0's binary_logloss: 0.224256\n",
      "[93]\tvalid_0's binary_logloss: 0.223477\n",
      "[94]\tvalid_0's binary_logloss: 0.226366\n",
      "[95]\tvalid_0's binary_logloss: 0.227169\n",
      "[96]\tvalid_0's binary_logloss: 0.225648\n",
      "[97]\tvalid_0's binary_logloss: 0.2285\n",
      "[98]\tvalid_0's binary_logloss: 0.226896\n",
      "[99]\tvalid_0's binary_logloss: 0.225761\n",
      "[100]\tvalid_0's binary_logloss: 0.226138\n",
      "[101]\tvalid_0's binary_logloss: 0.226974\n",
      "[102]\tvalid_0's binary_logloss: 0.227435\n",
      "[103]\tvalid_0's binary_logloss: 0.225568\n",
      "[104]\tvalid_0's binary_logloss: 0.226204\n",
      "[105]\tvalid_0's binary_logloss: 0.225158\n",
      "[106]\tvalid_0's binary_logloss: 0.223897\n",
      "[107]\tvalid_0's binary_logloss: 0.22464\n",
      "[108]\tvalid_0's binary_logloss: 0.225429\n",
      "[109]\tvalid_0's binary_logloss: 0.223984\n",
      "[110]\tvalid_0's binary_logloss: 0.224659\n",
      "[111]\tvalid_0's binary_logloss: 0.223179\n",
      "[112]\tvalid_0's binary_logloss: 0.222268\n",
      "[113]\tvalid_0's binary_logloss: 0.2213\n",
      "[114]\tvalid_0's binary_logloss: 0.220444\n",
      "[115]\tvalid_0's binary_logloss: 0.219603\n",
      "[116]\tvalid_0's binary_logloss: 0.21891\n",
      "[117]\tvalid_0's binary_logloss: 0.217941\n",
      "[118]\tvalid_0's binary_logloss: 0.218484\n",
      "[119]\tvalid_0's binary_logloss: 0.218964\n",
      "[120]\tvalid_0's binary_logloss: 0.217964\n",
      "[121]\tvalid_0's binary_logloss: 0.218349\n",
      "[122]\tvalid_0's binary_logloss: 0.217521\n",
      "[123]\tvalid_0's binary_logloss: 0.21673\n",
      "[124]\tvalid_0's binary_logloss: 0.217222\n",
      "[125]\tvalid_0's binary_logloss: 0.217496\n",
      "[126]\tvalid_0's binary_logloss: 0.216627\n",
      "[127]\tvalid_0's binary_logloss: 0.215846\n",
      "[128]\tvalid_0's binary_logloss: 0.216195\n",
      "[129]\tvalid_0's binary_logloss: 0.216615\n",
      "[130]\tvalid_0's binary_logloss: 0.217135\n",
      "[131]\tvalid_0's binary_logloss: 0.216674\n",
      "[132]\tvalid_0's binary_logloss: 0.216027\n",
      "[133]\tvalid_0's binary_logloss: 0.215379\n",
      "[134]\tvalid_0's binary_logloss: 0.215764\n",
      "[135]\tvalid_0's binary_logloss: 0.21608\n",
      "[136]\tvalid_0's binary_logloss: 0.215481\n",
      "[137]\tvalid_0's binary_logloss: 0.215299\n",
      "[138]\tvalid_0's binary_logloss: 0.215024\n",
      "[139]\tvalid_0's binary_logloss: 0.214507\n",
      "[140]\tvalid_0's binary_logloss: 0.214715\n",
      "[141]\tvalid_0's binary_logloss: 0.214217\n",
      "[142]\tvalid_0's binary_logloss: 0.214498\n",
      "[143]\tvalid_0's binary_logloss: 0.213767\n",
      "[144]\tvalid_0's binary_logloss: 0.213572\n",
      "[145]\tvalid_0's binary_logloss: 0.21384\n",
      "[146]\tvalid_0's binary_logloss: 0.21321\n",
      "[147]\tvalid_0's binary_logloss: 0.213396\n",
      "[148]\tvalid_0's binary_logloss: 0.212873\n",
      "[149]\tvalid_0's binary_logloss: 0.213736\n",
      "[150]\tvalid_0's binary_logloss: 0.214059\n",
      "[151]\tvalid_0's binary_logloss: 0.214342\n",
      "[152]\tvalid_0's binary_logloss: 0.214614\n",
      "[153]\tvalid_0's binary_logloss: 0.214896\n",
      "[154]\tvalid_0's binary_logloss: 0.214334\n",
      "[155]\tvalid_0's binary_logloss: 0.213764\n",
      "[156]\tvalid_0's binary_logloss: 0.214057\n",
      "[157]\tvalid_0's binary_logloss: 0.214353\n",
      "[158]\tvalid_0's binary_logloss: 0.213821\n",
      "[159]\tvalid_0's binary_logloss: 0.21342\n",
      "[160]\tvalid_0's binary_logloss: 0.212928\n",
      "[161]\tvalid_0's binary_logloss: 0.2124\n",
      "[162]\tvalid_0's binary_logloss: 0.212624\n",
      "[163]\tvalid_0's binary_logloss: 0.212906\n",
      "[164]\tvalid_0's binary_logloss: 0.213163\n",
      "[165]\tvalid_0's binary_logloss: 0.213097\n",
      "[166]\tvalid_0's binary_logloss: 0.213341\n",
      "[167]\tvalid_0's binary_logloss: 0.213612\n",
      "[168]\tvalid_0's binary_logloss: 0.212899\n",
      "[169]\tvalid_0's binary_logloss: 0.212645\n",
      "[170]\tvalid_0's binary_logloss: 0.21287\n",
      "[171]\tvalid_0's binary_logloss: 0.212524\n",
      "[172]\tvalid_0's binary_logloss: 0.212394\n",
      "[173]\tvalid_0's binary_logloss: 0.212168\n",
      "[174]\tvalid_0's binary_logloss: 0.212287\n",
      "[175]\tvalid_0's binary_logloss: 0.211985\n",
      "[176]\tvalid_0's binary_logloss: 0.211816\n",
      "[177]\tvalid_0's binary_logloss: 0.21199\n",
      "[178]\tvalid_0's binary_logloss: 0.211699\n",
      "[179]\tvalid_0's binary_logloss: 0.211534\n",
      "[180]\tvalid_0's binary_logloss: 0.211639\n",
      "[181]\tvalid_0's binary_logloss: 0.211275\n",
      "[182]\tvalid_0's binary_logloss: 0.212001\n",
      "[183]\tvalid_0's binary_logloss: 0.212194\n",
      "[184]\tvalid_0's binary_logloss: 0.211801\n",
      "[185]\tvalid_0's binary_logloss: 0.211361\n",
      "[186]\tvalid_0's binary_logloss: 0.211534\n",
      "[187]\tvalid_0's binary_logloss: 0.210878\n",
      "[188]\tvalid_0's binary_logloss: 0.210622\n",
      "[189]\tvalid_0's binary_logloss: 0.211129\n",
      "[190]\tvalid_0's binary_logloss: 0.210999\n",
      "[191]\tvalid_0's binary_logloss: 0.211791\n",
      "[192]\tvalid_0's binary_logloss: 0.211533\n",
      "[193]\tvalid_0's binary_logloss: 0.21094\n",
      "[194]\tvalid_0's binary_logloss: 0.210755\n",
      "[195]\tvalid_0's binary_logloss: 0.210868\n",
      "[196]\tvalid_0's binary_logloss: 0.210421\n",
      "[197]\tvalid_0's binary_logloss: 0.21041\n",
      "[198]\tvalid_0's binary_logloss: 0.210281\n",
      "[199]\tvalid_0's binary_logloss: 0.210187\n",
      "[200]\tvalid_0's binary_logloss: 0.210145\n",
      "[201]\tvalid_0's binary_logloss: 0.209742\n",
      "[202]\tvalid_0's binary_logloss: 0.209313\n",
      "[203]\tvalid_0's binary_logloss: 0.209353\n",
      "[204]\tvalid_0's binary_logloss: 0.209235\n",
      "[205]\tvalid_0's binary_logloss: 0.209298\n",
      "[206]\tvalid_0's binary_logloss: 0.209426\n",
      "[207]\tvalid_0's binary_logloss: 0.209382\n",
      "[208]\tvalid_0's binary_logloss: 0.209434\n",
      "[209]\tvalid_0's binary_logloss: 0.209432\n",
      "[210]\tvalid_0's binary_logloss: 0.209734\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2065: UserWarning: Using categorical_feature in Dataset.\n",
      "  _log_warning('Using categorical_feature in Dataset.')\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2068: UserWarning: categorical_feature in Dataset is overridden.\n",
      "New categorical_feature is []\n",
      "  _log_warning('categorical_feature in Dataset is overridden.\\n'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] boosting is set=dart, boosting_type=gbdt will be ignored. Current value: boosting=dart\n",
      "[1]\tvalid_0's binary_logloss: 0.261852\n",
      "[2]\tvalid_0's binary_logloss: 0.256866\n",
      "[3]\tvalid_0's binary_logloss: 0.252577\n",
      "[4]\tvalid_0's binary_logloss: 0.249516\n",
      "[5]\tvalid_0's binary_logloss: 0.246849\n",
      "[6]\tvalid_0's binary_logloss: 0.244189\n",
      "[7]\tvalid_0's binary_logloss: 0.244431\n",
      "[8]\tvalid_0's binary_logloss: 0.242343\n",
      "[9]\tvalid_0's binary_logloss: 0.273143\n",
      "[10]\tvalid_0's binary_logloss: 0.270135\n",
      "[11]\tvalid_0's binary_logloss: 0.267158\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/callback.py:223: UserWarning: Early stopping is not available in dart mode\n",
      "  _log_warning('Early stopping is not available in dart mode')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[12]\tvalid_0's binary_logloss: 0.267158\n",
      "[13]\tvalid_0's binary_logloss: 0.267414\n",
      "[14]\tvalid_0's binary_logloss: 0.263962\n",
      "[15]\tvalid_0's binary_logloss: 0.261404\n",
      "[16]\tvalid_0's binary_logloss: 0.258636\n",
      "[17]\tvalid_0's binary_logloss: 0.256449\n",
      "[18]\tvalid_0's binary_logloss: 0.253798\n",
      "[19]\tvalid_0's binary_logloss: 0.302642\n",
      "[20]\tvalid_0's binary_logloss: 0.297533\n",
      "[21]\tvalid_0's binary_logloss: 0.29841\n",
      "[22]\tvalid_0's binary_logloss: 0.292838\n",
      "[23]\tvalid_0's binary_logloss: 0.287847\n",
      "[24]\tvalid_0's binary_logloss: 0.286759\n",
      "[25]\tvalid_0's binary_logloss: 0.281855\n",
      "[26]\tvalid_0's binary_logloss: 0.277267\n",
      "[27]\tvalid_0's binary_logloss: 0.27316\n",
      "[28]\tvalid_0's binary_logloss: 0.270781\n",
      "[29]\tvalid_0's binary_logloss: 0.267521\n",
      "[30]\tvalid_0's binary_logloss: 0.264115\n",
      "[31]\tvalid_0's binary_logloss: 0.260902\n",
      "[32]\tvalid_0's binary_logloss: 0.259783\n",
      "[33]\tvalid_0's binary_logloss: 0.261092\n",
      "[34]\tvalid_0's binary_logloss: 0.260655\n",
      "[35]\tvalid_0's binary_logloss: 0.261686\n",
      "[36]\tvalid_0's binary_logloss: 0.258726\n",
      "[37]\tvalid_0's binary_logloss: 0.258248\n",
      "[38]\tvalid_0's binary_logloss: 0.255249\n",
      "[39]\tvalid_0's binary_logloss: 0.256217\n",
      "[40]\tvalid_0's binary_logloss: 0.25712\n",
      "[41]\tvalid_0's binary_logloss: 0.257852\n",
      "[42]\tvalid_0's binary_logloss: 0.274116\n",
      "[43]\tvalid_0's binary_logloss: 0.270284\n",
      "[44]\tvalid_0's binary_logloss: 0.266953\n",
      "[45]\tvalid_0's binary_logloss: 0.26325\n",
      "[46]\tvalid_0's binary_logloss: 0.259912\n",
      "[47]\tvalid_0's binary_logloss: 0.261159\n",
      "[48]\tvalid_0's binary_logloss: 0.261273\n",
      "[49]\tvalid_0's binary_logloss: 0.26268\n",
      "[50]\tvalid_0's binary_logloss: 0.263299\n",
      "[51]\tvalid_0's binary_logloss: 0.264762\n",
      "[52]\tvalid_0's binary_logloss: 0.26658\n",
      "[53]\tvalid_0's binary_logloss: 0.266026\n",
      "[54]\tvalid_0's binary_logloss: 0.262785\n",
      "[55]\tvalid_0's binary_logloss: 0.259265\n",
      "[56]\tvalid_0's binary_logloss: 0.256395\n",
      "[57]\tvalid_0's binary_logloss: 0.25348\n",
      "[58]\tvalid_0's binary_logloss: 0.250999\n",
      "[59]\tvalid_0's binary_logloss: 0.252068\n",
      "[60]\tvalid_0's binary_logloss: 0.253355\n",
      "[61]\tvalid_0's binary_logloss: 0.254594\n",
      "[62]\tvalid_0's binary_logloss: 0.255385\n",
      "[63]\tvalid_0's binary_logloss: 0.255743\n",
      "[64]\tvalid_0's binary_logloss: 0.253085\n",
      "[65]\tvalid_0's binary_logloss: 0.250189\n",
      "[66]\tvalid_0's binary_logloss: 0.247687\n",
      "[67]\tvalid_0's binary_logloss: 0.245579\n",
      "[68]\tvalid_0's binary_logloss: 0.243442\n",
      "[69]\tvalid_0's binary_logloss: 0.241667\n",
      "[70]\tvalid_0's binary_logloss: 0.242505\n",
      "[71]\tvalid_0's binary_logloss: 0.243032\n",
      "[72]\tvalid_0's binary_logloss: 0.243846\n",
      "[73]\tvalid_0's binary_logloss: 0.241799\n",
      "[74]\tvalid_0's binary_logloss: 0.240087\n",
      "[75]\tvalid_0's binary_logloss: 0.238783\n",
      "[76]\tvalid_0's binary_logloss: 0.239312\n",
      "[77]\tvalid_0's binary_logloss: 0.240364\n",
      "[78]\tvalid_0's binary_logloss: 0.238413\n",
      "[79]\tvalid_0's binary_logloss: 0.239102\n",
      "[80]\tvalid_0's binary_logloss: 0.237666\n",
      "[81]\tvalid_0's binary_logloss: 0.238469\n",
      "[82]\tvalid_0's binary_logloss: 0.239248\n",
      "[83]\tvalid_0's binary_logloss: 0.239932\n",
      "[84]\tvalid_0's binary_logloss: 0.238891\n",
      "[85]\tvalid_0's binary_logloss: 0.23747\n",
      "[86]\tvalid_0's binary_logloss: 0.236156\n",
      "[87]\tvalid_0's binary_logloss: 0.235207\n",
      "[88]\tvalid_0's binary_logloss: 0.233968\n",
      "[89]\tvalid_0's binary_logloss: 0.234551\n",
      "[90]\tvalid_0's binary_logloss: 0.235087\n",
      "[91]\tvalid_0's binary_logloss: 0.233709\n",
      "[92]\tvalid_0's binary_logloss: 0.232467\n",
      "[93]\tvalid_0's binary_logloss: 0.231355\n",
      "[94]\tvalid_0's binary_logloss: 0.233935\n",
      "[95]\tvalid_0's binary_logloss: 0.234752\n",
      "[96]\tvalid_0's binary_logloss: 0.233397\n",
      "[97]\tvalid_0's binary_logloss: 0.236021\n",
      "[98]\tvalid_0's binary_logloss: 0.234577\n",
      "[99]\tvalid_0's binary_logloss: 0.233336\n",
      "[100]\tvalid_0's binary_logloss: 0.233631\n",
      "[101]\tvalid_0's binary_logloss: 0.234354\n",
      "[102]\tvalid_0's binary_logloss: 0.234839\n",
      "[103]\tvalid_0's binary_logloss: 0.233424\n",
      "[104]\tvalid_0's binary_logloss: 0.234022\n",
      "[105]\tvalid_0's binary_logloss: 0.23262\n",
      "[106]\tvalid_0's binary_logloss: 0.231349\n",
      "[107]\tvalid_0's binary_logloss: 0.232\n",
      "[108]\tvalid_0's binary_logloss: 0.232772\n",
      "[109]\tvalid_0's binary_logloss: 0.231861\n",
      "[110]\tvalid_0's binary_logloss: 0.232427\n",
      "[111]\tvalid_0's binary_logloss: 0.231034\n",
      "[112]\tvalid_0's binary_logloss: 0.230092\n",
      "[113]\tvalid_0's binary_logloss: 0.229501\n",
      "[114]\tvalid_0's binary_logloss: 0.228372\n",
      "[115]\tvalid_0's binary_logloss: 0.227584\n",
      "[116]\tvalid_0's binary_logloss: 0.227003\n",
      "[117]\tvalid_0's binary_logloss: 0.226268\n",
      "[118]\tvalid_0's binary_logloss: 0.226754\n",
      "[119]\tvalid_0's binary_logloss: 0.227076\n",
      "[120]\tvalid_0's binary_logloss: 0.226645\n",
      "[121]\tvalid_0's binary_logloss: 0.226966\n",
      "[122]\tvalid_0's binary_logloss: 0.225941\n",
      "[123]\tvalid_0's binary_logloss: 0.225327\n",
      "[124]\tvalid_0's binary_logloss: 0.22565\n",
      "[125]\tvalid_0's binary_logloss: 0.225867\n",
      "[126]\tvalid_0's binary_logloss: 0.225113\n",
      "[127]\tvalid_0's binary_logloss: 0.224508\n",
      "[128]\tvalid_0's binary_logloss: 0.224749\n",
      "[129]\tvalid_0's binary_logloss: 0.225119\n",
      "[130]\tvalid_0's binary_logloss: 0.225544\n",
      "[131]\tvalid_0's binary_logloss: 0.225019\n",
      "[132]\tvalid_0's binary_logloss: 0.224381\n",
      "[133]\tvalid_0's binary_logloss: 0.223396\n",
      "[134]\tvalid_0's binary_logloss: 0.223702\n",
      "[135]\tvalid_0's binary_logloss: 0.223998\n",
      "[136]\tvalid_0's binary_logloss: 0.223446\n",
      "[137]\tvalid_0's binary_logloss: 0.222983\n",
      "[138]\tvalid_0's binary_logloss: 0.222464\n",
      "[139]\tvalid_0's binary_logloss: 0.221789\n",
      "[140]\tvalid_0's binary_logloss: 0.221937\n",
      "[141]\tvalid_0's binary_logloss: 0.221657\n",
      "[142]\tvalid_0's binary_logloss: 0.221828\n",
      "[143]\tvalid_0's binary_logloss: 0.221163\n",
      "[144]\tvalid_0's binary_logloss: 0.220826\n",
      "[145]\tvalid_0's binary_logloss: 0.22098\n",
      "[146]\tvalid_0's binary_logloss: 0.220528\n",
      "[147]\tvalid_0's binary_logloss: 0.220538\n",
      "[148]\tvalid_0's binary_logloss: 0.22041\n",
      "[149]\tvalid_0's binary_logloss: 0.220944\n",
      "[150]\tvalid_0's binary_logloss: 0.2211\n",
      "[151]\tvalid_0's binary_logloss: 0.221236\n",
      "[152]\tvalid_0's binary_logloss: 0.221348\n",
      "[153]\tvalid_0's binary_logloss: 0.221555\n",
      "[154]\tvalid_0's binary_logloss: 0.221168\n",
      "[155]\tvalid_0's binary_logloss: 0.220535\n",
      "[156]\tvalid_0's binary_logloss: 0.220691\n",
      "[157]\tvalid_0's binary_logloss: 0.220914\n",
      "[158]\tvalid_0's binary_logloss: 0.22028\n",
      "[159]\tvalid_0's binary_logloss: 0.220064\n",
      "[160]\tvalid_0's binary_logloss: 0.219932\n",
      "[161]\tvalid_0's binary_logloss: 0.219552\n",
      "[162]\tvalid_0's binary_logloss: 0.219698\n",
      "[163]\tvalid_0's binary_logloss: 0.219764\n",
      "[164]\tvalid_0's binary_logloss: 0.219945\n",
      "[165]\tvalid_0's binary_logloss: 0.219629\n",
      "[166]\tvalid_0's binary_logloss: 0.219743\n",
      "[167]\tvalid_0's binary_logloss: 0.219878\n",
      "[168]\tvalid_0's binary_logloss: 0.219868\n",
      "[169]\tvalid_0's binary_logloss: 0.219702\n",
      "[170]\tvalid_0's binary_logloss: 0.219868\n",
      "[171]\tvalid_0's binary_logloss: 0.219431\n",
      "[172]\tvalid_0's binary_logloss: 0.219178\n",
      "[173]\tvalid_0's binary_logloss: 0.218496\n",
      "[174]\tvalid_0's binary_logloss: 0.218597\n",
      "[175]\tvalid_0's binary_logloss: 0.218332\n",
      "[176]\tvalid_0's binary_logloss: 0.218284\n",
      "[177]\tvalid_0's binary_logloss: 0.218402\n",
      "[178]\tvalid_0's binary_logloss: 0.218194\n",
      "[179]\tvalid_0's binary_logloss: 0.217823\n",
      "[180]\tvalid_0's binary_logloss: 0.217853\n",
      "[181]\tvalid_0's binary_logloss: 0.21755\n",
      "[182]\tvalid_0's binary_logloss: 0.217868\n",
      "[183]\tvalid_0's binary_logloss: 0.217969\n",
      "[184]\tvalid_0's binary_logloss: 0.217684\n",
      "[185]\tvalid_0's binary_logloss: 0.217483\n",
      "[186]\tvalid_0's binary_logloss: 0.217556\n",
      "[187]\tvalid_0's binary_logloss: 0.217434\n",
      "[188]\tvalid_0's binary_logloss: 0.217287\n",
      "[189]\tvalid_0's binary_logloss: 0.217506\n",
      "[190]\tvalid_0's binary_logloss: 0.217818\n",
      "[191]\tvalid_0's binary_logloss: 0.218157\n",
      "[192]\tvalid_0's binary_logloss: 0.217935\n",
      "[193]\tvalid_0's binary_logloss: 0.217314\n",
      "[194]\tvalid_0's binary_logloss: 0.217408\n",
      "[195]\tvalid_0's binary_logloss: 0.217506\n",
      "[196]\tvalid_0's binary_logloss: 0.217385\n",
      "[197]\tvalid_0's binary_logloss: 0.217202\n",
      "[198]\tvalid_0's binary_logloss: 0.21717\n",
      "[199]\tvalid_0's binary_logloss: 0.217206\n",
      "[200]\tvalid_0's binary_logloss: 0.217053\n",
      "[201]\tvalid_0's binary_logloss: 0.216808\n",
      "[202]\tvalid_0's binary_logloss: 0.216719\n",
      "[203]\tvalid_0's binary_logloss: 0.216661\n",
      "[204]\tvalid_0's binary_logloss: 0.216506\n",
      "[205]\tvalid_0's binary_logloss: 0.216463\n",
      "[206]\tvalid_0's binary_logloss: 0.21643\n",
      "[207]\tvalid_0's binary_logloss: 0.216449\n",
      "[208]\tvalid_0's binary_logloss: 0.216366\n",
      "[209]\tvalid_0's binary_logloss: 0.216561\n",
      "[210]\tvalid_0's binary_logloss: 0.216565\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2065: UserWarning: Using categorical_feature in Dataset.\n",
      "  _log_warning('Using categorical_feature in Dataset.')\n",
      "/opt/conda/lib/python3.7/site-packages/lightgbm/basic.py:2068: UserWarning: categorical_feature in Dataset is overridden.\n",
      "New categorical_feature is []\n",
      "  _log_warning('categorical_feature in Dataset is overridden.\\n'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] boosting is set=dart, boosting_type=gbdt will be ignored. Current value: boosting=dart\n",
      "[1]\tvalid_0's binary_logloss: 0.260453\n",
      "[2]\tvalid_0's binary_logloss: 0.255429\n",
      "[3]\tvalid_0's binary_logloss: 0.250328\n",
      "[4]\tvalid_0's binary_logloss: 0.247365\n",
      "[5]\tvalid_0's binary_logloss: 0.243974\n",
      "[6]\tvalid_0's binary_logloss: 0.240649\n",
      "[7]\tvalid_0's binary_logloss: 0.240618\n",
      "[8]\tvalid_0's binary_logloss: 0.237913\n",
      "[9]\tvalid_0's binary_logloss: 0.268093\n",
      "[10]\tvalid_0's binary_logloss: 0.264253\n",
      "[11]\tvalid_0's binary_logloss: 0.260544\n",
      "[12]\tvalid_0's binary_logloss: 0.260544\n",
      "[13]\tvalid_0's binary_logloss: 0.26145\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/lightgbm/callback.py:223: UserWarning: Early stopping is not available in dart mode\n",
      "  _log_warning('Early stopping is not available in dart mode')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[14]\tvalid_0's binary_logloss: 0.258251\n",
      "[15]\tvalid_0's binary_logloss: 0.255536\n",
      "[16]\tvalid_0's binary_logloss: 0.252621\n",
      "[17]\tvalid_0's binary_logloss: 0.250082\n",
      "[18]\tvalid_0's binary_logloss: 0.247454\n",
      "[19]\tvalid_0's binary_logloss: 0.294372\n",
      "[20]\tvalid_0's binary_logloss: 0.288461\n",
      "[21]\tvalid_0's binary_logloss: 0.289398\n",
      "[22]\tvalid_0's binary_logloss: 0.283696\n",
      "[23]\tvalid_0's binary_logloss: 0.278645\n",
      "[24]\tvalid_0's binary_logloss: 0.278334\n",
      "[25]\tvalid_0's binary_logloss: 0.273566\n",
      "[26]\tvalid_0's binary_logloss: 0.268985\n",
      "[27]\tvalid_0's binary_logloss: 0.265028\n",
      "[28]\tvalid_0's binary_logloss: 0.260862\n",
      "[29]\tvalid_0's binary_logloss: 0.257435\n",
      "[30]\tvalid_0's binary_logloss: 0.254088\n",
      "[31]\tvalid_0's binary_logloss: 0.250922\n",
      "[32]\tvalid_0's binary_logloss: 0.249788\n",
      "[33]\tvalid_0's binary_logloss: 0.251168\n",
      "[34]\tvalid_0's binary_logloss: 0.250715\n",
      "[35]\tvalid_0's binary_logloss: 0.251799\n",
      "[36]\tvalid_0's binary_logloss: 0.248273\n",
      "[37]\tvalid_0's binary_logloss: 0.247761\n",
      "[38]\tvalid_0's binary_logloss: 0.244883\n",
      "[39]\tvalid_0's binary_logloss: 0.246048\n",
      "[40]\tvalid_0's binary_logloss: 0.247347\n",
      "[41]\tvalid_0's binary_logloss: 0.248214\n",
      "[42]\tvalid_0's binary_logloss: 0.263932\n",
      "[43]\tvalid_0's binary_logloss: 0.260126\n",
      "[44]\tvalid_0's binary_logloss: 0.256522\n",
      "[45]\tvalid_0's binary_logloss: 0.253054\n",
      "[46]\tvalid_0's binary_logloss: 0.249885\n",
      "[47]\tvalid_0's binary_logloss: 0.251054\n",
      "[48]\tvalid_0's binary_logloss: 0.25148\n",
      "[49]\tvalid_0's binary_logloss: 0.253229\n",
      "[50]\tvalid_0's binary_logloss: 0.254204\n",
      "[51]\tvalid_0's binary_logloss: 0.255651\n",
      "[52]\tvalid_0's binary_logloss: 0.257371\n",
      "[53]\tvalid_0's binary_logloss: 0.256816\n",
      "[54]\tvalid_0's binary_logloss: 0.253576\n",
      "[55]\tvalid_0's binary_logloss: 0.250475\n",
      "[56]\tvalid_0's binary_logloss: 0.247535\n",
      "[57]\tvalid_0's binary_logloss: 0.244548\n",
      "[58]\tvalid_0's binary_logloss: 0.241905\n",
      "[59]\tvalid_0's binary_logloss: 0.242906\n",
      "[60]\tvalid_0's binary_logloss: 0.244192\n",
      "[61]\tvalid_0's binary_logloss: 0.245395\n",
      "[62]\tvalid_0's binary_logloss: 0.246261\n",
      "[63]\tvalid_0's binary_logloss: 0.246683\n",
      "[64]\tvalid_0's binary_logloss: 0.243877\n",
      "[65]\tvalid_0's binary_logloss: 0.24123\n",
      "[66]\tvalid_0's binary_logloss: 0.239013\n",
      "[67]\tvalid_0's binary_logloss: 0.23663\n",
      "[68]\tvalid_0's binary_logloss: 0.233784\n",
      "[69]\tvalid_0's binary_logloss: 0.231741\n",
      "[70]\tvalid_0's binary_logloss: 0.232693\n",
      "[71]\tvalid_0's binary_logloss: 0.233213\n",
      "[72]\tvalid_0's binary_logloss: 0.234154\n",
      "[73]\tvalid_0's binary_logloss: 0.231822\n",
      "[74]\tvalid_0's binary_logloss: 0.230182\n",
      "[75]\tvalid_0's binary_logloss: 0.228408\n",
      "[76]\tvalid_0's binary_logloss: 0.229027\n",
      "[77]\tvalid_0's binary_logloss: 0.230141\n",
      "[78]\tvalid_0's binary_logloss: 0.228239\n",
      "[79]\tvalid_0's binary_logloss: 0.229097\n",
      "[80]\tvalid_0's binary_logloss: 0.227683\n",
      "[81]\tvalid_0's binary_logloss: 0.22866\n",
      "[82]\tvalid_0's binary_logloss: 0.229473\n",
      "[83]\tvalid_0's binary_logloss: 0.230262\n",
      "[84]\tvalid_0's binary_logloss: 0.228678\n",
      "[85]\tvalid_0's binary_logloss: 0.226776\n",
      "[86]\tvalid_0's binary_logloss: 0.225298\n",
      "[87]\tvalid_0's binary_logloss: 0.224087\n",
      "[88]\tvalid_0's binary_logloss: 0.222401\n",
      "[89]\tvalid_0's binary_logloss: 0.223188\n",
      "[90]\tvalid_0's binary_logloss: 0.223711\n",
      "[91]\tvalid_0's binary_logloss: 0.222562\n",
      "[92]\tvalid_0's binary_logloss: 0.221167\n",
      "[93]\tvalid_0's binary_logloss: 0.220038\n",
      "[94]\tvalid_0's binary_logloss: 0.222621\n",
      "[95]\tvalid_0's binary_logloss: 0.223469\n",
      "[96]\tvalid_0's binary_logloss: 0.222093\n",
      "[97]\tvalid_0's binary_logloss: 0.224749\n",
      "[98]\tvalid_0's binary_logloss: 0.223586\n",
      "[99]\tvalid_0's binary_logloss: 0.222104\n",
      "[100]\tvalid_0's binary_logloss: 0.222471\n",
      "[101]\tvalid_0's binary_logloss: 0.223359\n",
      "[102]\tvalid_0's binary_logloss: 0.224014\n",
      "[103]\tvalid_0's binary_logloss: 0.222925\n",
      "[104]\tvalid_0's binary_logloss: 0.22363\n",
      "[105]\tvalid_0's binary_logloss: 0.222422\n",
      "[106]\tvalid_0's binary_logloss: 0.221323\n",
      "[107]\tvalid_0's binary_logloss: 0.222041\n",
      "[108]\tvalid_0's binary_logloss: 0.222782\n",
      "[109]\tvalid_0's binary_logloss: 0.221472\n",
      "[110]\tvalid_0's binary_logloss: 0.222117\n",
      "[111]\tvalid_0's binary_logloss: 0.221082\n",
      "[112]\tvalid_0's binary_logloss: 0.219774\n",
      "[113]\tvalid_0's binary_logloss: 0.218464\n",
      "[114]\tvalid_0's binary_logloss: 0.21767\n",
      "[115]\tvalid_0's binary_logloss: 0.216662\n",
      "[116]\tvalid_0's binary_logloss: 0.215785\n",
      "[117]\tvalid_0's binary_logloss: 0.214787\n",
      "[118]\tvalid_0's binary_logloss: 0.215384\n",
      "[119]\tvalid_0's binary_logloss: 0.215893\n",
      "[120]\tvalid_0's binary_logloss: 0.214991\n",
      "[121]\tvalid_0's binary_logloss: 0.215374\n",
      "[122]\tvalid_0's binary_logloss: 0.21435\n",
      "[123]\tvalid_0's binary_logloss: 0.213322\n",
      "[124]\tvalid_0's binary_logloss: 0.213775\n",
      "[125]\tvalid_0's binary_logloss: 0.214111\n",
      "[126]\tvalid_0's binary_logloss: 0.213333\n",
      "[127]\tvalid_0's binary_logloss: 0.212995\n",
      "[128]\tvalid_0's binary_logloss: 0.213279\n",
      "[129]\tvalid_0's binary_logloss: 0.213688\n",
      "[130]\tvalid_0's binary_logloss: 0.214194\n",
      "[131]\tvalid_0's binary_logloss: 0.213752\n",
      "[132]\tvalid_0's binary_logloss: 0.213089\n",
      "[133]\tvalid_0's binary_logloss: 0.212354\n",
      "[134]\tvalid_0's binary_logloss: 0.212739\n",
      "[135]\tvalid_0's binary_logloss: 0.213088\n",
      "[136]\tvalid_0's binary_logloss: 0.212547\n",
      "[137]\tvalid_0's binary_logloss: 0.211883\n",
      "[138]\tvalid_0's binary_logloss: 0.211121\n",
      "[139]\tvalid_0's binary_logloss: 0.210498\n",
      "[140]\tvalid_0's binary_logloss: 0.210805\n",
      "[141]\tvalid_0's binary_logloss: 0.210701\n",
      "[142]\tvalid_0's binary_logloss: 0.210925\n",
      "[143]\tvalid_0's binary_logloss: 0.210673\n",
      "[144]\tvalid_0's binary_logloss: 0.210218\n",
      "[145]\tvalid_0's binary_logloss: 0.210524\n",
      "[146]\tvalid_0's binary_logloss: 0.209971\n",
      "[147]\tvalid_0's binary_logloss: 0.210122\n",
      "[148]\tvalid_0's binary_logloss: 0.209795\n",
      "[149]\tvalid_0's binary_logloss: 0.210518\n",
      "[150]\tvalid_0's binary_logloss: 0.210755\n",
      "[151]\tvalid_0's binary_logloss: 0.211023\n",
      "[152]\tvalid_0's binary_logloss: 0.211321\n",
      "[153]\tvalid_0's binary_logloss: 0.211519\n",
      "[154]\tvalid_0's binary_logloss: 0.210794\n",
      "[155]\tvalid_0's binary_logloss: 0.210455\n",
      "[156]\tvalid_0's binary_logloss: 0.210724\n",
      "[157]\tvalid_0's binary_logloss: 0.211063\n",
      "[158]\tvalid_0's binary_logloss: 0.210602\n",
      "[159]\tvalid_0's binary_logloss: 0.210278\n",
      "[160]\tvalid_0's binary_logloss: 0.209573\n",
      "[161]\tvalid_0's binary_logloss: 0.209114\n",
      "[162]\tvalid_0's binary_logloss: 0.20935\n",
      "[163]\tvalid_0's binary_logloss: 0.209618\n",
      "[164]\tvalid_0's binary_logloss: 0.209854\n",
      "[165]\tvalid_0's binary_logloss: 0.209649\n",
      "[166]\tvalid_0's binary_logloss: 0.209912\n",
      "[167]\tvalid_0's binary_logloss: 0.210145\n",
      "[168]\tvalid_0's binary_logloss: 0.209958\n",
      "[169]\tvalid_0's binary_logloss: 0.209403\n",
      "[170]\tvalid_0's binary_logloss: 0.209685\n",
      "[171]\tvalid_0's binary_logloss: 0.209036\n",
      "[172]\tvalid_0's binary_logloss: 0.208681\n",
      "[173]\tvalid_0's binary_logloss: 0.208377\n",
      "[174]\tvalid_0's binary_logloss: 0.208519\n",
      "[175]\tvalid_0's binary_logloss: 0.208114\n",
      "[176]\tvalid_0's binary_logloss: 0.207784\n",
      "[177]\tvalid_0's binary_logloss: 0.207945\n",
      "[178]\tvalid_0's binary_logloss: 0.207644\n",
      "[179]\tvalid_0's binary_logloss: 0.207386\n",
      "[180]\tvalid_0's binary_logloss: 0.207526\n",
      "[181]\tvalid_0's binary_logloss: 0.20738\n",
      "[182]\tvalid_0's binary_logloss: 0.207849\n",
      "[183]\tvalid_0's binary_logloss: 0.207991\n",
      "[184]\tvalid_0's binary_logloss: 0.207721\n",
      "[185]\tvalid_0's binary_logloss: 0.207501\n",
      "[186]\tvalid_0's binary_logloss: 0.207617\n",
      "[187]\tvalid_0's binary_logloss: 0.207364\n",
      "[188]\tvalid_0's binary_logloss: 0.207234\n",
      "[189]\tvalid_0's binary_logloss: 0.207543\n",
      "[190]\tvalid_0's binary_logloss: 0.207445\n",
      "[191]\tvalid_0's binary_logloss: 0.207965\n",
      "[192]\tvalid_0's binary_logloss: 0.207543\n",
      "[193]\tvalid_0's binary_logloss: 0.207383\n",
      "[194]\tvalid_0's binary_logloss: 0.207359\n",
      "[195]\tvalid_0's binary_logloss: 0.207529\n",
      "[196]\tvalid_0's binary_logloss: 0.207281\n",
      "[197]\tvalid_0's binary_logloss: 0.20653\n",
      "[198]\tvalid_0's binary_logloss: 0.206262\n",
      "[199]\tvalid_0's binary_logloss: 0.205935\n",
      "[200]\tvalid_0's binary_logloss: 0.206127\n",
      "[201]\tvalid_0's binary_logloss: 0.206021\n",
      "[202]\tvalid_0's binary_logloss: 0.206165\n",
      "[203]\tvalid_0's binary_logloss: 0.20621\n",
      "[204]\tvalid_0's binary_logloss: 0.206152\n",
      "[205]\tvalid_0's binary_logloss: 0.206154\n",
      "[206]\tvalid_0's binary_logloss: 0.20615\n",
      "[207]\tvalid_0's binary_logloss: 0.206134\n",
      "[208]\tvalid_0's binary_logloss: 0.20616\n",
      "[209]\tvalid_0's binary_logloss: 0.20635\n",
      "[210]\tvalid_0's binary_logloss: 0.206482\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 26178 (\\N{CJK UNIFIED IDEOGRAPH-6642}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 20195 (\\N{CJK UNIFIED IDEOGRAPH-4EE3}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 25152 (\\N{CJK UNIFIED IDEOGRAPH-6240}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 26377 (\\N{CJK UNIFIED IDEOGRAPH-6709}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 32773 (\\N{CJK UNIFIED IDEOGRAPH-8005}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 21517 (\\N{CJK UNIFIED IDEOGRAPH-540D}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 31278 (\\N{CJK UNIFIED IDEOGRAPH-7A2E}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 21029 (\\N{CJK UNIFIED IDEOGRAPH-5225}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 32066 (\\N{CJK UNIFIED IDEOGRAPH-7D42}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 20102 (\\N{CJK UNIFIED IDEOGRAPH-4E86}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 38283 (\\N{CJK UNIFIED IDEOGRAPH-958B}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 22987 (\\N{CJK UNIFIED IDEOGRAPH-59CB}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 22312 (\\N{CJK UNIFIED IDEOGRAPH-5728}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 22320 (\\N{CJK UNIFIED IDEOGRAPH-5730}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 37117 (\\N{CJK UNIFIED IDEOGRAPH-90FD}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 36947 (\\N{CJK UNIFIED IDEOGRAPH-9053}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 24220 (\\N{CJK UNIFIED IDEOGRAPH-5E9C}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 30476 (\\N{CJK UNIFIED IDEOGRAPH-770C}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 32654 (\\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 24037 (\\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 21697 (\\N{CJK UNIFIED IDEOGRAPH-54C1}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 12399 (\\N{HIRAGANA LETTER HA}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 12300 (\\N{LEFT CORNER BRACKET}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 20303 (\\N{CJK UNIFIED IDEOGRAPH-4F4F}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 65288 (\\N{FULLWIDTH LEFT PARENTHESIS}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 65289 (\\N{FULLWIDTH RIGHT PARENTHESIS}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 12301 (\\N{RIGHT CORNER BRACKET}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 20445 (\\N{CJK UNIFIED IDEOGRAPH-4FDD}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 31649 (\\N{CJK UNIFIED IDEOGRAPH-7BA1}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 26045 (\\N{CJK UNIFIED IDEOGRAPH-65BD}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 35373 (\\N{CJK UNIFIED IDEOGRAPH-8A2D}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 12398 (\\N{HIRAGANA LETTER NO}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 31216 (\\N{CJK UNIFIED IDEOGRAPH-79F0}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 32239 (\\N{CJK UNIFIED IDEOGRAPH-7DEF}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:57: UserWarning: Glyph 24230 (\\N{CJK UNIFIED IDEOGRAPH-5EA6}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 26178 (\\N{CJK UNIFIED IDEOGRAPH-6642}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 20195 (\\N{CJK UNIFIED IDEOGRAPH-4EE3}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 25152 (\\N{CJK UNIFIED IDEOGRAPH-6240}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 26377 (\\N{CJK UNIFIED IDEOGRAPH-6709}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 32773 (\\N{CJK UNIFIED IDEOGRAPH-8005}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 21517 (\\N{CJK UNIFIED IDEOGRAPH-540D}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 31278 (\\N{CJK UNIFIED IDEOGRAPH-7A2E}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 21029 (\\N{CJK UNIFIED IDEOGRAPH-5225}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 32066 (\\N{CJK UNIFIED IDEOGRAPH-7D42}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 20102 (\\N{CJK UNIFIED IDEOGRAPH-4E86}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 38283 (\\N{CJK UNIFIED IDEOGRAPH-958B}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 22987 (\\N{CJK UNIFIED IDEOGRAPH-59CB}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 22312 (\\N{CJK UNIFIED IDEOGRAPH-5728}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 22320 (\\N{CJK UNIFIED IDEOGRAPH-5730}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 37117 (\\N{CJK UNIFIED IDEOGRAPH-90FD}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 36947 (\\N{CJK UNIFIED IDEOGRAPH-9053}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 24220 (\\N{CJK UNIFIED IDEOGRAPH-5E9C}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 30476 (\\N{CJK UNIFIED IDEOGRAPH-770C}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 32654 (\\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 24037 (\\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 21697 (\\N{CJK UNIFIED IDEOGRAPH-54C1}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 12399 (\\N{HIRAGANA LETTER HA}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 12300 (\\N{LEFT CORNER BRACKET}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 20303 (\\N{CJK UNIFIED IDEOGRAPH-4F4F}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 65288 (\\N{FULLWIDTH LEFT PARENTHESIS}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 65289 (\\N{FULLWIDTH RIGHT PARENTHESIS}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 12301 (\\N{RIGHT CORNER BRACKET}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 20445 (\\N{CJK UNIFIED IDEOGRAPH-4FDD}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 31649 (\\N{CJK UNIFIED IDEOGRAPH-7BA1}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 26045 (\\N{CJK UNIFIED IDEOGRAPH-65BD}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 35373 (\\N{CJK UNIFIED IDEOGRAPH-8A2D}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 12398 (\\N{HIRAGANA LETTER NO}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 31216 (\\N{CJK UNIFIED IDEOGRAPH-79F0}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 32239 (\\N{CJK UNIFIED IDEOGRAPH-7DEF}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/nyaggle/util/plot_importance.py:59: UserWarning: Glyph 24230 (\\N{CJK UNIFIED IDEOGRAPH-5EA6}) missing from current font.\n",
      "  plt.savefig(path)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 26178 (\\N{CJK UNIFIED IDEOGRAPH-6642}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 20195 (\\N{CJK UNIFIED IDEOGRAPH-4EE3}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 25152 (\\N{CJK UNIFIED IDEOGRAPH-6240}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 26377 (\\N{CJK UNIFIED IDEOGRAPH-6709}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 32773 (\\N{CJK UNIFIED IDEOGRAPH-8005}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 21517 (\\N{CJK UNIFIED IDEOGRAPH-540D}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 31278 (\\N{CJK UNIFIED IDEOGRAPH-7A2E}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 21029 (\\N{CJK UNIFIED IDEOGRAPH-5225}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 32066 (\\N{CJK UNIFIED IDEOGRAPH-7D42}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 20102 (\\N{CJK UNIFIED IDEOGRAPH-4E86}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 38283 (\\N{CJK UNIFIED IDEOGRAPH-958B}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 22987 (\\N{CJK UNIFIED IDEOGRAPH-59CB}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 22312 (\\N{CJK UNIFIED IDEOGRAPH-5728}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 22320 (\\N{CJK UNIFIED IDEOGRAPH-5730}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 37117 (\\N{CJK UNIFIED IDEOGRAPH-90FD}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 36947 (\\N{CJK UNIFIED IDEOGRAPH-9053}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 24220 (\\N{CJK UNIFIED IDEOGRAPH-5E9C}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 30476 (\\N{CJK UNIFIED IDEOGRAPH-770C}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 32654 (\\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 24037 (\\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 21697 (\\N{CJK UNIFIED IDEOGRAPH-54C1}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 12399 (\\N{HIRAGANA LETTER HA}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 12300 (\\N{LEFT CORNER BRACKET}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 20303 (\\N{CJK UNIFIED IDEOGRAPH-4F4F}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 65288 (\\N{FULLWIDTH LEFT PARENTHESIS}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 65289 (\\N{FULLWIDTH RIGHT PARENTHESIS}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 12301 (\\N{RIGHT CORNER BRACKET}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 20445 (\\N{CJK UNIFIED IDEOGRAPH-4FDD}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 31649 (\\N{CJK UNIFIED IDEOGRAPH-7BA1}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 26045 (\\N{CJK UNIFIED IDEOGRAPH-65BD}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 35373 (\\N{CJK UNIFIED IDEOGRAPH-8A2D}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 12398 (\\N{HIRAGANA LETTER NO}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 31216 (\\N{CJK UNIFIED IDEOGRAPH-79F0}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 32239 (\\N{CJK UNIFIED IDEOGRAPH-7DEF}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "/opt/conda/lib/python3.7/site-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 24230 (\\N{CJK UNIFIED IDEOGRAPH-5EA6}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x1152 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "params = {\n",
    "    'boosting': \"dart\", # dart, gbdt, goss, rf\n",
    "    'objective': \"binary\", # binary, multiclass, regression\n",
    "    'metric': \"binary_logloss\", # mae, mse, auc, binary_logloss, multi_logloss, rmse, average_precision\n",
    "    'learning_rate':0.05, #0.05\n",
    "    'n_estimators': 210,\n",
    "    'max_depth': -1,\n",
    "    'random_state' : 42531\n",
    "    #'num_leaves' : 100, #カーディナリティが高い場合は大きくする\n",
    "    #'max_bin' : 500,\n",
    "}\n",
    "\n",
    "result = run_experiment(params,\n",
    "                        train[num_cols + cat_cols],\n",
    "                        train[target_col],\n",
    "                        test[num_cols + cat_cols],\n",
    "                        cv=StratifiedKFold(5, shuffle=True, random_state=42531), #cv=5,\n",
    "                        #groups=train.group\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.7908172302737521, 0.7850322061191626, 0.8267069243156199, 0.8143357487922706, 0.8313959518433713, 0.8081273433162665]\n"
     ]
    }
   ],
   "source": [
    "print(result.metrics) # Evalulation metrics for each fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.7830434782608695, 0.7822745571658616, 0.8266908212560387, 0.8261433172302737, 0.8351138095276086, 0.8096638097616974]\n"
     ]
    }
   ],
   "source": [
    "print(result.metrics) # Evalulation metrics for each fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.7725885668276973, 0.7818236714975845, 0.8162882447665056, 0.817194041867955, 0.8370046513112228, 0.8053098184368901]\n"
     ]
    }
   ],
   "source": [
    "print(result.metrics) # Evalulation metrics for each fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "IM = result.importance     # Feature importance for each fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>feature</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>754</th>\n",
       "      <td>時代</td>\n",
       "      <td>5533.974880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>758</th>\n",
       "      <td>所有者名</td>\n",
       "      <td>1655.755015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>752</th>\n",
       "      <td>種別2</td>\n",
       "      <td>1583.162767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>756</th>\n",
       "      <td>所在地</td>\n",
       "      <td>429.101809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>label_所有者名</td>\n",
       "      <td>352.651622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>504</th>\n",
       "      <td>265</td>\n",
       "      <td>349.998780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289</th>\n",
       "      <td>50</td>\n",
       "      <td>324.430539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>98</td>\n",
       "      <td>316.240280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>496</th>\n",
       "      <td>257</td>\n",
       "      <td>291.254490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>11</td>\n",
       "      <td>283.377229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>427</th>\n",
       "      <td>188</td>\n",
       "      <td>280.394289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>29</td>\n",
       "      <td>267.115792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>661</th>\n",
       "      <td>422</td>\n",
       "      <td>253.805949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>477</th>\n",
       "      <td>238</td>\n",
       "      <td>242.050458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>381</th>\n",
       "      <td>142</td>\n",
       "      <td>240.584730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>609</th>\n",
       "      <td>370</td>\n",
       "      <td>239.705039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>729</th>\n",
       "      <td>490</td>\n",
       "      <td>239.619091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>307</th>\n",
       "      <td>68</td>\n",
       "      <td>236.937880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>時代_経度_max</td>\n",
       "      <td>236.706360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>677</th>\n",
       "      <td>438</td>\n",
       "      <td>225.313049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        feature   importance\n",
       "754          時代  5533.974880\n",
       "758        所有者名  1655.755015\n",
       "752         種別2  1583.162767\n",
       "756         所在地   429.101809\n",
       "13   label_所有者名   352.651622\n",
       "504         265   349.998780\n",
       "289          50   324.430539\n",
       "337          98   316.240280\n",
       "496         257   291.254490\n",
       "250          11   283.377229\n",
       "427         188   280.394289\n",
       "268          29   267.115792\n",
       "661         422   253.805949\n",
       "477         238   242.050458\n",
       "381         142   240.584730\n",
       "609         370   239.705039\n",
       "729         490   239.619091\n",
       "307          68   236.937880\n",
       "101   時代_経度_max   236.706360\n",
       "677         438   225.313049"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "IM[2].sort_values('importance', ascending=False).head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.01686719 0.26354591 0.38695678 ... 0.11004132 0.02079375 0.01656444]\n",
      "[0.01023762 0.10690403 0.44758661 ... 0.10618172 0.21969019 0.02287255]\n"
     ]
    }
   ],
   "source": [
    "print(result.test_prediction)  # Test prediction in numpy array\n",
    "print(result.oof_prediction)   # Out-of-fold prediction in numpy array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 1, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[target_col].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best threshold: 0.21212121212121213\n",
      "Best F1 Score: 0.36044362292051757\n"
     ]
    }
   ],
   "source": [
    "# データ全体のF1スコアを計算 & 閾値の最適化\n",
    "from sklearn.metrics import f1_score\n",
    "\n",
    "def find_optimal_threshold(y_true, y_scores):\n",
    "    thresholds = np.linspace(0, 1, 100)\n",
    "    best_threshold = 0\n",
    "    best_score = 0\n",
    "\n",
    "    for threshold in thresholds:\n",
    "        y_pred = (y_scores > threshold).astype(int)\n",
    "        score = f1_score(y_true, y_pred)\n",
    "\n",
    "        if score > best_score:\n",
    "            best_score = score\n",
    "            best_threshold = threshold\n",
    "\n",
    "    return best_threshold, best_score\n",
    "\n",
    "# 例として使用するデータ（実際にはモデルからの予測値と実際のラベルを使用）\n",
    "y_true = train[target_col].values\n",
    "y_scores = result.oof_prediction\n",
    "\n",
    "best_threshold, best_f1_score = find_optimal_threshold(y_true, y_scores)\n",
    "print(f\"Best threshold: {best_threshold}\")\n",
    "print(f\"Best F1 Score: {best_f1_score}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 1 ... 0 0 0]\n",
      "(10065,)\n"
     ]
    }
   ],
   "source": [
    "# 閾値を超える確率を持つ要素を1に、それ以外を0に変換\n",
    "y_pred = (result.test_prediction > best_threshold).astype(int)\n",
    "print(y_pred)\n",
    "print(y_pred.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>is_kokuhou</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10060</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10061</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10062</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10063</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10064</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10065 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       is_kokuhou\n",
       "0               0\n",
       "1               1\n",
       "2               1\n",
       "3               0\n",
       "4               0\n",
       "...           ...\n",
       "10060           0\n",
       "10061           0\n",
       "10062           0\n",
       "10063           0\n",
       "10064           0\n",
       "\n",
       "[10065 rows x 1 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_submission['is_kokuhou'] = y_pred\n",
    "sample_submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    9148\n",
       "1     917\n",
       "Name: is_kokuhou, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_submission[\"is_kokuhou\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_submission.to_csv('lightgbm_ver5.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 6))\n",
    "\n",
    "vmax = .02\n",
    "bins = np.linspace(0, 10, 100)\n",
    "ax.hist(result.test_prediction, bins=bins, density=True, alpha=.5, label=\"Test\")\n",
    "ax.hist(result.oof_prediction, bins=bins, density=True, alpha=.5, label=\"OutOfFold\")\n",
    "ax.grid()\n",
    "ax.legend()\n",
    "# ax.set_title(\"テストと学習時の予測傾向差分\")\n",
    "\n",
    "fig.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.01686719 0.26354591 0.38695678 ... 0.11004132 0.02079375 0.01656444]\n",
      "[0.01023762 0.10690403 0.44758661 ... 0.10618172 0.21969019 0.02287255]\n"
     ]
    }
   ],
   "source": [
    "print(result.test_prediction)\n",
    "print(result.oof_prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_oof = pd.DataFrame({'lgb_ver5_oof':result.oof_prediction})\n",
    "df_test = pd.DataFrame({'lgb_ver5_test' : result.test_prediction})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(6710, 1)\n",
      "(10065, 1)\n"
     ]
    }
   ],
   "source": [
    "print(df_oof.shape)\n",
    "print(df_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_oof.to_csv('lgb_ver5_oof.csv', index=False)\n",
    "df_test.to_csv('lgb_ver5_test.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}