submit_central_booster_0718.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pickle\n",
    "from sklearn.metrics import roc_curve\n",
    "import matplotlib.pyplot as plt\n",
    "from catboost import CatBoostClassifier\n",
    "from functools import reduce\n",
    "from skopt import BayesSearchCV\n",
    "from sklearn.model_selection import StratifiedKFold, GroupKFold, ShuffleSplit\n",
    "import xgboost as xgb\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "def wauc(true, scores):\n",
    "    fpr, tpr, thresholds = roc_curve(true, scores, drop_intermediate=False)\n",
    "    tpr_thresholds = [0.0, 0.4, 1.0]\n",
    "    weights = [2.0, 1.0]\n",
    "    auc_x = 0.0\n",
    "    for idx in range(len(tpr_thresholds)-1):\n",
    "        mask = tpr>=tpr_thresholds[idx]\n",
    "        x = fpr[mask]\n",
    "        y = tpr[mask]  \n",
    "        mask = y>tpr_thresholds[idx+1]\n",
    "        y[mask] = tpr_thresholds[idx+1]\n",
    "        y = y-tpr_thresholds[idx]\n",
    "        auc_x = auc_x + weights[idx]*np.trapz(y,x)\n",
    "    areas = np.array(tpr_thresholds[1:]) - np.array(tpr_thresholds[:-1])\n",
    "    normalization = np.dot(areas, np.array(weights))    \n",
    "    return auc_x/normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "probabilities_zoo_holdout = pd.read_csv('probabilities_zoo_holdout_0718.csv', index_col=0)\n",
    "FEATURES = ['NAME','QF', \n",
    "            'DCTR', 'JRM',\n",
    "            #'SRNet_pc', 'SRNet_pjm', 'SRNet_pjuni', 'SRNet_puerd',\n",
    "            'efn_b4_mish_pc',# 'efn_b4_mish_pjm','efn_b4_mish_pjuni','efn_b4_mish_puerd',\n",
    "            'efn_b2_NR_mish_pc', 'efn_b2_NR_mish_pjm', 'efn_b2_NR_mish_pjuni', 'efn_b2_NR_mish_puerd',\n",
    "            'MixNet_xl_mish_pc', #'MixNet_xl_mish_pjm','MixNet_xl_mish_pjuni','MixNet_xl_mish_puerd',\n",
    "            'efn_b5_mish_pc', #'efn_b5_mish_pjm','efn_b5_mish_pjuni','efn_b5_mish_puerd',\n",
    "            'efn_b6_NR_mish_pc',#'efn_b6_NR_mish_pjm','efn_b6_NR_mish_pjuni','efn_b6_NR_mish_puerd',\n",
    "            'SRNet_noPC70_mckpt_pc',#'SRNet_noPC70_pjm','SRNet_noPC70_pjuni','SRNet_noPC70_puerd'\n",
    "             ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:134: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n"
     ]
    }
   ],
   "source": [
    "class DownsampledShuffleSplit:\n",
    "    def __init__(self, n_splits=3):\n",
    "        self.n_splits = n_splits\n",
    "    def split(self, X, y, groups=None):\n",
    "        for rx, tx in ShuffleSplit(self.n_splits, test_size=.25,).split(X,y):\n",
    "            nix = np.where(y[tx]==0)[0]\n",
    "            pix = np.where(y[tx]>0)[0]\n",
    "            #pixu = pix[:nix.shape[0]]\n",
    "            #np.random.seed(1994)\n",
    "            pixu = np.random.choice(pix, size=nix.shape[0], replace=False)\n",
    "            ix = np.append(nix, pixu)\n",
    "            txm = tx[ix]\n",
    "            yield rx, txm\n",
    "    def get_n_splits(self, X, y, groups=None):\n",
    "        return self.n_splits\n",
    "\n",
    "probabilities_zoo_holdout = probabilities_zoo_holdout[probabilities_zoo_holdout.FOLD == 'TST']\n",
    "labels = ~np.array(probabilities_zoo_holdout.CLASS == 'COVER') + 0\n",
    "df_features = probabilities_zoo_holdout[FEATURES]\n",
    "names = df_features.NAME\n",
    "class gkf:\n",
    "    def __init__(self, groups, n_splits=4):\n",
    "        self.n_splits = n_splits\n",
    "        self.groups=groups\n",
    "    def split(self, X, y, groups=None):\n",
    "        for rx, tx in GroupKFold(n_splits=self.n_splits).split(X,y,groups=self.groups):\n",
    "            yield rx, tx\n",
    "    def get_n_splits(self, X, y, groups=None):\n",
    "        return self.n_splits\n",
    "\n",
    "def scoring(estimator,X,y):\n",
    "    scores = 1-estimator.predict_proba(X)[:,0]\n",
    "    return wauc(y>0, scores)\n",
    "\n",
    "bayes_cv_tuner = BayesSearchCV(\n",
    "     estimator = CatBoostClassifier(\n",
    "         #thread_count=10,\n",
    "         objective = 'Logloss',\n",
    "         cat_features = [0],\n",
    "         iterations=150,\n",
    "         scale_pos_weight=0.3333),\n",
    "    \n",
    "   # estimator = xgb.XGBClassifier(seed=42, tree_method='exact', booster='gblinear', num_boost_round=5000),\n",
    "    \n",
    "    search_spaces = {\n",
    "        'learning_rate': (0.0001, 1.0, 'log-uniform'),\n",
    "        'depth': (2, 13),\n",
    "        #'subsample': (0.5, 1.0, 'uniform'),\n",
    "        'colsample_bylevel': (0.5, 1.0, 'uniform'),\n",
    "        'l2_leaf_reg': (1.0, 50.0, 'uniform'),\n",
    "        #'scale_pos_weight': (0.1, 5.0),\n",
    "        ## XGB\n",
    "        #'colsample_bylevel': (0.3, 1.0, 'uniform'),\n",
    "        #'colsample_bytree': (0.3, 1.0, 'uniform'),\n",
    "        #'colsample_bynode': (0.3, 1.0, 'uniform'),\n",
    "        #'subsample': (0.3, 1.0, 'uniform'),\n",
    "        #'learning_rate': (0.0001, 1.0, 'log-uniform'),\n",
    "        #'eta': (0.01, 1.0, 'uniform'),\n",
    "        #'max_depth': (2, 20),\n",
    "        #'alpha': (0.0001, 1.0, 'log-uniform'),\n",
    "        #'lambda':  (0.0001, 1.0, 'log-uniform'),\n",
    "        #'n_estimators': (100, 300, 'uniform'),\n",
    "        #'min_child_weight': (1, 10),\n",
    "        #'scale_pos_weight': (0.1, 1.0, 'uniform'),\n",
    "        \n",
    "    },    \n",
    "    scoring = scoring,\n",
    "    cv = DownsampledShuffleSplit(\n",
    "        n_splits=10,\n",
    "    ),\n",
    "    n_jobs = 10,\n",
    "    n_iter = 60,   \n",
    "    verbose = 0,\n",
    "    refit = True,\n",
    "    random_state = None\n",
    ")\n",
    "\n",
    "def status_print(optim_result):\n",
    "    \"\"\"Status callback durring bayesian hyperparameter search\"\"\"\n",
    "    # Get all the models tested so far in DataFrame format\n",
    "    all_models = bayes_cv_tuner.cv_results_  \n",
    "    # Get current parameters and the best parameters    \n",
    "    best_params = bayes_cv_tuner.best_params_\n",
    "    print('Model #{}\\nBest WAUC: {}\\nBest params: {}\\n'.format(\n",
    "        len(all_models),\n",
    "        np.round(bayes_cv_tuner.best_score_, 4),\n",
    "        bayes_cv_tuner.best_params_\n",
    "    ))\n",
    "    \n",
    "df_features['QF'] = df_features['QF'].astype('int').astype('str')\n",
    "\n",
    "if False:\n",
    "    oh = OneHotEncoder()\n",
    "    oh.fit(df_features['QF'].values.reshape(-1, 1))\n",
    "    qf_oh = oh.transform(df_features['QF'].values.reshape(-1, 1)).toarray()\n",
    "    X = df_features.values[:,2:]\n",
    "    X = np.c_[qf_oh,X]\n",
    "else: \n",
    "    X = df_features.values[:,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model #22\n",
      "Best WAUC: 0.9374\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.8874215816471416), ('depth', 6), ('l2_leaf_reg', 1.1479870990738419), ('learning_rate', 0.004673743210819987)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/skopt/optimizer/optimizer.py:409: UserWarning: The objective has been evaluated at this point before.\n",
      "  warnings.warn(\"The objective has been evaluated \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9409\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.9486733496674717), ('depth', 6), ('l2_leaf_reg', 40.990403393063545), ('learning_rate', 0.01786771119017783)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/skopt/optimizer/optimizer.py:409: UserWarning: The objective has been evaluated at this point before.\n",
      "  warnings.warn(\"The objective has been evaluated \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/skopt/optimizer/optimizer.py:409: UserWarning: The objective has been evaluated at this point before.\n",
      "  warnings.warn(\"The objective has been evaluated \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "Model #22\n",
      "Best WAUC: 0.9413\n",
      "Best params: OrderedDict([('colsample_bylevel', 0.5116707529466404), ('depth', 3), ('l2_leaf_reg', 49.91897773793486), ('learning_rate', 0.00010010626734225567)])\n",
      "\n",
      "0:\tlearn: 0.6931004\ttotal: 3.92ms\tremaining: 585ms\n",
      "1:\tlearn: 0.6930519\ttotal: 7.95ms\tremaining: 588ms\n",
      "2:\tlearn: 0.6930025\ttotal: 11.4ms\tremaining: 557ms\n",
      "3:\tlearn: 0.6929540\ttotal: 14.8ms\tremaining: 541ms\n",
      "4:\tlearn: 0.6929058\ttotal: 18ms\tremaining: 521ms\n",
      "5:\tlearn: 0.6928569\ttotal: 20.7ms\tremaining: 498ms\n",
      "6:\tlearn: 0.6928096\ttotal: 23.4ms\tremaining: 478ms\n",
      "7:\tlearn: 0.6927625\ttotal: 26.4ms\tremaining: 468ms\n",
      "8:\tlearn: 0.6927142\ttotal: 29.4ms\tremaining: 461ms\n",
      "9:\tlearn: 0.6926648\ttotal: 32.2ms\tremaining: 451ms\n",
      "10:\tlearn: 0.6926181\ttotal: 35ms\tremaining: 442ms\n",
      "11:\tlearn: 0.6925691\ttotal: 38ms\tremaining: 437ms\n",
      "12:\tlearn: 0.6925204\ttotal: 40.8ms\tremaining: 430ms\n",
      "13:\tlearn: 0.6924727\ttotal: 43.4ms\tremaining: 422ms\n",
      "14:\tlearn: 0.6924253\ttotal: 46.3ms\tremaining: 417ms\n",
      "15:\tlearn: 0.6923775\ttotal: 49ms\tremaining: 410ms\n",
      "16:\tlearn: 0.6923284\ttotal: 51.5ms\tremaining: 403ms\n",
      "17:\tlearn: 0.6922805\ttotal: 54.1ms\tremaining: 397ms\n",
      "18:\tlearn: 0.6922314\ttotal: 57.1ms\tremaining: 394ms\n",
      "19:\tlearn: 0.6921834\ttotal: 60ms\tremaining: 390ms\n",
      "20:\tlearn: 0.6921343\ttotal: 62.7ms\tremaining: 385ms\n",
      "21:\tlearn: 0.6920848\ttotal: 65.4ms\tremaining: 380ms\n",
      "22:\tlearn: 0.6920364\ttotal: 68ms\tremaining: 375ms\n",
      "23:\tlearn: 0.6919897\ttotal: 70.7ms\tremaining: 371ms\n",
      "24:\tlearn: 0.6919417\ttotal: 73.4ms\tremaining: 367ms\n",
      "25:\tlearn: 0.6918922\ttotal: 76.3ms\tremaining: 364ms\n",
      "26:\tlearn: 0.6918441\ttotal: 78.8ms\tremaining: 359ms\n",
      "27:\tlearn: 0.6917951\ttotal: 81.5ms\tremaining: 355ms\n",
      "28:\tlearn: 0.6917469\ttotal: 84.4ms\tremaining: 352ms\n",
      "29:\tlearn: 0.6916984\ttotal: 87.3ms\tremaining: 349ms\n",
      "30:\tlearn: 0.6916496\ttotal: 90.2ms\tremaining: 346ms\n",
      "31:\tlearn: 0.6916011\ttotal: 93ms\tremaining: 343ms\n",
      "32:\tlearn: 0.6915517\ttotal: 95.5ms\tremaining: 339ms\n",
      "33:\tlearn: 0.6915030\ttotal: 98.3ms\tremaining: 335ms\n",
      "34:\tlearn: 0.6914564\ttotal: 101ms\tremaining: 332ms\n",
      "35:\tlearn: 0.6914091\ttotal: 104ms\tremaining: 329ms\n",
      "36:\tlearn: 0.6913620\ttotal: 107ms\tremaining: 326ms\n",
      "37:\tlearn: 0.6913156\ttotal: 110ms\tremaining: 323ms\n",
      "38:\tlearn: 0.6912689\ttotal: 112ms\tremaining: 320ms\n",
      "39:\tlearn: 0.6912207\ttotal: 115ms\tremaining: 316ms\n",
      "40:\tlearn: 0.6911727\ttotal: 118ms\tremaining: 313ms\n",
      "41:\tlearn: 0.6911254\ttotal: 121ms\tremaining: 310ms\n",
      "42:\tlearn: 0.6910791\ttotal: 123ms\tremaining: 307ms\n",
      "43:\tlearn: 0.6910306\ttotal: 126ms\tremaining: 304ms\n",
      "44:\tlearn: 0.6909840\ttotal: 129ms\tremaining: 301ms\n",
      "45:\tlearn: 0.6909356\ttotal: 132ms\tremaining: 298ms\n",
      "46:\tlearn: 0.6908871\ttotal: 134ms\tremaining: 294ms\n",
      "47:\tlearn: 0.6908380\ttotal: 137ms\tremaining: 291ms\n",
      "48:\tlearn: 0.6907914\ttotal: 140ms\tremaining: 288ms\n",
      "49:\tlearn: 0.6907421\ttotal: 143ms\tremaining: 285ms\n",
      "50:\tlearn: 0.6906945\ttotal: 145ms\tremaining: 282ms\n",
      "51:\tlearn: 0.6906491\ttotal: 148ms\tremaining: 279ms\n",
      "52:\tlearn: 0.6906010\ttotal: 150ms\tremaining: 275ms\n",
      "53:\tlearn: 0.6905548\ttotal: 153ms\tremaining: 272ms\n",
      "54:\tlearn: 0.6905089\ttotal: 156ms\tremaining: 269ms\n",
      "55:\tlearn: 0.6904601\ttotal: 161ms\tremaining: 271ms\n",
      "56:\tlearn: 0.6904112\ttotal: 165ms\tremaining: 270ms\n",
      "57:\tlearn: 0.6903655\ttotal: 171ms\tremaining: 271ms\n",
      "58:\tlearn: 0.6903169\ttotal: 177ms\tremaining: 273ms\n",
      "59:\tlearn: 0.6902705\ttotal: 183ms\tremaining: 275ms\n",
      "60:\tlearn: 0.6902249\ttotal: 189ms\tremaining: 276ms\n",
      "61:\tlearn: 0.6901785\ttotal: 192ms\tremaining: 272ms\n",
      "62:\tlearn: 0.6901303\ttotal: 194ms\tremaining: 268ms\n",
      "63:\tlearn: 0.6900811\ttotal: 197ms\tremaining: 265ms\n",
      "64:\tlearn: 0.6900342\ttotal: 200ms\tremaining: 262ms\n",
      "65:\tlearn: 0.6899863\ttotal: 203ms\tremaining: 258ms\n",
      "66:\tlearn: 0.6899395\ttotal: 206ms\tremaining: 255ms\n",
      "67:\tlearn: 0.6898915\ttotal: 208ms\tremaining: 251ms\n",
      "68:\tlearn: 0.6898439\ttotal: 211ms\tremaining: 248ms\n",
      "69:\tlearn: 0.6897955\ttotal: 214ms\tremaining: 244ms\n",
      "70:\tlearn: 0.6897481\ttotal: 216ms\tremaining: 241ms\n",
      "71:\tlearn: 0.6897012\ttotal: 219ms\tremaining: 237ms\n",
      "72:\tlearn: 0.6896539\ttotal: 222ms\tremaining: 235ms\n",
      "73:\tlearn: 0.6896048\ttotal: 232ms\tremaining: 238ms\n",
      "74:\tlearn: 0.6895586\ttotal: 241ms\tremaining: 241ms\n",
      "75:\tlearn: 0.6895097\ttotal: 245ms\tremaining: 238ms\n",
      "76:\tlearn: 0.6894616\ttotal: 253ms\tremaining: 240ms\n",
      "77:\tlearn: 0.6894133\ttotal: 261ms\tremaining: 240ms\n",
      "78:\tlearn: 0.6893654\ttotal: 263ms\tremaining: 237ms\n",
      "79:\tlearn: 0.6893169\ttotal: 266ms\tremaining: 233ms\n",
      "80:\tlearn: 0.6892723\ttotal: 269ms\tremaining: 229ms\n",
      "81:\tlearn: 0.6892273\ttotal: 272ms\tremaining: 225ms\n",
      "82:\tlearn: 0.6891798\ttotal: 274ms\tremaining: 221ms\n",
      "83:\tlearn: 0.6891345\ttotal: 277ms\tremaining: 218ms\n",
      "84:\tlearn: 0.6890881\ttotal: 280ms\tremaining: 214ms\n",
      "85:\tlearn: 0.6890403\ttotal: 282ms\tremaining: 210ms\n",
      "86:\tlearn: 0.6889938\ttotal: 285ms\tremaining: 206ms\n",
      "87:\tlearn: 0.6889472\ttotal: 288ms\tremaining: 203ms\n",
      "88:\tlearn: 0.6888997\ttotal: 291ms\tremaining: 199ms\n",
      "89:\tlearn: 0.6888546\ttotal: 293ms\tremaining: 195ms\n",
      "90:\tlearn: 0.6888054\ttotal: 296ms\tremaining: 192ms\n",
      "91:\tlearn: 0.6887569\ttotal: 298ms\tremaining: 188ms\n",
      "92:\tlearn: 0.6887108\ttotal: 301ms\tremaining: 185ms\n",
      "93:\tlearn: 0.6886631\ttotal: 304ms\tremaining: 181ms\n",
      "94:\tlearn: 0.6886179\ttotal: 306ms\tremaining: 177ms\n",
      "95:\tlearn: 0.6885713\ttotal: 309ms\tremaining: 174ms\n",
      "96:\tlearn: 0.6885255\ttotal: 312ms\tremaining: 170ms\n",
      "97:\tlearn: 0.6884770\ttotal: 314ms\tremaining: 167ms\n",
      "98:\tlearn: 0.6884284\ttotal: 317ms\tremaining: 163ms\n",
      "99:\tlearn: 0.6883808\ttotal: 320ms\tremaining: 160ms\n",
      "100:\tlearn: 0.6883322\ttotal: 323ms\tremaining: 156ms\n",
      "101:\tlearn: 0.6882830\ttotal: 325ms\tremaining: 153ms\n",
      "102:\tlearn: 0.6882364\ttotal: 328ms\tremaining: 150ms\n",
      "103:\tlearn: 0.6881902\ttotal: 331ms\tremaining: 146ms\n",
      "104:\tlearn: 0.6881423\ttotal: 333ms\tremaining: 143ms\n",
      "105:\tlearn: 0.6880955\ttotal: 336ms\tremaining: 139ms\n",
      "106:\tlearn: 0.6880476\ttotal: 339ms\tremaining: 136ms\n",
      "107:\tlearn: 0.6880010\ttotal: 341ms\tremaining: 133ms\n",
      "108:\tlearn: 0.6879534\ttotal: 344ms\tremaining: 129ms\n",
      "109:\tlearn: 0.6879045\ttotal: 347ms\tremaining: 126ms\n",
      "110:\tlearn: 0.6878570\ttotal: 350ms\tremaining: 123ms\n",
      "111:\tlearn: 0.6878087\ttotal: 355ms\tremaining: 121ms\n",
      "112:\tlearn: 0.6877618\ttotal: 362ms\tremaining: 119ms\n",
      "113:\tlearn: 0.6877139\ttotal: 366ms\tremaining: 116ms\n",
      "114:\tlearn: 0.6876666\ttotal: 369ms\tremaining: 112ms\n",
      "115:\tlearn: 0.6876185\ttotal: 371ms\tremaining: 109ms\n",
      "116:\tlearn: 0.6875709\ttotal: 374ms\tremaining: 105ms\n",
      "117:\tlearn: 0.6875248\ttotal: 377ms\tremaining: 102ms\n",
      "118:\tlearn: 0.6874778\ttotal: 379ms\tremaining: 98.8ms\n",
      "119:\tlearn: 0.6874302\ttotal: 382ms\tremaining: 95.5ms\n",
      "120:\tlearn: 0.6873820\ttotal: 385ms\tremaining: 92.2ms\n",
      "121:\tlearn: 0.6873370\ttotal: 387ms\tremaining: 88.9ms\n",
      "122:\tlearn: 0.6872888\ttotal: 390ms\tremaining: 85.6ms\n",
      "123:\tlearn: 0.6872415\ttotal: 393ms\tremaining: 82.3ms\n",
      "124:\tlearn: 0.6871931\ttotal: 395ms\tremaining: 79.1ms\n",
      "125:\tlearn: 0.6871486\ttotal: 398ms\tremaining: 75.8ms\n",
      "126:\tlearn: 0.6871027\ttotal: 401ms\tremaining: 72.5ms\n",
      "127:\tlearn: 0.6870572\ttotal: 403ms\tremaining: 69.3ms\n",
      "128:\tlearn: 0.6870099\ttotal: 406ms\tremaining: 66.1ms\n",
      "129:\tlearn: 0.6869610\ttotal: 409ms\tremaining: 62.9ms\n",
      "130:\tlearn: 0.6869133\ttotal: 411ms\tremaining: 59.6ms\n",
      "131:\tlearn: 0.6868662\ttotal: 414ms\tremaining: 56.5ms\n",
      "132:\tlearn: 0.6868179\ttotal: 417ms\tremaining: 53.2ms\n",
      "133:\tlearn: 0.6867698\ttotal: 419ms\tremaining: 50.1ms\n",
      "134:\tlearn: 0.6867229\ttotal: 422ms\tremaining: 46.9ms\n",
      "135:\tlearn: 0.6866752\ttotal: 425ms\tremaining: 43.7ms\n",
      "136:\tlearn: 0.6866299\ttotal: 428ms\tremaining: 40.6ms\n",
      "137:\tlearn: 0.6865837\ttotal: 430ms\tremaining: 37.4ms\n",
      "138:\tlearn: 0.6865365\ttotal: 433ms\tremaining: 34.3ms\n",
      "139:\tlearn: 0.6864901\ttotal: 436ms\tremaining: 31.1ms\n",
      "140:\tlearn: 0.6864430\ttotal: 439ms\tremaining: 28ms\n",
      "141:\tlearn: 0.6863977\ttotal: 442ms\tremaining: 24.9ms\n",
      "142:\tlearn: 0.6863500\ttotal: 444ms\tremaining: 21.8ms\n",
      "143:\tlearn: 0.6863020\ttotal: 447ms\tremaining: 18.6ms\n",
      "144:\tlearn: 0.6862542\ttotal: 450ms\tremaining: 15.5ms\n",
      "145:\tlearn: 0.6862101\ttotal: 452ms\tremaining: 12.4ms\n",
      "146:\tlearn: 0.6861629\ttotal: 455ms\tremaining: 9.29ms\n",
      "147:\tlearn: 0.6861176\ttotal: 458ms\tremaining: 6.19ms\n",
      "148:\tlearn: 0.6860717\ttotal: 460ms\tremaining: 3.09ms\n",
      "149:\tlearn: 0.6860246\ttotal: 463ms\tremaining: 0us\n"
     ]
    }
   ],
   "source": [
    "result = bayes_cv_tuner.fit(X, \n",
    "                            labels, \n",
    "                            callback=status_print)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "means = np.array(result.cv_results_['mean_test_score'])\n",
    "means_cv = [np.array(result.cv_results_['split'+str(k)+'_test_score']) for k in range(10)]\n",
    "\n",
    "stds = np.array(result.cv_results_['std_test_score'])\n",
    "ranks = np.array(result.cv_results_['rank_test_score'])\n",
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "for k in range(10):\n",
    "    plt.plot(means_cv[k][:], alpha=0.5)\n",
    "    \n",
    "#plt.fill_between(range(len(means)),means-stds,means+stds, alpha=.2, label='+/- std')\n",
    "plt.scatter(np.where(ranks==1)[0],means[ranks==1], label='optimum')\n",
    "plt.ylabel('10 folds shuffle split (test size=1.5k+1.5k) wauc')\n",
    "plt.xlabel('Hyper parameter optimization steps')\n",
    "plt.ylim([0.932,0.95])\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "probabilities_zoo_lb = pd.read_csv('probabilities_zoo_lb_0718.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_features = probabilities_zoo_lb[FEATURES]\n",
    "df_features['QF'] = df_features['QF'].astype('int').astype('str')\n",
    "\n",
    "if False:\n",
    "    qf_oh = oh.transform(df_features['QF'].values.reshape(-1, 1)).toarray()\n",
    "    X_lb = df_features.values[:,2:]\n",
    "    X_lb = np.c_[qf_oh,X_lb]\n",
    "else:\n",
    "    X_lb = df_features.values[:,1:]\n",
    "\n",
    "estimator = result.best_estimator_                                     \n",
    "score_lb = estimator.predict_proba(X_lb)[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4127    0.989025\n",
      "Name: Label, dtype: float64\n",
      "1799    0.979589\n",
      "Name: Label, dtype: float64\n",
      "1715    0.974008\n",
      "Name: Label, dtype: float64\n",
      "2866    0.97226\n",
      "Name: Label, dtype: float64\n",
      "3244    0.837765\n",
      "Name: Label, dtype: float64\n",
      "1043    0.983609\n",
      "Name: Label, dtype: float64\n",
      "4454    0.98464\n",
      "Name: Label, dtype: float64\n",
      "373    0.973416\n",
      "Name: Label, dtype: float64\n",
      "4791    0.955109\n",
      "Name: Label, dtype: float64\n",
      "1770    0.987538\n",
      "Name: Label, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "plt.hist(score_lb,bins=50)\n",
    "plt.show()\n",
    "sub = pd.DataFrame({\"Id\":df_features.NAME, \"Label\":score_lb})\n",
    "oor = ['2395.jpg' ,'0412.jpg', '0146.jpg', '3578.jpg', '4747.jpg', '3128.jpg', '3395.jpg', '1208.jpg', '4415.jpg', '0327.jpg']\n",
    "for im_name in oor:\n",
    "    print(sub.loc[sub.Id==im_name, 'Label'])\n",
    "    sub.loc[sub.Id==im_name, 'Label'] = 1.01\n",
    "#sub.to_csv('submission_v28_all_xgb.csv', index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}