notebooks/CrossModelFabric_class.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CrossModelFabric class"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`CrossModelFabric` - abstract fabric for easy crossvalidation\n",
    "\n",
    "Notebook contains examples for `CrossLightgbmModel`, but there are `CrossXgboostModel` option as well!\n",
    "\n",
    "* CrossModelFabric.fit() fits models for each fold with early_stopping. Finds optimal num_boost parameter\n",
    "* CrossModelFabric.transform() - calculates OOF predictions for train data (with num_boost parameter)\n",
    "* CrossModelFabric.predict() - calculates predictions test data (with num_boost parameter)\n",
    "* CrossModelFabric.get_hyperopt_space() - returns hyperopt space\n",
    "* CrossModelFabric.shap() - returns figure with barplot of avg shap values for all folds\n",
    "* CrossModelFabric.shap_summary_plot() - returns figure with shap_summary for specific dataset of model[0]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:20:29.593396Z",
     "start_time": "2020-01-21T14:20:28.203623Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import requests\n",
    "import numpy as np\n",
    "import yaml\n",
    "import logging\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from crosspredict.iterator import Iterator\n",
    "from crosspredict.crossval import CrossLightgbmModel\n",
    "logging.basicConfig(level=logging.INFO)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:20:30.348914Z",
     "start_time": "2020-01-21T14:20:29.595148Z"
    }
   },
   "outputs": [],
   "source": [
    "file_url = 'https://boosters.pro/api/ch/files/pub/onetwotrip_challenge_train.csv'\n",
    "file_path = '../tests/onetwotrip_challenge_train.csv'\n",
    "if os.path.isfile(file_path)!=True:\n",
    "    myfile = requests.get(file_url)\n",
    "    open(file_path, 'wb').write(myfile.content)\n",
    "    \n",
    "df = pd.read_csv('../tests/onetwotrip_challenge_train.csv')\n",
    "\n",
    "\n",
    "unique_clients = pd.Series(df['userid'].unique())\n",
    "test_users = unique_clients.sample(frac=0.2,random_state=0)\n",
    "val_idx = df['userid'].isin(test_users)\n",
    "test = df[val_idx].copy()\n",
    "train = df[~val_idx].copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:20:30.356023Z",
     "start_time": "2020-01-21T14:20:30.350497Z"
    }
   },
   "outputs": [],
   "source": [
    "feature_name = df.columns.values\n",
    "feature_name = np.delete(feature_name, np.argwhere(feature_name == 'goal1'))\n",
    "feature_name = np.delete(feature_name, np.argwhere(feature_name == 'orderid'))\n",
    "feature_name = np.delete(feature_name, np.argwhere(feature_name == 'userid'))\n",
    "\n",
    "params= {'bagging_fraction': 0.849285747554019,\n",
    "  'bagging_freq': 5,\n",
    "  'bagging_seed': 0,\n",
    "  'boosting_type': 'gbdt',\n",
    "  'data_random_seed': 0,\n",
    "  'drop_seed': 0,\n",
    "  'feature_fraction': 0.8212766928844304,\n",
    "  'feature_fraction_seed': 0,\n",
    "  'lambda_l1': 0.8955546599539566,\n",
    "  'lambda_l2': 1.4423261095989717,\n",
    "  'learning_rate': 0.03,\n",
    "  'max_bin': 255,\n",
    "  'max_depth': 43,\n",
    "  'metric': 'auc',\n",
    "  'min_data_in_leaf': 149,\n",
    "  'min_sum_hessian_in_leaf': 1.804477623298885,\n",
    "  'num_leaves': 363,\n",
    "  'objective': 'binary',\n",
    "  'seed': 0,\n",
    "  'verbose': -1}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Crossvalidation by `col_client` column (rows from each users goes in specific fold)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### fit model on train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:22:58.673036Z",
     "start_time": "2020-01-21T14:20:30.357298Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using RepeatedKFold by column group \"userid\"\n",
      "INFO:crosspredict.crossval:{'bagging_fraction': 0.849285747554019, 'bagging_freq': 5, 'bagging_seed': 0, 'boosting_type': 'gbdt', 'data_random_seed': 0, 'drop_seed': 0, 'feature_fraction': 0.8212766928844304, 'feature_fraction_seed': 0, 'lambda_l1': 0.8955546599539566, 'lambda_l2': 1.4423261095989717, 'learning_rate': 0.03, 'max_bin': 255, 'max_depth': 43, 'metric': 'auc', 'min_data_in_leaf': 149, 'min_sum_hessian_in_leaf': 1.804477623298885, 'num_leaves': 363, 'objective': 'binary', 'seed': 0, 'verbose': -1}\n",
      "INFO:crosspredict.crossval:REPEAT FOLDS 0 START\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 0 ENDS with best ROCAUC = 0.697301828294343\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 1 ENDS with best ROCAUC = 0.6972290051903695\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 2 ENDS with best ROCAUC = 0.7077976895321761\n",
      "INFO:crosspredict.crossval:REPEAT FOLDS 1 START\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 0 ENDS with best ROCAUC = 0.6977572363632991\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 1 ENDS with best ROCAUC = 0.7060994308412251\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 2 ENDS with best ROCAUC = 0.6963671472236923\n",
      "INFO:crosspredict.crossval:REPEAT FOLDS 2 START\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 0 ENDS with best ROCAUC = 0.698132614504838\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 1 ENDS with best ROCAUC = 0.6969982720983842\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 2 ENDS with best ROCAUC = 0.7176314682411897\n",
      "INFO:crosspredict.crossval:{'loss': -0.7009149200802658, 'status': 'ok', 'std': 0.00707587736681883, 'score_max': 0.7009149200802658, 'scores_all': [0.697301828294343, 0.6972290051903695, 0.7077976895321761, 0.6977572363632991, 0.7060994308412251, 0.6963671472236923, 0.698132614504838, 0.6969982720983842, 0.7176314682411897], 'num_boost': 60}\n"
     ]
    }
   ],
   "source": [
    "iter_df = Iterator(n_repeats=3,\n",
    "                    n_splits=3,\n",
    "                    random_state = 0,\n",
    "                    col_client = 'userid',\n",
    "                    cv_byclient=True)\n",
    "\n",
    "model_class = CrossLightgbmModel(iterator=iter_df, \n",
    "                                 feature_name=feature_name,\n",
    "                                params=params,\n",
    "                                cols_cat = ['field3', 'field2', 'field11', 'field23', 'field18', 'field20'],\n",
    "                                num_boost_round = 9999,\n",
    "                                early_stopping_rounds = 50,\n",
    "                                valid = True,\n",
    "                                random_state = 0,\n",
    "                                col_target = 'goal1',)\n",
    "result = model_class.fit(train)\n",
    "\n",
    "# `fit` method returns loss from optimal num_boost parameter\n",
    "# optimal num_boost for cross validation:\n",
    "# - removes roc_auc metrics if at least one fold stopped before (early_stopping)\n",
    "# - calculates avg roc_auc on all folds on each num_boost\n",
    "# - argmax value"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### predict on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:22:59.647460Z",
     "start_time": "2020-01-21T14:22:58.674560Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3         0.029087\n",
       "6         0.021241\n",
       "9         0.024901\n",
       "12        0.009496\n",
       "14        0.011062\n",
       "            ...   \n",
       "196039    0.021094\n",
       "196042    0.027103\n",
       "196045    0.010621\n",
       "196046    0.012720\n",
       "196048    0.021069\n",
       "Name: PREDICT, Length: 38859, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# returns predctions from optimal num_boost parameter (mean prediction from all models)\n",
    "model_class.predict(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:00.630360Z",
     "start_time": "2020-01-21T14:22:59.650407Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7189406147162991"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(test['goal1'],model_class.predict(test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T13:43:18.402606Z",
     "start_time": "2020-01-21T13:40:56.713Z"
    }
   },
   "source": [
    "### predict on train data. (OOF - out of fold prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:02.726509Z",
     "start_time": "2020-01-21T14:23:00.632222Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0         0.018500\n",
       "1         0.031481\n",
       "2         0.011632\n",
       "4         0.020763\n",
       "5         0.016286\n",
       "            ...   \n",
       "196051    0.018088\n",
       "196052    0.012538\n",
       "196053    0.019197\n",
       "196054    0.042271\n",
       "196055    0.014101\n",
       "Name: PREDICT, Length: 157197, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_class.transform(train)\n",
    "\n",
    "# returns OOF predctions from optimal num_boost parameter \n",
    "# (mean prediction if n_repeats>1, because there are more than one model that can predict each row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:04.606945Z",
     "start_time": "2020-01-21T14:23:02.728030Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7083991804553007"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(train['goal1'],model_class.transform(train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### hyperopt hyperparameter optimisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:04.610871Z",
     "start_time": "2020-01-21T14:23:04.608397Z"
    }
   },
   "outputs": [],
   "source": [
    "# get standart hyperopt space for lightgbm\n",
    "space = model_class.get_hyperopt_space()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:29:44.325566Z",
     "start_time": "2020-01-21T14:29:44.320389Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'num_leaves': <hyperopt.pyll.base.Apply at 0x7f89bec70e90>,\n",
       " 'max_depth': <hyperopt.pyll.base.Apply at 0x7f89bec6c2d0>,\n",
       " 'min_data_in_leaf': <hyperopt.pyll.base.Apply at 0x7f89bec6c4d0>,\n",
       " 'feature_fraction': <hyperopt.pyll.base.Apply at 0x7f89bec6c650>,\n",
       " 'bagging_fraction': <hyperopt.pyll.base.Apply at 0x7f89bec6c7d0>,\n",
       " 'min_sum_hessian_in_leaf': <hyperopt.pyll.base.Apply at 0x7f89bec6c950>,\n",
       " 'lambda_l1': <hyperopt.pyll.base.Apply at 0x7f89bec6cad0>,\n",
       " 'lambda_l2': <hyperopt.pyll.base.Apply at 0x7f89bec6cc50>,\n",
       " 'seed': 0,\n",
       " 'feature_fraction_seed': 0,\n",
       " 'bagging_seed': 0,\n",
       " 'drop_seed': 0,\n",
       " 'data_random_seed': 0,\n",
       " 'verbose': -1,\n",
       " 'bagging_freq': 5,\n",
       " 'max_bin': 255,\n",
       " 'learning_rate': 0.03,\n",
       " 'boosting_type': 'gbdt',\n",
       " 'objective': 'binary',\n",
       " 'metric': 'auc'}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:04.619873Z",
     "start_time": "2020-01-21T14:23:04.612318Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 int\n",
      "1   float\n",
      "2     hyperopt_param\n",
      "3       Literal{num_leaves}\n",
      "4       quniform\n",
      "5         Literal{300}\n",
      "6         Literal{500}\n",
      "7         Literal{1}\n"
     ]
    }
   ],
   "source": [
    "print(str(space['num_leaves']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:04.627162Z",
     "start_time": "2020-01-21T14:23:04.621232Z"
    }
   },
   "outputs": [],
   "source": [
    "from hyperopt import fmin, hp, tpe, Trials, space_eval, STATUS_OK, STATUS_RUNNING\n",
    "from hyperopt.pyll import scope\n",
    "\n",
    "# get standart hyperopt space for lightgbm with custom options\n",
    "space = model_class.get_hyperopt_space({'num_leaves': scope.int(hp.quniform('num_leaves', 100, 200, 1))})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:04.637995Z",
     "start_time": "2020-01-21T14:23:04.628780Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 int\n",
      "1   float\n",
      "2     hyperopt_param\n",
      "3       Literal{num_leaves}\n",
      "4       quniform\n",
      "5         Literal{100}\n",
      "6         Literal{200}\n",
      "7         Literal{1}\n"
     ]
    }
   ],
   "source": [
    "print(str(space['num_leaves']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:30.176677Z",
     "start_time": "2020-01-21T14:23:04.639459Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/2 [00:00<?, ?trial/s, best loss=?]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using RepeatedKFold by column group \"userid\"\n",
      "INFO:hyperopt.tpe:build_posterior_wrapper took 0.002975 seconds\n",
      "INFO:hyperopt.tpe:TPE using 0 trials\n",
      "INFO:crosspredict.crossval:{'bagging_fraction': 0.9230857926783654, 'bagging_freq': 5, 'bagging_seed': 0, 'boosting_type': 'gbdt', 'data_random_seed': 0, 'drop_seed': 0, 'feature_fraction': 0.9920158580306366, 'feature_fraction_seed': 0, 'lambda_l1': 0.6803407372529847, 'lambda_l2': 0.3187092362014174, 'learning_rate': 0.03, 'max_bin': 255, 'max_depth': 53, 'metric': 'auc', 'min_data_in_leaf': 91, 'min_sum_hessian_in_leaf': 4.753894156420975, 'num_leaves': 119, 'objective': 'binary', 'seed': 0, 'verbose': -1}\n",
      "INFO:crosspredict.crossval:REPEAT FOLDS 0 START\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 0 ENDS with best ROCAUC = 0.6917724736953605\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 1 ENDS with best ROCAUC = 0.6981503974145464\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 2 ENDS with best ROCAUC = 0.6998738725344477\n",
      "INFO:crosspredict.crossval:{'loss': -0.6963528220111866, 'status': 'ok', 'std': 0.00402749360740368, 'score_max': 0.6963528220111866, 'scores_all': [0.6917724736953605, 0.6981503974145464, 0.6998738725344477], 'num_boost': 96}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 1/2 [00:12<00:12, 12.80s/trial, best loss: -0.6963528220111866]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:hyperopt.tpe:build_posterior_wrapper took 0.004642 seconds\n",
      "INFO:hyperopt.tpe:TPE using 1/1 trials with best loss -0.696353\n",
      "INFO:crosspredict.crossval:{'bagging_fraction': 0.9905357561494915, 'bagging_freq': 5, 'bagging_seed': 0, 'boosting_type': 'gbdt', 'data_random_seed': 0, 'drop_seed': 0, 'feature_fraction': 0.8737863546856102, 'feature_fraction_seed': 0, 'lambda_l1': 0.975105842345082, 'lambda_l2': 1.6686950583315598, 'learning_rate': 0.03, 'max_bin': 255, 'max_depth': 13, 'metric': 'auc', 'min_data_in_leaf': 86, 'min_sum_hessian_in_leaf': 1.428615700376287, 'num_leaves': 152, 'objective': 'binary', 'seed': 0, 'verbose': -1}\n",
      "INFO:crosspredict.crossval:REPEAT FOLDS 0 START\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 0 ENDS with best ROCAUC = 0.6985478224903406\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 1 ENDS with best ROCAUC = 0.7016942586308145\n",
      "INFO:crosspredict.crossval:\tCROSSVALIDATION FOLD 2 ENDS with best ROCAUC = 0.7102799688122039\n",
      "INFO:crosspredict.crossval:{'loss': -0.7022219455448453, 'status': 'ok', 'std': 0.005249014913812784, 'score_max': 0.7022219455448453, 'scores_all': [0.6985478224903406, 0.7016942586308145, 0.7102799688122039], 'num_boost': 64}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2/2 [00:25<00:00, 12.76s/trial, best loss: -0.7022219455448453]\n",
      "{'bagging_fraction': 0.9905357561494915, 'bagging_freq': 5, 'bagging_seed': 0, 'boosting_type': 'gbdt', 'data_random_seed': 0, 'drop_seed': 0, 'feature_fraction': 0.8737863546856102, 'feature_fraction_seed': 0, 'lambda_l1': 0.975105842345082, 'lambda_l2': 1.6686950583315598, 'learning_rate': 0.03, 'max_bin': 255, 'max_depth': 13, 'metric': 'auc', 'min_data_in_leaf': 86, 'min_sum_hessian_in_leaf': 1.428615700376287, 'num_leaves': 152, 'objective': 'binary', 'seed': 0, 'verbose': -1}\n"
     ]
    }
   ],
   "source": [
    "iter_df = Iterator(n_repeats=1,\n",
    "                    n_splits=3,\n",
    "                    random_state = 0,\n",
    "                    col_client = 'userid',\n",
    "                    cv_byclient=True)\n",
    "\n",
    "def score(params):\n",
    "    cv_score = CrossLightgbmModel(iterator=iter_df, \n",
    "                                 feature_name=feature_name,\n",
    "                                params=params,\n",
    "                                cols_cat = ['field3', 'field2', 'field11', 'field23', 'field18', 'field20'],\n",
    "                                num_boost_round = 9999,\n",
    "                                early_stopping_rounds = 50,\n",
    "                                valid = True,\n",
    "                                random_state = 0,\n",
    "                                col_target = 'goal1',)\n",
    "    return cv_score.fit(df=train)\n",
    "\n",
    "trials = Trials()\n",
    "best = fmin(fn=score,\n",
    "            space=space,\n",
    "            algo=tpe.suggest,\n",
    "            trials=trials,\n",
    "            max_evals=2\n",
    "            )\n",
    "results = space_eval(space, best)\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### .shap() method - returns dataframe with avg shap values from each model in crossvalidation and its std + barplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:43.495726Z",
     "start_time": "2020-01-21T14:23:30.178160Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Setting feature_perturbation = \"tree_path_dependent\" because no background data was given.\n",
      "LightGBM binary classifier with TreeExplainer shap values output has changed to a list of ndarray\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap_df, fig = model_class.shap(train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### .shap_summary_plot() returns figure of shap's summary_plot() from model[0] on any dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-21T14:23:44.904820Z",
     "start_time": "2020-01-21T14:23:43.497114Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = model_class.shap_summary_plot(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "kedro_test",
   "language": "python",
   "name": "kedro_test"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}