predict-the-quality-of-freshwater-model.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"### Introduction\n### Build a machine model to predict whether the freshwater is safe to drink or not.Based on the measures like pH, TDS, etc","metadata":{}},{"cell_type":"markdown","source":"### Import necessary packages and libraries","metadata":{}},{"cell_type":"code","source":"\n!pip install -U modin[all] -q --progress-bar off","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport seaborn as sns\nfrom matplotlib import pyplot as plt\n%matplotlib inline\n\nimport warnings\nwarnings.filterwarnings('ignore')\n\nfrom sklearnex import patch_sklearn\npatch_sklearn()\n\nfrom timeit import default_timer as timer\n\nimport wandb\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.preprocessing import LabelEncoder\nfrom sklearn.model_selection import  train_test_split,StratifiedKFold\nimport optuna\nfrom catboost import CatBoostClassifier\nfrom lightgbm import LGBMClassifier\nfrom optuna.integration.wandb import WeightsAndBiasesCallback\nfrom optuna.samplers import RandomSampler,TPESampler,MOTPESampler,CmaEsSampler\nfrom sklearn.metrics import f1_score,classification_report,confusion_matrix,log_loss,ConfusionMatrixDisplay,roc_curve,precision_recall_curve\nfrom xgboost import XGBClassifier,plot_tree\nimport xgboost as xgb\nfrom optuna.integration import XGBoostPruningCallback\nimport math\nimport daal4py as d4p","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:18:06.448638Z","iopub.execute_input":"2023-03-04T07:18:06.449092Z","iopub.status.idle":"2023-03-04T07:18:10.794925Z","shell.execute_reply.started":"2023-03-04T07:18:06.449057Z","shell.execute_reply":"2023-03-04T07:18:10.793830Z"},"trusted":true},"execution_count":8,"outputs":[{"name":"stderr","text":"Intel(R) Extension for Scikit-learn* enabled (https://github.com/intel/scikit-learn-intelex)\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","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"},"metadata":{}}]},{"cell_type":"markdown","source":"### Import dataset","metadata":{}},{"cell_type":"code","source":"%time data=pd.read_csv(\"/kaggle/input/intel-oneapi-predict-the-quality-of-freshwater/dataset.csv\",)\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:18:10.796528Z","iopub.execute_input":"2023-03-04T07:18:10.797351Z","iopub.status.idle":"2023-03-04T07:19:26.707786Z","shell.execute_reply.started":"2023-03-04T07:18:10.797302Z","shell.execute_reply":"2023-03-04T07:19:26.706487Z"},"trusted":true},"execution_count":9,"outputs":[{"name":"stdout","text":"CPU times: user 34.4 s, sys: 3.8 s, total: 38.2 s\nWall time: 1min 15s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Data Pre-processing Fuction","metadata":{}},{"cell_type":"code","source":"def pre_process(df):\n    \n    imputer = SimpleImputer(strategy='most_frequent')\n    lbl_encode=LabelEncoder()\n    \n    \n    \"\"\"categorical columns list\"\"\"\n    grpcol=['Color', 'Source']\n    \n    df['Color']= imputer.fit_transform(df[['Color']])\n    df['Source']=imputer.fit_transform(df[['Source']])\n    \n    \n    df['Color']= lbl_encode.fit_transform(df[['Color']])\n    df['Source']=lbl_encode.fit_transform(df[['Source']])\n    \n    \n    \n\n    \"\"\"numerical columns list\"\"\"    \n    numcol=['pH', 'Iron', 'Nitrate', 'Chloride', 'Lead', 'Zinc',\n       'Turbidity', 'Fluoride', 'Copper', 'Odor', 'Sulfate', 'Conductivity',\n       'Chlorine', 'Manganese', 'Total Dissolved Solids', \n       'Water Temperature', 'Air Temperature']\n    \n    def fill_missing_value(df):\n        for col in numcol:\n            if col in ['Odor','Total Dissolved Solids']:\n                df[col]=df[col].fillna(df[col].mean())\n            df[col]=df[col].fillna(df[col].median())\n        return df\n    \n    df=fill_missing_value(df)\n                   \n    \n    \"\"\"group by numerical summary of each numerical column\"\"\"\n    def feature_eng(df):\n        for cols in numcol:\n            df1=(df.groupby([grpcol])[cols].agg({'min','median','mean','max'}).reset_index())\n            df1=(df1.rename(columns={c:f\"color_source_{cols}_{c}\" for c in df1.loc[:,df1.columns.str.match(\"(min|median|mean|max)\")]}))\n            df=df.merge(df1,on=grpcol,how='left')\n        return df    \n\n\n    \n    #df=feature_eng(df)\n    \n   \n    \n    return df","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:19:26.709803Z","iopub.execute_input":"2023-03-04T07:19:26.710232Z","iopub.status.idle":"2023-03-04T07:19:26.724726Z","shell.execute_reply.started":"2023-03-04T07:19:26.710198Z","shell.execute_reply":"2023-03-04T07:19:26.723316Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"markdown","source":"### Apply the pre-process steps to the data","metadata":{}},{"cell_type":"code","source":"data=pre_process(data)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:19:26.726236Z","iopub.execute_input":"2023-03-04T07:19:26.726971Z","iopub.status.idle":"2023-03-04T07:19:36.908238Z","shell.execute_reply.started":"2023-03-04T07:19:26.726934Z","shell.execute_reply":"2023-03-04T07:19:36.907142Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"markdown","source":"### Split predictor and outcome variable","metadata":{}},{"cell_type":"code","source":"for c, i in enumerate(data.columns.values):\n    print(f\"{c}_{i}\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:33:32.305812Z","iopub.execute_input":"2023-03-04T08:33:32.306197Z","iopub.status.idle":"2023-03-04T08:33:32.313481Z","shell.execute_reply.started":"2023-03-04T08:33:32.306164Z","shell.execute_reply":"2023-03-04T08:33:32.312198Z"},"trusted":true},"execution_count":23,"outputs":[{"name":"stdout","text":"0_Index\n1_pH\n2_Iron\n3_Nitrate\n4_Chloride\n5_Lead\n6_Zinc\n7_Color\n8_Turbidity\n9_Fluoride\n10_Copper\n11_Odor\n12_Sulfate\n13_Conductivity\n14_Chlorine\n15_Manganese\n16_Total Dissolved Solids\n17_Source\n18_Water Temperature\n19_Air Temperature\n20_Month\n21_Day\n22_Time of Day\n23_Target\n","output_type":"stream"}]},{"cell_type":"code","source":"X=data.iloc[:,np.r_[1:7,8:17,18,19]]\ny=data['Target']","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:19:48.978821Z","iopub.execute_input":"2023-03-04T07:19:48.979252Z","iopub.status.idle":"2023-03-04T07:19:49.396503Z","shell.execute_reply.started":"2023-03-04T07:19:48.979213Z","shell.execute_reply":"2023-03-04T07:19:49.395056Z"},"trusted":true},"execution_count":13,"outputs":[]},{"cell_type":"markdown","source":"### Split train dataset into train and test data","metadata":{}},{"cell_type":"code","source":"X_train, X_test, y_train, y_test = train_test_split(\n    X, y, test_size=0.33, random_state=42,stratify=y)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T07:19:51.616819Z","iopub.execute_input":"2023-03-04T07:19:51.617669Z","iopub.status.idle":"2023-03-04T07:19:55.089584Z","shell.execute_reply.started":"2023-03-04T07:19:51.617609Z","shell.execute_reply":"2023-03-04T07:19:55.088397Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"markdown","source":"### Create an XGBoost model with default parameters","metadata":{}},{"cell_type":"code","source":"xgb_model=XGBClassifier(n_estimators=500,random_state=45)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:36:04.784285Z","iopub.execute_input":"2023-03-04T08:36:04.785228Z","iopub.status.idle":"2023-03-04T08:36:04.791312Z","shell.execute_reply.started":"2023-03-04T08:36:04.785182Z","shell.execute_reply":"2023-03-04T08:36:04.789633Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"code","source":"def f1_eval(y_pred, y_true):\n    y_true = y_true.get_label()\n    y_pred = y_pred.round().astype(int)\n    return 'f1', f1_score(y_true, y_pred)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:36:04.794549Z","iopub.execute_input":"2023-03-04T08:36:04.795083Z","iopub.status.idle":"2023-03-04T08:36:04.804706Z","shell.execute_reply.started":"2023-03-04T08:36:04.795030Z","shell.execute_reply":"2023-03-04T08:36:04.803416Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"code","source":"xgb_model.fit(X_train,y_train,eval_set=[(X_test,y_test)],\n              early_stopping_rounds=30,eval_metric=f1_eval,\n              verbose=True)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:36:04.806837Z","iopub.execute_input":"2023-03-04T08:36:04.807389Z","iopub.status.idle":"2023-03-04T08:42:49.900105Z","shell.execute_reply.started":"2023-03-04T08:36:04.807337Z","shell.execute_reply":"2023-03-04T08:42:49.897987Z"},"trusted":true},"execution_count":28,"outputs":[{"name":"stdout","text":"[0]\tvalidation_0-logloss:0.55433\tvalidation_0-f1:0.78004\n[1]\tvalidation_0-logloss:0.47553\tvalidation_0-f1:0.72672\n[2]\tvalidation_0-logloss:0.42433\tvalidation_0-f1:0.79155\n[3]\tvalidation_0-logloss:0.39102\tvalidation_0-f1:0.76598\n[4]\tvalidation_0-logloss:0.36716\tvalidation_0-f1:0.79844\n[5]\tvalidation_0-logloss:0.34990\tvalidation_0-f1:0.78554\n[6]\tvalidation_0-logloss:0.33790\tvalidation_0-f1:0.78694\n[7]\tvalidation_0-logloss:0.32872\tvalidation_0-f1:0.78215\n[8]\tvalidation_0-logloss:0.32207\tvalidation_0-f1:0.78331\n[9]\tvalidation_0-logloss:0.31713\tvalidation_0-f1:0.77937\n[10]\tvalidation_0-logloss:0.31319\tvalidation_0-f1:0.79220\n[11]\tvalidation_0-logloss:0.31030\tvalidation_0-f1:0.78725\n[12]\tvalidation_0-logloss:0.30777\tvalidation_0-f1:0.78201\n[13]\tvalidation_0-logloss:0.30631\tvalidation_0-f1:0.78278\n[14]\tvalidation_0-logloss:0.30438\tvalidation_0-f1:0.78279\n[15]\tvalidation_0-logloss:0.30327\tvalidation_0-f1:0.78338\n[16]\tvalidation_0-logloss:0.30222\tvalidation_0-f1:0.78343\n[17]\tvalidation_0-logloss:0.30124\tvalidation_0-f1:0.78350\n[18]\tvalidation_0-logloss:0.30037\tvalidation_0-f1:0.78346\n[19]\tvalidation_0-logloss:0.29933\tvalidation_0-f1:0.79040\n[20]\tvalidation_0-logloss:0.29874\tvalidation_0-f1:0.78447\n[21]\tvalidation_0-logloss:0.29802\tvalidation_0-f1:0.79068\n[22]\tvalidation_0-logloss:0.29758\tvalidation_0-f1:0.78447\n[23]\tvalidation_0-logloss:0.29706\tvalidation_0-f1:0.78451\n[24]\tvalidation_0-logloss:0.29677\tvalidation_0-f1:0.78473\n[25]\tvalidation_0-logloss:0.29653\tvalidation_0-f1:0.78478\n[26]\tvalidation_0-logloss:0.29633\tvalidation_0-f1:0.78915\n[27]\tvalidation_0-logloss:0.29577\tvalidation_0-f1:0.79210\n[28]\tvalidation_0-logloss:0.29522\tvalidation_0-f1:0.78743\n[29]\tvalidation_0-logloss:0.29479\tvalidation_0-f1:0.78788\n[30]\tvalidation_0-logloss:0.29435\tvalidation_0-f1:0.79099\n","output_type":"stream"},{"execution_count":28,"output_type":"execute_result","data":{"text/plain":"XGBClassifier(base_score=0.5, booster='gbtree', callbacks=None,\n              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,\n              early_stopping_rounds=None, enable_categorical=False,\n              eval_metric=None, gamma=0, gpu_id=-1, grow_policy='depthwise',\n              importance_type=None, interaction_constraints='',\n              learning_rate=0.300000012, max_bin=256, max_cat_to_onehot=4,\n              max_delta_step=0, max_depth=6, max_leaves=0, min_child_weight=1,\n              missing=nan, monotone_constraints='()', n_estimators=500,\n              n_jobs=0, num_parallel_tree=1, predictor='auto', random_state=45,\n              reg_alpha=0, reg_lambda=1, ...)"},"metadata":{}}]},{"cell_type":"markdown","source":"### Feature importance plot.","metadata":{}},{"cell_type":"code","source":"def feature_imp_plot(df,feat_imp_score,title=\"\" ):\n    _=plt.figure(figsize=(12,6))\n    feat_importance=pd.DataFrame({'columns':df.columns,'score':feat_imp_score})\n    feat_importance=feat_importance.sort_values('score',ascending=False)\n    _=sns.barplot(data=feat_importance.nlargest(10,'score'),x='score',y='columns')\n    plt.title(f\"{title}\",fontsize=16);","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:49.901415Z","iopub.execute_input":"2023-03-04T08:42:49.902014Z","iopub.status.idle":"2023-03-04T08:42:49.909545Z","shell.execute_reply.started":"2023-03-04T08:42:49.901977Z","shell.execute_reply":"2023-03-04T08:42:49.908354Z"},"trusted":true},"execution_count":29,"outputs":[]},{"cell_type":"code","source":"feature_imp_plot(X,xgb_model.feature_importances_,\"XGBoost Model Top Feature Importances\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:49.913007Z","iopub.execute_input":"2023-03-04T08:42:49.913733Z","iopub.status.idle":"2023-03-04T08:42:50.283339Z","shell.execute_reply.started":"2023-03-04T08:42:49.913632Z","shell.execute_reply":"2023-03-04T08:42:50.282477Z"},"trusted":true},"execution_count":30,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1200x600 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"### XGBoost inference time","metadata":{}},{"cell_type":"code","source":"%time test_prediction=xgb_model.predict(X_test)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:50.284688Z","iopub.execute_input":"2023-03-04T08:42:50.285052Z","iopub.status.idle":"2023-03-04T08:42:50.573991Z","shell.execute_reply.started":"2023-03-04T08:42:50.285017Z","shell.execute_reply":"2023-03-04T08:42:50.572683Z"},"trusted":true},"execution_count":31,"outputs":[{"name":"stdout","text":"CPU times: user 1.08 s, sys: 3.02 ms, total: 1.08 s\nWall time: 283 ms\n","output_type":"stream"}]},{"cell_type":"code","source":"#f1_score(y_test,daal_xgb_model_prediction)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:50.575759Z","iopub.execute_input":"2023-03-04T08:42:50.576490Z","iopub.status.idle":"2023-03-04T08:42:50.581879Z","shell.execute_reply.started":"2023-03-04T08:42:50.576441Z","shell.execute_reply":"2023-03-04T08:42:50.580435Z"},"trusted":true},"execution_count":32,"outputs":[]},{"cell_type":"markdown","source":"### Convert an XGBoost model to oneDAL\n\n","metadata":{}},{"cell_type":"code","source":"daal_xgb_model=d4p.get_gbt_model_from_xgboost(xgb_model.get_booster())","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:50.583657Z","iopub.execute_input":"2023-03-04T08:42:50.584087Z","iopub.status.idle":"2023-03-04T08:42:50.637636Z","shell.execute_reply.started":"2023-03-04T08:42:50.584042Z","shell.execute_reply":"2023-03-04T08:42:50.636565Z"},"trusted":true},"execution_count":33,"outputs":[]},{"cell_type":"markdown","source":"### oneDAL XGBoost inference time","metadata":{}},{"cell_type":"code","source":"%time daal_xgb_model_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_test,daal_xgb_model).prediction","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:50.639048Z","iopub.execute_input":"2023-03-04T08:42:50.639834Z","iopub.status.idle":"2023-03-04T08:42:51.334463Z","shell.execute_reply.started":"2023-03-04T08:42:50.639781Z","shell.execute_reply":"2023-03-04T08:42:51.333549Z"},"trusted":true},"execution_count":34,"outputs":[{"name":"stdout","text":"CPU times: user 1.33 s, sys: 3.98 ms, total: 1.33 s\nWall time: 687 ms\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Confustion matrix for train and validation data","metadata":{}},{"cell_type":"code","source":"val_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_test,\n                                                             daal_xgb_model_prediction)\n                                            ,display_labels=['Safe to drink', 'Not safe to drink'])","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:51.335819Z","iopub.execute_input":"2023-03-04T08:42:51.336424Z","iopub.status.idle":"2023-03-04T08:42:51.880806Z","shell.execute_reply.started":"2023-03-04T08:42:51.336389Z","shell.execute_reply":"2023-03-04T08:42:51.879554Z"},"trusted":true},"execution_count":35,"outputs":[]},{"cell_type":"code","source":"val_confusion_matrix.plot();","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:51.882108Z","iopub.execute_input":"2023-03-04T08:42:51.882430Z","iopub.status.idle":"2023-03-04T08:42:52.170732Z","shell.execute_reply.started":"2023-03-04T08:42:51.882392Z","shell.execute_reply":"2023-03-04T08:42:52.169432Z"},"trusted":true},"execution_count":36,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_test,daal_xgb_model_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:52.172645Z","iopub.execute_input":"2023-03-04T08:42:52.173429Z","iopub.status.idle":"2023-03-04T08:42:57.491238Z","shell.execute_reply.started":"2023-03-04T08:42:52.173375Z","shell.execute_reply":"2023-03-04T08:42:57.490334Z"},"trusted":true},"execution_count":37,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.94      0.86      0.90   1370025\nNot safe to drink       0.73      0.87      0.79    595733\n\n         accuracy                           0.86   1965758\n        macro avg       0.83      0.87      0.85   1965758\n     weighted avg       0.88      0.86      0.87   1965758\n\n","output_type":"stream"}]},{"cell_type":"code","source":"%time daal_xgb_model__train_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_train,daal_xgb_model).prediction\ntrain_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_train,\n                                                               daal_xgb_model__train_prediction),\n                                              display_labels=['Safe to drink', 'Not safe to drink'])","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:42:57.492531Z","iopub.execute_input":"2023-03-04T08:42:57.493070Z","iopub.status.idle":"2023-03-04T08:43:00.047222Z","shell.execute_reply.started":"2023-03-04T08:42:57.493035Z","shell.execute_reply":"2023-03-04T08:43:00.045687Z"},"trusted":true},"execution_count":38,"outputs":[{"name":"stdout","text":"CPU times: user 2.73 s, sys: 4.96 ms, total: 2.74 s\nWall time: 1.38 s\n","output_type":"stream"}]},{"cell_type":"code","source":"train_confusion_matrix.plot();","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:00.052930Z","iopub.execute_input":"2023-03-04T08:43:00.053344Z","iopub.status.idle":"2023-03-04T08:43:00.362996Z","shell.execute_reply.started":"2023-03-04T08:43:00.053293Z","shell.execute_reply":"2023-03-04T08:43:00.361552Z"},"trusted":true},"execution_count":39,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_train,daal_xgb_model__train_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:00.364359Z","iopub.execute_input":"2023-03-04T08:43:00.364879Z","iopub.status.idle":"2023-03-04T08:43:11.727040Z","shell.execute_reply.started":"2023-03-04T08:43:00.364838Z","shell.execute_reply":"2023-03-04T08:43:11.725788Z"},"trusted":true},"execution_count":40,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.94      0.86      0.90   2781565\nNot safe to drink       0.73      0.87      0.79   1209519\n\n         accuracy                           0.86   3991084\n        macro avg       0.83      0.86      0.85   3991084\n     weighted avg       0.88      0.86      0.87   3991084\n\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Let's use lightgbm model","metadata":{}},{"cell_type":"code","source":"lgbm_model=LGBMClassifier(verbose=-1)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:11.728658Z","iopub.execute_input":"2023-03-04T08:43:11.729004Z","iopub.status.idle":"2023-03-04T08:43:11.733987Z","shell.execute_reply.started":"2023-03-04T08:43:11.728972Z","shell.execute_reply":"2023-03-04T08:43:11.732951Z"},"trusted":true},"execution_count":41,"outputs":[]},{"cell_type":"code","source":"\nlgbm_model.fit(X_train,y_train,eval_set=[(X_test,y_test)],\n                     early_stopping_rounds=30,eval_metric='f1',\n                     )","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:11.735337Z","iopub.execute_input":"2023-03-04T08:43:11.735663Z","iopub.status.idle":"2023-03-04T08:43:51.023067Z","shell.execute_reply.started":"2023-03-04T08:43:11.735606Z","shell.execute_reply":"2023-03-04T08:43:51.022227Z"},"trusted":true},"execution_count":42,"outputs":[{"name":"stdout","text":"[1]\tvalid_0's binary_logloss: 0.558271\n[2]\tvalid_0's binary_logloss: 0.514431\n[3]\tvalid_0's binary_logloss: 0.478759\n[4]\tvalid_0's binary_logloss: 0.44929\n[5]\tvalid_0's binary_logloss: 0.424679\n[6]\tvalid_0's binary_logloss: 0.403981\n[7]\tvalid_0's binary_logloss: 0.386454\n[8]\tvalid_0's binary_logloss: 0.37153\n[9]\tvalid_0's binary_logloss: 0.358782\n[10]\tvalid_0's binary_logloss: 0.347857\n[11]\tvalid_0's binary_logloss: 0.338472\n[12]\tvalid_0's binary_logloss: 0.330393\n[13]\tvalid_0's binary_logloss: 0.323425\n[14]\tvalid_0's binary_logloss: 0.31741\n[15]\tvalid_0's binary_logloss: 0.312212\n[16]\tvalid_0's binary_logloss: 0.307711\n[17]\tvalid_0's binary_logloss: 0.303802\n[18]\tvalid_0's binary_logloss: 0.300424\n[19]\tvalid_0's binary_logloss: 0.297473\n[20]\tvalid_0's binary_logloss: 0.294937\n[21]\tvalid_0's binary_logloss: 0.292735\n[22]\tvalid_0's binary_logloss: 0.290804\n[23]\tvalid_0's binary_logloss: 0.289155\n[24]\tvalid_0's binary_logloss: 0.287708\n[25]\tvalid_0's binary_logloss: 0.286428\n[26]\tvalid_0's binary_logloss: 0.285368\n[27]\tvalid_0's binary_logloss: 0.284432\n[28]\tvalid_0's binary_logloss: 0.283605\n[29]\tvalid_0's binary_logloss: 0.282914\n[30]\tvalid_0's binary_logloss: 0.282322\n[31]\tvalid_0's binary_logloss: 0.281813\n[32]\tvalid_0's binary_logloss: 0.28137\n[33]\tvalid_0's binary_logloss: 0.280978\n[34]\tvalid_0's binary_logloss: 0.280655\n[35]\tvalid_0's binary_logloss: 0.280364\n[36]\tvalid_0's binary_logloss: 0.280129\n[37]\tvalid_0's binary_logloss: 0.279926\n[38]\tvalid_0's binary_logloss: 0.279741\n[39]\tvalid_0's binary_logloss: 0.27958\n[40]\tvalid_0's binary_logloss: 0.279458\n[41]\tvalid_0's binary_logloss: 0.279334\n[42]\tvalid_0's binary_logloss: 0.279242\n[43]\tvalid_0's binary_logloss: 0.279156\n[44]\tvalid_0's binary_logloss: 0.27909\n[45]\tvalid_0's binary_logloss: 0.279026\n[46]\tvalid_0's binary_logloss: 0.278973\n[47]\tvalid_0's binary_logloss: 0.278931\n[48]\tvalid_0's binary_logloss: 0.278891\n[49]\tvalid_0's binary_logloss: 0.278862\n[50]\tvalid_0's binary_logloss: 0.278836\n[51]\tvalid_0's binary_logloss: 0.278814\n[52]\tvalid_0's binary_logloss: 0.278795\n[53]\tvalid_0's binary_logloss: 0.278773\n[54]\tvalid_0's binary_logloss: 0.278761\n[55]\tvalid_0's binary_logloss: 0.27875\n[56]\tvalid_0's binary_logloss: 0.278733\n[57]\tvalid_0's binary_logloss: 0.278729\n[58]\tvalid_0's binary_logloss: 0.278722\n[59]\tvalid_0's binary_logloss: 0.278718\n[60]\tvalid_0's binary_logloss: 0.278713\n[61]\tvalid_0's binary_logloss: 0.278714\n[62]\tvalid_0's binary_logloss: 0.278714\n[63]\tvalid_0's binary_logloss: 0.278713\n[64]\tvalid_0's binary_logloss: 0.278714\n[65]\tvalid_0's binary_logloss: 0.278715\n[66]\tvalid_0's binary_logloss: 0.278715\n[67]\tvalid_0's binary_logloss: 0.278714\n[68]\tvalid_0's binary_logloss: 0.278714\n[69]\tvalid_0's binary_logloss: 0.278712\n[70]\tvalid_0's binary_logloss: 0.278713\n[71]\tvalid_0's binary_logloss: 0.278714\n[72]\tvalid_0's binary_logloss: 0.278713\n[73]\tvalid_0's binary_logloss: 0.278712\n[74]\tvalid_0's binary_logloss: 0.278713\n[75]\tvalid_0's binary_logloss: 0.278714\n[76]\tvalid_0's binary_logloss: 0.278714\n[77]\tvalid_0's binary_logloss: 0.278712\n[78]\tvalid_0's binary_logloss: 0.278712\n[79]\tvalid_0's binary_logloss: 0.278712\n[80]\tvalid_0's binary_logloss: 0.278711\n[81]\tvalid_0's binary_logloss: 0.278712\n[82]\tvalid_0's binary_logloss: 0.278713\n[83]\tvalid_0's binary_logloss: 0.278713\n[84]\tvalid_0's binary_logloss: 0.278714\n[85]\tvalid_0's binary_logloss: 0.278713\n[86]\tvalid_0's binary_logloss: 0.278711\n[87]\tvalid_0's binary_logloss: 0.278714\n[88]\tvalid_0's binary_logloss: 0.278714\n[89]\tvalid_0's binary_logloss: 0.278714\n[90]\tvalid_0's binary_logloss: 0.278713\n[91]\tvalid_0's binary_logloss: 0.278715\n[92]\tvalid_0's binary_logloss: 0.278715\n[93]\tvalid_0's binary_logloss: 0.278715\n[94]\tvalid_0's binary_logloss: 0.278715\n[95]\tvalid_0's binary_logloss: 0.278716\n[96]\tvalid_0's binary_logloss: 0.278716\n[97]\tvalid_0's binary_logloss: 0.278716\n[98]\tvalid_0's binary_logloss: 0.278716\n[99]\tvalid_0's binary_logloss: 0.278716\n[100]\tvalid_0's binary_logloss: 0.278716\n","output_type":"stream"},{"execution_count":42,"output_type":"execute_result","data":{"text/plain":"LGBMClassifier(verbose=-1)"},"metadata":{}}]},{"cell_type":"markdown","source":"### Feature importances plot","metadata":{}},{"cell_type":"code","source":"feature_imp_plot(X,lgbm_model.feature_importances_,\"Lightgbm Model Top Feature Importances\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:51.024375Z","iopub.execute_input":"2023-03-04T08:43:51.025340Z","iopub.status.idle":"2023-03-04T08:43:51.343543Z","shell.execute_reply.started":"2023-03-04T08:43:51.025304Z","shell.execute_reply":"2023-03-04T08:43:51.342412Z"},"trusted":true},"execution_count":43,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1200x600 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"### Lightgbm inference time\n","metadata":{}},{"cell_type":"code","source":"%time lgbm_test_prediction=lgbm_model.predict(X_test)\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:51.345055Z","iopub.execute_input":"2023-03-04T08:43:51.345490Z","iopub.status.idle":"2023-03-04T08:43:55.142985Z","shell.execute_reply.started":"2023-03-04T08:43:51.345444Z","shell.execute_reply":"2023-03-04T08:43:55.141716Z"},"trusted":true},"execution_count":44,"outputs":[{"name":"stdout","text":"CPU times: user 13.8 s, sys: 72.6 ms, total: 13.9 s\nWall time: 3.79 s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"\n### Convert an Lightgbm model to oneDAL","metadata":{}},{"cell_type":"code","source":"daal_lgbm_model=d4p.get_gbt_model_from_lightgbm(lgbm_model.booster_)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:43:55.144562Z","iopub.execute_input":"2023-03-04T08:43:55.144986Z","iopub.status.idle":"2023-03-04T08:44:20.724633Z","shell.execute_reply.started":"2023-03-04T08:43:55.144952Z","shell.execute_reply":"2023-03-04T08:44:20.723654Z"},"trusted":true},"execution_count":45,"outputs":[]},{"cell_type":"markdown","source":"### oneDAL lightgbm inference time","metadata":{}},{"cell_type":"code","source":"%time daal_lgbm_model_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_test,daal_lgbm_model).prediction","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:44:20.726013Z","iopub.execute_input":"2023-03-04T08:44:20.728198Z","iopub.status.idle":"2023-03-04T08:44:28.588705Z","shell.execute_reply.started":"2023-03-04T08:44:20.728156Z","shell.execute_reply":"2023-03-04T08:44:28.587588Z"},"trusted":true},"execution_count":46,"outputs":[{"name":"stdout","text":"CPU times: user 15.6 s, sys: 7.37 ms, total: 15.7 s\nWall time: 7.85 s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Confustion matrix for train and validation data","metadata":{}},{"cell_type":"code","source":"lgbm_val_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_test,\n                                                             daal_lgbm_model_prediction)\n                                            ,display_labels=['Safe to drink', 'Not safe to drink'])\n\nlgbm_val_confusion_matrix.plot();","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:44:28.590382Z","iopub.execute_input":"2023-03-04T08:44:28.590902Z","iopub.status.idle":"2023-03-04T08:44:29.455806Z","shell.execute_reply.started":"2023-03-04T08:44:28.590854Z","shell.execute_reply":"2023-03-04T08:44:29.454481Z"},"trusted":true},"execution_count":47,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_test,daal_lgbm_model_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:44:29.457437Z","iopub.execute_input":"2023-03-04T08:44:29.457825Z","iopub.status.idle":"2023-03-04T08:44:34.985903Z","shell.execute_reply.started":"2023-03-04T08:44:29.457784Z","shell.execute_reply":"2023-03-04T08:44:34.984535Z"},"trusted":true},"execution_count":48,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.99      0.84      0.91   1370025\nNot safe to drink       0.73      0.97      0.83    595733\n\n         accuracy                           0.88   1965758\n        macro avg       0.86      0.91      0.87   1965758\n     weighted avg       0.91      0.88      0.89   1965758\n\n","output_type":"stream"}]},{"cell_type":"code","source":"%time daal_lgbm_model__train_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_train,daal_lgbm_model).prediction\n\nlgbm_train_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_train,\n                                                               daal_lgbm_model__train_prediction),\n                                              display_labels=['Safe to drink', 'Not safe to drink'])\n\n\nlgbm_train_confusion_matrix.plot();\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:44:34.987559Z","iopub.execute_input":"2023-03-04T08:44:34.988052Z","iopub.status.idle":"2023-03-04T08:44:52.477106Z","shell.execute_reply.started":"2023-03-04T08:44:34.988003Z","shell.execute_reply":"2023-03-04T08:44:52.475670Z"},"trusted":true},"execution_count":49,"outputs":[{"name":"stdout","text":"CPU times: user 32 s, sys: 6.55 ms, total: 32 s\nWall time: 16 s\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_train,daal_lgbm_model__train_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:44:52.478777Z","iopub.execute_input":"2023-03-04T08:44:52.479211Z","iopub.status.idle":"2023-03-04T08:45:03.944633Z","shell.execute_reply.started":"2023-03-04T08:44:52.479165Z","shell.execute_reply":"2023-03-04T08:45:03.943509Z"},"trusted":true},"execution_count":50,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.99      0.84      0.91   2781565\nNot safe to drink       0.73      0.97      0.83   1209519\n\n         accuracy                           0.88   3991084\n        macro avg       0.86      0.91      0.87   3991084\n     weighted avg       0.91      0.88      0.89   3991084\n\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Let's create catboost model","metadata":{}},{"cell_type":"code","source":"cat_clf=CatBoostClassifier(random_seed=45,eval_metric='F1',verbose=False,)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:45:03.945984Z","iopub.execute_input":"2023-03-04T08:45:03.946836Z","iopub.status.idle":"2023-03-04T08:45:03.955776Z","shell.execute_reply.started":"2023-03-04T08:45:03.946799Z","shell.execute_reply":"2023-03-04T08:45:03.954668Z"},"trusted":true},"execution_count":51,"outputs":[]},{"cell_type":"code","source":"cat_clf.fit(X_train,y_train,eval_set=(X_test,y_test),early_stopping_rounds=100)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:45:03.956895Z","iopub.execute_input":"2023-03-04T08:45:03.957221Z","iopub.status.idle":"2023-03-04T08:55:41.958453Z","shell.execute_reply.started":"2023-03-04T08:45:03.957188Z","shell.execute_reply":"2023-03-04T08:55:41.956741Z"},"trusted":true},"execution_count":52,"outputs":[{"execution_count":52,"output_type":"execute_result","data":{"text/plain":"<catboost.core.CatBoostClassifier at 0x7ff810429a10>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Feature importance plot.","metadata":{}},{"cell_type":"code","source":"feature_imp_plot(X,cat_clf.get_feature_importance(),\"Catboost Model Top Feature Importances\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:55:41.961108Z","iopub.execute_input":"2023-03-04T08:55:41.967042Z","iopub.status.idle":"2023-03-04T08:55:42.317495Z","shell.execute_reply.started":"2023-03-04T08:55:41.966963Z","shell.execute_reply":"2023-03-04T08:55:42.316128Z"},"trusted":true},"execution_count":53,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1200x600 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"### Catboost inference time","metadata":{}},{"cell_type":"code","source":"%time cat_test_prediction=cat_clf.predict(X_test)\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:55:42.319209Z","iopub.execute_input":"2023-03-04T08:55:42.319774Z","iopub.status.idle":"2023-03-04T08:55:45.281688Z","shell.execute_reply.started":"2023-03-04T08:55:42.319717Z","shell.execute_reply":"2023-03-04T08:55:45.280634Z"},"trusted":true},"execution_count":54,"outputs":[{"name":"stdout","text":"CPU times: user 2.93 s, sys: 23 ms, total: 2.95 s\nWall time: 2.95 s\n","output_type":"stream"}]},{"cell_type":"code","source":"#f1_score(y_test,daal_catb_model_prediction)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:55:45.283098Z","iopub.execute_input":"2023-03-04T08:55:45.283423Z","iopub.status.idle":"2023-03-04T08:55:45.288275Z","shell.execute_reply.started":"2023-03-04T08:55:45.283392Z","shell.execute_reply":"2023-03-04T08:55:45.287019Z"},"trusted":true},"execution_count":55,"outputs":[]},{"cell_type":"markdown","source":"### Convert an catboost model to oneDAL","metadata":{}},{"cell_type":"code","source":"daal_catb_model=d4p.get_gbt_model_from_catboost(cat_clf)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:55:45.289498Z","iopub.execute_input":"2023-03-04T08:55:45.289860Z","iopub.status.idle":"2023-03-04T08:55:45.550718Z","shell.execute_reply.started":"2023-03-04T08:55:45.289828Z","shell.execute_reply":"2023-03-04T08:55:45.549410Z"},"trusted":true},"execution_count":56,"outputs":[]},{"cell_type":"markdown","source":"\n### oneDAL catboost inference time","metadata":{}},{"cell_type":"code","source":"%time daal_catb_model_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_test,daal_catb_model).prediction\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:55:45.552102Z","iopub.execute_input":"2023-03-04T08:55:45.553002Z","iopub.status.idle":"2023-03-04T08:56:04.992557Z","shell.execute_reply.started":"2023-03-04T08:55:45.552951Z","shell.execute_reply":"2023-03-04T08:56:04.991175Z"},"trusted":true},"execution_count":57,"outputs":[{"name":"stdout","text":"CPU times: user 38.8 s, sys: 7.29 ms, total: 38.8 s\nWall time: 19.4 s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Confustion matrix for train and validation data","metadata":{}},{"cell_type":"code","source":"catb_val_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_test,\n                                                             daal_catb_model_prediction)\n                                            ,display_labels=['Safe to drink', 'Not safe to drink'])\n\ncatb_val_confusion_matrix.plot();","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:56:04.994567Z","iopub.execute_input":"2023-03-04T08:56:04.995379Z","iopub.status.idle":"2023-03-04T08:56:05.839396Z","shell.execute_reply.started":"2023-03-04T08:56:04.995308Z","shell.execute_reply":"2023-03-04T08:56:05.838153Z"},"trusted":true},"execution_count":58,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_test,daal_catb_model_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:56:05.841129Z","iopub.execute_input":"2023-03-04T08:56:05.841588Z","iopub.status.idle":"2023-03-04T08:56:11.108529Z","shell.execute_reply.started":"2023-03-04T08:56:05.841537Z","shell.execute_reply":"2023-03-04T08:56:11.107390Z"},"trusted":true},"execution_count":59,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.97      0.85      0.90   1370025\nNot safe to drink       0.73      0.93      0.82    595733\n\n         accuracy                           0.87   1965758\n        macro avg       0.85      0.89      0.86   1965758\n     weighted avg       0.89      0.87      0.88   1965758\n\n","output_type":"stream"}]},{"cell_type":"code","source":"%time daal_cat_model__train_prediction=d4p.gbt_classification_prediction(nClasses=2).compute(X_train,daal_xgb_model).prediction\ncatb_train_confusion_matrix=ConfusionMatrixDisplay(confusion_matrix(y_train,\n                                                               daal_cat_model__train_prediction),\n                                              display_labels=['Safe to drink', 'Not safe to drink'])\n\n\ncatb_train_confusion_matrix.plot();\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:56:11.112409Z","iopub.execute_input":"2023-03-04T08:56:11.113658Z","iopub.status.idle":"2023-03-04T08:56:13.967901Z","shell.execute_reply.started":"2023-03-04T08:56:11.113582Z","shell.execute_reply":"2023-03-04T08:56:13.966838Z"},"trusted":true},"execution_count":60,"outputs":[{"name":"stdout","text":"CPU times: user 2.71 s, sys: 1.02 ms, total: 2.71 s\nWall time: 1.36 s\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(y_train,daal_cat_model__train_prediction,target_names=['Safe to drink', 'Not safe to drink']))","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:56:13.969184Z","iopub.execute_input":"2023-03-04T08:56:13.969503Z","iopub.status.idle":"2023-03-04T08:56:25.512274Z","shell.execute_reply.started":"2023-03-04T08:56:13.969472Z","shell.execute_reply":"2023-03-04T08:56:25.510841Z"},"trusted":true},"execution_count":61,"outputs":[{"name":"stdout","text":"                   precision    recall  f1-score   support\n\n    Safe to drink       0.94      0.86      0.90   2781565\nNot safe to drink       0.73      0.87      0.79   1209519\n\n         accuracy                           0.86   3991084\n        macro avg       0.83      0.86      0.85   3991084\n     weighted avg       0.88      0.86      0.87   3991084\n\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Let's compare the models F1 score.","metadata":{}},{"cell_type":"code","source":"resul_df=pd.DataFrame(columns=['Model_Name','Type_of_Data','F1_score'])\nfor name,model in zip(['XGBoost','Lightgbm','Catboost'],[daal_xgb_model,daal_lgbm_model,daal_catb_model]):\n        for datatype in ['train','validation']:\n            if datatype =='train':\n                score=f1_score(y_train,d4p.gbt_classification_prediction(nClasses=2).compute(X_train,model).prediction)\n                resul_df.loc[len(resul_df)]=[name,datatype,score]\n            elif datatype =='validation':\n                score=f1_score(y_test,d4p.gbt_classification_prediction(nClasses=2).compute(X_test,model).prediction)\n                resul_df.loc[len(resul_df)]=[name,datatype,score]\n            \n                \n            \n            ","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:56:25.513668Z","iopub.execute_input":"2023-03-04T08:56:25.513992Z","iopub.status.idle":"2023-03-04T08:58:00.688054Z","shell.execute_reply.started":"2023-03-04T08:56:25.513961Z","shell.execute_reply":"2023-03-04T08:58:00.686554Z"},"trusted":true},"execution_count":62,"outputs":[]},{"cell_type":"code","source":"resul_df","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:00.689544Z","iopub.execute_input":"2023-03-04T08:58:00.690016Z","iopub.status.idle":"2023-03-04T08:58:00.705211Z","shell.execute_reply.started":"2023-03-04T08:58:00.689966Z","shell.execute_reply":"2023-03-04T08:58:00.703766Z"},"trusted":true},"execution_count":63,"outputs":[{"execution_count":63,"output_type":"execute_result","data":{"text/plain":"  Model_Name Type_of_Data  F1_score\n0    XGBoost        train  0.793467\n1    XGBoost   validation  0.793757\n2   Lightgbm        train  0.832525\n3   Lightgbm   validation  0.832876\n4   Catboost        train  0.819155\n5   Catboost   validation  0.817306","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>Model_Name</th>\n      <th>Type_of_Data</th>\n      <th>F1_score</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>XGBoost</td>\n      <td>train</td>\n      <td>0.793467</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>XGBoost</td>\n      <td>validation</td>\n      <td>0.793757</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>Lightgbm</td>\n      <td>train</td>\n      <td>0.832525</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>Lightgbm</td>\n      <td>validation</td>\n      <td>0.832876</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>Catboost</td>\n      <td>train</td>\n      <td>0.819155</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>Catboost</td>\n      <td>validation</td>\n      <td>0.817306</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"_=plt.figure(figsize=(8,5))\nsns.barplot(data=resul_df,x='Model_Name',y='F1_score',hue='Type_of_Data');\n_=plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\nax = plt.gca()\nfor p in ax.patches:\n    ax.annotate(\"%.2f\" % p.get_height(), (p.get_x() + p.get_width() / 2., p.get_height()),\n                ha='center', va='center', fontsize=11, color='black', xytext=(0, 7),\n                textcoords='offset points')\n_=plt.title(\"F1 Score Comparison\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:00.706700Z","iopub.execute_input":"2023-03-04T08:58:00.707070Z","iopub.status.idle":"2023-03-04T08:58:01.029979Z","shell.execute_reply.started":"2023-03-04T08:58:00.707036Z","shell.execute_reply":"2023-03-04T08:58:01.028880Z"},"trusted":true},"execution_count":64,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 800x500 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"#### The above plot explains that the lightgbm model performs better on both train and test data.","metadata":{}},{"cell_type":"markdown","source":"### Let's understand the lightgbm model with shapash","metadata":{}},{"cell_type":"code","source":"!pip install shapash -q","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:01.031321Z","iopub.execute_input":"2023-03-04T08:58:01.031675Z","iopub.status.idle":"2023-03-04T08:58:22.223689Z","shell.execute_reply.started":"2023-03-04T08:58:01.031641Z","shell.execute_reply":"2023-03-04T08:58:22.222457Z"},"trusted":true},"execution_count":65,"outputs":[{"name":"stdout","text":"\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n\u001b[0m","output_type":"stream"}]},{"cell_type":"code","source":"from shapash import SmartExplainer\nfrom shapash.explainer.smart_predictor import SmartPredictor\nfrom shapash.utils.load_smartpredictor import load_smartpredictor","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:22.235792Z","iopub.execute_input":"2023-03-04T08:58:22.236999Z","iopub.status.idle":"2023-03-04T08:58:28.636267Z","shell.execute_reply.started":"2023-03-04T08:58:22.236947Z","shell.execute_reply":"2023-03-04T08:58:28.634989Z"},"trusted":true},"execution_count":66,"outputs":[]},{"cell_type":"markdown","source":"### Let's declare and compile SmartExplainer for lightgbm model","metadata":{}},{"cell_type":"code","source":"feature_dict = {col:f\"about_{col}\" for col in X_train.columns}\nlabel_dict = {0: \"Not safe to drink\", 1: \"Safe to drink\"}","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:28.638327Z","iopub.execute_input":"2023-03-04T08:58:28.638815Z","iopub.status.idle":"2023-03-04T08:58:28.645450Z","shell.execute_reply.started":"2023-03-04T08:58:28.638764Z","shell.execute_reply":"2023-03-04T08:58:28.643635Z"},"trusted":true},"execution_count":67,"outputs":[]},{"cell_type":"code","source":"lgbm_xpl=SmartExplainer(\n    model=lgbm_model,label_dict = {0: \"Not safe to drink\", 1: \"Safe to drink\"}\n    )\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:28.647560Z","iopub.execute_input":"2023-03-04T08:58:28.648042Z","iopub.status.idle":"2023-03-04T08:58:28.890236Z","shell.execute_reply.started":"2023-03-04T08:58:28.647995Z","shell.execute_reply":"2023-03-04T08:58:28.888815Z"},"trusted":true},"execution_count":68,"outputs":[]},{"cell_type":"code","source":"\n\nlgbm_xpl.compile(x=X_test,\n y_pred=pd.DataFrame(lgbm_test_prediction,columns=['Target'],index=y_test.index),\n y_target=y_test, # Optional: allows to display True Values vs Predicted Values\n )","metadata":{"execution":{"iopub.status.busy":"2023-03-04T08:58:28.891925Z","iopub.execute_input":"2023-03-04T08:58:28.892430Z","iopub.status.idle":"2023-03-04T09:39:07.009077Z","shell.execute_reply.started":"2023-03-04T08:58:28.892376Z","shell.execute_reply":"2023-03-04T09:39:07.007677Z"},"trusted":true},"execution_count":69,"outputs":[]},{"cell_type":"markdown","source":"### Lightgbm Model Features importances SHAP","metadata":{}},{"cell_type":"code","source":"lgbm_xpl.plot.features_importance()","metadata":{"execution":{"iopub.status.busy":"2023-03-04T09:51:13.654807Z","iopub.execute_input":"2023-03-04T09:51:13.655288Z","iopub.status.idle":"2023-03-04T09:51:14.619009Z","shell.execute_reply.started":"2023-03-04T09:51:13.655245Z","shell.execute_reply":"2023-03-04T09:51:14.617656Z"},"trusted":true},"execution_count":78,"outputs":[{"output_type":"display_data","data":{"text/html":"        <script type=\"text/javascript\">\n        window.PlotlyConfig = {MathJaxConfig: 'local'};\n        if (window.MathJax && window.MathJax.Hub && window.MathJax.Hub.Config) {window.MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n        if (typeof require !== 'undefined') {\n        require.undef(\"plotly\");\n        requirejs.config({\n            paths: {\n                'plotly': ['https://cdn.plot.ly/plotly-2.18.0.min']\n            }\n        });\n        require(['plotly'], function(Plotly) {\n            window._Plotly = Plotly;\n        });\n        }\n        </script>\n        "},"metadata":{}},{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"4ce661ec-f32c-4917-ab89-819daca0eb25\" class=\"plotly-graph-div\" style=\"height:500px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"4ce661ec-f32c-4917-ab89-819daca0eb25\")) {                    Plotly.newPlot(                        \"4ce661ec-f32c-4917-ab89-819daca0eb25\",                        [{\"customdata\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":[\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\"],\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"name\":\"Global\",\"orientation\":\"h\",\"x\":[0.0002,0.0003,0.0004,0.0046,0.0133,0.0273,0.032,0.0402,0.0439,0.0462,0.0477,0.0769,0.0889,0.106,0.1255,0.1468,0.2],\"y\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"type\":\"bar\"}],                        {\"autosize\":false,\"barmode\":\"group\",\"height\":500,\"hovermode\":\"closest\",\"margin\":{\"b\":50,\"l\":160,\"r\":0,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Features Importance<br><sup>Response: <b>Safe to drink</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Mean absolute Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"ticktext\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"tickvals\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"tickmode\":\"array\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('4ce661ec-f32c-4917-ab89-819daca0eb25');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Subset importances","metadata":{}},{"cell_type":"markdown","source":"### Let's take 10 samples from each class and see the feature importance.","metadata":{}},{"cell_type":"code","source":"y_test[y_test==1].index","metadata":{"execution":{"iopub.status.busy":"2023-03-04T09:53:03.284407Z","iopub.execute_input":"2023-03-04T09:53:03.285029Z","iopub.status.idle":"2023-03-04T09:53:03.335431Z","shell.execute_reply.started":"2023-03-04T09:53:03.284974Z","shell.execute_reply":"2023-03-04T09:53:03.334066Z"},"trusted":true},"execution_count":79,"outputs":[{"execution_count":79,"output_type":"execute_result","data":{"text/plain":"Int64Index([4642920, 1982184,  741199, 5882817, 5951701, 3111599, 2723276,\n            1166197, 4707813, 3099090,\n            ...\n            3906414, 5109363, 3150618, 5906744,  344108, 3918971, 5955590,\n            5841087, 4646348, 1516916],\n           dtype='int64', length=595733)"},"metadata":{}}]},{"cell_type":"code","source":"y_test[y_test==0]%%!.index","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:00:21.917852Z","iopub.execute_input":"2023-03-04T10:00:21.918751Z","iopub.status.idle":"2023-03-04T10:00:21.956128Z","shell.execute_reply.started":"2023-03-04T10:00:21.918683Z","shell.execute_reply":"2023-03-04T10:00:21.954873Z"},"trusted":true},"execution_count":88,"outputs":[{"execution_count":88,"output_type":"execute_result","data":{"text/plain":"Int64Index([2843577, 1610572, 1429181, 2405952,  521610,  454228, 2807684,\n            4064876, 2480755, 2239755,\n            ...\n            5610876, 1224392, 4813231, 1843943, 1048427,   23691, 4128941,\n            2252212, 3743848, 2857888],\n           dtype='int64', length=1370025)"},"metadata":{}}]},{"cell_type":"code","source":"subset = [4642920, 1982184,  741199, 5882817, 5951701, 3111599, 2723276,1166197, \n     4707813, 3099090,2843577, 1610572, 1429181, 2405952,  521610,  454228,\n     2807684, 4064876, 2480755, 2239755]\nlgbm_xpl.plot.features_importance(selection=subset)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T09:53:39.390809Z","iopub.execute_input":"2023-03-04T09:53:39.391328Z","iopub.status.idle":"2023-03-04T09:53:40.225368Z","shell.execute_reply.started":"2023-03-04T09:53:39.391271Z","shell.execute_reply":"2023-03-04T09:53:40.223931Z"},"trusted":true},"execution_count":81,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"89d12240-1f3c-4eac-9ac4-3383799a4b88\" class=\"plotly-graph-div\" style=\"height:500px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"89d12240-1f3c-4eac-9ac4-3383799a4b88\")) {                    Plotly.newPlot(                        \"89d12240-1f3c-4eac-9ac4-3383799a4b88\",                        [{\"customdata\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(52, 55, 54, 0.7)\"},\"name\":\"Subset\",\"orientation\":\"h\",\"x\":[0.0001,0.0002,0.0004,0.0005,0.0281,0.0646,0.0929,0.0537,0.0426,0.0357,0.0208,0.0673,0.1079,0.0894,0.0701,0.1771,0.1487],\"y\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"type\":\"bar\"},{\"customdata\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":[\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\"],\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"name\":\"Global\",\"orientation\":\"h\",\"x\":[0.0002,0.0003,0.0004,0.0046,0.0133,0.0273,0.032,0.0402,0.0439,0.0462,0.0477,0.0769,0.0889,0.106,0.1255,0.1468,0.2],\"y\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"type\":\"bar\"}],                        {\"autosize\":false,\"barmode\":\"group\",\"height\":500,\"hovermode\":\"closest\",\"margin\":{\"b\":50,\"l\":160,\"r\":0,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Features Importance<br><sup>Response: <b>Safe to drink</b> - Subset length: 20 (0%)</sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Mean absolute Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"ticktext\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"tickvals\":[\"Water Temperature\",\"Air Temperature\",\"Conductivity\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Fluoride\",\"Chlorine\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Chloride\",\"Turbidity\",\"pH\",\"Manganese\"],\"tickmode\":\"array\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('89d12240-1f3c-4eac-9ac4-3383799a4b88');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Local explantion for class 1(safe to drink)","metadata":{}},{"cell_type":"code","source":"lgbm_xpl.plot.local_plot(index=4642920,label=1)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:02:56.664687Z","iopub.execute_input":"2023-03-04T10:02:56.665100Z","iopub.status.idle":"2023-03-04T10:02:56.709421Z","shell.execute_reply.started":"2023-03-04T10:02:56.665064Z","shell.execute_reply":"2023-03-04T10:02:56.708079Z"},"trusted":true},"execution_count":93,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"b71738fa-bf2e-4af4-904c-2448cd2f95dd\" class=\"plotly-graph-div\" style=\"height:550px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"b71738fa-bf2e-4af4-904c-2448cd2f95dd\")) {                    Plotly.newPlot(                        \"b71738fa-bf2e-4af4-904c-2448cd2f95dd\",                        [{\"customdata\":[\"<b>Manganese :</b><br />3.098688657658745e-10\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.13602989635194535],\"y\":[\"<b>Manganese</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Turbidity :</b><br />0.0095333982886721\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.07380718838004213],\"y\":[\"<b>Turbidity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Copper :</b><br />0.1864565906380453\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.044401359441786595],\"y\":[\"<b>Copper</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Odor :</b><br />2.087222297073667\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.0323384229434395],\"y\":[\"<b>Odor</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Nitrate :</b><br />2.04394462105232\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.01758152612248989],\"y\":[\"<b>Nitrate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Iron :</b><br />0.0832012567872399\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.012794972180916952],\"y\":[\"<b>Iron</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chlorine :</b><br />3.333113166670345\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.011534768792573629],\"y\":[\"<b>Chlorine</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Fluoride :</b><br />0.1345994952349928\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.008966360204154328],\"y\":[\"<b>Fluoride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Total Dissolved Solids :</b><br />225.40785498160088\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.006262648693952326],\"y\":[\"<b>Total Dissolved Solids</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Zinc :</b><br />1.280131687352417\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.005453952833175827],\"y\":[\"<b>Zinc</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Lead :</b><br />2.3327709869653323e-18\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.0012220136170197063],\"y\":[\"<b>Lead</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Water Temperature :</b><br />10.341549580677556\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.0001236815433428788],\"y\":[\"<b>Water Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Air Temperature :</b><br />45.451641437331375\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.00007870440520326688],\"y\":[\"<b>Air Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Conductivity :</b><br />873.7807482002731\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0006866455434223607],\"y\":[\"<b>Conductivity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>pH :</b><br />8.889125047607246\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[1.1560640474858888],\"y\":[\"<b>pH</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chloride :</b><br />318.33330623326464\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[1.160784039405201],\"y\":[\"<b>Chloride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Sulfate :</b><br />272.51187095824605\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[1.1611711377667775],\"y\":[\"<b>Sulfate</b>\"],\"type\":\"bar\"}],                        {\"barmode\":\"group\",\"height\":550,\"hovermode\":\"closest\",\"margin\":{\"b\":70,\"l\":150,\"r\":20,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Local Explanation - Id: <b>4642920</b><br><sup>Response: <b>Safe to drink</b> - Proba: <b>0.7253</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"type\":\"category\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('b71738fa-bf2e-4af4-904c-2448cd2f95dd');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Local explantion for class 0(not safe to drink)","metadata":{}},{"cell_type":"code","source":"lgbm_xpl.plot.local_plot(index=23691,label=0)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:02:46.416146Z","iopub.execute_input":"2023-03-04T10:02:46.416581Z","iopub.status.idle":"2023-03-04T10:02:46.463881Z","shell.execute_reply.started":"2023-03-04T10:02:46.416546Z","shell.execute_reply":"2023-03-04T10:02:46.462503Z"},"trusted":true},"execution_count":92,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"0b686b9e-49ae-41ee-843f-ed0bba50ca6d\" class=\"plotly-graph-div\" style=\"height:550px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"0b686b9e-49ae-41ee-843f-ed0bba50ca6d\")) {                    Plotly.newPlot(                        \"0b686b9e-49ae-41ee-843f-ed0bba50ca6d\",                        [{\"customdata\":[\"<b>Water Temperature :</b><br />14.633729663268864\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.000011104536608427822],\"y\":[\"<b>Water Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Conductivity :</b><br />242.58198280871952\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0001872898090711461],\"y\":[\"<b>Conductivity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Air Temperature :</b><br />78.97384274877871\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0004310620230856361],\"y\":[\"<b>Air Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Lead :</b><br />5.088895176402236e-150\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0035388968079420133],\"y\":[\"<b>Lead</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Zinc :</b><br />2.146188242032322\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.013465344795845087],\"y\":[\"<b>Zinc</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Sulfate :</b><br />153.842016668393\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.023400546354919817],\"y\":[\"<b>Sulfate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Total Dissolved Solids :</b><br />260.6520159346165\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.03249751511692767],\"y\":[\"<b>Total Dissolved Solids</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Fluoride :</b><br />0.5174591220123643\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0425857458383781],\"y\":[\"<b>Fluoride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chlorine :</b><br />3.8494302721358498\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.05453589511859562],\"y\":[\"<b>Chlorine</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Iron :</b><br />2.5426463256435315e-05\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.06311641258202762],\"y\":[\"<b>Iron</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Nitrate :</b><br />3.3652545136741234\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.06992946177991287],\"y\":[\"<b>Nitrate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Odor :</b><br />0.8849928325338988\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.1656104070209201],\"y\":[\"<b>Odor</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Copper :</b><br />0.7500051696241399\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.21116087392682925],\"y\":[\"<b>Copper</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chloride :</b><br />167.85224884007812\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.2669994355505031],\"y\":[\"<b>Chloride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>pH :</b><br />6.770109945164787\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.3131040551245505],\"y\":[\"<b>pH</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Turbidity :</b><br />0.1228084764765649\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.34987140534817335],\"y\":[\"<b>Turbidity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Manganese :</b><br />8.012709144565779e-12\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.6284330053525509],\"y\":[\"<b>Manganese</b>\"],\"type\":\"bar\"}],                        {\"barmode\":\"group\",\"height\":550,\"hovermode\":\"closest\",\"margin\":{\"b\":70,\"l\":150,\"r\":20,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Local Explanation - Id: <b>23691</b><br><sup>Response: <b>Not safe to drink</b> - Proba: <b>0.9878</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"type\":\"category\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('0b686b9e-49ae-41ee-843f-ed0bba50ca6d');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Let's declare and compile SmartExplainer for catboost model","metadata":{}},{"cell_type":"code","source":"catboost_xpl=SmartExplainer(\n    model=cat_clf,\n    label_dict = {0: \"Not safe to drink\", 1: \"Safe to drink\"}\n    )\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:06:07.696876Z","iopub.execute_input":"2023-03-04T10:06:07.697327Z","iopub.status.idle":"2023-03-04T10:06:08.460585Z","shell.execute_reply.started":"2023-03-04T10:06:07.697289Z","shell.execute_reply":"2023-03-04T10:06:08.459221Z"},"trusted":true},"execution_count":96,"outputs":[]},{"cell_type":"code","source":"\ncatboost_xpl.compile(x=X_test,\n y_pred=pd.DataFrame(cat_test_prediction,columns=['Target'],index=y_test.index),\n y_target=y_test, # Optional: allows to display True Values vs Predicted Values\n )","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:06:12.134856Z","iopub.execute_input":"2023-03-04T10:06:12.135907Z","iopub.status.idle":"2023-03-04T10:15:19.596080Z","shell.execute_reply.started":"2023-03-04T10:06:12.135851Z","shell.execute_reply":"2023-03-04T10:15:19.594799Z"},"trusted":true},"execution_count":97,"outputs":[]},{"cell_type":"markdown","source":"### Catboost Model Features importances SHAP","metadata":{}},{"cell_type":"code","source":"catboost_xpl.plot.features_importance()","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:16:02.021725Z","iopub.execute_input":"2023-03-04T10:16:02.022467Z","iopub.status.idle":"2023-03-04T10:16:02.823861Z","shell.execute_reply.started":"2023-03-04T10:16:02.022392Z","shell.execute_reply":"2023-03-04T10:16:02.822713Z"},"trusted":true},"execution_count":98,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"7e4a15f0-6808-489d-b84b-e208d0b999cd\" class=\"plotly-graph-div\" style=\"height:500px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"7e4a15f0-6808-489d-b84b-e208d0b999cd\")) {                    Plotly.newPlot(                        \"7e4a15f0-6808-489d-b84b-e208d0b999cd\",                        [{\"customdata\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":[\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\"],\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"name\":\"Global\",\"orientation\":\"h\",\"x\":[0.0032,0.0034,0.0035,0.0068,0.0179,0.037,0.0424,0.0535,0.0541,0.0577,0.0581,0.0901,0.0912,0.0921,0.0952,0.133,0.1607],\"y\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"type\":\"bar\"}],                        {\"autosize\":false,\"barmode\":\"group\",\"height\":500,\"hovermode\":\"closest\",\"margin\":{\"b\":50,\"l\":160,\"r\":0,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Features Importance<br><sup>Response: <b>Safe to drink</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Mean absolute Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"ticktext\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"tickvals\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"tickmode\":\"array\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('7e4a15f0-6808-489d-b84b-e208d0b999cd');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Subset importances","metadata":{}},{"cell_type":"code","source":"catboost_xpl.plot.features_importance(selection=subset)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:16:32.854784Z","iopub.execute_input":"2023-03-04T10:16:32.855230Z","iopub.status.idle":"2023-03-04T10:16:33.437794Z","shell.execute_reply.started":"2023-03-04T10:16:32.855192Z","shell.execute_reply":"2023-03-04T10:16:33.436363Z"},"trusted":true},"execution_count":99,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"270d95b1-41b0-441d-9e06-e51fa81ff1be\" class=\"plotly-graph-div\" style=\"height:500px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"270d95b1-41b0-441d-9e06-e51fa81ff1be\")) {                    Plotly.newPlot(                        \"270d95b1-41b0-441d-9e06-e51fa81ff1be\",                        [{\"customdata\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(52, 55, 54, 0.7)\"},\"name\":\"Subset\",\"orientation\":\"h\",\"x\":[0.0043,0.0024,0.005,0.0026,0.0326,0.0607,0.0793,0.0509,0.0692,0.0428,0.0248,0.0781,0.1156,0.0514,0.084,0.1153,0.1809],\"y\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"type\":\"bar\"},{\"customdata\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"hovertemplate\":\"Feature: %{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":[\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\",\"rgba(244, 192, 0, 1.0)\"],\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"name\":\"Global\",\"orientation\":\"h\",\"x\":[0.0032,0.0034,0.0035,0.0068,0.0179,0.037,0.0424,0.0535,0.0541,0.0577,0.0581,0.0901,0.0912,0.0921,0.0952,0.133,0.1607],\"y\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"type\":\"bar\"}],                        {\"autosize\":false,\"barmode\":\"group\",\"height\":500,\"hovermode\":\"closest\",\"margin\":{\"b\":50,\"l\":160,\"r\":0,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Features Importance<br><sup>Response: <b>Safe to drink</b> - Subset length: 20 (0%)</sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Mean absolute Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"ticktext\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"tickvals\":[\"Conductivity\",\"Air Temperature\",\"Water Temperature\",\"Lead\",\"Zinc\",\"Sulfate\",\"Total Dissolved Solids\",\"Chlorine\",\"Fluoride\",\"Iron\",\"Nitrate\",\"Odor\",\"Copper\",\"Turbidity\",\"Chloride\",\"Manganese\",\"pH\"],\"tickmode\":\"array\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('270d95b1-41b0-441d-9e06-e51fa81ff1be');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Local explantion for class 1(safe to drink)","metadata":{}},{"cell_type":"code","source":"catboost_xpl.plot.local_plot(index=4642920,label=1)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:16:59.721353Z","iopub.execute_input":"2023-03-04T10:16:59.722181Z","iopub.status.idle":"2023-03-04T10:17:10.131230Z","shell.execute_reply.started":"2023-03-04T10:16:59.722136Z","shell.execute_reply":"2023-03-04T10:17:10.130101Z"},"trusted":true},"execution_count":100,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"848d592f-9232-4d2f-ab15-e9e0d5b3a9d7\" class=\"plotly-graph-div\" style=\"height:550px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"848d592f-9232-4d2f-ab15-e9e0d5b3a9d7\")) {                    Plotly.newPlot(                        \"848d592f-9232-4d2f-ab15-e9e0d5b3a9d7\",                        [{\"customdata\":[\"<b>Manganese :</b><br />3.098688657658745e-10\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.03644452067833432],\"y\":[\"<b>Manganese</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Water Temperature :</b><br />10.341549580677556\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.012908653919324876],\"y\":[\"<b>Water Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Copper :</b><br />0.1864565906380453\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.012294017426979229],\"y\":[\"<b>Copper</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Turbidity :</b><br />0.0095333982886721\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.011505618087475225],\"y\":[\"<b>Turbidity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Conductivity :</b><br />873.7807482002731\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.00463698118951555],\"y\":[\"<b>Conductivity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Lead :</b><br />2.3327709869653323e-18\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.000779146906261531],\"y\":[\"<b>Lead</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Odor :</b><br />2.087222297073667\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0066367021768281706],\"y\":[\"<b>Odor</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Air Temperature :</b><br />45.451641437331375\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.007936991054735682],\"y\":[\"<b>Air Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Nitrate :</b><br />2.04394462105232\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.008590493277493394],\"y\":[\"<b>Nitrate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Zinc :</b><br />1.280131687352417\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.00912486664941591],\"y\":[\"<b>Zinc</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Iron :</b><br />0.0832012567872399\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.020104836684638725],\"y\":[\"<b>Iron</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Total Dissolved Solids :</b><br />225.40785498160088\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.025704383374930157],\"y\":[\"<b>Total Dissolved Solids</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Fluoride :</b><br />0.1345994952349928\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.047169296929148966],\"y\":[\"<b>Fluoride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chlorine :</b><br />3.333113166670345\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.047432855824432206],\"y\":[\"<b>Chlorine</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Sulfate :</b><br />272.51187095824605\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.9467002095743067],\"y\":[\"<b>Sulfate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>pH :</b><br />8.889125047607246\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.9657707973700552],\"y\":[\"<b>pH</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chloride :</b><br />318.33330623326464\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.9807377637872553],\"y\":[\"<b>Chloride</b>\"],\"type\":\"bar\"}],                        {\"barmode\":\"group\",\"height\":550,\"hovermode\":\"closest\",\"margin\":{\"b\":70,\"l\":150,\"r\":20,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Local Explanation - Id: <b>4642920</b><br><sup>Response: <b>Safe to drink</b> - Proba: <b>0.7521</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"type\":\"category\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('848d592f-9232-4d2f-ab15-e9e0d5b3a9d7');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Local explantion for class 1(safe to drink)","metadata":{}},{"cell_type":"code","source":"catboost_xpl.plot.local_plot(index=23691,label=0)","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:17:38.404927Z","iopub.execute_input":"2023-03-04T10:17:38.405353Z","iopub.status.idle":"2023-03-04T10:17:38.452043Z","shell.execute_reply.started":"2023-03-04T10:17:38.405307Z","shell.execute_reply":"2023-03-04T10:17:38.450700Z"},"trusted":true},"execution_count":101,"outputs":[{"output_type":"display_data","data":{"text/html":"<div>                            <div id=\"b8cae237-9214-4671-8c57-a11fa820fd2a\" class=\"plotly-graph-div\" style=\"height:550px; width:900px;\"></div>            <script type=\"text/javascript\">                require([\"plotly\"], function(Plotly) {                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"b8cae237-9214-4671-8c57-a11fa820fd2a\")) {                    Plotly.newPlot(                        \"b8cae237-9214-4671-8c57-a11fa820fd2a\",                        [{\"customdata\":[\"<b>Conductivity :</b><br />242.58198280871952\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.005188076671338825],\"y\":[\"<b>Conductivity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Air Temperature :</b><br />78.97384274877871\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.0033259154701629737],\"y\":[\"<b>Air Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Water Temperature :</b><br />14.633729663268864\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(74, 99, 138, 0.7)\",\"line\":{\"color\":\"rgba(27, 28, 28, 1.0)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[-0.003305215344817331],\"y\":[\"<b>Water Temperature</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Lead :</b><br />5.088895176402236e-150\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.010618074108236389],\"y\":[\"<b>Lead</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Zinc :</b><br />2.146188242032322\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.022955997026995474],\"y\":[\"<b>Zinc</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Sulfate :</b><br />153.842016668393\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.0511269631427597],\"y\":[\"<b>Sulfate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Total Dissolved Solids :</b><br />260.6520159346165\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.07333020985488177],\"y\":[\"<b>Total Dissolved Solids</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chlorine :</b><br />3.8494302721358498\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.09657449394838864],\"y\":[\"<b>Chlorine</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Fluoride :</b><br />0.5174591220123643\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.1161290114834475],\"y\":[\"<b>Fluoride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Iron :</b><br />2.5426463256435315e-05\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.12623855188764443],\"y\":[\"<b>Iron</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Nitrate :</b><br />3.3652545136741234\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.1423703212451807],\"y\":[\"<b>Nitrate</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Turbidity :</b><br />0.1228084764765649\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.19144812610368517],\"y\":[\"<b>Turbidity</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Copper :</b><br />0.7500051696241399\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.1997446830568323],\"y\":[\"<b>Copper</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Chloride :</b><br />167.85224884007812\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.2109205296330048],\"y\":[\"<b>Chloride</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Odor :</b><br />0.8849928325338988\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.22490293146623527],\"y\":[\"<b>Odor</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>Manganese :</b><br />8.012709144565779e-12\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.2747725634335806],\"y\":[\"<b>Manganese</b>\"],\"type\":\"bar\"},{\"customdata\":[\"<b>pH :</b><br />6.770109945164787\"],\"hovertemplate\":\"%{customdata}<br />Contribution: %{x:.4f}<extra></extra>\",\"marker\":{\"color\":\"rgba(244, 192, 0, 1.0)\",\"line\":{\"color\":\"rgba(52, 55, 54, 0.8)\",\"width\":0.5}},\"orientation\":\"h\",\"showlegend\":false,\"x\":[0.339439418348295],\"y\":[\"<b>pH</b>\"],\"type\":\"bar\"}],                        {\"barmode\":\"group\",\"height\":550,\"hovermode\":\"closest\",\"margin\":{\"b\":70,\"l\":150,\"r\":20,\"t\":95},\"template\":{\"data\":{\"scatter\":[{\"type\":\"scatter\"}]}},\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial\",\"size\":24},\"text\":\"Local Explanation - Id: <b>23691</b><br><sup>Response: <b>Not safe to drink</b> - Proba: <b>0.9810</b></sup>\",\"x\":0.5,\"xanchor\":\"center\",\"y\":0.9,\"yanchor\":\"middle\"},\"width\":900,\"xaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16},\"text\":\"Contribution\"}},\"yaxis\":{\"title\":{\"font\":{\"color\":\"rgb(50, 50, 50)\",\"family\":\"Arial Black\",\"size\":16}},\"type\":\"category\",\"dtick\":1,\"automargin\":true}},                        {\"responsive\": true}                    ).then(function(){\n                            \nvar gd = document.getElementById('b8cae237-9214-4671-8c57-a11fa820fd2a');\nvar x = new MutationObserver(function (mutations, observer) {{\n        var display = window.getComputedStyle(gd).display;\n        if (!display || display === 'none') {{\n            console.log([gd, 'removed!']);\n            Plotly.purge(gd);\n            observer.disconnect();\n        }}\n}});\n\n// Listen for the removal of the full notebook cells\nvar notebookContainer = gd.closest('#notebook-container');\nif (notebookContainer) {{\n    x.observe(notebookContainer, {childList: true});\n}}\n\n// Listen for the clearing of the current output cell\nvar outputEl = gd.closest('.output');\nif (outputEl) {{\n    x.observe(outputEl, {childList: true});\n}}\n\n                        })                };                });            </script>        </div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### Catboost model used for the streamlit model demo.","metadata":{}},{"cell_type":"markdown","source":"### Convert the catboost SmartExplainer to SmartPredictor","metadata":{}},{"cell_type":"code","source":"catboost_predictor = catboost_xpl.to_smartpredictor()","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:18:05.275896Z","iopub.execute_input":"2023-03-04T10:18:05.276490Z","iopub.status.idle":"2023-03-04T10:18:05.284281Z","shell.execute_reply.started":"2023-03-04T10:18:05.276435Z","shell.execute_reply":"2023-03-04T10:18:05.282888Z"},"trusted":true},"execution_count":102,"outputs":[]},{"cell_type":"markdown","source":"### Save the catboost SmartPredictor in Pickle File","metadata":{}},{"cell_type":"code","source":"catboost_predictor.save(\"shapash_catboost_smart_predictor.pkl\")","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:18:08.814771Z","iopub.execute_input":"2023-03-04T10:18:08.815722Z","iopub.status.idle":"2023-03-04T10:18:09.237555Z","shell.execute_reply.started":"2023-03-04T10:18:08.815672Z","shell.execute_reply":"2023-03-04T10:18:09.236301Z"},"trusted":true},"execution_count":103,"outputs":[]},{"cell_type":"markdown","source":"### Let's save oneDAL catboost model","metadata":{}},{"cell_type":"code","source":"import pickle\n","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:19:23.486985Z","iopub.execute_input":"2023-03-04T10:19:23.487434Z","iopub.status.idle":"2023-03-04T10:19:23.494314Z","shell.execute_reply.started":"2023-03-04T10:19:23.487396Z","shell.execute_reply":"2023-03-04T10:19:23.492637Z"},"trusted":true},"execution_count":104,"outputs":[]},{"cell_type":"code","source":"pickle.dump(daal_catb_model, open('Catboost_oneDAL_model.pkl', 'wb'))\n ","metadata":{"execution":{"iopub.status.busy":"2023-03-04T10:22:10.676534Z","iopub.execute_input":"2023-03-04T10:22:10.677037Z","iopub.status.idle":"2023-03-04T10:22:10.698698Z","shell.execute_reply.started":"2023-03-04T10:22:10.676996Z","shell.execute_reply":"2023-03-04T10:22:10.697443Z"},"trusted":true},"execution_count":108,"outputs":[]}]}