{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"

Default Prediction - Gradient Boosting Decision Trees

\n\n![credit card 2.jpg](attachment:324b2d76-c038-4ae8-b4de-2ddab9b5ae68.jpg)\n\n\n
\n\nBasic Intro\n\nCredit default prediction is central to managing risk in a consumer lending business. The objective of this competition is to predict the probability that a customer does not pay back their credit card balance amount in the future based on their monthly customer profile. The target binary variable is calculated by observing 18 months performance window after the latest credit card statement, and if the customer does not pay due amount in 120 days after their latest statement date it is considered a default event.\n\nPreprocessing\n\nWe preprocessed the data by:\n* removing categorical features\n* removing colmuns that have more then threshold precent of missing values\n* keeping only colmuns above variance threshold \n\nFeatures engineering\n\nWe checked two kinds of features, which are the median and the mean on each of the customer columns. From the validation set results we conclude that the mean features are more suitable for this task.\n\nOur model\n\nWe used gradient boosting decision trees. specifically, we trained and evaluated the LGBMClassifier, XGBClassifier and the CatBoostClassifier models. Our best performing model got accuracy of 0.891 on the validation set.\nAlso, we ploted the ROC curve and Confusion Matrices for each of our models.\nAs our submission for this challenge, we used the mean of the tree models predictions.\n\nGradient Boosting Decision Trees\n\nGradient Boosted Trees is a ensembling method that perform regression or classification by combining the outputs from individual trees. In this method, we combines weak learners (usually decision trees with only one split, called decision stumps) sequentially, so that each new tree corrects the errors of the previous one.\n\n\n![The-architecture-of-Gradient-Boosting-Decision-Tree 2.png](attachment:a88a6e99-0f1d-40a8-bab2-330afdd7eae1.png)\n\n\nExplainability with SHAP\n\nSHAP is a mathematical method to explain the predictions of ML models. It is based on the concepts of game theory and can be used to explain the predictions of any machine learning model by calculating the contribution of each feature to the prediction.\n\nAcknowledgements\n\n* [Introduction to Boosted Trees](https://xgboost.readthedocs.io/en/stable/tutorials/model.html).\n\n* [You Are Missing Out on LightGBM. It Crushes XGBoost in Every Aspect](https://towardsdatascience.com/how-to-beat-the-heck-out-of-xgboost-with-lightgbm-comprehensive-tutorial-5eba52195997).\n\n* [AMEX: Bruteforce Feature Engineering](https://www.kaggle.com/code/thedevastator/amex-bruteforce-feature-engineering).\n\n* [Ensemble xg_boost&catboost&lightgbm | Credit default](https://www.kaggle.com/code/amant555/ensemble-xg-boost-catboost-lightgbm-credit-default?scriptVersionId=103255966).\n\n* [XGBoost explainability with SHAP](https://www.kaggle.com/code/bryanb/xgboost-explainability-with-shap/notebook).","metadata":{},"attachments":{"324b2d76-c038-4ae8-b4de-2ddab9b5ae68.jpg":{"image/jpeg":"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"},"a88a6e99-0f1d-40a8-bab2-330afdd7eae1.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"# 1. Imports","metadata":{}},{"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 time\nfrom sklearn.feature_selection import VarianceThreshold\nfrom itertools import compress\nfrom sklearn.metrics import accuracy_score \nfrom sklearn.metrics import precision_score \nfrom sklearn.metrics import recall_score\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_curve, auc, confusion_matrix\n\nfrom matplotlib import pyplot\nimport seaborn as sns\n\nfrom lightgbm import LGBMClassifier\nfrom xgboost import XGBClassifier\nfrom catboost import CatBoostClassifier\nfrom IPython.display import FileLink\n\nimport shap\n\n#CONSTS\nCATEGORICAL_FEATURES = [\"B_30\", \"B_38\", \"D_114\", \"D_116\", \"D_117\", \"D_120\", \"D_126\", \"D_63\", \"D_64\", \"D_66\", \"D_68\"]\nBATCH_SIZE = 1_000_000\nVARIANCE_THRESHOLD = 0.03\nNULL_VALUES_THRESHOLD = 0.95","metadata":{"execution":{"iopub.status.busy":"2022-09-27T08:47:09.711672Z","iopub.execute_input":"2022-09-27T08:47:09.71215Z","iopub.status.idle":"2022-09-27T08:47:15.734816Z","shell.execute_reply.started":"2022-09-27T08:47:09.71206Z","shell.execute_reply":"2022-09-27T08:47:15.733737Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 2. Helper Code","metadata":{}},{"cell_type":"code","source":"#function for getting best indexes for batches, so we won't split the information of the same customer_ID\n#we are processing the data in batches due memory constraints \ndef get_indexes_for_batching(df):\n indexes = []\n size = df.shape[0]\n for i in range(BATCH_SIZE, size, BATCH_SIZE):\n temp = currCustomer = df['customer_ID'][i]\n index = i\n while temp == currCustomer:\n index += 1\n temp = df['customer_ID'][index]\n indexes.append(index)\n if indexes[-1] != size - 1: \n indexes.append(size - 1)\n \n return indexes","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:43:16.666359Z","iopub.execute_input":"2022-09-22T12:43:16.668226Z","iopub.status.idle":"2022-09-22T12:43:16.676551Z","shell.execute_reply.started":"2022-09-22T12:43:16.668163Z","shell.execute_reply":"2022-09-22T12:43:16.675113Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#function that combines customer entries to one entry based on median of each colmun\ndef combine_customer_entries(df, indexes, criteria, clean_rows=False):\n prev_index = 0\n train_data_features = pd.DataFrame()\n for i, curr_index in enumerate(indexes):\n if (criteria == 'mean'):\n cur_df = df.iloc[prev_index:curr_index,:].groupby(['customer_ID']).mean()\n elif (criteria == 'median'):\n cur_df = df.iloc[prev_index:curr_index,:].groupby(['customer_ID']).median()\n else:\n print(\"Error, must choose mean or median\")\n return\n if (clean_rows):\n cur_df = cur_df.dropna(how='all')\n train_data_features = train_data_features.append(cur_df)\n prev_index = curr_index\n print(i, train_data_features.shape)\n return train_data_features","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:43:16.678055Z","iopub.execute_input":"2022-09-22T12:43:16.678698Z","iopub.status.idle":"2022-09-22T12:43:16.690657Z","shell.execute_reply.started":"2022-09-22T12:43:16.678638Z","shell.execute_reply":"2022-09-22T12:43:16.689036Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#function that keeps only the colmuns above variance threshold \ndef apply_variance_threshold(df):\n selector = VarianceThreshold(VARIANCE_THRESHOLD)\n selector.fit(df)\n return selector.get_support(indices=True)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:43:16.69421Z","iopub.execute_input":"2022-09-22T12:43:16.694639Z","iopub.status.idle":"2022-09-22T12:43:16.70944Z","shell.execute_reply.started":"2022-09-22T12:43:16.694602Z","shell.execute_reply":"2022-09-22T12:43:16.707822Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 3. Loading the data","metadata":{}},{"cell_type":"code","source":"#load train data\ntrain_data = pd.read_parquet('../input/amex-parquet/train_data.parquet')\nprint(\"The shape with the categorical features is:\", train_data.shape)\n\n#get train labels\ntrain_labels = pd.read_csv('../input/amex-default-prediction/train_labels.csv').iloc[:,1]\nprint(\"The shape of the train labels is:\", train_labels.shape)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:43:16.711362Z","iopub.execute_input":"2022-09-22T12:43:16.711814Z","iopub.status.idle":"2022-09-22T12:43:55.932969Z","shell.execute_reply.started":"2022-09-22T12:43:16.711776Z","shell.execute_reply":"2022-09-22T12:43:55.931379Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#remove catagorical features & target colmun\ntrain_data = train_data.drop(CATEGORICAL_FEATURES, axis=1)\ntrain_data = train_data.drop('target', axis=1)\nprint(\"The shape without the categorical features is:\", train_data.shape)\ntrain_data.head()","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:43:55.934336Z","iopub.execute_input":"2022-09-22T12:43:55.934704Z","iopub.status.idle":"2022-09-22T12:44:00.760806Z","shell.execute_reply.started":"2022-09-22T12:43:55.934672Z","shell.execute_reply":"2022-09-22T12:44:00.759432Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#remove colmuns that have more then THRESHOLD precent of missing values\nprint(train_data.shape)\ntrain_data = train_data.dropna(axis=1, thresh=int(NULL_VALUES_THRESHOLD*train_data.shape[0]))\nprint(train_data.shape)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:00.762824Z","iopub.execute_input":"2022-09-22T12:44:00.76318Z","iopub.status.idle":"2022-09-22T12:44:04.686641Z","shell.execute_reply.started":"2022-09-22T12:44:00.763147Z","shell.execute_reply":"2022-09-22T12:44:04.685211Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#get indexes for processing batches of data\nindexes_for_batching = get_indexes_for_batching(train_data)\nprint(indexes_for_batching)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:04.688296Z","iopub.execute_input":"2022-09-22T12:44:04.688651Z","iopub.status.idle":"2022-09-22T12:44:04.711821Z","shell.execute_reply.started":"2022-09-22T12:44:04.68862Z","shell.execute_reply":"2022-09-22T12:44:04.710578Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 4. Preprocessing the data","metadata":{}},{"cell_type":"code","source":"#creates the features vector for each 'customer_ID' according to criteria (mean or median)\n#splits the dataset to train set and test set\ndef preprocessing_and_split_data(df_train_data, df_train_labels, indexes_for_batching, criteria):\n #get combined entries for each customer based on median of each clomun \n data_features = combine_customer_entries(df_train_data, indexes_for_batching, criteria)\n \n #setting train and test from training data received from processing\n x_train_split, x_test_split, y_train_split, y_test_split = train_test_split(data_features, df_train_labels, test_size=0.1, random_state=26)\n \n #get only data with colmuns with variance above threshold\n #the colmuns to keep are determined only by train set \n columns_to_keep = apply_variance_threshold(x_train_split)\n x_train_split_final = x_train_split[x_train_split.columns[columns_to_keep]]\n x_test_split_final = x_test_split[x_test_split.columns[columns_to_keep]]\n \n return columns_to_keep, x_train_split_final, x_test_split_final, y_train_split, y_test_split","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:04.713676Z","iopub.execute_input":"2022-09-22T12:44:04.714035Z","iopub.status.idle":"2022-09-22T12:44:04.722437Z","shell.execute_reply.started":"2022-09-22T12:44:04.714003Z","shell.execute_reply":"2022-09-22T12:44:04.721377Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#creating the median features\ncolumns_to_keep_median, x_train_median, x_test_median, y_train_median, y_test_median \\\n= preprocessing_and_split_data(train_data, train_labels, indexes_for_batching, 'median')\nprint(\"The median features shape:\", x_train_median.shape)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:04.728666Z","iopub.execute_input":"2022-09-22T12:44:04.72916Z","iopub.status.idle":"2022-09-22T12:44:29.427824Z","shell.execute_reply.started":"2022-09-22T12:44:04.729101Z","shell.execute_reply":"2022-09-22T12:44:29.426555Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#creating the mean features\ncolumns_to_keep_mean, x_train_mean, x_test_mean, y_train_mean, y_test_mean \\\n= preprocessing_and_split_data(train_data, train_labels, indexes_for_batching, 'mean')\nprint(\"The mean features shape:\", x_train_mean.shape)\nprint(\"The chosen columns:\", x_train_mean.columns)\nx_train_mean.head()","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:29.439657Z","iopub.execute_input":"2022-09-22T12:44:29.439986Z","iopub.status.idle":"2022-09-22T12:44:43.978856Z","shell.execute_reply.started":"2022-09-22T12:44:29.439956Z","shell.execute_reply":"2022-09-22T12:44:43.977959Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 5. Training","metadata":{}},{"cell_type":"code","source":"#set up lgbm model, train on train part in training data and test vs test part in training data!\ndef train_and_eval_model(model, model_name, x_train, x_test, y_train, y_test, criteria, is_silent=False):\n start = time.time()\n if (is_silent):\n model.fit(x_train, y_train, silent=is_silent) #for canceling the training prints\n else:\n model.fit(x_train, y_train)\n end = time.time()\n print(\"The training took\", '{:.2f}'.format(end - start) ,\"seconds\")\n \n y_predict_train=model.predict(x_train)\n y_predict_test=model.predict(x_test)\n \n print('Results with the {} features:'.format(criteria))\n print('Classifier Accuracy on the train set: {:.3f}'.format(accuracy_score(y_train, y_predict_train)))\n print('Classifier Accuracy on the test set: {:.3f}'.format(accuracy_score(y_test, y_predict_test)))\n print('Classifier Recall on the test set: {:.3f}'.format(recall_score(y_test, y_predict_test)))\n print('Classifier Precision on the test set: {:.3f}'.format(precision_score(y_test, y_predict_test)))\n \n # Create a DataFrame from our training statistics\n training_stats = []\n training_stats.append(\n {\n 'Model Name' : model_name + \"_\" + criteria,\n 'Train. Accuracy.': accuracy_score(y_train, y_predict_train),\n 'Test. Accuracy.': accuracy_score(y_test, y_predict_test),\n 'Test. Recall.': recall_score(y_test, y_predict_test),\n 'Test. Precision.': precision_score(y_test, y_predict_test),\n 'Training Time [s]': end - start,\n }\n )\n pd.set_option('precision', 3)\n df_stats = pd.DataFrame(data=training_stats)\n df_stats = df_stats.set_index('Model Name')\n return model, df_stats","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:43.983928Z","iopub.execute_input":"2022-09-22T12:44:43.984867Z","iopub.status.idle":"2022-09-22T12:44:43.999738Z","shell.execute_reply.started":"2022-09-22T12:44:43.984814Z","shell.execute_reply":"2022-09-22T12:44:43.998418Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#training and evaluating lightgbm, xgboost and catboost, with median and mean features\ncriterias = ['median', 'mean']\nmodels_types = [['catboost', True], ['lightgbm', False], ['xgboost', False]]\nmodels_dict = {'lightgbm':LGBMClassifier, 'xgboost':XGBClassifier, 'catboost':CatBoostClassifier}\ntrained_models = []\nresults = pd.DataFrame()\n# results_data_type = pd.concat([models_schedulers[0].df_stats, models_schedulers[1].df_stats, models_schedulers[2].df_stats, models_schedulers[3].df_stats], ignore_index=True)\n\n\nfor model_type_elem in models_types:\n for criteria in criterias:\n model_type = model_type_elem[0]\n is_silent = model_type_elem[1]\n print('Using the {} model with the {} features'.format(model_type, criteria))\n model = models_dict[model_type]()\n if (criteria == 'median'):\n model, df_stats = train_and_eval_model(model, model_type, x_train_median, x_test_median, y_train_median, y_test_median, criteria, is_silent)\n trained_models.append(model)\n results = pd.concat([results, df_stats])\n if (criteria == 'mean'):\n model, df_stats = train_and_eval_model(model, model_type, x_train_mean, x_test_mean, y_train_mean, y_test_mean, criteria, is_silent)\n trained_models.append(model)\n results = pd.concat([results, df_stats])\n print()","metadata":{"execution":{"iopub.status.busy":"2022-09-22T12:44:44.006591Z","iopub.execute_input":"2022-09-22T12:44:44.007135Z","iopub.status.idle":"2022-09-22T13:15:26.493397Z","shell.execute_reply.started":"2022-09-22T12:44:44.007086Z","shell.execute_reply":"2022-09-22T13:15:26.491477Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Lets look at the results\nresults","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:26.496282Z","iopub.execute_input":"2022-09-22T13:15:26.496758Z","iopub.status.idle":"2022-09-22T13:15:26.516333Z","shell.execute_reply.started":"2022-09-22T13:15:26.496717Z","shell.execute_reply":"2022-09-22T13:15:26.515252Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#plotting the ROC curve of our xgboost classifier with mean features ROC curve, \n#ROC is a graphical plot that illustrates the diagnostic ability of a binary classifier system as its discrimination threshold is varied.\ncat_model = trained_models[1]\nlight_model = trained_models[3]\nxgb_model = trained_models[5]\ny_predict_cat = cat_model.predict(x_test_mean)\ny_predict_light = light_model.predict(x_test_mean)\ny_predict_xgb = xgb_model.predict(x_test_mean)\n\nx = np.linspace(0, 1, 5)\ncat_fpr, cat_tpr, _ = roc_curve(y_test_mean, y_predict_cat)\nlight_fpr, light_tpr, _ = roc_curve(y_test_mean, y_predict_light)\nxgb_fpr, xgb_tpr, _ = roc_curve(y_test_mean, y_predict_xgb)\n\nprint('The AUC of the catboost model is {:.3f}'.format(auc(cat_fpr, cat_tpr)))\nprint('The AUC of the lightgbm model is {:.3f}'.format(auc(light_fpr, light_tpr)))\nprint('The AUC of the xgboost model is {:.3f}'.format(auc(xgb_fpr, xgb_tpr)))\n\npyplot.plot(cat_fpr, cat_tpr, marker='.', label='catboost')\npyplot.plot(light_fpr, light_tpr, marker='.', label='lightgbm')\npyplot.plot(xgb_fpr, xgb_tpr, marker='.', label='xgboost')\npyplot.plot(x, x, linestyle='--', label='No Skill')\npyplot.xlabel('False Positive Rate')\npyplot.ylabel('True Positive Rate')\npyplot.title('ROC curve')\npyplot.legend()\npyplot.show()","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:26.517941Z","iopub.execute_input":"2022-09-22T13:15:26.519133Z","iopub.status.idle":"2022-09-22T13:15:27.448521Z","shell.execute_reply.started":"2022-09-22T13:15:26.519092Z","shell.execute_reply":"2022-09-22T13:15:27.447137Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Showing Confusion Matrix\ndef plot_cm(y_true, y_pred, title, figsize=(5,5)):\n cm = confusion_matrix(y_true, y_pred, labels=np.unique(y_true))\n cm_sum = np.sum(cm, axis=1, keepdims=True)\n cm_perc = cm / cm_sum.astype(float) * 100\n annot = np.empty_like(cm).astype(str)\n nrows, ncols = cm.shape\n for i in range(nrows):\n for j in range(ncols):\n c = cm[i, j]\n p = cm_perc[i, j]\n if i == j:\n s = cm_sum[i]\n annot[i, j] = '%.1f%%\\n%d/%d' % (p, c, s)\n elif c == 0:\n annot[i, j] = ''\n else:\n annot[i, j] = '%.1f%%\\n%d' % (p, c)\n cm = pd.DataFrame(cm, index=np.unique(y_true), columns=np.unique(y_true))\n cm.index.name = 'Actual'\n cm.columns.name = 'Predicted'\n fig, ax = pyplot.subplots(figsize=figsize)\n pyplot.title(title)\n sns.heatmap(cm, cmap= \"YlGnBu\", annot=annot, fmt='', ax=ax)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:27.450166Z","iopub.execute_input":"2022-09-22T13:15:27.451325Z","iopub.status.idle":"2022-09-22T13:15:27.46308Z","shell.execute_reply.started":"2022-09-22T13:15:27.451284Z","shell.execute_reply":"2022-09-22T13:15:27.461476Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_cm(y_predict_cat, y_test_mean, 'Confusion Matrix for Catboost, mean features', figsize=(7,7))\nplot_cm(y_predict_light, y_test_mean, 'Confusion Matrix for Lightgbm, mean features', figsize=(7,7))\nplot_cm(y_predict_xgb, y_test_mean, 'Confusion Matrix for Xgboost, mean features', figsize=(7,7))","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:27.464948Z","iopub.execute_input":"2022-09-22T13:15:27.465408Z","iopub.status.idle":"2022-09-22T13:15:28.376628Z","shell.execute_reply.started":"2022-09-22T13:15:27.465362Z","shell.execute_reply":"2022-09-22T13:15:28.375366Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 6. Explainability with SHAP","metadata":{}},{"cell_type":"markdown","source":"We will use SHAP for our Xgboost mean model, to understand better the features effect on the output.","metadata":{}},{"cell_type":"code","source":"#Using a random sample of the dataframe for better time computation\nX_sampled = x_train_mean.sample(3_000, random_state=10)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:28.378505Z","iopub.execute_input":"2022-09-22T13:15:28.379204Z","iopub.status.idle":"2022-09-22T13:15:28.426113Z","shell.execute_reply.started":"2022-09-22T13:15:28.379129Z","shell.execute_reply":"2022-09-22T13:15:28.424593Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#explain the model's predictions using SHAP values\n#(same syntax works for LightGBM, CatBoost, and scikit-learn models)\nexplainer = shap.TreeExplainer(trained_models[5])\nshap_values = explainer.shap_values(X_sampled)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:28.427724Z","iopub.execute_input":"2022-09-22T13:15:28.428174Z","iopub.status.idle":"2022-09-22T13:15:30.869997Z","shell.execute_reply.started":"2022-09-22T13:15:28.428134Z","shell.execute_reply":"2022-09-22T13:15:30.868973Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"shap.summary_plot(shap_values, X_sampled, plot_type=\"bar\")","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:30.871565Z","iopub.execute_input":"2022-09-22T13:15:30.872216Z","iopub.status.idle":"2022-09-22T13:15:31.343091Z","shell.execute_reply.started":"2022-09-22T13:15:30.87216Z","shell.execute_reply":"2022-09-22T13:15:31.341699Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# load JS visualization code to notebook\nshap.initjs()\n\n# visualize the first prediction's explanation\nshap.force_plot(explainer.expected_value, shap_values[0,:], X_sampled.iloc[0,:])","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:31.345225Z","iopub.execute_input":"2022-09-22T13:15:31.345794Z","iopub.status.idle":"2022-09-22T13:15:31.376856Z","shell.execute_reply.started":"2022-09-22T13:15:31.345733Z","shell.execute_reply":"2022-09-22T13:15:31.37532Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The above explanation shows features each contributing to push the model output from the base value (the average model output over the training dataset we passed) to the model output. Features pushing the prediction higher are shown in red, those pushing the prediction lower are in blue.","metadata":{}},{"cell_type":"code","source":"# summarize the effects of all the features\nshap.summary_plot(shap_values, X_sampled)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:31.378546Z","iopub.execute_input":"2022-09-22T13:15:31.37904Z","iopub.status.idle":"2022-09-22T13:15:33.084026Z","shell.execute_reply.started":"2022-09-22T13:15:31.379001Z","shell.execute_reply":"2022-09-22T13:15:33.082675Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 7. Submission","metadata":{}},{"cell_type":"code","source":"columns_to_load = list(x_train_mean.columns)\ncolumns_to_load = ['customer_ID'] + columns_to_load\nprint(\"The columns we use for our model: \", columns_to_load)\ndel(train_data)\ndel(train_labels)\ndel(x_train_median, x_test_median, y_train_median, y_test_median)\ndel(x_train_mean, x_test_mean, y_train_mean, y_test_mean)\ndel(y_predict_cat, y_predict_light, y_predict_xgb)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:33.085367Z","iopub.execute_input":"2022-09-22T13:15:33.085745Z","iopub.status.idle":"2022-09-22T13:15:33.352842Z","shell.execute_reply.started":"2022-09-22T13:15:33.085713Z","shell.execute_reply":"2022-09-22T13:15:33.350855Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#creates a prediction for submission, from catboost, lightgbm and xgboost models\ndef create_submission(data, trained_models):\n y_test_predict = trained_models[1].predict_proba(data)\n pred_final_cbm = y_test_predict[:,1]\n y_test_predict = trained_models[3].predict_proba(data)\n pred_final_lgbm = y_test_predict[:,1]\n y_test_predict = trained_models[5].predict_proba(data)\n pred_final_xgbm = y_test_predict[:,1]\n \n pred_df= pd.DataFrame({'lgbm':pred_final_lgbm,'xgbm':pred_final_xgbm,'catboost':pred_final_cbm})\n pred_df['mean'] = (pred_df['lgbm'] + pred_df['xgbm'] + pred_df['catboost'])/3\n pred_mean_values=pred_df['mean'].values\n \n # Merge the prediction and customer_ID into submission dataframe\n submission = pd.DataFrame({\"customer_ID\":data.index,\"prediction\":pred_final_cbm})\n return submission, pred_df","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:33.359397Z","iopub.execute_input":"2022-09-22T13:15:33.35981Z","iopub.status.idle":"2022-09-22T13:15:33.368568Z","shell.execute_reply.started":"2022-09-22T13:15:33.359776Z","shell.execute_reply":"2022-09-22T13:15:33.367108Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#load test data\ntest_data = pd.read_parquet('../input/amex-parquet/test_data.parquet',columns=columns_to_load)\nprint(\"The shape of the testdata is:\", test_data.shape)\nprint(\"The number of uniqe 'customer_IDs' is: \", len(pd.unique(test_data['customer_ID'])))","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:15:33.386657Z","iopub.execute_input":"2022-09-22T13:15:33.387398Z","iopub.status.idle":"2022-09-22T13:16:39.946148Z","shell.execute_reply.started":"2022-09-22T13:15:33.387342Z","shell.execute_reply":"2022-09-22T13:16:39.944939Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#get indexes for processing batches of data\nindexes_for_batching_sub = get_indexes_for_batching(test_data)\nprint(indexes_for_batching_sub)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:16:39.947912Z","iopub.execute_input":"2022-09-22T13:16:39.948369Z","iopub.status.idle":"2022-09-22T13:16:39.956767Z","shell.execute_reply.started":"2022-09-22T13:16:39.948321Z","shell.execute_reply":"2022-09-22T13:16:39.955481Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#creating the mean features for submission\ndata_features = combine_customer_entries(test_data, indexes_for_batching_sub, 'mean', clean_rows=True)\nprint(\"The test features shape is: \", data_features.shape)\ndel(test_data)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:16:39.958766Z","iopub.execute_input":"2022-09-22T13:16:39.959162Z","iopub.status.idle":"2022-09-22T13:17:31.008818Z","shell.execute_reply.started":"2022-09-22T13:16:39.959126Z","shell.execute_reply":"2022-09-22T13:17:31.007092Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission, pred_df = create_submission(data_features, trained_models)\nprint(\"The predictions of all the three models shape: \", pred_df.shape)\nprint(pred_df)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:17:31.011051Z","iopub.execute_input":"2022-09-22T13:17:31.011503Z","iopub.status.idle":"2022-09-22T13:17:37.479204Z","shell.execute_reply.started":"2022-09-22T13:17:31.011466Z","shell.execute_reply":"2022-09-22T13:17:37.477503Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"The submission shape: \", submission.shape)\nprint(submission)","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:17:37.481369Z","iopub.execute_input":"2022-09-22T13:17:37.481903Z","iopub.status.idle":"2022-09-22T13:17:37.499281Z","shell.execute_reply.started":"2022-09-22T13:17:37.481855Z","shell.execute_reply":"2022-09-22T13:17:37.497006Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.to_csv('submission2.csv', index=False)\nFileLink('submission2.csv')","metadata":{"execution":{"iopub.status.busy":"2022-09-22T13:17:37.501316Z","iopub.execute_input":"2022-09-22T13:17:37.501864Z","iopub.status.idle":"2022-09-22T13:17:41.43708Z","shell.execute_reply.started":"2022-09-22T13:17:37.501801Z","shell.execute_reply":"2022-09-22T13:17:41.435707Z"},"trusted":true},"execution_count":null,"outputs":[]}]}