{"metadata":{"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":50160,"databundleVersionId":7602123,"sourceType":"competition"}],"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":true},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.11.5"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# Example Notebook\n\nWelcome to the example notebook for the Home Credit Kaggle competition. The goal of this competition is to determine how likely a customer is going to default on an issued loan. The main difference between the [first](https://www.kaggle.com/c/home-credit-default-risk) and this competition is that now your submission will be scored with a custom metric that will take into account how well the model performs in future. A decline in performance will be penalized. The goal is to create a model that is stable and performs well in the future.\n\n목적: 고객의 대출 불이행 가능성\n\nIn this notebook you will see how to:\n* Load the data\n* Join tables with Polars - a DataFrame library implemented in Rust language, designed to be blazingy fast and memory efficient.  \n* Create simple aggregation features\n* Train a LightGBM model\n* Create a submission table\n\n## Load the data","metadata":{"_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5"}},{"cell_type":"code","source":"import polars as pl\nimport numpy as np\nimport pandas as pd\nimport lightgbm as lgb\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_auc_score \n\nimport polars.selectors as cs\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport warnings\nimport missingno as mn\n\ndataPath = \"/kaggle/input/home-credit-credit-risk-model-stability/\"\n# dataPath = \"../data/\"","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:51:57.158053Z","iopub.status.busy":"2024-02-19T05:51:57.157786Z","iopub.status.idle":"2024-02-19T05:52:06.234948Z","shell.execute_reply":"2024-02-19T05:52:06.234145Z","shell.execute_reply.started":"2024-02-19T05:51:57.158028Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def set_table_dtypes(df: pl.DataFrame) -> pl.DataFrame:\n    # implement here all desired dtypes for tables\n    # the following is just an example\n    for col in df.columns:\n        # last letter of column name will help you determine the type\n        if col[-1] in (\"P\", \"A\"):\n            df = df.with_columns(pl.col(col).cast(pl.Float64).alias(col))\n    return df\n\ndef convert_strings(df: pd.DataFrame) -> pd.DataFrame:\n    for col in df.columns:  \n        if df[col].dtype.name in ['object', 'string']:\n            df[col] = df[col].astype(\"string\").astype('category')\n            current_categories = df[col].cat.categories\n            new_categories = current_categories.to_list() + [\"Unknown\"]\n            new_dtype = pd.CategoricalDtype(categories=new_categories, ordered=True)\n            df[col] = df[col].astype(new_dtype)\n    return df","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:06.236869Z","iopub.status.busy":"2024-02-19T05:52:06.236583Z","iopub.status.idle":"2024-02-19T05:52:06.244219Z","shell.execute_reply":"2024-02-19T05:52:06.243376Z","shell.execute_reply.started":"2024-02-19T05:52:06.236843Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Load feature_def","metadata":{}},{"cell_type":"code","source":"def get_feature_definitions(columns):\n    return pl.DataFrame({'Variable': columns}).join(\n        feature_def,\n        on = 'Variable',\n        how = 'left',\n    )\n\nfeature_def = pl.read_csv(dataPath + \"feature_definitions.csv\")","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:06.245574Z","iopub.status.busy":"2024-02-19T05:52:06.245259Z","iopub.status.idle":"2024-02-19T05:52:06.361586Z","shell.execute_reply":"2024-02-19T05:52:06.360705Z","shell.execute_reply.started":"2024-02-19T05:52:06.245540Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Load Basetable","metadata":{}},{"cell_type":"code","source":"# train\ntrain_basetable = pl.read_csv(dataPath + \"csv_files/train/train_base.csv\")\n\n# test\ntest_basetable = pl.read_csv(dataPath + \"csv_files/test/test_base.csv\")","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:06.364509Z","iopub.status.busy":"2024-02-19T05:52:06.363943Z","iopub.status.idle":"2024-02-19T05:52:06.698452Z","shell.execute_reply":"2024-02-19T05:52:06.697584Z","shell.execute_reply.started":"2024-02-19T05:52:06.364475Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(train_basetable.columns)\nprint('\\n')\nprint(train_basetable.shape)\nprint('\\n')\ndisplay(train_basetable.head())","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:06.699847Z","iopub.status.busy":"2024-02-19T05:52:06.699551Z","iopub.status.idle":"2024-02-19T05:52:06.737021Z","shell.execute_reply":"2024-02-19T05:52:06.736192Z","shell.execute_reply.started":"2024-02-19T05:52:06.699822Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# depth=0","metadata":{}},{"cell_type":"code","source":"# train\ntrain_static = pl.concat(\n    [\n        pl.read_csv(dataPath + \"csv_files/train/train_static_0_0.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/train/train_static_0_1.csv\").pipe(set_table_dtypes),\n    ],\n    how=\"vertical_relaxed\",\n)\ntrain_static_cb = pl.read_csv(dataPath + \"csv_files/train/train_static_cb_0.csv\").pipe(set_table_dtypes)\n\n# test\ntest_static = pl.concat(\n    [\n        pl.read_csv(dataPath + \"csv_files/test/test_static_0_0.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/test/test_static_0_1.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/test/test_static_0_2.csv\").pipe(set_table_dtypes),\n    ],\n    how=\"vertical_relaxed\",\n)\ntest_static_cb = pl.read_csv(dataPath + \"csv_files/test/test_static_cb_0.csv\").pipe(set_table_dtypes)","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:06.738353Z","iopub.status.busy":"2024-02-19T05:52:06.738029Z","iopub.status.idle":"2024-02-19T05:52:18.281135Z","shell.execute_reply":"2024-02-19T05:52:18.280335Z","shell.execute_reply.started":"2024-02-19T05:52:06.738304Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(train_static.columns)\nprint('\\n')\nprint(train_static.shape)\nprint('\\n')\ndisplay(train_static.head())","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:18.283044Z","iopub.status.busy":"2024-02-19T05:52:18.282539Z","iopub.status.idle":"2024-02-19T05:52:18.301565Z","shell.execute_reply":"2024-02-19T05:52:18.300626Z","shell.execute_reply.started":"2024-02-19T05:52:18.283006Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(train_static_cb.columns)\nprint('\\n')\nprint(train_static_cb.shape)\nprint('\\n')\ndisplay(train_static_cb.head())","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:18.303000Z","iopub.status.busy":"2024-02-19T05:52:18.302690Z","iopub.status.idle":"2024-02-19T05:52:18.336679Z","shell.execute_reply":"2024-02-19T05:52:18.335638Z","shell.execute_reply.started":"2024-02-19T05:52:18.302976Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# depth=1에 해당하는 Internal file \n- `debitcard_1`\n- `deposit_1`","metadata":{}},{"cell_type":"code","source":"# train\ntrain_debitcard_1 = pl.read_csv(dataPath + \"csv_files/train/train_debitcard_1.csv\").pipe(set_table_dtypes)\ntrain_deposit_1 = pl.read_csv(dataPath + \"csv_files/train/train_deposit_1.csv\").pipe(set_table_dtypes)\n\n# test\ntest_debitcard_1 = pl.read_csv(dataPath + \"csv_files/test/test_debitcard_1.csv\").pipe(set_table_dtypes)\ntest_deposit_1 = pl.read_csv(dataPath + \"csv_files/test/test_deposit_1.csv\").pipe(set_table_dtypes)","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:18.338095Z","iopub.status.busy":"2024-02-19T05:52:18.337776Z","iopub.status.idle":"2024-02-19T05:52:18.407212Z","shell.execute_reply":"2024-02-19T05:52:18.406303Z","shell.execute_reply.started":"2024-02-19T05:52:18.338061Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(train_debitcard_1.shape, train_deposit_1.shape)","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:18.409911Z","iopub.status.busy":"2024-02-19T05:52:18.409638Z","iopub.status.idle":"2024-02-19T05:52:18.415796Z","shell.execute_reply":"2024-02-19T05:52:18.414897Z","shell.execute_reply.started":"2024-02-19T05:52:18.409887Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Feature engineering\n\nIn this part, we can see a simple example of joining tables via `case_id`. Here the loading and joining is done with polars library. Polars library is blazingly fast and has much smaller memory footprint than pandas. ","metadata":{}},{"cell_type":"markdown","source":"## Custom\n- `last180dayaveragebalance_704A` -> `max_last180dayaveragebalance_704A`\n- `amount_416A` -> `sum_deposit_amount_A`","metadata":{}},{"cell_type":"code","source":"### debitcard ###\n# case_id를 기준으로 그룹화\n# aggregation functions: 체크카드의 평균 잔액의 최대치\ntrain_debitcard_1_feats = train_debitcard_1.group_by(\"case_id\").agg(\n    pl.col(\"last180dayaveragebalance_704A\").max().alias(\"max_last180dayaveragebalance_704A\")\n)\n\n### deposit ###\n# aggregation functions : 예/적금의 합계 \ntrain_deposit_1_feats = train_deposit_1.group_by('case_id').agg(\n    pl.sum('amount_416A').alias('sum_deposit_amount_A')\n).sort(by='case_id')\n\n# A, D 유형만 선택\nselected_static_cols = []\nfor col in train_static.columns:\n    if col[-1] in (\"A\", \"D\"):\n        selected_static_cols.append(col)\nprint(selected_static_cols)\n\nselected_static_cb_cols = []\nfor col in train_static_cb.columns:\n    if col[-1] in (\"A\", \"D\"):\n        selected_static_cb_cols.append(col)\nprint(selected_static_cb_cols)\n\n\n# Join all tables together\ndata = train_basetable.join(\n    train_static.select([\"case_id\"]+selected_static_cols), how=\"left\", on=\"case_id\"\n).join(\n    train_static_cb.select([\"case_id\"]+selected_static_cb_cols), how=\"left\", on=\"case_id\"\n).join(\n    train_debitcard_1_feats, how=\"left\", on=\"case_id\"\n).join(\n    train_deposit_1_feats, how=\"left\", on=\"case_id\"\n)\n","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:18.417296Z","iopub.status.busy":"2024-02-19T05:52:18.417004Z","iopub.status.idle":"2024-02-19T05:52:19.368251Z","shell.execute_reply":"2024-02-19T05:52:19.367490Z","shell.execute_reply.started":"2024-02-19T05:52:18.417270Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"##","metadata":{}},{"cell_type":"code","source":"### debitcard ###\ntest_debitcard_1_feats_1 = test_debitcard_1.group_by(\"case_id\").agg(\n    pl.col(\"last180dayaveragebalance_704A\").max().alias(\"max_last180dayaveragebalance_704A\")\n)\n\n### deposit ###\ntest_deposit_1_feats = test_deposit_1.group_by('case_id').agg(\n    pl.sum('amount_416A').alias('sum_deposit_amount_A')\n).sort(by='case_id')\n\ndata_submission = test_basetable.join(\n    test_static.select([\"case_id\"]+selected_static_cols), how=\"left\", on=\"case_id\"\n).join(\n    test_static_cb.select([\"case_id\"]+selected_static_cb_cols), how=\"left\", on=\"case_id\"\n).join(\n    test_debitcard_1_feats_1, how=\"left\", on=\"case_id\"\n).join(\n    test_deposit_1_feats, how=\"left\", on=\"case_id\"\n)","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:19.369587Z","iopub.status.busy":"2024-02-19T05:52:19.369313Z","iopub.status.idle":"2024-02-19T05:52:19.381530Z","shell.execute_reply":"2024-02-19T05:52:19.380645Z","shell.execute_reply.started":"2024-02-19T05:52:19.369565Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"case_ids = data[\"case_id\"].unique().shuffle(seed=1)\ncase_ids_train, case_ids_test = train_test_split(case_ids, train_size=0.6, random_state=1)\ncase_ids_valid, case_ids_test = train_test_split(case_ids_test, train_size=0.5, random_state=1)\n\ncols_pred = []\nfor col in data.columns:\n    if col[-1].isupper() and col[:-1].islower():\n        cols_pred.append(col)\n\nprint(cols_pred)\n\ndef from_polars_to_pandas(case_ids: pl.DataFrame) -> pl.DataFrame:\n    return (\n        data.filter(pl.col(\"case_id\").is_in(case_ids))[[\"case_id\", \"WEEK_NUM\", \"target\"]].to_pandas(),\n        data.filter(pl.col(\"case_id\").is_in(case_ids))[cols_pred].to_pandas(),\n        data.filter(pl.col(\"case_id\").is_in(case_ids))[\"target\"].to_pandas()\n    )\n\nbase_train, X_train, y_train = from_polars_to_pandas(case_ids_train)\nbase_valid, X_valid, y_valid = from_polars_to_pandas(case_ids_valid)\nbase_test, X_test, y_test = from_polars_to_pandas(case_ids_test)\n\nfor df in [X_train, X_valid, X_test]:\n    df = convert_strings(df)\n    \nprint('\\n')\nprint(f\"Train: {X_train.shape}\")\nprint(f\"Valid: {X_valid.shape}\")\nprint(f\"Test: {X_test.shape}\")","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:52:19.383112Z","iopub.status.busy":"2024-02-19T05:52:19.382736Z","iopub.status.idle":"2024-02-19T05:52:30.030387Z","shell.execute_reply":"2024-02-19T05:52:30.029190Z","shell.execute_reply.started":"2024-02-19T05:52:19.383080Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Training LightGBM\n\nMinimal example of LightGBM training is shown below.","metadata":{}},{"cell_type":"code","source":"lgb_train = lgb.Dataset(X_train, label=y_train)\nlgb_valid = lgb.Dataset(X_valid, label=y_valid, reference=lgb_train)\n\nparams = {\n    \"boosting_type\": \"gbdt\",\n    \"objective\": \"binary\",\n    \"metric\": \"auc\",\n    \"max_depth\": 3,\n    \"num_leaves\": 31,\n    \"learning_rate\": 0.05,\n    \"feature_fraction\": 0.9,\n    \"bagging_fraction\": 0.8,\n    \"bagging_freq\": 5,\n    \"n_estimators\": 1000,\n    \"verbose\": -1,\n}\n\ngbm = lgb.train(\n    params,\n    lgb_train,\n    valid_sets=lgb_valid,\n    callbacks=[lgb.log_evaluation(50), lgb.early_stopping(10)]\n)","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:53:14.938680Z","iopub.status.busy":"2024-02-19T05:53:14.937913Z","iopub.status.idle":"2024-02-19T05:54:28.320159Z","shell.execute_reply":"2024-02-19T05:54:28.319459Z","shell.execute_reply.started":"2024-02-19T05:53:14.938645Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Evaluation with AUC and then comparison with the stability metric is shown below.","metadata":{}},{"cell_type":"code","source":"for base, X in [(base_train, X_train), (base_valid, X_valid), (base_test, X_test)]:\n    y_pred = gbm.predict(X, num_iteration=gbm.best_iteration)\n    base[\"score\"] = y_pred\n\nprint(f'The AUC score on the train set is: {roc_auc_score(base_train[\"target\"], base_train[\"score\"])}') \nprint(f'The AUC score on the valid set is: {roc_auc_score(base_valid[\"target\"], base_valid[\"score\"])}') \nprint(f'The AUC score on the test set is: {roc_auc_score(base_test[\"target\"], base_test[\"score\"])}')  \n\n\ndef gini_stability(base, w_fallingrate=88.0, w_resstd=-0.5):\n    gini_in_time = base.loc[:, [\"WEEK_NUM\", \"target\", \"score\"]]\\\n        .sort_values(\"WEEK_NUM\")\\\n        .groupby(\"WEEK_NUM\")[[\"target\", \"score\"]]\\\n        .apply(lambda x: 2*roc_auc_score(x[\"target\"], x[\"score\"])-1).tolist()\n    \n    x = np.arange(len(gini_in_time))\n    y = gini_in_time\n    a, b = np.polyfit(x, y, 1)\n    y_hat = a*x + b\n    residuals = y - y_hat\n    res_std = np.std(residuals)\n    avg_gini = np.mean(gini_in_time)\n    return avg_gini + w_fallingrate * min(0, a) + w_resstd * res_std\n\nstability_score_train = gini_stability(base_train)\nstability_score_valid = gini_stability(base_valid)\nstability_score_test = gini_stability(base_test)\n\nprint('\\n')\nprint(f'The stability score on the train set is: {stability_score_train}') \nprint(f'The stability score on the valid set is: {stability_score_valid}') \nprint(f'The stability score on the test set is: {stability_score_test}')  ","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:54:47.773338Z","iopub.status.busy":"2024-02-19T05:54:47.772484Z","iopub.status.idle":"2024-02-19T05:55:08.291894Z","shell.execute_reply":"2024-02-19T05:55:08.290871Z","shell.execute_reply.started":"2024-02-19T05:54:47.773286Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Submission\n\nScoring the submission dataset is below, we need to take care of new categories. Then we save the score as a last step. ","metadata":{}},{"cell_type":"code","source":"X_submission = data_submission[cols_pred].to_pandas()\nX_submission = convert_strings(X_submission)\ncategorical_cols = X_train.select_dtypes(include=['category']).columns\n\nfor col in categorical_cols:\n    train_categories = set(X_train[col].cat.categories)\n    submission_categories = set(X_submission[col].cat.categories)\n    new_categories = submission_categories - train_categories\n    X_submission.loc[X_submission[col].isin(new_categories), col] = \"Unknown\"\n    new_dtype = pd.CategoricalDtype(categories=train_categories, ordered=True)\n    X_train[col] = X_train[col].astype(new_dtype)\n    X_submission[col] = X_submission[col].astype(new_dtype)\n\ny_submission_pred = gbm.predict(X_submission, num_iteration=gbm.best_iteration)\n\n\nsubmission = pd.DataFrame({\n    \"case_id\": data_submission[\"case_id\"].to_numpy(),\n    \"score\": y_submission_pred\n}).set_index('case_id')\nsubmission.to_csv(\"./submission.csv\")","metadata":{"execution":{"iopub.execute_input":"2024-02-19T05:55:24.790421Z","iopub.status.busy":"2024-02-19T05:55:24.790050Z","iopub.status.idle":"2024-02-19T05:55:25.052236Z","shell.execute_reply":"2024-02-19T05:55:25.051367Z","shell.execute_reply.started":"2024-02-19T05:55:24.790389Z"},"trusted":true},"execution_count":null,"outputs":[]}]}