{"metadata":{"kernelspec":{"name":"python3","display_name":"Python 3","language":"python"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.13"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":50160,"databundleVersionId":7602123,"sourceType":"competition"}],"dockerImageVersionId":30648,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# Loading Lib\nimport 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","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:19:51.892725Z","iopub.execute_input":"2024-02-19T16:19:51.893069Z","iopub.status.idle":"2024-02-19T16:19:58.092725Z","shell.execute_reply.started":"2024-02-19T16:19:51.893035Z","shell.execute_reply":"2024-02-19T16:19:58.091821Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def load_and_transform_parquet(path: str, set_dtypes: bool = True) -> pl.DataFrame:\n    # Load Parquet file into a Polars DataFrame\n    df = pl.read_parquet(path)\n    # Optionally apply data type conversions\n    if set_dtypes:\n        df = set_table_dtypes(df)\n\n    return df\n\ndef set_table_dtypes(df: pl.DataFrame) -> pl.DataFrame:\n    # Implement desired dtypes for tables\n    for col in df.columns:\n        # Last letter of column name determines the type\n        if col[-1] in (\"P\", \"A\"):\n            df = df.with_column(col, pl.col(col).cast(pl.Float64))\n    return df\n    \ndef convert_strings(df: pd.DataFrame) -> pd.DataFrame:\n    # Convert object columns to categorical\n    for col in df.columns:\n        if df[col].dtype == 'object':\n            df[col] = df[col].astype(\"category\")\n            # Add \"Unknown\" category if it's not already present\n            if \"Unknown\" not in df[col].cat.categories:\n                df[col] = df[col].cat.add_categories(\"Unknown\")\n            # Replace null values with \"Unknown\"\n            df[col] = df[col].fillna(\"Unknown\")\n    return df","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:19:58.094498Z","iopub.execute_input":"2024-02-19T16:19:58.094775Z","iopub.status.idle":"2024-02-19T16:19:58.104677Z","shell.execute_reply.started":"2024-02-19T16:19:58.094751Z","shell.execute_reply":"2024-02-19T16:19:58.103699Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Defining DataPath\ndataPath = \"/kaggle/input/home-credit-credit-risk-model-stability/parquet_files/\"","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:19:58.105784Z","iopub.execute_input":"2024-02-19T16:19:58.106068Z","iopub.status.idle":"2024-02-19T16:19:58.116109Z","shell.execute_reply.started":"2024-02-19T16:19:58.106044Z","shell.execute_reply":"2024-02-19T16:19:58.115216Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Preparing Training Data\ntrain_basetable = load_and_transform_parquet(dataPath + \"train/train_base.parquet\", set_dtypes=False)\ntrain_static = pl.concat(\n    [\n        load_and_transform_parquet(dataPath + \"train/train_static_0_0.parquet\", set_dtypes=False),\n        load_and_transform_parquet(dataPath + \"train/train_static_0_1.parquet\", set_dtypes=False),\n    ],\n    how=\"vertical_relaxed\",\n)\ntrain_static_cb = load_and_transform_parquet(dataPath + \"train/train_static_cb_0.parquet\", set_dtypes=False)\ntrain_person_1 = load_and_transform_parquet(dataPath + \"train/train_person_1.parquet\", set_dtypes=False)\ntrain_credit_bureau_b_2 = load_and_transform_parquet(dataPath + \"train/train_credit_bureau_b_2.parquet\", set_dtypes=False)","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:19:58.118131Z","iopub.execute_input":"2024-02-19T16:19:58.118384Z","iopub.status.idle":"2024-02-19T16:20:03.019858Z","shell.execute_reply.started":"2024-02-19T16:19:58.118362Z","shell.execute_reply":"2024-02-19T16:20:03.018819Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Preparing TestData\ntest_basetable = load_and_transform_parquet(dataPath + \"test/test_base.parquet\", set_dtypes=False)\ntest_static = pl.concat(\n    [\n        load_and_transform_parquet(dataPath + \"test/test_static_0_0.parquet\", set_dtypes=False),\n        load_and_transform_parquet(dataPath + \"test/test_static_0_1.parquet\", set_dtypes=False),\n    ],\n    how=\"vertical_relaxed\",\n)\ntest_static_cb = load_and_transform_parquet(dataPath + \"test/test_static_cb_0.parquet\", set_dtypes=False)\ntest_person_1 = load_and_transform_parquet(dataPath + \"test/test_person_1.parquet\", set_dtypes=False)\ntest_credit_bureau_b_2 = load_and_transform_parquet(dataPath + \"test/test_credit_bureau_b_2.parquet\", set_dtypes=False)","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:20:03.021054Z","iopub.execute_input":"2024-02-19T16:20:03.021382Z","iopub.status.idle":"2024-02-19T16:20:03.093740Z","shell.execute_reply.started":"2024-02-19T16:20:03.021354Z","shell.execute_reply":"2024-02-19T16:20:03.092760Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Perform operations\ntrain_person_1_feats_1 = train_person_1.group_by(\"case_id\").agg(\n    pl.col(\"mainoccupationinc_384A\").max().alias(\"mainoccupationinc_384A_max\"),\n    (pl.col(\"incometype_1044T\") == \"SELFEMPLOYED\").max().alias(\"mainoccupationinc_384A_any_selfemployed\")\n)\n\ntrain_person_1_feats_2 = train_person_1.select([\"case_id\", \"num_group1\", \"housetype_905L\"]).filter(\n    pl.col(\"num_group1\") == 0\n).drop(\"num_group1\").rename({\"housetype_905L\": \"person_housetype\"})\n\ntrain_credit_bureau_b_2_feats = train_credit_bureau_b_2.group_by(\"case_id\").agg(\n    pl.col(\"pmts_pmtsoverdue_635A\").max().alias(\"pmts_pmtsoverdue_635A_max\"),\n    (pl.col(\"pmts_dpdvalue_108P\") > 31).max().alias(\"pmts_dpdvalue_108P_over31\")\n)\n\nselected_static_cols = []\nfor col in train_static.columns:\n    if col[-1] in (\"A\", \"M\"):\n        selected_static_cols.append(col)\n\nselected_static_cb_cols = []\nfor col in train_static_cb.columns:\n    if col[-1] in (\"A\", \"M\"):\n        selected_static_cb_cols.append(col)\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_person_1_feats_1, how=\"left\", on=\"case_id\"\n).join(\n    train_person_1_feats_2, how=\"left\", on=\"case_id\"\n).join(\n    train_credit_bureau_b_2_feats, how=\"left\", on=\"case_id\"\n)\n","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:20:03.094884Z","iopub.execute_input":"2024-02-19T16:20:03.095185Z","iopub.status.idle":"2024-02-19T16:20:04.503297Z","shell.execute_reply.started":"2024-02-19T16:20:03.095158Z","shell.execute_reply":"2024-02-19T16:20:04.502265Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Perform operations\ntest_person_1_feats_1 = test_person_1.group_by(\"case_id\").agg(\n    pl.col(\"mainoccupationinc_384A\").max().alias(\"mainoccupationinc_384A_max\"),\n    (pl.col(\"incometype_1044T\") == \"SELFEMPLOYED\").max().alias(\"mainoccupationinc_384A_any_selfemployed\")\n)\n\ntest_person_1_feats_2 = test_person_1.select([\"case_id\", \"num_group1\", \"housetype_905L\"]).filter(\n    pl.col(\"num_group1\") == 0\n).drop(\"num_group1\").rename({\"housetype_905L\": \"person_housetype\"})\n\ntest_credit_bureau_b_2_feats = test_credit_bureau_b_2.group_by(\"case_id\").agg(\n    pl.col(\"pmts_pmtsoverdue_635A\").max().alias(\"pmts_pmtsoverdue_635A_max\"),\n    (pl.col(\"pmts_dpdvalue_108P\") > 31).max().alias(\"pmts_dpdvalue_108P_over31\")\n)\n\n# Join all tables together\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_person_1_feats_1, how=\"left\", on=\"case_id\"\n).join(\n    test_person_1_feats_2, how=\"left\", on=\"case_id\"\n).join(\n    test_credit_bureau_b_2_feats, how=\"left\", on=\"case_id\"\n)\n","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:20:04.525653Z","iopub.execute_input":"2024-02-19T16:20:04.525945Z","iopub.status.idle":"2024-02-19T16:20:04.537538Z","shell.execute_reply.started":"2024-02-19T16:20:04.525913Z","shell.execute_reply":"2024-02-19T16:20:04.536815Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"case_ids = data[\"case_id\"].unique()\ncase_ids = case_ids.to_numpy()\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)\nprint(cols_pred)\n\ndef from_polars_to_pandas(data: pl.DataFrame, case_ids: pl.Series, cols_pred: list) -> tuple:\n    selected_columns = [\"case_id\", \"WEEK_NUM\", \"target\"] + cols_pred\n    filtered_data = data.filter(pl.col(\"case_id\").is_in(case_ids)).select(selected_columns)\n\n    base_df = filtered_data[[\"case_id\", \"WEEK_NUM\", \"target\"]].to_pandas()\n    feature_df = filtered_data[cols_pred].to_pandas()\n    target_df = filtered_data[\"target\"].to_pandas()\n\n    return base_df, feature_df, target_df\n\n# Assuming `data`, `case_ids_train`, `case_ids_valid`, `case_ids_test`, and `cols_pred` are already defined\nbase_train, X_train, y_train = from_polars_to_pandas(data, case_ids_train, cols_pred)\nbase_valid, X_valid, y_valid = from_polars_to_pandas(data, case_ids_valid, cols_pred)\nbase_test, X_test, y_test = from_polars_to_pandas(data, case_ids_test, cols_pred)\n\n# Convert string columns to categorical for train, validation, and test sets\nfor df in [X_train, X_valid, X_test]:\n    df = convert_strings(df)\n","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:20:04.538674Z","iopub.execute_input":"2024-02-19T16:20:04.539008Z","iopub.status.idle":"2024-02-19T16:20:08.384697Z","shell.execute_reply.started":"2024-02-19T16:20:04.538964Z","shell.execute_reply":"2024-02-19T16:20:08.383882Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f\"Train: {X_train.shape}\")\nprint(f\"Valid: {X_valid.shape}\")\nprint(f\"Test: {X_test.shape}\")","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:20:08.387658Z","iopub.execute_input":"2024-02-19T16:20:08.387963Z","iopub.status.idle":"2024-02-19T16:20:08.393004Z","shell.execute_reply.started":"2024-02-19T16:20:08.387932Z","shell.execute_reply":"2024-02-19T16:20:08.391944Z"},"trusted":true},"execution_count":null,"outputs":[]},{"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\": 2,\n    \"num_leaves\": 28,\n    \"learning_rate\": 0.01,\n    \"feature_fraction\": 0.9,\n    \"bagging_fraction\": 0.8,\n    \"bagging_freq\": 5,\n    \"n_estimators\": 500,\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.status.busy":"2024-02-19T16:30:05.876265Z","iopub.execute_input":"2024-02-19T16:30:05.876897Z","iopub.status.idle":"2024-02-19T16:30:53.858852Z","shell.execute_reply.started":"2024-02-19T16:30:05.876864Z","shell.execute_reply":"2024-02-19T16:30:53.858121Z"},"trusted":true},"execution_count":null,"outputs":[]},{"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\"])}')","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:30:56.474364Z","iopub.execute_input":"2024-02-19T16:30:56.475209Z","iopub.status.idle":"2024-02-19T16:31:06.157115Z","shell.execute_reply.started":"2024-02-19T16:30:56.475181Z","shell.execute_reply":"2024-02-19T16:31:06.156052Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def 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(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.status.busy":"2024-02-19T16:31:09.311042Z","iopub.execute_input":"2024-02-19T16:31:09.311634Z","iopub.status.idle":"2024-02-19T16:31:10.336551Z","shell.execute_reply.started":"2024-02-19T16:31:09.311602Z","shell.execute_reply":"2024-02-19T16:31:10.335548Z"},"trusted":true},"execution_count":null,"outputs":[]},{"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)","metadata":{"execution":{"iopub.status.busy":"2024-02-19T16:31:17.176785Z","iopub.execute_input":"2024-02-19T16:31:17.177698Z","iopub.status.idle":"2024-02-19T16:31:17.253385Z","shell.execute_reply.started":"2024-02-19T16:31:17.177663Z","shell.execute_reply":"2024-02-19T16:31:17.252561Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission = 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.status.busy":"2024-02-19T16:31:18.144332Z","iopub.execute_input":"2024-02-19T16:31:18.144986Z","iopub.status.idle":"2024-02-19T16:31:18.151590Z","shell.execute_reply.started":"2024-02-19T16:31:18.144955Z","shell.execute_reply":"2024-02-19T16:31:18.150805Z"},"trusted":true},"execution_count":null,"outputs":[]}]}