{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":50160,"databundleVersionId":7921029,"sourceType":"competition"}],"dockerImageVersionId":30635,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"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\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":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19"}},{"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\ndataPath = \"/kaggle/input/home-credit-credit-risk-model-stability/\"","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:38:16.409337Z","iopub.execute_input":"2024-05-19T12:38:16.409835Z","iopub.status.idle":"2024-05-19T12:38:20.261054Z","shell.execute_reply.started":"2024-05-19T12:38:16.409786Z","shell.execute_reply":"2024-05-19T12:38:20.260047Z"},"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\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.status.busy":"2024-05-19T12:38:34.370785Z","iopub.execute_input":"2024-05-19T12:38:34.371181Z","iopub.status.idle":"2024-05-19T12:38:34.380813Z","shell.execute_reply.started":"2024-05-19T12:38:34.371149Z","shell.execute_reply":"2024-05-19T12:38:34.379094Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_basetable = pl.read_csv(dataPath + \"csv_files/train/train_base.csv\")\ntrain_applprev_1 = pl.concat(\n    [\n        pl.read_csv(dataPath + \"csv_files/train/train_applprev_1_0.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/train/train_applprev_1_0.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)\ntrain_deposit_1 = pl.read_csv(dataPath + \"csv_files/train/train_deposit_1.csv\").pipe(set_table_dtypes) \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)","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:39:13.828904Z","iopub.execute_input":"2024-05-19T12:39:13.829317Z","iopub.status.idle":"2024-05-19T12:39:40.724398Z","shell.execute_reply.started":"2024-05-19T12:39:13.829285Z","shell.execute_reply":"2024-05-19T12:39:40.722916Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_deposit_1","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:40:27.303819Z","iopub.execute_input":"2024-05-19T12:40:27.304259Z","iopub.status.idle":"2024-05-19T12:40:27.315684Z","shell.execute_reply.started":"2024-05-19T12:40:27.304222Z","shell.execute_reply":"2024-05-19T12:40:27.314259Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_basetable = pl.read_csv(dataPath + \"csv_files/test/test_base.csv\")\ntest_applprev_1 = pl.concat(\n    [\n        pl.read_csv(dataPath + \"csv_files/test/test_applprev_1_0.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/test/test_applprev_1_1.csv\").pipe(set_table_dtypes),\n        pl.read_csv(dataPath + \"csv_files/test/test_applprev_1_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)\ntest_deposit_1 = pl.read_csv(dataPath + \"csv_files/test/test_deposit_1.csv\").pipe(set_table_dtypes) \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)","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:41:50.27637Z","iopub.execute_input":"2024-05-19T12:41:50.27684Z","iopub.status.idle":"2024-05-19T12:41:50.422918Z","shell.execute_reply.started":"2024-05-19T12:41:50.276808Z","shell.execute_reply":"2024-05-19T12:41:50.420887Z"},"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":"code","source":"# Aggregating features from applprev table\ntrain_applprev_feats = train_applprev_1.group_by(\"case_id\").agg(\n    [\n        pl.col(\"actualdpd_943P\").max().alias(\"max_actualdpd\"),\n        pl.col(\"annuity_853A\").mean().alias(\"mean_annuity\"),\n        pl.col(\"credacc_actualbalance_314A\").max().alias(\"max_actualbalance\"),\n        pl.col(\"credacc_credlmt_575A\").max().alias(\"max_credlmt\"),\n        pl.col(\"currdebt_94A\").sum().alias(\"sum_currdebt\"),\n        pl.col(\"outstandingdebt_522A\").sum().alias(\"sum_outstandingdebt\")\n    ]\n)\n\n# Filter and rename columns for num_group1 = 0 in applprev table\nfiltered_applprev = train_applprev_1.filter(pl.col(\"num_group1\") == 0)\n\n# Merging the aggregated features and filtered columns\napplprev_features = train_applprev_feats.join(filtered_applprev.select([\"case_id\"]), on=\"case_id\", how=\"left\")\n\n# Aggregating features from deposit table\ntrain_deposit_feats = train_deposit_1.group_by(\"case_id\").agg(\n    [\n        pl.col(\"amount_416A\").sum().alias(\"total_amount\"),\n        pl.col(\"contractenddate_991D\").max().alias(\"last_contractenddate\"),\n        pl.col(\"openingdate_313D\").min().alias(\"first_openingdate\")\n    ]\n)\n\n# Filter and rename columns for num_group1 = 0 in deposit table\nfiltered_deposit = train_deposit_1.filter(pl.col(\"num_group1\") == 0)\n\n# Merging the aggregated features and filtered columns\ndeposit_features = train_deposit_feats.join(filtered_deposit.select([\"case_id\"]), on=\"case_id\", how=\"left\")\n\n# Select columns for static and static_cb tables\nselected_static_cols = [col for col in train_static.columns if col[-1] in (\"A\", \"M\")]\nselected_static_cb_cols = [col for col in train_static_cb.columns if col[-1] in (\"A\", \"M\")]\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    applprev_features, how=\"left\", on=\"case_id\"\n).join(\n    deposit_features, how=\"left\", on=\"case_id\"\n)\ndata.columns","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:52:25.835311Z","iopub.execute_input":"2024-05-19T12:52:25.835817Z","iopub.status.idle":"2024-05-19T12:52:31.599832Z","shell.execute_reply.started":"2024-05-19T12:52:25.835776Z","shell.execute_reply":"2024-05-19T12:52:31.598518Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\n# Aggregating features from applprev table\ntest_applprev_feats = test_applprev_1.group_by(\"case_id\").agg(\n    [\n        pl.col(\"actualdpd_943P\").max().alias(\"max_actualdpd\"),\n        pl.col(\"annuity_853A\").mean().alias(\"mean_annuity\"),\n        pl.col(\"credacc_actualbalance_314A\").max().alias(\"max_actualbalance\"),\n        pl.col(\"credacc_credlmt_575A\").max().alias(\"max_credlmt\"),\n        pl.col(\"currdebt_94A\").sum().alias(\"sum_currdebt\"),\n        pl.col(\"outstandingdebt_522A\").sum().alias(\"sum_outstandingdebt\")\n    ]\n)\n\n# Filter and rename columns for num_group1 = 0 in applprev table\nfiltered_applprev = test_applprev_1.filter(pl.col(\"num_group1\") == 0)\n\n# Merging the aggregated features and filtered columns\napplprev_features = test_applprev_feats.join(filtered_applprev.select([\"case_id\"]), on=\"case_id\", how=\"left\")\n\n# Aggregating features from deposit table\ntest_deposit_feats = test_deposit_1.group_by(\"case_id\").agg(\n    [\n        pl.col(\"amount_416A\").sum().alias(\"total_amount\"),\n        pl.col(\"contractenddate_991D\").max().alias(\"last_contractenddate\"),\n        pl.col(\"openingdate_313D\").min().alias(\"first_openingdate\")\n    ]\n)\n\n# Filter and rename columns for num_group1 = 0 in deposit table\nfiltered_deposit = test_deposit_1.filter(pl.col(\"num_group1\") == 0)\n\n# Merging the aggregated features and filtered columns\ndeposit_features = test_deposit_feats.join(filtered_deposit.select([\"case_id\"]), on=\"case_id\", how=\"left\")\n\n# Select columns for static and static_cb tables\nselected_static_cols = [col for col in test_static.columns if col[-1] in (\"A\", \"M\")]\nselected_static_cb_cols = [col for col in test_static_cb.columns if col[-1] in (\"A\", \"M\")]\n\n# Join all tables together\ndata = 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    applprev_features, how=\"left\", on=\"case_id\"\n).join(\n    deposit_features, how=\"left\", on=\"case_id\"\n)\n\n# Print the column names of the final DataFrame\nprint(data.columns)","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:54:41.869742Z","iopub.execute_input":"2024-05-19T12:54:41.870306Z","iopub.status.idle":"2024-05-19T12:54:42.053242Z","shell.execute_reply.started":"2024-05-19T12:54:41.870258Z","shell.execute_reply":"2024-05-19T12:54:42.052086Z"},"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)","metadata":{"execution":{"iopub.status.busy":"2024-05-19T12:55:13.094948Z","iopub.execute_input":"2024-05-19T12:55:13.095791Z","iopub.status.idle":"2024-05-19T12:55:13.194092Z","shell.execute_reply.started":"2024-05-19T12:55:13.095742Z","shell.execute_reply":"2024-05-19T12:55:13.192222Z"},"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":"2024-02-07T21:25:28.52813Z","iopub.execute_input":"2024-02-07T21:25:28.529014Z","iopub.status.idle":"2024-02-07T21:25:28.534479Z","shell.execute_reply.started":"2024-02-07T21:25:28.528913Z","shell.execute_reply":"2024-02-07T21:25:28.53334Z"},"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.status.busy":"2024-02-07T21:25:28.535756Z","iopub.execute_input":"2024-02-07T21:25:28.536641Z","iopub.status.idle":"2024-02-07T21:26:56.543315Z","shell.execute_reply.started":"2024-02-07T21:25:28.5366Z","shell.execute_reply":"2024-02-07T21:26:56.541627Z"},"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\"])}')  ","metadata":{"execution":{"iopub.status.busy":"2024-02-07T21:26:56.547044Z","iopub.execute_input":"2024-02-07T21:26:56.547522Z","iopub.status.idle":"2024-02-07T21:27:22.5052Z","shell.execute_reply.started":"2024-02-07T21:26:56.547479Z","shell.execute_reply":"2024-02-07T21:27:22.503776Z"},"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":"2024-02-07T21:27:22.506645Z","iopub.execute_input":"2024-02-07T21:27:22.507036Z","iopub.status.idle":"2024-02-07T21:27:23.796452Z","shell.execute_reply.started":"2024-02-07T21:27:22.507Z","shell.execute_reply":"2024-02-07T21:27:23.79501Z"},"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)","metadata":{"execution":{"iopub.status.busy":"2024-02-07T21:27:23.798139Z","iopub.execute_input":"2024-02-07T21:27:23.798639Z","iopub.status.idle":"2024-02-07T21:27:23.946242Z","shell.execute_reply.started":"2024-02-07T21:27:23.798595Z","shell.execute_reply":"2024-02-07T21:27:23.944996Z"},"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":"2024-02-07T21:27:23.947771Z","iopub.execute_input":"2024-02-07T21:27:23.948164Z","iopub.status.idle":"2024-02-07T21:27:23.96104Z","shell.execute_reply.started":"2024-02-07T21:27:23.948128Z","shell.execute_reply":"2024-02-07T21:27:23.959969Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Best of luck, and most importantly, enjoy the process of learning and discovery! \n\n<img src=\"https://i.imgur.com/obVWIBh.png\" alt=\"Image\" width=\"700\"/>","metadata":{}}]}