{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","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":30664,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"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\nimport matplotlib.pyplot as plt\n\ndataPath = \"/kaggle/input/home-credit-credit-risk-model-stability/\"","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-03-17T19:18:29.920282Z","iopub.execute_input":"2024-03-17T19:18:29.920717Z","iopub.status.idle":"2024-03-17T19:18:29.929347Z","shell.execute_reply.started":"2024-03-17T19:18:29.920687Z","shell.execute_reply":"2024-03-17T19:18:29.927741Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Главная информация\n\n- Выполнил: Швецов Артемий Романович БПМИ208\n\n- Используемый бейзлайн: https://www.kaggle.com/code/jetakow/home-credit-2024-starter-notebook/notebook\n\n### Что было сделано:\n\n##### Фичи\n\n- выбрал дополнительные файлы, которые увеличивают объем данных, а также имеют полезную информацию\n\n- некоторые из признаков, которые несут в себе один и тот же смысл, были объединены в один с использованием различных техник\n\n- добавлены свои признаки\n\n- проведена фильтрация признаков на основе их важности\n\n##### Тюнинг параметров\n\n- с использованием GridSearchCV был произведен перебор некоторых из признаков","metadata":{}},{"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":"2024-03-17T19:18:29.931039Z","iopub.execute_input":"2024-03-17T19:18:29.931805Z","iopub.status.idle":"2024-03-17T19:18:29.947339Z","shell.execute_reply.started":"2024-03-17T19:18:29.931775Z","shell.execute_reply":"2024-03-17T19:18:29.946415Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"добавим данные из таблиц train_deposit_1 и train_other_1. \n\nИспользование train_deposit_1 мне кажется крайне полезным, так как в данной таблице хранится информмация касательно cуммы депозита, а также дата открытия и закрытия счета\n\nИспользование train_other_1 обусловлено тем, что данная таблица хранит информацию о входящих и исходящих транзакциях на счет","metadata":{}},{"cell_type":"code","source":"# ORIGINAL NOTEBOOK\ntrain_basetable = pl.read_csv(dataPath + \"csv_files/train/train_base.csv\")\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)\ntrain_person_1 = pl.read_csv(dataPath + \"csv_files/train/train_person_1.csv\").pipe(set_table_dtypes)\ntrain_credit_bureau_b_2 = pl.read_csv(dataPath + \"csv_files/train/train_credit_bureau_b_2.csv\").pipe(set_table_dtypes) \n# NEW DATA ADDED\ntrain_deposit_1 = pl.read_csv(dataPath + \"csv_files/train/train_deposit_1.csv\").pipe(set_table_dtypes)\ntrain_other_1 = pl.read_csv(dataPath + \"csv_files/train/train_other_1.csv\").pipe(set_table_dtypes)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:18:29.948425Z","iopub.execute_input":"2024-03-17T19:18:29.949169Z","iopub.status.idle":"2024-03-17T19:18:51.542922Z","shell.execute_reply.started":"2024-03-17T19:18:29.949136Z","shell.execute_reply":"2024-03-17T19:18:51.542017Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# ORIGINAL NOTEBOOK\ntest_basetable = pl.read_csv(dataPath + \"csv_files/test/test_base.csv\")\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)\ntest_person_1 = pl.read_csv(dataPath + \"csv_files/test/test_person_1.csv\").pipe(set_table_dtypes) \ntest_credit_bureau_b_2 = pl.read_csv(dataPath + \"csv_files/test/test_credit_bureau_b_2.csv\").pipe(set_table_dtypes)\n# NEW DATA ADDED\ntest_deposit_1 = pl.read_csv(dataPath + \"csv_files/test/test_deposit_1.csv\").pipe(set_table_dtypes)\ntest_other_1 = pl.read_csv(dataPath + \"csv_files/test/test_other_1.csv\").pipe(set_table_dtypes)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:18:51.547976Z","iopub.execute_input":"2024-03-17T19:18:51.548756Z","iopub.status.idle":"2024-03-17T19:18:51.625390Z","shell.execute_reply.started":"2024-03-17T19:18:51.548715Z","shell.execute_reply":"2024-03-17T19:18:51.624305Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# ORIGINAL NOTEBOOK\n# We need to use aggregation functions in tables with depth > 1, so tables that contain num_group1 column or \n# also num_group2 column.\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\n# Here num_group1=0 has special meaning, it is the person who applied for the loan.\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\n# Here we have num_goup1 and num_group2, so we need to aggregate again.\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\n# We will process in this examples only A-type and M-type columns, so we need to select them.\nselected_static_cols = []\nfor col in train_static.columns:\n    if col[-1] in (\"A\", \"M\"):\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\", \"M\"):\n        selected_static_cb_cols.append(col)\nprint(selected_static_cb_cols)\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).join(\n    train_deposit_1, how=\"left\", on=\"case_id\"\n).join(\n    train_other_1, how=\"left\", on=\"case_id\"\n)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:18:51.627200Z","iopub.execute_input":"2024-03-17T19:18:51.627592Z","iopub.status.idle":"2024-03-17T19:18:55.131456Z","shell.execute_reply.started":"2024-03-17T19:18:51.627555Z","shell.execute_reply":"2024-03-17T19:18:55.130493Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# ORIGINAL NOTEBOOK\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\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).join(\n    test_deposit_1, how=\"left\", on=\"case_id\"\n).join(\n    test_other_1, how=\"left\", on=\"case_id\"\n)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:18:55.132491Z","iopub.execute_input":"2024-03-17T19:18:55.132815Z","iopub.status.idle":"2024-03-17T19:18:55.149367Z","shell.execute_reply.started":"2024-03-17T19:18:55.132790Z","shell.execute_reply":"2024-03-17T19:18:55.148424Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data = data.unique(subset='case_id', keep='last')\ndata_submission = data_submission.unique(subset='case_id', keep='last')","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:18:55.153065Z","iopub.execute_input":"2024-03-17T19:18:55.154942Z","iopub.status.idle":"2024-03-17T19:18:56.549919Z","shell.execute_reply.started":"2024-03-17T19:18:55.154908Z","shell.execute_reply":"2024-03-17T19:18:56.549064Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"В ячейке ниже представлен код, который изменяет существующие признаки. Быстро пробежимся по каждому из нововведений:\n\n- избавился от признака maritalst_893M так как он обладает той же информацией, что и maritalst_385M, но второй оказался гораздо полезнее\n\n- pmtaverage_total_avg_A представляет собой усредненное значение признаков pmtaverage_3A, pmtaverage_4527227A и pmtaverage_4955615A\n\n- sumoutstandtotal_avg_A построен аналогично pmtaverage_total_avg_A\n\nДля того, чтобы понять, что признаки несут в себе одну и ту же информацию, я изучил их описание. Если у признаков одно и то же описание, то считаем их одинаковыми. \n\nПомимо этого, добавлены новые признаки:\n\n- pay_zero - у человека нет обязанностей перед банком и ему не нужно платить в следующем месяце\n\n- max_payment_next_month представляет собой разницу между обязательствами и доходами человека. Если он потратит все свои доходы, но не сможет покрыть обязательства, это может сигнализировать о том, что его лучше не выбирать\n\nДополнительно выкинул 5 признаков, которые плохо себя показали во время обучения с точки зрения полезности","metadata":{}},{"cell_type":"code","source":"# ORIGINAL NOTEBOOK\ncase_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\n#print(cols_pred)\n\ndef from_polars_to_pandas(case_ids: pl.DataFrame) -> pl.DataFrame:\n    base = data.filter(pl.col(\"case_id\").is_in(case_ids))[[\"case_id\", \"WEEK_NUM\", \"target\"]].to_pandas()\n    X = data.filter(pl.col(\"case_id\").is_in(case_ids))[cols_pred].to_pandas()\n    y = data.filter(pl.col(\"case_id\").is_in(case_ids))[\"target\"].to_pandas()\n    \n    X['pmtaverage_total_avg_A'] = X[['pmtaverage_3A', 'pmtaverage_4527227A', 'pmtaverage_4955615A']].mean(axis=1)\n    X['sumoutstandtotal_avg_A'] = X[['sumoutstandtotal_3546847A', 'sumoutstandtotalest_4493215A']].mean(axis=1)\n    \n    X[\"pay_zero\"] = (X[\"annuitynextmonth_57A\"] == 0).astype(int)\n    X[\"max_payment_next_month\"] = X[\"currdebt_22A\"] - X[\"maininc_215A\"] - X['lastotherinc_902A']\n    return (\n        base, \n        X.drop(columns=['maritalst_893M', 'pmtaverage_3A', 'pmtaverage_4527227A', 'pmtaverage_4955615A',\n                        'sumoutstandtotal_3546847A', 'sumoutstandtotalest_4493215A', 'lastapprcommoditytypec_5251766M',\n                        'lastotherlnsexpense_631A', 'amtdebitoutgoing_4809440A', 'amtdepositincoming_4809444A', 'lastrejectcommodtypec_5251769M']),\n        y\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":"2024-03-17T19:18:56.551955Z","iopub.execute_input":"2024-03-17T19:18:56.552933Z","iopub.status.idle":"2024-03-17T19:19:04.366597Z","shell.execute_reply.started":"2024-03-17T19:18:56.552879Z","shell.execute_reply":"2024-03-17T19:19:04.365558Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Также для улучшения кода закомментил не самый интересный процесс перебора гиперпараметров. Наиболее оптимальные выставлены далее","metadata":{}},{"cell_type":"code","source":"#from sklearn.model_selection import GridSearchCV\n#\n#param_grid = {\n#    \"boosting_type\": [\"gbdt\"],\n#    \"objective\": [\"binary\"],\n#    \"metric\": [\"auc\"],\n#    \"max_depth\": [3, 5, 7],\n#    \"num_leaves\": [31, 63, 127],\n#    \"learning_rate\": [0.05],\n#    \"feature_fraction\": [0.8],\n#    \"bagging_fraction\": [0.7],\n#    \"bagging_freq\": [5],\n#    \"n_estimators\": [750]\n#}\n#\n#create the LightGBM dataset objects\n#lgb_train = lgb.Dataset(X_train_3, label=y_train)\n#lgb_valid = lgb.Dataset(X_valid_3, label=y_valid, reference=lgb_train)\n#\n#lgb_model = lgb.LGBMClassifier(boosting_type='gbdt', objective='binary', metric='auc')\n#\n#grid_search = GridSearchCV(estimator=lgb_model, param_grid=param_grid, cv=3, scoring='roc_auc')\n#grid_search.fit(X_train_3, y_train, eval_set=[(X_valid_3, y_valid)], callbacks=[lgb.log_evaluation(50), lgb.early_stopping(10)])\n#\n#print(\"Best parameters found: \", grid_search.best_params_)\n#print(\"Best score found: \", grid_search.best_score_)\n#","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:19:04.368633Z","iopub.execute_input":"2024-03-17T19:19:04.369547Z","iopub.status.idle":"2024-03-17T19:19:04.375391Z","shell.execute_reply.started":"2024-03-17T19:19:04.369505Z","shell.execute_reply":"2024-03-17T19:19:04.374342Z"},"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\": 7,\n    \"num_leaves\": 31,\n    \"learning_rate\": 0.05,\n    \"feature_fraction\": 0.8,\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":"2024-03-17T19:19:04.379249Z","iopub.execute_input":"2024-03-17T19:19:04.379888Z","iopub.status.idle":"2024-03-17T19:20:09.087454Z","shell.execute_reply.started":"2024-03-17T19:19:04.379855Z","shell.execute_reply":"2024-03-17T19:20:09.086221Z"},"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":"2024-03-17T19:20:09.088974Z","iopub.execute_input":"2024-03-17T19:20:09.089485Z","iopub.status.idle":"2024-03-17T19:20:29.318651Z","shell.execute_reply.started":"2024-03-17T19:20:09.089437Z","shell.execute_reply":"2024-03-17T19:20:29.317068Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"у бейзлайна были метрики\n\nvalid_0's auc: 0.747468\n\nThe AUC score on the train set is: 0.7568533830862914\n\nThe AUC score on the valid set is: 0.7474683808032441\n\nThe AUC score on the test set is: 0.7435939368164314\n\nкак мы видим, мое решение побило бейзлайн, однако в обсуждениях я обнаружил некоторые вопросы к метрике, поэтому это может повлиять на то, почему я получил такой скор в лидерборде","metadata":{}},{"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":"2024-03-17T19:20:29.320137Z","iopub.execute_input":"2024-03-17T19:20:29.320487Z","iopub.status.idle":"2024-03-17T19:20:30.548660Z","shell.execute_reply.started":"2024-03-17T19:20:29.320458Z","shell.execute_reply":"2024-03-17T19:20:30.547414Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_submission = data_submission[cols_pred].to_pandas()\n\nX_submission.drop(columns=['maritalst_893M'], inplace=True)\nX_submission['pmtaverage_total_avg_A'] = X_submission[['pmtaverage_3A', 'pmtaverage_4527227A', 'pmtaverage_4955615A']].mean(axis=1)\nX_submission.drop(columns=['pmtaverage_3A', 'pmtaverage_4527227A', 'pmtaverage_4955615A'], inplace=True)\nX_submission['sumoutstandtotal_avg_A'] = X_submission[['sumoutstandtotal_3546847A', 'sumoutstandtotalest_4493215A']].mean(axis=1)\nX_submission.drop(columns=['sumoutstandtotal_3546847A', 'sumoutstandtotalest_4493215A'], inplace=True)\nX_submission[\"pay_zero\"] = (X_submission[\"annuitynextmonth_57A\"] == 0).astype(int)\nX_submission[\"max_payment_next_month\"] = X_submission[\"currdebt_22A\"] - X_submission[\"maininc_215A\"] - X_submission['lastotherinc_902A']\n\nX_submission.drop(columns=['lastapprcommoditytypec_5251766M',\n                                    'lastotherlnsexpense_631A',\n                                    'amtdebitoutgoing_4809440A',\n                                    'amtdepositincoming_4809444A',\n                                    'lastrejectcommodtypec_5251769M'], inplace=True)\n\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    \nother_cols = X_train.select_dtypes(include=['float64', 'int64']).columns\nfor col in other_cols:\n    new_dtype = X_train[col].dtype\n    X_submission[col].fillna(X_train[col].mean(), inplace=True)\n    X_submission[col] = X_submission[col].astype(new_dtype)\n\n#print(set(X_submission.columns) - set(X_valid.columns))\ny_submission_pred = gbm.predict(X_submission, num_iteration=gbm.best_iteration)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:20:30.550157Z","iopub.execute_input":"2024-03-17T19:20:30.550474Z","iopub.status.idle":"2024-03-17T19:20:31.295104Z","shell.execute_reply.started":"2024-03-17T19:20:30.550449Z","shell.execute_reply":"2024-03-17T19:20:31.293897Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"y_submission_pred","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:20:31.296458Z","iopub.execute_input":"2024-03-17T19:20:31.296803Z","iopub.status.idle":"2024-03-17T19:20:31.306270Z","shell.execute_reply.started":"2024-03-17T19:20:31.296773Z","shell.execute_reply":"2024-03-17T19:20:31.304811Z"},"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","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:20:31.307551Z","iopub.execute_input":"2024-03-17T19:20:31.307878Z","iopub.status.idle":"2024-03-17T19:20:31.320123Z","shell.execute_reply.started":"2024-03-17T19:20:31.307851Z","shell.execute_reply":"2024-03-17T19:20:31.319192Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.to_csv(\"submission.csv\")","metadata":{"execution":{"iopub.status.busy":"2024-03-17T19:20:31.321159Z","iopub.execute_input":"2024-03-17T19:20:31.322251Z","iopub.status.idle":"2024-03-17T19:20:31.334532Z","shell.execute_reply.started":"2024-03-17T19:20:31.322218Z","shell.execute_reply":"2024-03-17T19:20:31.333222Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}