{"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":"markdown","source":"# Imports","metadata":{}},{"cell_type":"markdown","source":"Добавлены временные факторы, анализ train_base, так же с весом 0.2 добавлен CatBoost","metadata":{}},{"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\nimport matplotlib\n\nfrom catboost import CatBoostClassifier\n\ndataPath = \"/kaggle/input/home-credit-credit-risk-model-stability/\"","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:23:49.518449Z","iopub.execute_input":"2024-03-17T20:23:49.518996Z","iopub.status.idle":"2024-03-17T20:23:52.320967Z","shell.execute_reply.started":"2024-03-17T20:23:49.518960Z","shell.execute_reply":"2024-03-17T20:23:52.319792Z"},"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        elif col[-1] in ['D']:\n#             df = df.with_columns(pl.col(col).str.to_datetime().dt.to_timestamp())\n            df = df.with_columns(pl.col(col).str.to_datetime().dt.timestamp(\"ms\"))\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":"2024-03-17T20:23:52.322957Z","iopub.execute_input":"2024-03-17T20:23:52.323451Z","iopub.status.idle":"2024-03-17T20:23:52.333763Z","shell.execute_reply.started":"2024-03-17T20:23:52.323405Z","shell.execute_reply":"2024-03-17T20:23:52.332729Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_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)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:23:52.335315Z","iopub.execute_input":"2024-03-17T20:23:52.335659Z","iopub.status.idle":"2024-03-17T20:24:20.920424Z","shell.execute_reply.started":"2024-03-17T20:23:52.335632Z","shell.execute_reply":"2024-03-17T20:24:20.918943Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### basetable contains decision dates and target","metadata":{}},{"cell_type":"code","source":"train_basetable.shape","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:20.923228Z","iopub.execute_input":"2024-03-17T20:24:20.923636Z","iopub.status.idle":"2024-03-17T20:24:20.931669Z","shell.execute_reply.started":"2024-03-17T20:24:20.923604Z","shell.execute_reply":"2024-03-17T20:24:20.930342Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_basetable.sample(10)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:20.933030Z","iopub.execute_input":"2024-03-17T20:24:20.933460Z","iopub.status.idle":"2024-03-17T20:24:20.979019Z","shell.execute_reply.started":"2024-03-17T20:24:20.933425Z","shell.execute_reply":"2024-03-17T20:24:20.977788Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.hist(train_basetable['target'])\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:20.980519Z","iopub.execute_input":"2024-03-17T20:24:20.980930Z","iopub.status.idle":"2024-03-17T20:24:21.299836Z","shell.execute_reply.started":"2024-03-17T20:24:20.980895Z","shell.execute_reply":"2024-03-17T20:24:21.298779Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"zeroes_ratio = train_basetable.filter(pl.col('target') == False).count()['target'][0] / train_basetable.shape[0]\nzeroes_ratio","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:21.301032Z","iopub.execute_input":"2024-03-17T20:24:21.303026Z","iopub.status.idle":"2024-03-17T20:24:21.384691Z","shell.execute_reply.started":"2024-03-17T20:24:21.302995Z","shell.execute_reply":"2024-03-17T20:24:21.383565Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Значит, нулевым строчкам при обучении следует отдавать меньший вес","metadata":{}},{"cell_type":"code","source":"train_basetable = train_basetable.with_columns(\n    date_decision_upd=pl.col(\"date_decision\").str.to_datetime()\n)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:21.386006Z","iopub.execute_input":"2024-03-17T20:24:21.387040Z","iopub.status.idle":"2024-03-17T20:24:22.409029Z","shell.execute_reply.started":"2024-03-17T20:24:21.387001Z","shell.execute_reply":"2024-03-17T20:24:22.408037Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, axes = plt.subplots(1, 2, figsize=(12, 4))\naxes[0].hist(train_basetable['case_id'], bins=100)\naxes[1].hist(train_basetable['date_decision_upd'], bins=100)\naxes[1].xaxis.set_major_formatter(matplotlib.dates.DateFormatter('%b %Y'))\nfig.autofmt_xdate()\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:22.410279Z","iopub.execute_input":"2024-03-17T20:24:22.411103Z","iopub.status.idle":"2024-03-17T20:24:23.413694Z","shell.execute_reply.started":"2024-03-17T20:24:22.411070Z","shell.execute_reply":"2024-03-17T20:24:23.412536Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Айдишники очень странно распределены, может быть стоит этому дать какое-то внимание. Так же видно что за Апрель 2020 года данных меньше всего. Примерно в это время начался ковид и был общий упад экономики, о чем тоже стоит помнить.\n\nВсего данных за промежуток в **1 год 9 месяцев**","metadata":{}},{"cell_type":"markdown","source":"### static table contains paid related features","metadata":{}},{"cell_type":"code","source":"print(train_static.shape)\ntrain_static.sample(10)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:23.417765Z","iopub.execute_input":"2024-03-17T20:24:23.418115Z","iopub.status.idle":"2024-03-17T20:24:23.455753Z","shell.execute_reply.started":"2024-03-17T20:24:23.418086Z","shell.execute_reply":"2024-03-17T20:24:23.454709Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Static_cb Contains personal information like DOB, education, marital status","metadata":{}},{"cell_type":"code","source":"print(train_static_cb.shape)\ntrain_static_cb.sample(10)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:23.459088Z","iopub.execute_input":"2024-03-17T20:24:23.459405Z","iopub.status.idle":"2024-03-17T20:24:23.589519Z","shell.execute_reply.started":"2024-03-17T20:24:23.459379Z","shell.execute_reply":"2024-03-17T20:24:23.588777Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for column in train_static_cb.columns:\n    if column[-1] != 'M':\n        continue\n    print(train_static_cb[column].unique())","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:23.590516Z","iopub.execute_input":"2024-03-17T20:24:23.591694Z","iopub.status.idle":"2024-03-17T20:24:23.713033Z","shell.execute_reply.started":"2024-03-17T20:24:23.591664Z","shell.execute_reply":"2024-03-17T20:24:23.711916Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"В каждой из колонок не особо большое количество уникальных категорий","metadata":{}},{"cell_type":"markdown","source":"### Person personal info like gender, zipcode district etc","metadata":{}},{"cell_type":"code","source":"print(train_person_1.shape)\ntrain_person_1.sample(5)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:23.714338Z","iopub.execute_input":"2024-03-17T20:24:23.714702Z","iopub.status.idle":"2024-03-17T20:24:24.137103Z","shell.execute_reply.started":"2024-03-17T20:24:23.714675Z","shell.execute_reply":"2024-03-17T20:24:24.136196Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### credit_bureau_b_2 ? need more information","metadata":{}},{"cell_type":"code","source":"print(train_credit_bureau_b_2.shape)\ntrain_credit_bureau_b_2.sample(5)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:24.140653Z","iopub.execute_input":"2024-03-17T20:24:24.141849Z","iopub.status.idle":"2024-03-17T20:24:24.149956Z","shell.execute_reply.started":"2024-03-17T20:24:24.141805Z","shell.execute_reply":"2024-03-17T20:24:24.149134Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Test data","metadata":{}},{"cell_type":"code","source":"test_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)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:24.151590Z","iopub.execute_input":"2024-03-17T20:24:24.152333Z","iopub.status.idle":"2024-03-17T20:24:24.227133Z","shell.execute_reply.started":"2024-03-17T20:24:24.152304Z","shell.execute_reply":"2024-03-17T20:24:24.226219Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Feature engineering","metadata":{}},{"cell_type":"code","source":"train_person_1.sample(10)\n# train_person_1.group_by('case_id')","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:24.228411Z","iopub.execute_input":"2024-03-17T20:24:24.229297Z","iopub.status.idle":"2024-03-17T20:24:24.657982Z","shell.execute_reply.started":"2024-03-17T20:24:24.229266Z","shell.execute_reply":"2024-03-17T20:24:24.657007Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# 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\", \"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\", \"M\", \"D\"):\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)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:24.659497Z","iopub.execute_input":"2024-03-17T20:24:24.660149Z","iopub.status.idle":"2024-03-17T20:24:27.246860Z","shell.execute_reply.started":"2024-03-17T20:24:24.660120Z","shell.execute_reply":"2024-03-17T20:24:27.245845Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_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)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:27.247848Z","iopub.execute_input":"2024-03-17T20:24:27.248144Z","iopub.status.idle":"2024-03-17T20:24:27.262632Z","shell.execute_reply.started":"2024-03-17T20:24:27.248119Z","shell.execute_reply":"2024-03-17T20:24:27.261551Z"},"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":"2024-03-17T20:24:27.265038Z","iopub.execute_input":"2024-03-17T20:24:27.265384Z","iopub.status.idle":"2024-03-17T20:24:36.638751Z","shell.execute_reply.started":"2024-03-17T20:24:27.265356Z","shell.execute_reply":"2024-03-17T20:24:36.637830Z"},"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":"2024-03-17T20:24:36.639898Z","iopub.execute_input":"2024-03-17T20:24:36.640237Z","iopub.status.idle":"2024-03-17T20:24:36.646077Z","shell.execute_reply.started":"2024-03-17T20:24:36.640211Z","shell.execute_reply":"2024-03-17T20:24:36.645048Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"Number of loan defaulters in train data : \",y_train.sum())\nprint(\"Percentage of loan defaulters in train data : \",(y_train.sum()/len(y_train)*100))\n\nprint(\"Number of loan defaulters in valid data : \",y_valid.sum())\nprint(\"Percentage of loan defaulters in valid data : \",(y_valid.sum()/len(y_valid)*100))\n\nprint(\"Number of loan defaulters in test data : \",y_test.sum())\nprint(\"Percentage of loan defaulters in test data : \",(y_test.sum()/len(y_test)*100))","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:24:36.647609Z","iopub.execute_input":"2024-03-17T20:24:36.648758Z","iopub.status.idle":"2024-03-17T20:24:36.660872Z","shell.execute_reply.started":"2024-03-17T20:24:36.648719Z","shell.execute_reply":"2024-03-17T20:24:36.659831Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Baseline model","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\": 5,\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\": 2000,\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":"2024-03-17T20:24:36.662350Z","iopub.execute_input":"2024-03-17T20:24:36.662764Z","iopub.status.idle":"2024-03-17T20:26:01.055098Z","shell.execute_reply.started":"2024-03-17T20:24:36.662729Z","shell.execute_reply":"2024-03-17T20:26:01.053194Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"увеличение глубины деревьев до 5 дало улучшение","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":"2024-03-17T20:26:01.056302Z","iopub.execute_input":"2024-03-17T20:26:01.056627Z","iopub.status.idle":"2024-03-17T20:26:21.967154Z","shell.execute_reply.started":"2024-03-17T20:26:01.056590Z","shell.execute_reply":"2024-03-17T20:26:21.965919Z"},"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":"2024-03-17T20:26:21.968940Z","iopub.execute_input":"2024-03-17T20:26:21.969236Z","iopub.status.idle":"2024-03-17T20:26:23.015796Z","shell.execute_reply.started":"2024-03-17T20:26:21.969211Z","shell.execute_reply":"2024-03-17T20:26:23.014705Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### Дополнительно обучим catboost","metadata":{}},{"cell_type":"code","source":"cat_features = [col for col in X_train.columns if X_train[col].dtype.name == 'category' or X_train[col].dtype.name == 'object']\n\nfor col in cat_features:\n    X_train[col] = X_train[col].cat.add_categories('Missing').fillna('Missing')\n    X_valid[col] = X_valid[col].cat.add_categories('Missing').fillna('Missing')\n    X_test[col] = X_test[col].cat.add_categories('Missing').fillna('Missing')","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:26:23.017356Z","iopub.execute_input":"2024-03-17T20:26:23.017714Z","iopub.status.idle":"2024-03-17T20:26:23.064975Z","shell.execute_reply.started":"2024-03-17T20:26:23.017686Z","shell.execute_reply":"2024-03-17T20:26:23.063898Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = CatBoostClassifier(\n    iterations=2000,\n    depth=5,\n    learning_rate=0.1,\n    eval_metric='AUC',\n    cat_features=cat_features,\n    bootstrap_type='Bayesian',\n    bagging_temperature=1,           \n    od_type='Iter',                  \n    od_wait=30,\n)\nmodel.fit(\n    X_train, y_train,\n    eval_set=(X_valid, y_valid),\n    use_best_model=True,\n#     verbose=True,\n    verbose_eval=10,\n)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:26:42.412308Z","iopub.execute_input":"2024-03-17T20:26:42.412770Z","iopub.status.idle":"2024-03-17T20:28:00.351010Z","shell.execute_reply.started":"2024-03-17T20:26:42.412731Z","shell.execute_reply":"2024-03-17T20:28:00.349848Z"},"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 = 0.2 * model.predict(X) + 0.8 * 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\nstability_score_train = gini_stability(base_train)\nstability_score_valid = gini_stability(base_valid)\nstability_score_test = gini_stability(base_test)\n\nprint(f'{stability_score_train=}') \nprint(f'{stability_score_valid=}') \nprint(f'{stability_score_test=}')","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:28:05.082016Z","iopub.execute_input":"2024-03-17T20:28:05.082436Z","iopub.status.idle":"2024-03-17T20:28:33.367523Z","shell.execute_reply.started":"2024-03-17T20:28:05.082402Z","shell.execute_reply":"2024-03-17T20:28:33.366276Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"При использовании исключительно CatBoost:\n```\nThe AUC score on the train set is: 0.5008791089129861\nThe AUC score on the valid set is: 0.5006529273113246\nThe AUC score on the test set is: 0.5006063052910206\nstability_score_train=0.00020535534474309123\nstability_score_valid=-0.0009845762197290364\nstability_score_test=-0.0006813595115867277```","metadata":{}},{"cell_type":"markdown","source":"# Submission","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 = 0.2 * model.predict(X_submission) + 0.8 * gbm.predict(X_submission, num_iteration=gbm.best_iteration)","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:28:46.206754Z","iopub.execute_input":"2024-03-17T20:28:46.207206Z","iopub.status.idle":"2024-03-17T20:28:46.313994Z","shell.execute_reply.started":"2024-03-17T20:28:46.207173Z","shell.execute_reply":"2024-03-17T20:28:46.312907Z"},"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":"2024-03-17T20:28:47.953440Z","iopub.execute_input":"2024-03-17T20:28:47.954434Z","iopub.status.idle":"2024-03-17T20:28:47.967587Z","shell.execute_reply.started":"2024-03-17T20:28:47.954397Z","shell.execute_reply":"2024-03-17T20:28:47.966066Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.head()","metadata":{"execution":{"iopub.status.busy":"2024-03-17T20:28:49.391739Z","iopub.execute_input":"2024-03-17T20:28:49.392578Z","iopub.status.idle":"2024-03-17T20:28:49.407754Z","shell.execute_reply.started":"2024-03-17T20:28:49.392537Z","shell.execute_reply":"2024-03-17T20:28:49.406666Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}