","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport polars as pl\nimport numpy as np\nimport lightgbm as lgb\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_auc_score\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom sklearn.preprocessing import StandardScaler\nimport os\nimport gc\nfrom glob import glob\nfrom pathlib import Path\nfrom datetime import datetime\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport joblib\n\npd.set_option('display.max_columns', None)\npd.set_option('display.max_rows', None)\n\nROOT = Path(\"/kaggle/input/home-credit-credit-risk-model-stability\")\nTRAIN_DIR = ROOT / \"parquet_files\" / \"train\"\nTEST_DIR = ROOT / \"parquet_files\" / \"test\"","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:18:20.008959Z","iopub.execute_input":"2024-02-11T20:18:20.010149Z","iopub.status.idle":"2024-02-11T20:18:22.952498Z","shell.execute_reply.started":"2024-02-11T20:18:20.010100Z","shell.execute_reply":"2024-02-11T20:18:22.951553Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n#
User Defined Functions
","metadata":{}},{"cell_type":"markdown","source":"## Transformation Functions","metadata":{}},{"cell_type":"code","source":"def set_table_dtypes(df):\n '''\n Set the data types based on the last letter of the column\n '''\n for col in df.columns:\n if col in [\"case_id\", \"WEEK_NUM\", \"num_group1\", \"num_group2\"]:\n df = df.with_columns(pl.col(col).cast(pl.Int64))\n elif col in [\"date_decision\"]:\n df = df.with_columns(pl.col(col).cast(pl.Date))\n elif col[-1] in (\"P\", \"A\"):\n df = df.with_columns(pl.col(col).cast(pl.Float64))\n elif col[-1] in (\"M\",):\n df = df.with_columns(pl.col(col).cast(pl.String))\n elif col[-1] in (\"D\",):\n df = df.with_columns(pl.col(col).cast(pl.Date)) \n return df\n\ndef handle_dates(df):\n '''\n Turn dates into the days since application\n '''\n for col in df.columns:\n if col[-1] in (\"D\",):\n df = df.with_columns(pl.col(col) - pl.col(\"date_decision\"))\n df = df.with_columns(pl.col(col).dt.total_days())\n df = df.drop(\"date_decision\", \"MONTH\")\n return df\n\ndef filter_cols(df):\n '''\n Remove all variable that are >90% missing or many classes\n '''\n for col in df.columns:\n if col not in [\"target\", \"case_id\", \"WEEK_NUM\"]:\n isnull = df[col].is_null().mean()\n if isnull > 0.90:\n df = df.drop(col)\n for col in df.columns:\n if (col not in [\"target\", \"case_id\", \"WEEK_NUM\"]) & (df[col].dtype == pl.String):\n freq = df[col].n_unique()\n if (freq == 1) | (freq > 100):\n df = df.drop(col)\n return df\n\ndef num_expr(df):\n '''\n Calculate the maximim for columns that end in (P=DPD) or (A=Amount)\n '''\n cols = [col for col in df.columns if col[-1] in (\"P\", \"A\")]\n expr_max = [pl.max(col).alias(f\"max_{col}\") for col in cols]\n return expr_max\n \ndef date_expr(df):\n '''\n Calculate the maximim date\n '''\n cols = [col for col in df.columns if col[-1] in (\"D\")]\n expr_max = [pl.max(col).alias(f\"max_{col}\") for col in cols]\n return expr_max\n\ndef str_expr(df):\n '''\n Calculate the maximim of categories\n '''\n cols = [col for col in df.columns if col[-1] in (\"M\")]\n expr_max = [pl.max(col).alias(f\"max_{col}\") for col in cols]\n return expr_max\n\ndef other_expr(df):\n '''\n Calculate the maximim of other varaibles\n '''\n cols = [col for col in df.columns if col[-1] in (\"T\", \"L\")]\n expr_max = [pl.max(col).alias(f\"max_{col}\") for col in cols]\n return expr_max\n\ndef count_expr(df):\n '''\n Count tune number of groups\n '''\n cols = [col for col in df.columns if \"num_group\" in col]\n expr_max = [pl.max(col).alias(f\"max_{col}\") for col in cols]\n return expr_max\n\ndef get_exprs(df):\n '''\n Aggregate all expressions\n '''\n exprs = num_expr(df) + \\\n date_expr(df) + \\\n str_expr(df) + \\\n other_expr(df) + \\\n count_expr(df)\n return exprs","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2024-02-11T20:18:22.954435Z","iopub.execute_input":"2024-02-11T20:18:22.954747Z","iopub.status.idle":"2024-02-11T20:18:22.978511Z","shell.execute_reply.started":"2024-02-11T20:18:22.954721Z","shell.execute_reply":"2024-02-11T20:18:22.977662Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Reading Data Functions","metadata":{}},{"cell_type":"code","source":"def read_file(path, depth=None):\n df = pl.read_parquet(path)\n df = df.pipe(set_table_dtypes)\n if depth in [1, 2]:\n df = df.group_by(\"case_id\").agg(get_exprs(df))\n return df\n\n\ndef read_files(regex_path, depth=None):\n chunks = []\n for path in glob(str(regex_path)):\n chunks.append(pl.read_parquet(path).pipe(set_table_dtypes))\n df = pl.concat(chunks, how=\"vertical_relaxed\")\n if depth in [1, 2]:\n df = df.group_by(\"case_id\").agg(get_exprs(df))\n return df\n\n\ndef feature_eng(df_base, depth_0, depth_1, depth_2):\n df_base = (df_base.with_columns(month_decision = pl.col(\"date_decision\").dt.month(),\n weekday_decision = pl.col(\"date_decision\").dt.weekday()))\n for i, df in enumerate(depth_0 + depth_1 + depth_2):\n df_base = df_base.join(df, how=\"left\", on=\"case_id\", suffix=f\"_{i}\")\n df_base = df_base.pipe(handle_dates)\n return df_base\n\n\ndef to_pandas(df_data, cat_cols=None):\n df_data = df_data.to_pandas()\n if cat_cols is None:\n cat_cols = list(df_data.select_dtypes(\"object\").columns)\n df_data[cat_cols] = df_data[cat_cols].astype(\"category\")\n return df_data, cat_cols","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:18:22.979892Z","iopub.execute_input":"2024-02-11T20:18:22.980783Z","iopub.status.idle":"2024-02-11T20:18:22.995209Z","shell.execute_reply.started":"2024-02-11T20:18:22.980742Z","shell.execute_reply":"2024-02-11T20:18:22.994245Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Other Functions","metadata":{}},{"cell_type":"code","source":"def 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\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\", \"date_decision\", \"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\ndef summary(df):\n summ = pd.DataFrame(df.dtypes, columns=['data type'])\n summ['#total'] = df.shape[0]\n summ['#missing'] = df.isnull().sum().values \n summ['%missing'] = df.isnull().sum().values / len(df)* 100\n summ['#unique'] = df.nunique().values\n summ['#duplicates'] = summ['#total'] - summ['#unique']\n desc = pd.DataFrame(df.describe(include='all').transpose())\n summ['min'] = desc['min'].values\n summ['max'] = desc['max'].values\n return summ\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\ndef drop_outliers(df, field_name):\n iqr = 1.5 * (np.percentile(df[field_name], 75) - np.percentile(df[field_name], 25))\n df.drop(df[df[field_name] > (iqr + np.percentile(df[field_name], 75))].index, inplace=True)\n df.drop(df[df[field_name] < (np.percentile(df[field_name], 25) - iqr)].index, inplace=True)","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:18:22.997262Z","iopub.execute_input":"2024-02-11T20:18:22.997684Z","iopub.status.idle":"2024-02-11T20:18:23.016961Z","shell.execute_reply.started":"2024-02-11T20:18:22.997655Z","shell.execute_reply":"2024-02-11T20:18:23.015906Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n#
Reading Data
","metadata":{}},{"cell_type":"markdown","source":"## Training Data","metadata":{}},{"cell_type":"code","source":"train_data_store = {\n \"df_base\": read_file(TRAIN_DIR / \"train_base.parquet\"),\n \"depth_0\": [\n read_file(TRAIN_DIR / \"train_static_cb_0.parquet\"),\n read_files(TRAIN_DIR / \"train_static_0_*.parquet\"),\n ],\n \"depth_1\": [\n read_files(TRAIN_DIR / \"train_applprev_1_*.parquet\", 1),\n read_file(TRAIN_DIR / \"train_tax_registry_a_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_tax_registry_b_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_tax_registry_c_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_credit_bureau_b_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_other_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_person_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_deposit_1.parquet\", 1),\n read_file(TRAIN_DIR / \"train_debitcard_1.parquet\", 1),\n ],\n \"depth_2\": [\n read_file(TRAIN_DIR / \"train_credit_bureau_b_2.parquet\", 2),\n ]\n}\n\ndf_train = feature_eng(**train_data_store)\nprint(\"Train data shape: {}\".format(df_train.shape))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:18:23.677236Z","iopub.execute_input":"2024-02-11T20:18:23.677649Z","iopub.status.idle":"2024-02-11T20:19:07.417916Z","shell.execute_reply.started":"2024-02-11T20:18:23.677619Z","shell.execute_reply":"2024-02-11T20:19:07.416961Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Testing Data","metadata":{}},{"cell_type":"code","source":"test_data_store = {\n \"df_base\": read_file(TEST_DIR / \"test_base.parquet\"),\n \"depth_0\": [\n read_file(TEST_DIR / \"test_static_cb_0.parquet\"),\n read_files(TEST_DIR / \"test_static_0_*.parquet\"),\n ],\n \"depth_1\": [\n read_files(TEST_DIR / \"test_applprev_1_*.parquet\", 1),\n read_file(TEST_DIR / \"test_tax_registry_a_1.parquet\", 1),\n read_file(TEST_DIR / \"test_tax_registry_b_1.parquet\", 1),\n read_file(TEST_DIR / \"test_tax_registry_c_1.parquet\", 1),\n read_file(TEST_DIR / \"test_credit_bureau_b_1.parquet\", 1),\n read_file(TEST_DIR / \"test_other_1.parquet\", 1),\n read_file(TEST_DIR / \"test_person_1.parquet\", 1),\n read_file(TEST_DIR / \"test_deposit_1.parquet\", 1),\n read_file(TEST_DIR / \"test_debitcard_1.parquet\", 1),\n ],\n \"depth_2\": [\n read_file(TEST_DIR / \"test_credit_bureau_b_2.parquet\", 2),\n ]\n}\n\ndf_test = feature_eng(**test_data_store)\nprint(\"Test data shape: {}\".format(df_test.shape))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:19:07.419530Z","iopub.execute_input":"2024-02-11T20:19:07.420494Z","iopub.status.idle":"2024-02-11T20:19:07.641926Z","shell.execute_reply.started":"2024-02-11T20:19:07.420461Z","shell.execute_reply":"2024-02-11T20:19:07.641044Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Filtering","metadata":{}},{"cell_type":"code","source":"df_train = df_train.pipe(filter_cols)\ndf_test = df_test.select([col for col in df_train.columns if col != \"target\"])\n\nprint(\"Train data shape: {}\".format(df_train.shape))\nprint(\"Test data shape: {}\".format(df_test.shape))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:19:07.642965Z","iopub.execute_input":"2024-02-11T20:19:07.643452Z","iopub.status.idle":"2024-02-11T20:19:10.336064Z","shell.execute_reply.started":"2024-02-11T20:19:07.643423Z","shell.execute_reply":"2024-02-11T20:19:10.334905Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Garbage Collection","metadata":{}},{"cell_type":"code","source":"df_train, cat_cols = to_pandas(df_train)\ndf_test, cat_cols = to_pandas(df_test, cat_cols)\n\ndel train_data_store\ndel test_data_store\ngc.collect()\n\nprint(\"Train Week Range:\\t\", (df_train[\"WEEK_NUM\"].min(), df_train[\"WEEK_NUM\"].max()))\nprint(\"Test Week Range:\\t\", (df_test[\"WEEK_NUM\"].min(), df_test[\"WEEK_NUM\"].max()))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:19:10.338804Z","iopub.execute_input":"2024-02-11T20:19:10.339897Z","iopub.status.idle":"2024-02-11T20:19:27.433944Z","shell.execute_reply.started":"2024-02-11T20:19:10.339853Z","shell.execute_reply":"2024-02-11T20:19:27.432822Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n#
Data Exploration - 10% Random Sample
","metadata":{}},{"cell_type":"code","source":"# df_plotting = df_train.sample(n=100000, random_state=888)\n# print(df_plotting.shape)","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:21:08.803007Z","iopub.execute_input":"2024-02-11T20:21:08.803461Z","iopub.status.idle":"2024-02-11T20:21:09.187784Z","shell.execute_reply.started":"2024-02-11T20:21:08.803422Z","shell.execute_reply":"2024-02-11T20:21:09.186728Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## View the training data","metadata":{}},{"cell_type":"code","source":"# df_plotting.head()","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:21:09.834553Z","iopub.execute_input":"2024-02-11T20:21:09.834979Z","iopub.status.idle":"2024-02-11T20:21:10.114859Z","shell.execute_reply.started":"2024-02-11T20:21:09.834945Z","shell.execute_reply":"2024-02-11T20:21:10.113745Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Plot the target and heatmap of missing data","metadata":{}},{"cell_type":"code","source":"# fig, ax = plt.subplots(figsize=(24,8), ncols=2, nrows=1)\n# sns.lineplot(data=df_plotting, x=\"WEEK_NUM\", y=\"target\", ax=ax[0])\n# ax[0].set(xlabel='Week Number', ylabel='Target', title='Target Time Series')\n\n# sns.heatmap(df_plotting.isna().transpose(), cbar=False, xticklabels=False, yticklabels=False, ax=ax[1])\n# ax[1].set(ylabel='Variables', xlabel='Observations', title='Missing Data Heatmap')\n\n# plt.tight_layout()\n# plt.show()","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:24:50.297261Z","iopub.execute_input":"2024-02-11T20:24:50.298168Z","iopub.status.idle":"2024-02-11T20:25:25.430390Z","shell.execute_reply.started":"2024-02-11T20:24:50.298128Z","shell.execute_reply":"2024-02-11T20:25:25.429498Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Amount Variables","metadata":{}},{"cell_type":"code","source":"cols = [col for col in df_train.columns if col[-1] in (\"A\")]\ndf_amount = df_train[cols].copy()\ndf_amount['target'] = df_train['target']\ndf_amount.head()","metadata":{"execution":{"iopub.status.busy":"2024-02-11T21:59:32.134480Z","iopub.execute_input":"2024-02-11T21:59:32.134971Z","iopub.status.idle":"2024-02-11T21:59:32.669135Z","shell.execute_reply.started":"2024-02-11T21:59:32.134927Z","shell.execute_reply":"2024-02-11T21:59:32.668037Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Annuaity\nVariable: `annuity_780A`\nDescription: Monthly annuity amount","metadata":{}},{"cell_type":"code","source":"df_annuity_low = df_amount[df_amount['annuity_780A']<=20000][['target','annuity_780A']].copy()\ndf_annuity_high = df_amount[df_amount['annuity_780A']>20000][['target','annuity_780A']].copy()\n\nfig, ax = plt.subplots(figsize=(24, 6), ncols=2, nrows=1)\n\n# Low annuity\nsns.histplot(data=df_annuity_low[df_annuity_low['target']==0], x='annuity_780A', bins=np.arange(0,20000,500), log_scale=False, stat=\"density\", ax=ax[0], label='Non-Default')\nsns.histplot(data=df_annuity_low[df_annuity_low['target']==1], x='annuity_780A', bins=np.arange(0,20000,500), log_scale=False, stat=\"density\", ax=ax[0], label='Default')\nax[0].set(xlabel='Annuity', ylabel='Density', title='Low Annuality Amount')\nax[0].legend()\n\n# High annuity\nsns.histplot(data=df_annuity_high[df_annuity_high['target']==0], x='annuity_780A', bins=np.arange(20000,60000,1000), log_scale=False, stat=\"density\", ax=ax[1], label='Non-Default')\nsns.histplot(data=df_annuity_high[df_annuity_high['target']==1], x='annuity_780A', bins=np.arange(20000,60000,1000), log_scale=False, stat=\"density\", ax=ax[1], label='Default')\nax[1].set(xlabel='Annuity', ylabel='Density', title='High Annuality Amount')\nax[1].legend()\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-02-11T22:19:46.624283Z","iopub.execute_input":"2024-02-11T22:19:46.624769Z","iopub.status.idle":"2024-02-11T22:19:48.842077Z","shell.execute_reply.started":"2024-02-11T22:19:46.624729Z","shell.execute_reply":"2024-02-11T22:19:48.841198Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Credit Amount\nVariable: `credamount_770A`\nDescription: Loan amount or credit card limit","metadata":{}},{"cell_type":"code","source":"df_low = df_amount[['target','credamount_770A']].copy()\n\nfig, ax = plt.subplots(figsize=(24, 6))\n\n# Low annuity\nsns.histplot(data=df_low[df_low['target']==0], x='credamount_770A', bins=np.arange(0,250000,1000), log_scale=False, stat=\"density\", ax=ax, label='Non-Default')\nsns.histplot(data=df_low[df_low['target']==1], x='credamount_770A', bins=np.arange(0,250000,1000), log_scale=False, stat=\"density\", ax=ax, label='Default')\nax.set(xlabel='Loan Amount or Credit Card Limit', ylabel='Density', title='Credit Amount')\nax.legend()\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-02-11T22:24:24.060808Z","iopub.execute_input":"2024-02-11T22:24:24.061288Z","iopub.status.idle":"2024-02-11T22:24:26.488533Z","shell.execute_reply.started":"2024-02-11T22:24:24.061252Z","shell.execute_reply":"2024-02-11T22:24:26.487402Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## To be continued...","metadata":{}},{"cell_type":"markdown","source":"\n#
Modelling
","metadata":{}},{"cell_type":"markdown","source":"![](\"https://iili.io/J14JIjI.png\")
","metadata":{}},{"cell_type":"markdown","source":"## Modelling Varaibles","metadata":{}},{"cell_type":"code","source":"# # Training Data\n# X = df_train.drop(columns=[\"target\", \"case_id\", \"WEEK_NUM\"])\n# y = df_train[\"target\"]\n# weeks = df_train[\"WEEK_NUM\"]\n\n# print(\"Shape of X:\\t\", (X.shape))\n# print(\"Shape of y:\\t\", (y.shape))\n\n# # Testing Data\n# test = df_test.drop(columns=[\"WEEK_NUM\", \"case_id\"])\n\n# print(\"Shape of test:\\t\", (test.shape))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:31:36.453946Z","iopub.execute_input":"2024-02-11T20:31:36.454616Z","iopub.status.idle":"2024-02-11T20:31:37.839591Z","shell.execute_reply.started":"2024-02-11T20:31:36.454575Z","shell.execute_reply":"2024-02-11T20:31:37.838362Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Hyperparameter Tuning","metadata":{}},{"cell_type":"code","source":"# # Define a parameter grid for hyperparameter tuning\n# param_grid = {\n# 'max_depth': [6, 8, 10]\n# }\n\n# # K-fold cross validation\n# cv = StratifiedGroupKFold(n_splits=10, shuffle=False)\n\n# # Create instance of model\n# lgb_model = lgb.LGBMClassifier(**params)\n\n# # Initialize GridSearchCV\n# grid_search = GridSearchCV(estimator=lgb_model, param_grid=param_grid, cv=cv, scoring='roc_auc', verbose=-1)\n\n# # Fit the GridSearchCV to the data\n# grid_search.fit(X, y)\n\n# # Get the best parameters\n# best_params = grid_search.best_params_\n# print(\"Best Parameters:\", best_params)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Fitting","metadata":{}},{"cell_type":"code","source":"# # Define the final hyperparameters\n# params = {\n# \"boosting_type\": \"gbdt\",\n# \"objective\": \"binary\",\n# \"metric\": \"auc\",\n# \"max_depth\": 8,\n# \"learning_rate\": 0.01,\n# \"n_estimators\": 1000,\n# \"colsample_bytree\": 0.8, \n# \"colsample_bynode\": 0.8,\n# \"verbose\": -1,\n# \"random_state\": 888\n# }\n\n# print(params)\n\n# # Create instance of model\n# lgb_model = lgb.LGBMClassifier(**params)\n\n# # Fit the model on the training data\n# lgb_model.fit(X, y)\n\n# # Predict model\n# y_pred = lgb_model.predict_proba(test)[:, 1]\n\n# # Submit results\n# df_submmission = pd.read_csv(ROOT / \"sample_submission.csv\")\n# df_submmission[\"score\"] = y_pred\n# df_submmission.to_csv(\"submission.csv\", index=False)\n\n# print(\"Complete!\")","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:34:02.803263Z","iopub.execute_input":"2024-02-11T20:34:02.804551Z","iopub.status.idle":"2024-02-11T20:45:00.563160Z","shell.execute_reply.started":"2024-02-11T20:34:02.804494Z","shell.execute_reply":"2024-02-11T20:45:00.561816Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"Score = 0.53\")","metadata":{"execution":{"iopub.status.busy":"2024-02-11T22:30:05.551357Z","iopub.execute_input":"2024-02-11T22:30:05.551833Z","iopub.status.idle":"2024-02-11T22:30:05.563286Z","shell.execute_reply.started":"2024-02-11T22:30:05.551798Z","shell.execute_reply":"2024-02-11T22:30:05.561613Z"},"trusted":true},"execution_count":null,"outputs":[]}]}
","metadata":{}},{"cell_type":"markdown","source":"## Modelling Varaibles","metadata":{}},{"cell_type":"code","source":"# # Training Data\n# X = df_train.drop(columns=[\"target\", \"case_id\", \"WEEK_NUM\"])\n# y = df_train[\"target\"]\n# weeks = df_train[\"WEEK_NUM\"]\n\n# print(\"Shape of X:\\t\", (X.shape))\n# print(\"Shape of y:\\t\", (y.shape))\n\n# # Testing Data\n# test = df_test.drop(columns=[\"WEEK_NUM\", \"case_id\"])\n\n# print(\"Shape of test:\\t\", (test.shape))","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:31:36.453946Z","iopub.execute_input":"2024-02-11T20:31:36.454616Z","iopub.status.idle":"2024-02-11T20:31:37.839591Z","shell.execute_reply.started":"2024-02-11T20:31:36.454575Z","shell.execute_reply":"2024-02-11T20:31:37.838362Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Hyperparameter Tuning","metadata":{}},{"cell_type":"code","source":"# # Define a parameter grid for hyperparameter tuning\n# param_grid = {\n# 'max_depth': [6, 8, 10]\n# }\n\n# # K-fold cross validation\n# cv = StratifiedGroupKFold(n_splits=10, shuffle=False)\n\n# # Create instance of model\n# lgb_model = lgb.LGBMClassifier(**params)\n\n# # Initialize GridSearchCV\n# grid_search = GridSearchCV(estimator=lgb_model, param_grid=param_grid, cv=cv, scoring='roc_auc', verbose=-1)\n\n# # Fit the GridSearchCV to the data\n# grid_search.fit(X, y)\n\n# # Get the best parameters\n# best_params = grid_search.best_params_\n# print(\"Best Parameters:\", best_params)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Fitting","metadata":{}},{"cell_type":"code","source":"# # Define the final hyperparameters\n# params = {\n# \"boosting_type\": \"gbdt\",\n# \"objective\": \"binary\",\n# \"metric\": \"auc\",\n# \"max_depth\": 8,\n# \"learning_rate\": 0.01,\n# \"n_estimators\": 1000,\n# \"colsample_bytree\": 0.8, \n# \"colsample_bynode\": 0.8,\n# \"verbose\": -1,\n# \"random_state\": 888\n# }\n\n# print(params)\n\n# # Create instance of model\n# lgb_model = lgb.LGBMClassifier(**params)\n\n# # Fit the model on the training data\n# lgb_model.fit(X, y)\n\n# # Predict model\n# y_pred = lgb_model.predict_proba(test)[:, 1]\n\n# # Submit results\n# df_submmission = pd.read_csv(ROOT / \"sample_submission.csv\")\n# df_submmission[\"score\"] = y_pred\n# df_submmission.to_csv(\"submission.csv\", index=False)\n\n# print(\"Complete!\")","metadata":{"execution":{"iopub.status.busy":"2024-02-11T20:34:02.803263Z","iopub.execute_input":"2024-02-11T20:34:02.804551Z","iopub.status.idle":"2024-02-11T20:45:00.563160Z","shell.execute_reply.started":"2024-02-11T20:34:02.804494Z","shell.execute_reply":"2024-02-11T20:45:00.561816Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"Score = 0.53\")","metadata":{"execution":{"iopub.status.busy":"2024-02-11T22:30:05.551357Z","iopub.execute_input":"2024-02-11T22:30:05.551833Z","iopub.status.idle":"2024-02-11T22:30:05.563286Z","shell.execute_reply.started":"2024-02-11T22:30:05.551798Z","shell.execute_reply":"2024-02-11T22:30:05.561613Z"},"trusted":true},"execution_count":null,"outputs":[]}]}