{"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":"gpu","dataSources":[{"sourceId":84493,"databundleVersionId":9871156,"sourceType":"competition"},{"sourceId":10386100,"sourceType":"datasetVersion","datasetId":6434219},{"sourceId":10366351,"sourceType":"datasetVersion","datasetId":6410005}],"dockerImageVersionId":30823,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import pandas as pd\nimport polars as pl\nimport numpy as np\nimport os, gc\nfrom tqdm import tqdm\n\nimport warnings\nwarnings.filterwarnings('ignore')\npd.options.display.max_columns = None\n\nimport lightgbm as lgb\nfrom sklearn.model_selection import *\nfrom sklearn.metrics import *\nimport joblib\nimport kaggle_evaluation.jane_street_inference_server","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:24.119313Z","iopub.execute_input":"2025-01-06T13:34:24.119636Z","iopub.status.idle":"2025-01-06T13:34:28.284734Z","shell.execute_reply.started":"2025-01-06T13:34:24.119608Z","shell.execute_reply":"2025-01-06T13:34:28.284083Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"feature_cols = [f\"feature_{idx:02d}\" for idx in range(79)]+ [f\"responder_{idx}_lag_1\" for idx in range(9)]\ntarget_col = \"responder_6\"\nselected_features = [\"symbol_id\", \"time_id\"] + feature_cols+[target_col]\nfeatures = [\"symbol_id\", \"time_id\"] + feature_cols","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:28.285651Z","iopub.execute_input":"2025-01-06T13:34:28.286157Z","iopub.status.idle":"2025-01-06T13:34:28.290223Z","shell.execute_reply.started":"2025-01-06T13:34:28.286135Z","shell.execute_reply":"2025-01-06T13:34:28.289220Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"valid = pl.scan_parquet(\"/kaggle/input/js-24-dataset-with-lags/validation.parquet\").collect().to_pandas()\n\nX_valid = valid[features]\ny_valid = valid['responder_6']\nw_valid = valid[\"weight\"]","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:28.292082Z","iopub.execute_input":"2025-01-06T13:34:28.292326Z","iopub.status.idle":"2025-01-06T13:34:31.779304Z","shell.execute_reply.started":"2025-01-06T13:34:28.292296Z","shell.execute_reply":"2025-01-06T13:34:31.778491Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"model_lgb = joblib.load(\"/kaggle/input/js-24-traind-models/lightgbm_model2.pkl\")","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:31.780388Z","iopub.execute_input":"2025-01-06T13:34:31.780688Z","iopub.status.idle":"2025-01-06T13:34:31.831804Z","shell.execute_reply.started":"2025-01-06T13:34:31.780659Z","shell.execute_reply":"2025-01-06T13:34:31.827000Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"y_pred_valid = model_lgb.predict(X_valid)\nvalid_score = r2_score(y_valid,y_pred_valid)\nvalid_score","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:31.832622Z","iopub.execute_input":"2025-01-06T13:34:31.833022Z","iopub.status.idle":"2025-01-06T13:34:50.391837Z","shell.execute_reply.started":"2025-01-06T13:34:31.832998Z","shell.execute_reply":"2025-01-06T13:34:50.390947Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"lags_ : pl.DataFrame | None = None\n    \ndef predict(test: pl.DataFrame, lags: pl.DataFrame | None) -> pl.DataFrame | pd.DataFrame:\n    global lags_\n    if lags is not None:\n        lags_ = lags\n\n    predictions = test.select(\n        'row_id',\n        pl.lit(0.0).alias('responder_6'),\n    )\n    symbol_ids = test.select('symbol_id').to_numpy()[:, 0]\n\n    lags = lags_.clone().group_by([\"date_id\", \"symbol_id\"], maintain_order=True).last() # pick up last record of previous date\n    test = test.join(lags, on=[\"date_id\", \"symbol_id\"],  how=\"left\")\n\n    \n    \"\"\" Pred LGB \"\"\"\n    preds = model_lgb.predict(test[features].to_pandas())\n\n    \"\"\" Finaly \"\"\"\n    predictions = \\\n    test.select('row_id').\\\n    with_columns(\n        pl.Series(\n            name   = 'responder_6', \n            values = np.clip(preds, a_min = -5, a_max = 5),\n            dtype  = pl.Float64,\n        )\n    )\n\n    assert isinstance(predictions, pl.DataFrame | pd.DataFrame)\n    assert list(predictions.columns) == ['row_id', 'responder_6']\n    assert len(predictions) == len(test)\n\n    return predictions","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:50.392732Z","iopub.execute_input":"2025-01-06T13:34:50.393066Z","iopub.status.idle":"2025-01-06T13:34:50.399537Z","shell.execute_reply.started":"2025-01-06T13:34:50.393038Z","shell.execute_reply":"2025-01-06T13:34:50.398858Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"inference_server = kaggle_evaluation.jane_street_inference_server.JSInferenceServer(predict)\n\nif os.getenv('KAGGLE_IS_COMPETITION_RERUN'):\n    inference_server.serve()\nelse:\n    inference_server.run_local_gateway(\n        (\n            '/kaggle/input/jane-street-real-time-market-data-forecasting/test.parquet',\n            '/kaggle/input/jane-street-real-time-market-data-forecasting/lags.parquet',\n        )\n    )\n\nprint('submitted')","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-06T13:34:50.400369Z","iopub.execute_input":"2025-01-06T13:34:50.400565Z","iopub.status.idle":"2025-01-06T13:34:50.600020Z","shell.execute_reply.started":"2025-01-06T13:34:50.400548Z","shell.execute_reply":"2025-01-06T13:34:50.599289Z"}},"outputs":[],"execution_count":null}]}