fit_predict

TabularCloudPredictor.fit_predict(train_data: str | Path | DataFrame, test_data: str | Path | DataFrame, *, predictor_init_args: Dict[str, Any], predictor_fit_args: Dict[str, Any] | None = None, leaderboard: bool = True, framework_version: str = 'latest', job_name: str | None = None, instance_type: str = 'ml.m5.2xlarge', instance_count: int = 1, volume_size: int = 256, custom_image_uri: str | None = None, wait: bool = True, predictions_path: str | None = None, backend_kwargs: Dict | None = None) → Series | None[source]

Fit and predict in a single SageMaker training job.

Fits a TabularPredictor on train_data and runs batch prediction on test_data inside the same training container. This avoids the overhead of a separate batch-transform job (one cold start, one data upload, no predictor-tarball round-trip). The predictor is left fitted afterward, so deploy() / predict() still work.

Parameters:

  • train_data (Union[str, pathlib.Path, pd.DataFrame]) – Training data, as a DataFrame or local/S3 path to a data file.

  • test_data (Union[str, pathlib.Path, pd.DataFrame]) – Data to predict on, as a DataFrame or local/S3 path to a data file. Must contain every feature column present in train_data (the label column is not required).

  • predictor_init_args (dict) – Init args for the predictor.

  • predictor_fit_args (Optional[dict], default = None) – Additional fit args forwarded to TabularPredictor.fit(). Must NOT contain train_data or tuning_data.

  • leaderboard (bool, default = True) – Whether to include the leaderboard in the output artifact.

  • framework_version (str, default = latest) – Training container version of autogluon. If latest, will use the latest available container version. If custom_image_uri is set, this argument will be ignored.

  • job_name (str, default = None) – Name of the launched training job. If None, CloudPredictor will create one with prefix ag-cloudpredictor.

  • instance_type (str, default = 'ml.m5.2xlarge') – Instance type the predictor will be trained on with SageMaker.

  • instance_count (int, default = 1) – Number of instances used to fit the predictor.

  • volume_size (int, default = 256) – Size in GB of the EBS volume to use for storing input data during training.

  • custom_image_uri (Optional[str], default = None) – Custom container image URI. If set, framework_version is ignored.

  • wait (bool, default = True) – Whether the call should wait until the job completes.

  • predictions_path (Optional[str]) – S3 URL where predictions will be written by the training container (e.g. s3://my-bucket/runs/2024-05-01/predictions.csv). Defaults to {cloud_output_path}/{job_name}/predictions.csv.

  • backend_kwargs (Optional[dict], default = None) – Backend-specific arguments. Same keys as fit().

Returns:

Predictions as a Series. Returns None when wait is False; fetch later via get_fit_predict_results().

Return type:

Optional[pd.Series]