Pipeline

Automation pipeline for end-to-end workflows.

Pipeline

Pipeline controller.

Pipeline is a DAG of base steps, each step will be executed.

Parameters

name : str, default=“modeva-pipeline”
Name of the pipeline.

save : bool, default=False
Whether to save the pipeline results.

__init__(name: str='modeva-pipeline', save: bool=False)

add_step(name: str, func: Callable, func_inputs: Optional[Dict[str, Any]]=None, parent: Optional[Sequence[str]]=None, save_data: bool=False, save_model: bool=False, save_testsuite: bool=False, title: str=None)

Add a step to pipeline.

Parameters

name : str or Path
The name of the pipeline.

func : Callable
The callable function for this step.

func_inputs : Optional[Dict[str, Any]], default = None
Inputs for the callable function.

parent : Optional[Sequence[str]] = None
Parent step of this step

save_data : bool, default = False
Whether to register the data object to database.

save_model : bool, default = False
Whether to register the model object to database.

save_testsuite : bool, default = False
Whether to register the validation result object to database.

title : str, default = None
Title of this step.

run()

Run the pipeline.