Workflow module

It is often crucial to perform large scale experiments to analyze the performance of anomaly detectors on large benchmark sets. The Workflow module offer such functionality and can be imported as follows:

>>> from dtaianomaly import workflow

We refer to the documentation for more information regarding the configuration and use of a Workflow.

Workflow related classes

Workflow(dataloaders, metrics, detectors[, ...])	Run anomaly detection experiments.
JobBasedWorkflow(jobs, metrics[, ...])	Run anomaly detection experiments.
Job(dataloader, detector[, preprocessor])	A job to execute within the JobBasedWorkflow.

Loading a workflow from a configuration file

workflow_from_config(path[, max_size])	Construct a Workflow using a configuration at a given path.
interpret_config(config)	Actual parsing/interpretation logic.