Custom Pipeline

Runner: pyasl.pipelines.custom_pipeline.run_custom_pipeline

Purpose. Power-user pipeline that can call any PyASL module or your own class by fully-qualified class name (FQCN). Supports two call styles:

  • kwargs modules: run(ctx, **params) (preclinical-style)

  • dict modules: run(data_descrip, params) (ASLtbx/MRICloud/Oxford-style)

Smart Parameters

  • Path resolution: any value on a “path-ish” key (path, savedir, etc.) or that looks like a path is resolved relative to data_dir.

  • Meta injection: any "from_meta" token is replaced by the value of the same key from the shared Context.

YAML Example

type: custom
steps:
  - module: BrukerLoader
    params: { root: "{root}", expno: 21 }
  - module: pyasl.modules.preclinical_motion_check.MotionCheck  # FQCN allowed
  - name: MRICloudCalculateDiffMap   # alias resolves to MRICloud module
    params: {}
  - module: SaveOutputs