API Reference¶

QSIParc is organised as a small set of focused modules. Each module has a single responsibility and exposes a clean public interface.

Module overview¶

qsiparc/
├── cli.py        # Click command-line entry point
├── discover.py   # BIDS file discovery and entity parsing
├── atlas.py      # Atlas LUT loading and region metadata
├── extract.py    # Per-region scalar extraction (numpy masking)
├── connectome.py # tck2connectome subprocess wrapper
└── output.py     # BIDS-derivative TSV/JSON writing

Data flow¶

flowchart TD
    cli[cli.py\nmain] --> discover[discover.py\ndiscover_dseg_files\ndiscover_scalar_maps\ndiscover_tractography]
    discover --> atlas[atlas.py\nload_lut_for_dseg]
    atlas --> extract[extract.py\nextract_scalar_map]
    extract --> output[output.py\nwrite_diffmap_tsv]
    discover --> connectome[connectome.py\nbuild_connectomes]
    atlas --> connectome
    connectome --> output2[output.py\nJSON sidecars]

Key types¶

Type	Module	Description
`BIDSFile`	`discover`	A discovered file with parsed BIDS entities
`AtlasDsegFile`	`discover`	Atlas dseg NIfTI paired with its LUT path
`RegionInfo`	`atlas`	Immutable metadata for one atlas region
`AtlasLUT`	`atlas`	Ordered container of `RegionInfo` objects
`ExtractionResult`	`extract`	Scalar name + atlas name + stats DataFrame
`ConnectomeResult`	`connectome`	Connectivity matrix + full provenance
`DiffmapProvenance`	`output`	Provenance metadata for a diffmap TSV

Module pages¶

discover — File discovery and BIDS entity parsing
atlas — Atlas LUT loading
extract — Scalar extraction
connectome — Connectivity matrix construction
output — Output writing