`habit.pipeline`: compose subject / table pipelines

Pipelines: subject-level and table-level component composition.

User guide: Habitat domain API · Python API guide (v2.0). Execution backends: Execution backends.

PoolMarker is the subject↔cohort watershed (pool stage). TablePipeline is the tabular-ML composer; see Tabular ML (bookmark).

Classes

`SubjectPipeline`	The subject-level chain composed into a single callable.
`TablePipeline`	Fitted preprocessing/selection chain plus model over feature tables.
`PoolMarker`	Marker component that declares the subject→cohort fan-in watershed.
`PooledUnits`	The fan-in product: one cohort matrix plus its subject index.
`PoolingRegistry`	Name-to-implementation registry for dataflow pooling markers.

`voxel_units`	Wrap a voxel feature field as single-voxel clustering units.
`fan_in`	Merge per-subject clustering units into one indexed cohort matrix.