nirs4all.pipeline.dagml.named_stacking module

Named model views of one native nested-stacking execution outcome.

class nirs4all.pipeline.dagml.named_stacking.NamedStackingResult(results: list[RunResult])

Bases: DagMLBatchResult

Preserve named base/meta predictions and select their own fitted artifacts.

The children are producer views of ONE DAG execution, not independently trained pipelines. Their canonical ScoreSet is the exact shared outcome.

export(output_path: str | Path, format: str = 'n4a', source: dict[str, Any] | None = None, chain_id: str | None = None, *, compatibility: str | None = None) → Path

Resolve a producer identity, then export that view’s captured REFIT.

export_model(output_path: str | Path, source: dict[str, Any] | None = None, format: str | None = None, fold: int | None = None, *, compatibility: str | None = None) → Path

Choose the actual producer before delegating model-format validation.

nirs4all.pipeline.dagml.named_stacking.project_named_stacking(outcome: dict[str, Any], *, branches: list[list[Any]], branch_names: list[str], base_model_ids: list[str], meta_node_id: str, meta_learner: Any, spectro: Any, identity: Any, metric: str, task_type: str, config_name: str, pipeline: list[Any], random_state: int | None, meta_per_branch: bool = False) → NamedStackingResult

Project real producer scores/arrays and bind only that producer’s capture.