nirs4all.pipeline.dagml.raw_training_lowerer module

Lower raw-array native Methods training into attested DAG-ML contracts.

The portable lane deliberately accepts only the Methods shapes that have an equivalent controller-owned execution: a linear PLS model, or an exact nested-OOF stack of two-or-more PLS base models followed by native Ridge. In particular, Ridge consumes scheduler-delivered OOF values only; the raw provider matrix remains an identity/target attestation and never becomes a meta-model feature matrix. Every wider branch, transform, estimator, or meta-model configuration fails before native execution.

class nirs4all.pipeline.dagml.raw_training_lowerer.RawArrayDagMLTrainingCompiler(selection_metric: str = 'rmse', selection_objective: str = 'minimize', request_id: str = 'training:nirs4all.raw_fit', plan_id: str = 'plan:nirs4all.raw_fit', outcome_id: str = 'outcome:nirs4all.raw_fit', run_id: str = 'run:nirs4all.raw_fit', bundle_id: str = 'bundle:nirs4all.raw_fit', seed: int = 12345, dagml_module: str = 'dag_ml', training_losses: tuple[~collections.abc.Mapping[str, ~typing.Any], ...] = (), local_implementations: ~typing.Any = None, methods_library_path: str | None = None, methods_hpo_operation: ~collections.abc.Mapping[str, ~typing.Any] | None = None, additional_diagnostics: ~collections.abc.Mapping[str, ~typing.Any] = <factory>)

Bases: object

Compile a minimal raw-array fit into native DAG-ML training contracts.

additional_diagnostics: Mapping[str, Any]

bundle_id: str = 'bundle:nirs4all.raw_fit'

compile_fit(estimator: DagMLPipelineEstimator, X: Any, y: Any, *, sample_ids: Any = None, groups: Any = None, metadata: Any = None, identity_frame: DagMLFitIdentityFrame) → DagMLTrainingExecution

Compile one estimator fit into signed native execution inputs.

dagml_module: str = 'dag_ml'

local_implementations: Any = None

methods_hpo_operation: Mapping[str, Any] | None = None

methods_library_path: str | None = None

outcome_id: str = 'outcome:nirs4all.raw_fit'

plan_id: str = 'plan:nirs4all.raw_fit'

request_id: str = 'training:nirs4all.raw_fit'

run_id: str = 'run:nirs4all.raw_fit'

seed: int = 12345

selection_metric: str = 'rmse'

selection_objective: str = 'minimize'

training_losses: tuple[Mapping[str, Any], ...] = ()

nirs4all.pipeline.dagml.raw_training_lowerer.lower_raw_array_training_contracts(pipeline: Any, X: Any, y: Any, *, identity_frame: DagMLFitIdentityFrame, selection_metric: str = 'rmse', selection_objective: str = 'minimize', request_id: str = 'training:nirs4all.raw_fit', plan_id: str = 'plan:nirs4all.raw_fit', outcome_id: str = 'outcome:nirs4all.raw_fit', run_id: str = 'run:nirs4all.raw_fit', bundle_id: str = 'bundle:nirs4all.raw_fit', seed: int = 12345, dagml_module: str = 'dag_ml', training_losses: tuple[Mapping[str, Any], ...] = (), local_implementations: Any = None, methods_library_path: str | None = None, methods_hpo_operation: Mapping[str, Any] | None = None, portable_methods: bool | None = None, target_names: Sequence[str] | None = None, native_role_profile: Mapping[str, Any] | None = None) → DagMLTrainingRequestContracts

Lower a linear raw-array pipeline into executable DAG-ML contracts.

nirs4all.pipeline.dagml.raw_training_lowerer.raw_arrays_to_spectro_dataset(X: Any, y: Any, *, identity_frame: DagMLFitIdentityFrame, name: str = 'nirs4all_raw_fit') → SpectroDataset

Build the minimal SpectroDataset representation for raw estimator arrays.