EnsembleProcessor

class sdm.processing.ensemble.EnsembleProcessor

Bases: Processor

Base processor for ensemble-aware table transformations.

An EnsembleProcessor defines a reusable transformation on EnsembleTable for feature, target and output pre/post-processing across ensemble members. An EnsembleProcessor learns any required state via fit_ensemble() and applies the transformation via transform_ensemble(). fit_ensemble(), transform_ensemble(), and fit_transform_ensemble() are no-ops for stypes outside of handles_stypes.

As a Processor, it also accepts a TableTensor and processes it as an ensemble with one member.

static as_processor(processor: object) → EnsembleProcessor

Normalize a processor-like object to a EnsembleProcessor.

Parameters:

processor (object) – A processor-like object. An EnsembleProcessor is returned as-is, and any ordinary processor-like object is adapted to ensemble processing.

Return type:

EnsembleProcessor

fit_ensemble(ensemble_table: EnsembleTable, *, generator: Generator | None = None) → Self

Fit the processor on an ensemble table.

Parameters:

• ensemble_table (EnsembleTable) – Ensemble table used to compute the processor state.

• generator (Generator | None) – Pseudorandom number generator used for sampling.

Return type:

Self

transform_ensemble(ensemble_table: EnsembleTable) → EnsembleTable

Transform an ensemble table with fitted state.

Parameters:

ensemble_table (EnsembleTable) – Ensemble table to transform.

Returns:

The transformed ensemble table.

Return type:

EnsembleTable

fit_transform_ensemble(ensemble_table: EnsembleTable, *, generator: Generator | None = None) → EnsembleTable

Fit the processor and transform an ensemble table.

Parameters:

• ensemble_table (EnsembleTable) – Ensemble table to fit on and transform.

• generator (Generator | None) – Pseudorandom number generator used for sampling.

Returns:

The transformed ensemble table.

Return type:

EnsembleTable