SoftmaxWeightedSumFitter.build_mapping

SoftmaxWeightedSumFitter.build_mapping(X, y, coords=None)

Construct a specialized mapping-input graph without sampling.

Like fit_mapping(), this hook requires an explicit subclass implementation. Mapping models should validate their components and delegate graph construction to build(), without running their eager fit_mapping implementation.

Parameters:

• X (dict[str, DataArray]) – Labeled predictor arrays for specialized model components.

• y (dict[str, DataArray]) – Labeled target arrays for specialized model components.

• coords (dict[str, Any] | None) – Coordinate metadata for the model.

Return type:

None