cascade.model.CASCADE.design_brute_force

CASCADE.design_brute_force(source, target, pool=None, design_size=1, k=30, counterfactual_kws=None, neighbor_kws=None)[source]

Intervention design with brute-force exhaustion

Parameters:

  • source (AnnData) – Source dataset

  • target (AnnData) – Target dataset representing desired outcome

  • pool (list[str] | None) – Optional list of variables as candidate pool

  • design_size (int) – Maximal combinatorial order to consider

  • k (int) – Number of samples to generate for each design

  • counterfactual_kws (Mapping[str, Any] | None) – Additional keyword arguments passed to counterfactual()

  • neighbor_kws (Mapping[str, Any] | None) – Additional keyword arguments passed to NearestNeighbors

Return type:

tuple[DataFrame, AnnData]

Returns:

  • DataFrame of intervention designs, sorted by descending vote counts

  • AnnData object with counterfactual predictions for all designs