setu.sliced_wasserstein# setu.sliced_wasserstein(samples_a, samples_b, n_projections=100, seed=42)[source]# Sliced Wasserstein distance. Projects samples onto random 1D directions and computes the average 1D Wasserstein distance across projections. Parameters: samples_a (Array) – First set of samples (n, dim) samples_b (Array) – Second set of samples (m, dim) n_projections (int) – Number of random projections seed (int) – Random seed for reproducibility Return type: float Returns: Sliced Wasserstein distance (lower = more similar)
