---
hsic() to compute the empirical Hilbert-Schmidt Independence Criterion with flexible kernel selection (Gaussian, Laplace, linear, polynomial).hsic.test() with four null-distribution methods: gamma approximation, permutation, eigenvalue (spectral), and bootstrap. Enables kernel-based independence testing integrated with DDA workflows.median\_bandwidth() utility for automated kernel bandwidth selection using the median heuristic.---
dda.indep, the bootstrap HSIC method argument is now hsic.method = "bootstrap" (previously hsic.method = "boot" in 0.1.0).dda.vardist and dda.resdist now consistently return error messages when the number of bootstrap replications (B) is too small for boot.type = "bca" (previously, behavior could be inconsistent).---
dda.dda functions and S3 generics where applicable (print, summary, plot).---