uadapy.dr.wgmm_uapca module

uadapy.dr.wgmm_uapca.create_gmm_from_components(means: ndarray, covariances: ndarray, weights: ndarray) → GaussianMixture

Creates a fitted GaussianMixture object from component parameters.

Parameters:

  • means (np.ndarray) – Array of component means.

  • covariances (np.ndarray) – Array of component covariances.

  • weights (np.ndarray) – Array of component weights.

Returns:

A fitted GaussianMixture object with the given parameters.

Return type:

GaussianMixture

uadapy.dr.wgmm_uapca.transform_wgmm_uapca(distributions: list, means: ndarray, covs: ndarray, weights: ndarray = None, n_dims: int = 2) → list[GaussianMixture]

Projects GMM distributions into a lower-dimensional space.

Parameters:

  • distributions (list) – List of Distribution objects wrapping MultivariateGMM models.

  • means (np.ndarray) – Array of overall mean vectors from each GMM.

  • covs (np.ndarray) – Array of overall covariance matrices from each GMM.

  • weights (np.ndarray, optional) – Array of weights for each distribution. If None, uniform weights are used.

  • n_dims (int) – Target dimension for projection.

Returns:

List of projected GaussianMixture objects.

Return type:

list[GaussianMixture]

uadapy.dr.wgmm_uapca.wgmm_uapca(distributions: list, weights: ndarray = None, n_dims: int = 2) → list

Applies weighted GMM UAPCA to Gaussian Mixture Model distributions and returns the projected distributions in lower-dimensional space.

Parameters:

  • distributions (list) – List of input Distribution objects wrapping MultivariateGMM models.

  • weights (np.ndarray, optional) – Array of weights for each distribution. If None, uniform weights are used.

  • n_dims (int) – Target dimension. Default is 2.

Returns:

List of Distribution objects wrapping projected MultivariateGMM models.

Return type:

list