paste3.helper.glmpca_distance

paste3.helper.glmpca_distance(a_exp_dissim, b_exp_dissim, latent_dim=50, filter=True, maxIter=1000, eps=0.0001, optimizeTheta=True)[source]

Computes the distance between two distributions after reducing dimensionality using GLM-PCA.

Parameters:

  • a_exp_dissim (torch.Tensor) -- A tensor representing first probability distribution.

  • b_exp_dissim (torch.Tensor) -- A tensor representing the second probability distribution.

  • latent_dim (int, Optional) -- Number of latent dimensions for GLM-PCA reduction.

  • filter (bool, Optional) -- Whether to filter features based on top gene counts before GLM-PCA.

  • maxIter (int, Optional) -- Maximum number of iterations for GLM-PCA.

  • eps (float, Optional) -- Convergence threshold for GLM-PCA.

  • optimizeTheta (bool, Optional) -- If True, optimizes theta during GLM-PCA.

Returns:

Distances between the two distributions after dimensionality reduction.

Return type:

np.ndarray