ehrapy.tools.diffmap

ehrapy.tools.diffmap(adata, n_comps=15, neighbors_key=None, random_state=0, copy=False)[source]

Diffusion Maps [Coifman05] [Haghverdi15] [Wolf18].

Diffusion maps [Coifman05] has been proposed for visualizing single-cell data by [Haghverdi15]. The tool uses the adapted Gaussian kernel suggested by [Haghverdi16] in the implementation of [Wolf18]. The width (“sigma”) of the connectivity kernel is implicitly determined by the number of neighbors used to compute the single-cell graph in neighbors(). To reproduce the original implementation using a Gaussian kernel, use method==’gauss’ in neighbors(). To use an exponential kernel, use the default method==’umap’. Differences between these options shouldn’t usually be dramatic.

Parameters:

  • adata (AnnData) – AnnData object containing all observations.

  • n_comps (int, default: 15) – The number of dimensions of the representation. neighbors_key: If not specified, diffmap looks .uns[‘neighbors’] for neighbors settings and .obsp[‘connectivities’], .obsp[‘distances’] for connectivities and distances respectively (default storage places for pp.neighbors). If specified, diffmap looks .uns[neighbors_key] for neighbors settings and .obsp[.uns[neighbors_key][‘connectivities_key’]], .obsp[.uns[neighbors_key][‘distances_key’]] for connectivities and distances respectively.

  • random_state (Union[int, RandomState, None], default: 0) – Random seed for the initialization.

  • copy (bool, default: False) – Whether to return a copy of the AnnData object.

Return type:

Optional[AnnData]

Returns:

Depending on copy, returns or updates adata with the following fields.

X_diffmap : numpy.ndarray (adata.obsm) Diffusion map representation of data, which is the right eigen basis of the transition matrix with eigenvectors as columns.

diffmap_evals : numpy.ndarray (adata.uns) Array of size (number of eigen vectors). Eigenvalues of transition matrix.