BAITS.st.gr.aggregate_neighbors

BAITS.st.gr.aggregate_neighbors#

BAITS.st.gr.aggregate_neighbors(adata, n_layers, aggregations='mean', connectivity_key='spatial_connectivities', use_rep='X_pca', sample_key='sample', out_key='X_cellcharter', copy=False)#

Aggregate the features from each neighborhood layers and concatenate them

Parameters:

  • adata (AnnData) – The annotated data matrix (single-cell dataset in AnnData format).

  • n_layers (int or list) – The number of neighborhood layers to aggregate. If an integer is given, it is converted into a list ranging from 0 to n_layers.

  • aggregations (str or list, default="mean") – The type of aggregation(s) to perform (e.g., “mean”, “sum”).

  • connectivity_key (str, optional) – Key in adata.obsp that contains the spatial connectivity matrix.

  • use_rep (str, optional) – Key in adata.obsm that specifies which representation (feature matrix) to use. If None, the function uses adata.X.

  • sample_key (str, optional) – Key in adata.obs that identifies sample/grouping information. If None, all cells are considered together.

  • out_key (str, default="X_cellcharter") – Key under which the aggregated features are stored in adata.obsm.

  • copy (bool, default=False) – If True, returns the aggregated feature matrix instead of modifying adata.

Returns:

    • If copy is True, returns the aggregated feature matrix as a numpy array or sparse matrix.

    • Otherwise, modifies adata.obsm in place, storing the aggregated matrix under out_key.

Contents