interscale.evaluation.plot_all_spatial_net_streams

interscale.evaluation.plot_all_spatial_net_streams#

interscale.evaluation.plot_all_spatial_net_streams(adata, fov_id, fov_key='fov', window_key='sliding_window_assignment', cell_type_col='cell_type_coarse', grid_res=50, max_dist=None, k_dist=0.05, density=1.5, ax=None, additional_embeddings=None, return_streams=False, cell_list=None, inter_only=False, **kwargs)#

Visualize spatial interaction flows using streamplots overlaid on a spatial scatter plot.

This function coordinates the extraction of spatial metadata, computation of vector fields based on attention matrices, and the final rendering of directional flows (streams) for different cell types.

Parameters:

  • adata (anndata.AnnData) – The complete AnnData object.

  • fov_id (str) – The specific Field of View to visualize.

  • fov_key (str) – Key in adata.obs identifying FOVs. Defaults to “fov”.

  • window_key (str) – Key identifying local windows for flow computation.

  • cell_type_col (str) – Column in adata.obs with cell type labels.

  • grid_res (int) – Resolution of the interpolation grid. Defaults to 50.

  • max_dist (float) – Maximum influence distance for vectors. If None, calculated via k_dist.

  • k_dist (float) – Fraction of the FOV width to use as max_dist if max_dist is None.

  • density (float) – Density of the streamplot lines. Defaults to 1.5.

  • ax (matplotlib.axes.Axes) – Pre-existing axes for plotting.

  • additional_embeddings (str) – Additional color mapping for the background scatter plot.

  • return_streams (bool) – If True, returns the computed vector fields instead of plotting.

  • cell_list (list) – Subset of cell types to include in the flow analysis.

  • inter_only (bool) – If True, excludes self-interactions from the vector fields.

  • **kwargs – Passed to squidpy.pl.spatial_scatter.

Returns:

ax : matplotlib.axes.Axes or dict The plot axes, or a dictionary of vector fields if return_streams is True.