sparrow.pl.cluster_cleanliness

sparrow.pl.cluster_cleanliness#

sparrow.pl.cluster_cleanliness(sdata, table_layer, img_layer=None, shapes_layer='segmentation_mask_boundaries', crd=None, color_dict=None, celltype_column='annotation', output=None)#

Generate plots that allow assessing the “cleanliness” or accuracy of the cell clustering:

  • a barplot with a bar for each cluster, showing the composition by cell type of that cluster;

  • a UMAP with cells colored by cell type;

  • an image of the tissue with cells colored by cell type.

Parameters:

  • sdata (SpatialData) – SpatialData object containing the spatial data and annotations.

  • table_layer (str) – The table layer in sdata to visualize.

  • img_layer (str | None (default: None)) – Name of the imgage layer in sdata (default is None).

  • shapes_layer (str (default: 'segmentation_mask_boundaries')) – Name of the shapes layer in sdata object (default is “segmentation_mask_boundaries”).

  • crd (tuple[int, int, int, int] | None (default: None)) – An optional rectangle [xmin, xmax, ymin, ymax] (default is None). If specified, the tissue image will be cropped to this rectangle, otherwise the full image will be displayed.

  • color_dict (dict | None (default: None)) – Custom colormap dictionary for coloring cell types in the barplot.

  • celltype_column (str (default: 'annotation')) – Name of the column in sdata.tables[table_layer] containing cell type annotations (default is _ANNOTATION_KEY).

  • output (str | None (default: None)) – The file path prefix for the plots (default is None). If provided, the plots will be saved to the specified output file path with “_barplot.png”, “_{celltype_column}.png”, “_{celltype_column}_crop.png” and “_{celltype_column}_umap.png” as suffixes. If None, the plots will be displayed directly without saving.

Return type:

None

Returns:

: None