tacco.plots.co_occurrence_matrix

co_occurrence_matrix(adata, analysis_key, score_key='log_occ', log_base=None, colors=None, show_only=None, show_only_center=None, axsize=None, hspace=None, wspace=None, x_padding=2.0, y_padding=2.0, value_cluster=False, group_cluster=False, restrict_intervals=None, p_corr='fdr_bh', cmap='bwr', cmap_vmin_vmax=None, legend=True, ax=None, scale_legend=1.0)[source]

Plot co-occurrence as determined by co_occurrence() or co_occurrence_matrix() as a matrix.

Parameters:

  • adata – An AnnData with the co-occurence analysis in .uns. Can also be a mapping of labels to AnnData to specify multiple datasets.

  • analysis_key – The .uns key with the co-occurence analysis result.

  • score_key –

    The .uns[analysis_key] key of the score to use. Available keys .uns[analysis_key] include:

    • ’occ’: co-occurrence

    • ’log_occ’: logarithm of the co-occurrence

    • ’log2_occ’: base-2-logarithm of the co-occurrence; this is a not a real key but a convenience function to rescale the ‘log_occ’ values

    • ’z’: z-score of the log of the neighbourship counts with respect to random neighbourships

    • ’composition’: distance dependent composition, p(anno|center, dist)

    • ’log_composition’: log of ‘composition’

    • ’distance_distribution’: distribution of distances between anno and center ´p(dist|anno,center)´

    • ’log_distance_distribution’: log of ‘distance_distribution’

    • ’relative_distance_distribution’: ‘distance_distribution’ normalized to p(dist|*,center), the distance distribution of any annotation to the center

    • ’log_relative_distance_distribution’: log of ‘relative_distance_distribution’

  • log_base – The base of the logarithm to use for plotting if score_key is a log quantity. If None, use the natural logarithm.

  • colors – The mapping of value names to colors. If None, a set of standard colors is used.

  • show_only – A subset of annotation values to restrict the plotting to.

  • show_only_center – A subset of the center annotation values to restrict the plotting to.

  • axsize – Tuple of width and size of a single axis. If None, some heuristic value is used.

  • hspace – Relative horizontal and vertical spacing between plots

  • vspace – Relative horizontal and vertical spacing between plots

  • x_padding – Absolute horizontal and vertical spacing between plots; this setting overrides hspace and vspace; if None, use the value from hspace vspace; if None, use the value from vspace

  • y_padding – Absolute horizontal and vertical spacing between plots; this setting overrides hspace and vspace; if None, use the value from hspace vspace; if None, use the value from vspace

  • value_cluster – Whether to cluster and reorder the values.

  • group_cluster – Whether to cluster and reorder the groups.

  • restrict_intervals – A list-like containing the indices of the intervals to plot. If None, all intervals are included.

  • cmap – A string/colormap to override the colors with globally.

  • cmap_vmin_vmax – A tuple giving the range of values for the colormap.

  • legend – Whether to include the legend

  • scale_legend – Set to scale height and width of legend.

  • ax – The Axes to plot on. If None, creates a fresh figure for plotting. Incompatible with dendrogram plotting.

Returns:

A Figure.