mantispy.pl.feature_correlation#

mantispy.pl.feature_correlation(adata, key='selected', groupby='feature_group', max_features=300, ax=None)[source]#

Correlation heatmap with features ordered by their annotation.

Features are sorted by groupby and then by channel, with a line at each group boundary. Ordering by annotation instead of clustering shows directly whether correlated features fall within the same measurement family.

Parameters:
adata AnnData

Object to draw. Usually well-level profiles.

key str | None (default: 'selected')

Restrict to features flagged by this boolean var column. None, or a column the object does not hold, uses every feature.

groupby str (default: 'feature_group')

var column to order and delimit by.

max_features int (default: 300)

Draw at most this many features, taken in the sorted order.

ax Axes | None (default: None)

Axes to draw into.

Return type:

Axes | None

Returns:

The axes when the caller passed ax, else None because the plot then owns the figure it created. When returned they hold the correlation matrix on a diverging scale fixed to [-1, 1], with a line at each group boundary and one tick per group.

Raises:

KeyError – var has no groupby column, or no channel column.