arviz_plots.PlotCollection.add_legend

arviz_plots.PlotCollection.add_legend#

PlotCollection.add_legend(dim, aes=None, artist_kwargs=None, title=None, text_only=False, labeller=None, **kwargs)[source]#

Add a legend for the given visual/aesthetic to the plot.

Parameters:
dimhashable or iterable of hashable

Dimension or dimensions for which to generate the legend. The pseudo-dimension __variable__ is allowed too. It should have at least one aesthetic mapped to it. Only the mappings that match will be taken into account; if a legend is requested for the “chain” dimension but there is only one aesthetic mapping for (“chain”, “group”) no legend can be generated.

aesstr or iterable of str, optional

Specific aesthetics to take into account when generating the legend. They should all be mapped to dim. Defaults to all aesthetics matching that mapping with the exception “x” and “y” which are never included.

artist_kwargsmapping, optional

Keyword arguments passed to the backend visual function used to generate the miniatures in the legend.

titlestr, optional

Legend title. Defaults to dim.

text_onlybool, optional

If True, creates a text-only legend without graphical markers.

labellerlabeller instance, optional

Labeller to generate the legend entries

position(int, int), default (0, -1)
**kwargsmapping, optional

Keyword arguments passed to the backend function that generates the legend.

Returns:
legendobject

The corresponding legend object for the backend of the PlotCollection.

Warning

This method is still in early stages of experimentation and anything beyond the basic usage add_legend("dim_name") will probably change in breaking ways.