Title. In the current plotter, using the zoom functionality from the matplotlib toolbar is a bit annoying, because the zoom, if set, is pretty much reset whenever anything is done with the clusters plotter (i.e., selecting an object or unselecting it).
The ax reset function is necessary, though. Otherwise, choosing different features as x/y-features wouldn't resize the axis and the data could end up way outside the biaplotter's FOV, making it practically unusable.
A solution could be to break out the axis resizing in its own little function and trigger it separately and explicitly in the feature_update function here in the clusters plotter, not everytime the _colorize() and _refresh() methods in the biaplotter are invoked.
Title. In the current plotter, using the zoom functionality from the matplotlib toolbar is a bit annoying, because the zoom, if set, is pretty much reset whenever anything is done with the clusters plotter (i.e., selecting an object or unselecting it).
The ax reset function is necessary, though. Otherwise, choosing different features as x/y-features wouldn't resize the axis and the data could end up way outside the biaplotter's FOV, making it practically unusable.
A solution could be to break out the axis resizing in its own little function and trigger it separately and explicitly in the
feature_updatefunction here in the clusters plotter, not everytime the_colorize()and_refresh()methods in the biaplotter are invoked.