You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I recently learned about the nbagg backend, and wanted to try it out in the notebook. My natural assumption on how to do this was %matplotlib nbagg, but this failed with a KeyError: u'nbagg'. (this is with mpl 1.4.0 and the latest ipython dev, btw)
I saw the discussion in #6424 that said it was decided not to default to nbagg, but is there a reason why the matplotlib magic shouldn't support this? I definitely have a workflow where it sometimes makes sense to mix in a few nbagg plots with mostly inline plots. Letting the %matplotlib magic work with nbagg in combination with #6439 would be a beautiful solution for this problem.
Activity
takluyver commentedon Sep 15, 2014
I think it certainly makes sense for
%matplotliband%pylabto support nbagg.minrk commentedon Oct 2, 2014
fixed by #6587