Closed jr2804 closed 2 years ago
Hi!
Thank you for your contribution!
I wasn't aware of this issue until now, and switching to plt.axes()
seems a good replacement.
However, the call ax = plt.axes(fig, projection='3d')
with 'fig' as the first argument is not working for me. It works just fine without, though.
Can you fix this, or should I go ahead and commit?
Actually, I just remembered why the call is fig.gca()
.
The reasoning is that this makes it possible to pass a custom figure fig
, e.g. with certain dimensions, which then gets reused.
Therefore, I think that calling a member function of fig is better. In fact, fig.gca()
calls Figure.add_subplot
, which seems to be the correct way to call explicitly.
So I would propose to use fig.add_subplot()
.
So I would propose to use
fig.add_subplot()
.
Yes, that seems to be the best way. My solution was just the one with the least amount of changes :-) I couldn't check yet why I proposed the non-working call to plt.axes(fig, ...)
... actually I only needed one plotting function and modified everything else with search & replace - maybe not the best idea ;-)
I will close this PR and make a new one, thanks!
Initialization of 3D plot via ax = fig.gca(projection='3d') was declared as deprecated in recent matplotlib version.
(usage of gca() should in general be avoided, when possible)