Error creating AxisGrid with non-default axis class #17069
Comments
|
This could probably be made to work by: a) renaming the |
|
Ok. It this begs the question of why one would use axes grid for cartopy axes? |
|
An alternative change here would be to put is the type check and raise an informative error that it is not going to work. |
|
OTOH it may be nice to slowly move axes_grid towards an API more consistent with the rest of mpl. From a very, very quick look, I guess its |
There's an example in the Cartopy docs. |
|
For that example I get |
|
So, I'm confused, is |
|
Sorry for the above, I see now. That example could be changed to fig, axgr = plt.subplots(3, 2, constrained_layout=True,
subplot_kw={'projection':projection})
axgr = axgr.flat
...
fig.colorbar(p, ax=axgr, shrink=0.6, extend='both') |
|
@jklymak the reason why I went to use AxesGrid was because it seemed the easiest for me to create multiple GeoAxes instances flexibly (i.e. with or without colorbar axes, and with flexible location of those) and with an easy control of both horizonal and vertical padding of GeoAxis instances and independently, of the colorbar axes, also because the aspect of maps (lat / lon range) tends to mess with the alignment. The code I used to create initiate this figure is based on the example from the cartopy website that @QuLogic mentions above: fig = plt.figure(figsize=(18, 7))
axes_class = (GeoAxes, dict(map_projection=ccrs.PlateCarree()))
axgr = AxesGrid(fig, 111, axes_class=axes_class,
nrows_ncols=(3, 4),
axes_pad=(0.6, 0.5), # control padding separately for e.g. colorbar labels, axes titles, etc.
cbar_location='right',
cbar_mode="each",
cbar_pad="5%",
cbar_size='3%',
label_mode='')
# here follows the plotting code of the displayed climate data using pyaerocom by loading a 2010 monthly example model dataset, looping over the (GeoAxes, cax) instances of the grid and calling pyaerocom.plot.mapping.plot_griddeddata_on_map on the monthly subsets.However, @jklymak I was not aware of the Also, constrained layout itself is declared experimental in the above guide and may be deprecated, so it may be a bit uncertain for users and developers that build upon matplotlib, what option to go for. |
|
@jgliss Yeah, I think I agree that I agree with @anntzer comment above (#17069 (comment)), but feel that we should move axes_grid into core matplotlib and change the API as we see fit when we do so. I also agree that its time Back on topic, though, this seems to be a regression and we should fix it. |
|
Re-milestoned to 3.2.2 as this seems like it is a regression, not "always broken"? |
|
Right after I re-milestoned I see that this is with 3.1.2 so I'm confused if this ever worked? |
|
I re-milestoned this to 3.4 as I don't think this has ever worked without setting the kwarg |
|
Actually, looks simple enough to just check what kind of object ax.axis is: diff --git i/lib/mpl_toolkits/axes_grid1/axes_grid.py w/lib/mpl_toolkits/axes_grid1/axes_grid.py
index 2b1b1d3200..8b947a5836 100644
--- i/lib/mpl_toolkits/axes_grid1/axes_grid.py
+++ w/lib/mpl_toolkits/axes_grid1/axes_grid.py
@@ -1,5 +1,6 @@
from numbers import Number
import functools
+from types import MethodType
import numpy as np
@@ -7,14 +8,20 @@ from matplotlib import _api, cbook
from matplotlib.gridspec import SubplotSpec
from .axes_divider import Size, SubplotDivider, Divider
-from .mpl_axes import Axes
+from .mpl_axes import Axes, SimpleAxisArtist
def _tick_only(ax, bottom_on, left_on):
bottom_off = not bottom_on
left_off = not left_on
- ax.axis["bottom"].toggle(ticklabels=bottom_off, label=bottom_off)
- ax.axis["left"].toggle(ticklabels=left_off, label=left_off)
+ if isinstance(ax.axis, MethodType):
+ bottom = SimpleAxisArtist(ax.xaxis, 1, ax.spines["bottom"])
+ left = SimpleAxisArtist(ax.yaxis, 1, ax.spines["left"])
+ else:
+ bottom = ax.axis["bottom"]
+ left = ax.axis["left"]
+ bottom.toggle(ticklabels=bottom_off, label=bottom_off)
+ left.toggle(ticklabels=left_off, label=left_off)
class CbarAxesBase:seems to be enough. |
Bug report
Bug summary
Creating
AxesGridusing cartopyGeoAxesasaxis_classraisesTypeError: 'method' object is not subscriptable. Seems to be due to different behaviour ofaxisattr. formpl_toolkits.axes_grid1.mpl_axes.Axesand other axes instances (likeGeoAxes) whereaxisis only a callable. The error is raised in methodmpl_toolkits.axes_grid1.axes_grid._tick_onlywhen trying to access keys fromaxisattr.Code for reproduction
Actual outcome
Expected outcome
Matplotlib version
The text was updated successfully, but these errors were encountered: