GeezerVison

Copy link

Copy Markdown

Contributor

eric-wieser requested a review from anntzer February 11, 2020 16:27

eric-wieser force-pushed the fix-mplot3d-projection branch from 8ffc57b to ce08435 Compare February 11, 2020 16:51

jklymak approved these changes Feb 11, 2020

View reviewed changes

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

anntzer approved these changes Feb 11, 2020

View reviewed changes

To make bbox_inches='tight' work correctly, we cache the positions of
all of the axes, change their aspect to 'auto', adjust the transforms
on the figure, render the output, and then restore the previous
setting of aspect to each of the figures.  This prevent the Axes from
trying to adjust their size on draw. This matters because for the
render that is emitted the figure transforms are out of sync with the
figure size.

As part of cleaning fixing the rendering of Axes3D in matplotlib#8896 / matplotlib#16472
we started to use apply_aspect to re-size the area the Axes3D takes up
to maintain a fixed ratio between the axes when the aspect is "auto".
However this conflicts with the expectation in tight_bbox.adjust_bbox
as it assumes setting the aspect to "auto" will prevent any axes
resizing.

This commit addresses this by:
 - exiting the Axes3D.apply_aspect early if aspect is auto
 - adding a new aspect mode 'auto_pb' which is the default
   for Axes3D which maintains the current master branch behavior
   of maintaining a fixed ratio between the sizes of the x, y z axis
   independent of the data limits.
 - re implement several functions on Axes3D to make sure they handle
   sharez correctly.

closes matplotlib#16463.

The way that bbox_inches='tight' is implemented we need to ensure that
we do not try to adjust the aspect during the draw (because we have
temporarily de-coupled the reported figure size from the transforms
which results in the being distorted).  Previously we did not have a
way to fix the aspect ratio in screen space of the Axes (only the
aspect ratio in dataspace) however in 3.3 we gained this ability for
both Axes (matplotlib#14917) and Axes3D (matplotlib#8896 / matplotlib#16472).

Rather than add an aspect value to `set_aspect` to handle this case,
in the tight_bbox code we monkey-patch the `apply_aspect` method with
a no-op function and then restore it when we are done.  Previously we
would set the aspect to "auto" and restore it in the same places.

closes matplotlib#16463.

The way that bbox_inches='tight' is implemented we need to ensure that
we do not try to adjust the aspect during the draw (because we have
temporarily de-coupled the reported figure size from the transforms
which results in the being distorted).  Previously we did not have a
way to fix the aspect ratio in screen space of the Axes (only the
aspect ratio in dataspace) however in 3.3 we gained this ability for
both Axes (matplotlib#14917) and Axes3D (matplotlib#8896 / matplotlib#16472).

Rather than add an aspect value to `set_aspect` to handle this case,
in the tight_bbox code we monkey-patch the `apply_aspect` method with
a no-op function and then restore it when we are done.  Previously we
would set the aspect to "auto" and restore it in the same places.

closes matplotlib#16463.