Support ax.grid(visible=<bool>).
#18769
Conversation
Remove the separate _gridOnMajor/_gridOnMinor attributes and store visibility together with all other grid attributes in the _major_tick_kw/_minor_tick_kw dicts under the "gridOn" key. This makes it possible to update just that entry in `Axis.grid`. Also, don't normalize `b` to True `if len(kwargs)` in `Axes.grid`: we need to keep it as `None`, normalizing it only to True/False in `Axis.grid` (which handles that just fine), so that `Axis.grid` can handle the consistency checks between `b` and `visible`.
|
That issue was a regression from 3.2.2. Is this entire change necessary to fix it? I'm trying to determine whether this whole PR would need to go back, or some smaller version only. |
|
I didn't know it was a regression, so I have no idea whether there's a simpler fix... |
| self.set_tick_params(which='minor', gridOn=self._gridOnMinor, | ||
| **gridkw) | ||
| if b is None and not forced_visibility: | ||
| gridkw['gridOn'] = not self._minor_tick_kw['gridOn'] |
There was a problem hiding this comment.
Should this toggle the value in _minor_tick_kw?
There was a problem hiding this comment.
set_tick_params (called just below) does that.
| """ | ||
| if len(kwargs): | ||
| if not b and b is not None: # something false-like but not None |
There was a problem hiding this comment.
Shouldn't we instead rename b to visible?
There was a problem hiding this comment.
b=None (== toggle) and visible=None (== False) behave differently. I think all these toggling APIs are pretty terrible, but breaking all the scripts in the world that do ax.grid() to toggle the grid on is just not worth it.
I guess we could aim to have the end API be def grid(visible=True, ...): (and deprecate b) though. Then the only case we'd break is people doing ax.grid() to switch off the grid when the grid is already on, which is probably much less common (and ax.grid(False) just seems much clearer). But let's keep that discussion for another time.
There was a problem hiding this comment.
Uh, it's worse than I thought.
|
Owee, I'm MrMeeseeks, Look at me. There seem to be a conflict, please backport manually. Here are approximate instructions:
And apply the correct labels and milestones. Congratulation you did some good work ! Hopefully your backport PR will be tested by the continuous integration and merged soon! If these instruction are inaccurate, feel free to suggest an improvement. |
…3.3.x Backport PR #18769 on branch v3.3.x
|
This change has broken other packages. Is it really necessary to remove _gridOnMajor/_gridOnMinor? |
|
Do we really have no public interface to check whether the grid is on? Downstream libraries are now replacing the private |
|
It would be helpful if other packages open an issue if they are stuck using a private attribute to access information they need. |
Remove the separate _gridOnMajor/_gridOnMinor attributes and store
visibility together with all other grid attributes in the
_major_tick_kw/_minor_tick_kw dicts under the "gridOn" key. This makes
it possible to update just that entry in
Axis.grid.Also, don't normalize
bto Trueif len(kwargs)inAxes.grid:we need to keep it as
None, normalizing it only to True/False inAxis.grid(which handles that just fine), so thatAxis.gridcanhandle the consistency checks between
bandvisible.Closes #18758.
PR Summary
PR Checklist
pytestpasses).flake8on changed files to check).flake8-docstringsandpydocstyle<4and runflake8 --docstring-convention=all).doc/users/next_whats_new/(follow instructions in README.rst there).doc/api/next_api_changes/(follow instructions in README.rst there).