FIX: Handle no-offsets in collection datalim #22945
Conversation
This adds a check for whether the collection has offsets passed in initially. This is necessary to handle the (0, 0) offset case that is desired and distinguishing that from the initial empty offset of (0, 0). In particular, this is necessary for the non-transData IdentityTransform passed in during tight_layout calculations.
|
Is the extra variable really necessary? One could internally keep _offsets = None. The advantage is that the logical information ( no offset | zero offset | finite offset ) is kept in one variable. But I didn't look into the code maybe it's overly complicated to handle None everywhere in the code. |
The solution to this may be to introduce a |
|
See #22946 for an alternative to this. I don't have a strong preference either way. |
|
I'm confused; both of these are merged? |
|
I was too, but I didn't see multiple commits in main. I think that what happened is GitHub saw that this commit hash was added to main as a part of the other PR (I branched off of this to keep the test) and then said that this was merged already, so updated the notice here. Maybe @timhoffm did manually merge both though? |
|
Oh, I didn't notice there were two commits in the other. Yea, GitHub will mark this merged since that commit is now in |
PR Summary
This adds a check for whether the collection has offsets
passed in initially. This is necessary to handle the (0, 0) offset
case that is desired and distinguishing that from the initial empty
offset of (0, 0). In particular, this is necessary for the
non-transData IdentityTransform passed in during tight_layout calculations.
closes #22921
PR Checklist
Tests and Styling
pytestpasses).flake8-docstringsand runflake8 --docstring-convention=all).Documentation
doc/users/next_whats_new/(follow instructions in README.rst there).doc/api/next_api_changes/(follow instructions in README.rst there).