MNT: Separate property cycle handling from _process_plot_var_args #29469
Conversation
By encapsulating the cycling into a dedicated class, we prepare for making cycling sharable between multiple _process_plot_var_args instances, which is the prerequisite for unifying the separate cyclers for lines and patches (matplotlib#29468) and sharing cyclers between multiple Axes (matplotlib#19479).
|
|
||
| def _setdefaults(self, defaults, kw): | ||
| @staticmethod |
There was a problem hiding this comment.
This is not strictly related, but noticed that _setdefaults does not depend on the class, so making that explicit via staticmethod as an improvement on the side.
|
Hi, just wanted to chip in that we are facing similar issues that it is a bit opague how the cycling is handled. In our case we would also need a |
|
Thanks for the feedback. This PR is an internal refactoring and since we currently don’t need that function. I won’t add it to the PR. After all, this is all internal for now and we reserve the right to change everything. Even if I would add the function you could not rely that it will be available and have a defined API. That said, this is part of a larger plan to make cycling more flexible. I plan to make the cycle in the Axes accessible as public API, which also includes the ability to get the next element. |
|
Ok thanks for clarifying. Then I will proceed with our internal cycling for now. |
By encapsulating the cycling into a dedicated class, we prepare for making cycling sharable between multiple _process_plot_var_args instances, which is the prerequisite for unifying the separate cyclers for lines and patches (#29468) and sharing cyclers between multiple Axes (#19479).