DOC: MEP to improve the Axes API #5029
Conversation
|
|
||
| First we define our coordinate transformation functions: | ||
| axes_to_base(self, \*q) | ||
| base_to_axes(self, x, y) |
There was a problem hiding this comment.
These live in the transform in existing matplotlib speak. It might be worth mentioning this (or even providing a section on what the transform currently does).
There was a problem hiding this comment.
Good idea, I will do that! In terms of these specific lines in the MEP, I kind of knew that we did these by transforms, but it becomes really unclear (especially for the novice user, and in terms of the Axes API, I still consider myself one of them) which does what, so many transforms ;)... so here I thought having a clear self-explaining API would help here.
|
Nice start @OceanWolf. Obviously this is not complete, so I encourage you to continue iterating, and to ping us when you are ready for some substantial comment. 👍 |
|
Thanks @pelson, yes, I will continue iterating on this. As I say in my reply above. while I now have a lot of knowledge in the backends, I still feel very new to the Axes "logic". It feels like a maze (well more of a jungle) of callbacks and aliases to functions and weird function names that don't do what they say they do... and that happens before you get to the transform stack/swamp which recurses (verb of recursion?) many layers deep. Luckily one can mostly avoid having to go near that swamp, but if one has too, like I had to to debug #4318, eeek. But yes, now I have an idea of a few things to look at in this jungle, I can expand a lot more on this MEP :). |
|
...closing as stale.. |
A beginning of an MEP to start discussion...
As well as providing a more streamlined and logical API, this MEP will close the following bugs: