[MNT]: Cleanup ticklabel_format (style=) #26801

pedrompecanha · 2023-09-17T16:42:14Z

PR summary

Fixes #25974.
-->

PR checklist

"closes #0000" is in the body of the PR description to link the related issue
[N/A] new and changed code is tested
[N/A] Plotting related features are demonstrated in an example
New Features and API Changes are noted with a directive and release note
Documentation complies with general and docstring guidelines

…fixing extra logic.

Fixing if condition.

github-actions

Thank you for opening your first PR into Matplotlib!

If you have not heard from us in a week or so, please leave a new comment below and that should bring it to our attention. Most of our reviewers are volunteers and sometimes things fall through the cracks.

You can also join us on gitter for real-time discussion.

For details on testing, writing docs, and our review process, please see the developer guide

We strive to be a welcoming and open project. Please follow our Code of Conduct.

rcomer

Thank you for your work on this @pedrompecanha! Just a couple of comments on the implementation...

This line will break if style is None, so perhaps should be moved into the else of your new if loop below.

matplotlib/lib/matplotlib/axes/_base.py

Line 3277 in 6fa5045

style = style.lower()

rcomer · 2023-09-18T08:00:09Z

lib/matplotlib/axes/_base.py

-        is_sci_style = _api.check_getitem(STYLES, style=style)
+        STYLES = {'sci': True, 'scientific': True, 'plain': False}
+        if style is None:
+            is_sci_style = False


Suggested change

is_sci_style = False

is_sci_style = None

None here should mean that the style is not changed from what it previously was. Setting is_sci_style also to None achieves that.

I'm changing it so that everything is in the dictionary. I believe it solves it more effectively. About setting it to None or False, I thought this variable should be either True or False but, after reading the logic after it, I've seen it can be None, indeed.

rcomer · 2023-09-18T08:10:57Z

The typing stub file will also need updating to match the new signature

matplotlib/lib/matplotlib/axes/_base.pyi

Lines 282 to 291 in 6fa5045

    
           def ticklabel_format( 
        
               self, 
        
               *, 
        
               axis: Literal["both", "x", "y"] = ..., 
        
               style: Literal["", "sci", "scientific", "plain"] = ..., 
        
               scilimits: tuple[int, int] | None = ..., 
        
               useOffset: bool | float | None = ..., 
        
               useLocale: bool | None = ..., 
        
               useMathText: bool | None = ... 
        
           ) -> None: ...

oscargus · 2023-09-18T08:13:23Z

lib/matplotlib/axes/_base.py

@@ -3283,8 +3284,11 @@ def ticklabel_format(self, *, axis='both', style='', scilimits=None,
            except (ValueError, TypeError) as err:
                raise ValueError("scilimits must be a sequence of 2 integers"
                                 ) from err
-        STYLES = {'sci': True, 'scientific': True, 'plain': False, '': None}
-        is_sci_style = _api.check_getitem(STYLES, style=style)
+        STYLES = {'sci': True, 'scientific': True, 'plain': False}


An option here (and also in case someone has set style='' explicitly), is to do:

Suggested change

STYLES = {'sci': True, 'scientific': True, 'plain': False}

STYLES = {'sci': True, 'scientific': True, 'plain': False, '': None, None: None}

is_sci_style = _api.check_getitem(STYLES, style=style)

in case someone has set style='' explicitly

That's a good point. I suggest to also add a comment that '' is included for backwards-compatibility, as I do not think it would be obvious to future readers of the code.

I added the comment and added the empty string into the dictionary. Actually, I was playing tennis and thought about how my changes weren't backwards compatible. Nice to see we are in the same page :)

…ionary and added stub parameter alternative.

rcomer · 2023-09-18T15:39:17Z

lib/matplotlib/axes/_base.py

            Whether to use scientific notation.
            The formatter default is to use scientific notation.
+            Sci is equivalent to scientific.
+            The '' option is included solely for backwards-compatibility.


Apologies, I think I was unclear. I meant to add a code comment where the dictionary is defined, so that future developers know that it’s there for a reason. I don’t think it’s needed in the docstring.

Done in the latest commit!

rcomer

The last test failure comes with instructions (which I admit are new to me!)

rcomer · 2023-09-19T08:36:14Z

lib/matplotlib/axes/_base.py

@@ -3283,7 +3284,9 @@ def ticklabel_format(self, *, axis='both', style='', scilimits=None,
            except (ValueError, TypeError) as err:
                raise ValueError("scilimits must be a sequence of 2 integers"
                                 ) from err
-        STYLES = {'sci': True, 'scientific': True, 'plain': False, '': None}
+        STYLES = {'sci': True, 'scientific': True, 'plain': False, '': None, None: None}
+        # 'sci' is equivalent to 'scientific'.


I think the fact that 'sci' and ‘scientific' are equivalent should be in the docstring, as this is useful information for the user.

Just put them back there! As for the pyplot issue, I attempted following the instructions and it told me, in my system, that the file got reformatted. Let's see how it goes!

It seems the pyplot change has not appeared here on GitHub...

Yeah! I tried running the command again, but the file remained unchanged. I tried to change the file by hand, although this is not encouraged, and the commit went through on my fork, but it won't open here on Github. Just goes straight to a 404 page. Weird.

Here is the final change I made: pedrompecanha@3cf0234

I believe it has gone through now. Must have been a GitHub instability. The tests seem to be running well, although I'm kind of scared of having to change this file by hand. Thoughts?

Your manual change matches what I get by running the script, so I think it's OK.

…ng area.

rcomer

Thanks @pedrompecanha, I think this is in good shape now. I just have a tiny suggestion on the docstring.

lib/matplotlib/axes/_base.py

…ers that default to None. Co-authored-by: Ruth Comer <[email protected]>

pedrompecanha · 2023-09-19T18:56:08Z

@rcomer, @timhoffm and @oscargus, thank you for the support and patience during this process! I'd like to ask that, if you find any similar cleanup or refactoring tasks, if you could please tag me, as I'll be really active in the next 1-2 months in here. Of course, if that is allowed by the rules of the project. I'll try my best to solve the issues. Again, thank you!

timhoffm · 2023-09-20T22:56:37Z

Thanks @pedrompecanha and congratulations on your first contribution to Matplotlib! We hope to see you back.

pedrompecanha added 2 commits September 17, 2023 13:28

Changing the default value to None, adding docstring documenting and …

c67d1c3

…fixing extra logic.

Update _base.py

b660799

Fixing if condition.

github-actions bot reviewed Sep 17, 2023

View reviewed changes

rcomer reviewed Sep 18, 2023

View reviewed changes

oscargus reviewed Sep 18, 2023

View reviewed changes

rcomer mentioned this pull request Sep 18, 2023

[MNT]: Cleanup ticklabel_format(..., style=) #25991

Closed

3 tasks

Added backwards-compatibility comment, changing to a solo STYLES dict…

2096ae7

…ionary and added stub parameter alternative.

rcomer reviewed Sep 18, 2023

View reviewed changes

Changing comments from the docstring area to inline comments.

c6fdf59

rcomer reviewed Sep 19, 2023

View reviewed changes

pedrompecanha added 3 commits September 19, 2023 09:30

Reformatting pyplot.py and putting the sci equivalency in the docstri…

4c3355d

…ng area.

Attempting to fix pyplot.py by hand, although this is not encouraged.

558b2d8

Adding a comma to pyplot.py.

3cf0234

rcomer reviewed Sep 19, 2023

View reviewed changes

lib/matplotlib/axes/_base.py Outdated Show resolved Hide resolved

Editing the docstring, making it more consistent to the other paramet…

6bdc99f

…ers that default to None. Co-authored-by: Ruth Comer <[email protected]>

rcomer approved these changes Sep 19, 2023

View reviewed changes

rcomer added this to the v3.9.0 milestone Sep 20, 2023

rcomer added the topic: ticks axis labels label Sep 20, 2023

timhoffm approved these changes Sep 20, 2023

View reviewed changes

timhoffm merged commit 2cbdff8 into matplotlib:main Sep 20, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[MNT]: Cleanup ticklabel_format (style=) #26801

[MNT]: Cleanup ticklabel_format (style=) #26801

pedrompecanha commented Sep 17, 2023 •

edited

Loading

github-actions bot left a comment

rcomer left a comment

rcomer Sep 18, 2023

pedrompecanha Sep 18, 2023

rcomer commented Sep 18, 2023 •

edited

Loading

oscargus Sep 18, 2023 •

edited

Loading

rcomer Sep 18, 2023

pedrompecanha Sep 18, 2023

rcomer Sep 18, 2023

pedrompecanha Sep 18, 2023

rcomer left a comment

rcomer Sep 19, 2023

pedrompecanha Sep 19, 2023

rcomer Sep 19, 2023

pedrompecanha Sep 19, 2023 •

edited

Loading

pedrompecanha Sep 19, 2023 •

edited

Loading

rcomer Sep 19, 2023

rcomer left a comment

pedrompecanha commented Sep 19, 2023

timhoffm commented Sep 20, 2023

	STYLES = {'sci': True, 'scientific': True, 'plain': False}
	STYLES = {'sci': True, 'scientific': True, 'plain': False, '': None, None: None}
	is_sci_style = _api.check_getitem(STYLES, style=style)

[MNT]: Cleanup ticklabel_format (style=) #26801

[MNT]: Cleanup ticklabel_format (style=) #26801

Conversation

pedrompecanha commented Sep 17, 2023 • edited Loading

PR summary

PR checklist

github-actions bot left a comment

Choose a reason for hiding this comment

rcomer left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rcomer commented Sep 18, 2023 • edited Loading

oscargus Sep 18, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rcomer left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pedrompecanha Sep 19, 2023 • edited Loading

Choose a reason for hiding this comment

pedrompecanha Sep 19, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rcomer left a comment

Choose a reason for hiding this comment

pedrompecanha commented Sep 19, 2023

timhoffm commented Sep 20, 2023

pedrompecanha commented Sep 17, 2023 •

edited

Loading

rcomer commented Sep 18, 2023 •

edited

Loading

oscargus Sep 18, 2023 •

edited

Loading

pedrompecanha Sep 19, 2023 •

edited

Loading

pedrompecanha Sep 19, 2023 •

edited

Loading