Display r and theta on hover for polar plots #26271

klorine28 · 2023-07-06T10:36:28Z

PR summary

PR checklist

"Closes Add fmt_r and fmt_theta methods to polar axes #4568" is in the body of the PR description to link the related issue
new and changed code is tested
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

tacaswell · 2023-07-06T13:44:02Z

Thank you for your interest in contributing to Matplotilb, however this PR does not address the issue in #4568 .

The core issue is that for Cartesian plots there are functions you can set to control how the mouse over reports (see https://matplotlib.org/stable/gallery/misc/coords_report.html for the updated version of the docs, you will need to use an interactive backend (not matplotlib-inline in a notebook) to see it do anything) the x and y values. The issue reports that those functions are ignored on polar axes (see

matplotlib/lib/matplotlib/axes/_base.py

Lines 3905 to 3910 in b274f78

    
           def format_coord(self, x, y): 
        
               """Return a format string formatting the *x*, *y* coordinates.""" 
        
               return "x={} y={}".format( 
        
                   "???" if x is None else self.format_xdata(x), 
        
                   "???" if y is None else self.format_ydata(y), 
        
               )

and

matplotlib/lib/matplotlib/projections/polar.py

Lines 1396 to 1425 in b274f78

    
           def format_coord(self, theta, r): 
        
               # docstring inherited 
        
               screen_xy = self.transData.transform((theta, r)) 
        
               screen_xys = screen_xy + np.stack( 
        
                   np.meshgrid([-1, 0, 1], [-1, 0, 1])).reshape((2, -1)).T 
        
               ts, rs = self.transData.inverted().transform(screen_xys).T 
        
               delta_t = abs((ts - theta + np.pi) % (2 * np.pi) - np.pi).max() 
        
               delta_t_halfturns = delta_t / np.pi 
        
               delta_t_degrees = delta_t_halfturns * 180 
        
               delta_r = abs(rs - r).max() 
        
               if theta < 0: 
        
                   theta += 2 * np.pi 
        
               theta_halfturns = theta / np.pi 
        
               theta_degrees = theta_halfturns * 180 
        
               # See ScalarFormatter.format_data_short.  For r, use #g-formatting 
        
               # (as for linear axes), but for theta, use f-formatting as scientific 
        
               # notation doesn't make sense and the trailing dot is ugly. 
        
               def format_sig(value, delta, opt, fmt): 
        
                   # For "f", only count digits after decimal point. 
        
                   prec = (max(0, -math.floor(math.log10(delta))) if fmt == "f" else 
        
                           cbook._g_sig_digits(value, delta)) 
        
                   return f"{value:-{opt}.{prec}{fmt}}" 
        
               return ('\N{GREEK SMALL LETTER THETA}={}\N{GREEK SMALL LETTER PI} ' 
        
                       '({}\N{DEGREE SIGN}), r={}').format( 
        
                           format_sig(theta_halfturns, delta_t_halfturns, "", "f"), 
        
                           format_sig(theta_degrees, delta_t_degrees, "", "f"), 
        
                           format_sig(r, delta_r, "#", "g"), 
        
                       )

for updated perma-links from #4568 (comment) ).

The proposed fixes are to do one of the following:

add fmt_r and fmt_theta as analogs to the Cartesian versions
update the format_coords method on polar axes to look at the fmt_xdata and fmt_ydata methods

See the discussion in the issue for pros and cons of each. The rough consensus seemed to be towards the second option.

klorine28 · 2023-07-06T17:50:59Z

I am unusre if my solution is of any help as it was not clear to me if the goal was to leave a possibe example or make a standarization to solve an underlying issue. i did however manage to get the example code working with the bits that were asked for.

klorine28 · 2023-07-07T10:51:01Z

so the issue is that interactive ploting features are ignored when the plot is polar? and the goal of the issue is to modify the code above to make it so that these interactive plotting functions are not ignored?

This modification uses the fmt_xdata and fmt_ydata methods by retrieving the corresponding formatters (fmt_theta and fmt_r) from the x-axis and y-axis, respectively, to format the theta and r values in the format_coord method.

klorine28 · 2023-07-08T20:50:49Z

is the change I made better?

jklymak · 2023-07-09T15:32:28Z

@klorine28 ive not followed the suggestion here fully. However this is not in the form of either an example (galleries/examples) or what I think was being asked for, which is a modification to the polar axes that would display r and theta by default.

klorine28 · 2023-07-09T17:36:15Z

I apologise for this is my first interaction with open source that is not documentation. So what I am trying to do is modify the format coords so that r and theta are displayed by default? What I understood from the issue was that in cartesian plots, you can see the info of a point by interacting with it, but this is ignored in polar plots. As such, the solution would be to make it so that when you hover over a point in a polar plot, it displays theta and r. Is that correct?

klorine28 · 2023-07-17T17:07:32Z

any coment on my question? is the second commit what is wanted?

melissawm · 2023-07-20T17:39:13Z

Hi @klorine28 - you may get more attention here if you rename your PR to something more descriptive, for example something like "Display r and theta on hover for polar plots" (please choose the wording that makes more sense to you). In addition, you can then add "Closes #4568" to the description of your PR to make sure it is linked to the original issue here in the GitHub web interface.

Beyond that, there are also a couple of Linting errors that you can fix while you wait for a review: https://github.com/matplotlib/matplotlib/actions/runs/5486484810/job/14858128627?pr=26271

Thanks!

possible solve for issue matplotlib#4568

56af056

Update issue matplotlib#4568.py

e3cca9b

This modification uses the fmt_xdata and fmt_ydata methods by retrieving the corresponding formatters (fmt_theta and fmt_r) from the x-axis and y-axis, respectively, to format the theta and r values in the format_coord method.

melissawm added topic: polar API: consistency labels Jul 10, 2023

klorine28 changed the title ~~possible solve for issue #4568~~ Display r and theta on hover for polar plots Jul 22, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Display r and theta on hover for polar plots #26271

Display r and theta on hover for polar plots #26271

Uh oh!

klorine28 commented Jul 6, 2023 •

edited

Loading

Uh oh!

tacaswell commented Jul 6, 2023

Uh oh!

klorine28 commented Jul 6, 2023

Uh oh!

klorine28 commented Jul 7, 2023

Uh oh!

klorine28 commented Jul 8, 2023

Uh oh!

jklymak commented Jul 9, 2023

Uh oh!

klorine28 commented Jul 9, 2023

Uh oh!

klorine28 commented Jul 17, 2023

Uh oh!

melissawm commented Jul 20, 2023

Uh oh!

Uh oh!

Uh oh!

Display r and theta on hover for polar plots #26271

Are you sure you want to change the base?

Display r and theta on hover for polar plots #26271

Uh oh!

Conversation

klorine28 commented Jul 6, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR summary

PR checklist

Uh oh!

tacaswell commented Jul 6, 2023

Uh oh!

klorine28 commented Jul 6, 2023

Uh oh!

klorine28 commented Jul 7, 2023

Uh oh!

klorine28 commented Jul 8, 2023

Uh oh!

jklymak commented Jul 9, 2023

Uh oh!

klorine28 commented Jul 9, 2023

Uh oh!

klorine28 commented Jul 17, 2023

Uh oh!

melissawm commented Jul 20, 2023

Uh oh!

Uh oh!

klorine28 commented Jul 6, 2023 •

edited

Loading