FIX: indexed pandas bar

[jklymak]

PR Summary

Closes #15162

basically if you do df = pd.DataFrame({"x":[1,2,3],"width":[.2,.4,.6]},index=[1,2,3]), then df.x[0] errors because that index doesn't exist, which breaks bar.

While I appreciate this is a bit more code than trying to call iat(0), which is what we did before, this solution doesn't presuppose iat exists.

import pandas as pd
from matplotlib import pyplot as plt
import numpy as np

df = pd.DataFrame({"x":[1.,2.,3.],"width":[.2,.4,.6]}, index=[1, 2, 3])
plt.figure()
plt.bar(df.x, 1., width=df.width)
plt.savefig('/Users/jklymak/downloads/Test1.png')

df = pd.DataFrame({"x":[1,3,10]},index=[1000,2000,3000])

plt.figure()
plt.bar(df.x, 1, width=.2)
plt.savefig('/Users/jklymak/downloads/Test2.png')

Now yields the expected (and pre 3.1.0):

PR Checklist

• Has Pytest style unit tests
• Code is Flake 8 compliant
• New features are documented, with examples if plot related
• Documentation is sphinx and numpydoc compliant
• Added an entry to doc/users/next_whats_new/ if major new feature (follow instructions in README.rst there)
• Documented in doc/api/api_changes.rst if API changed in a backward-incompatible way

[ImportanceOfBeingErnest]

Is this function used anywhere else then with bars? Should bar also support dataframes, like

df = pd.DataFrame({0:[1,2,3]},index=[1,2,3])

This works with plt.scatter(df, [3,4,5]), but probably not with plt.bar(df, [3,4,5]), right?

[jklymak]

I don't think

df = pd.DataFrame({'boo':[1,2,3]},index=[1,2,3])
plt.figure()
plt.bar(df, 1, width=.2)

ever worked (TypeError: float() argument must be a string or a number, not 'NoneType' on 3.0.0). The error with this PR if the data name is called 0 is a bit more mysterious, but that's a relatively perverse data frame (I hope).

[ImportanceOfBeingErnest]

Ok, just take df = pd.DataFrame([1,2,3]) then. I think it would be good to define whether single-column dataframes are supposed to work or not. They do for scatter/plot, so one might expect that, but if they don't, maybe a message that is a bit more helpful than "Your linewidth is wrong" would be useful.

[jklymak]

This PR is trying to fix something I broke, not expand the duck typing for bar.

[jklymak]

This is great! Should have checked cbook....

[lumberbot-app]

Owee, I'm MrMeeseeks, Look at me.

There seem to be a conflict, please backport manually. Here are approximate instructions:

1. Checkout backport branch and update it.

$ git checkout v3.1.x
$ git pull

2. Cherry pick the first parent branch of the this PR on top of the older branch:

$ git cherry-pick -m1 e67c2fbfc257a1f3414eebfe873b78552ba9a7f3

3. You will likely have some merge/cherry-pick conflict here, fix them and commit:

$ git commit -am 'Backport PR #15166: FIX: indexed pandas bar'

4. Push to a named branch :

git push YOURFORK v3.1.x:auto-backport-of-pr-15166-on-v3.1.x

5. Create a PR against branch v3.1.x, I would have named this PR:

"Backport PR #15166 on branch v3.1.x"

And apply the correct labels and milestones.

Congratulation you did some good work ! Hopefully your backport PR will be tested by the continuous integration and merged soon!

If these instruction are inaccurate, feel free to suggest an improvement.

[tacaswell]

backported in #15172 / 169a9f7