fix: Correct variable assignment for limit band artists

[lukasheinrich]

Description

We discovered a little bug in the brazil band visualization where the two sigma band was assigned to the "one_sigma_band" attribute of our artist collection and vice versa.

Fixing this now.

cc @malin-horstmann

Amends PR #1377

Checklist Before Requesting Reviewer

• Tests are passing
• "WIP" removed from the title of the pull request
• Selected an Assignee for the PR to be responsible for the log summary

Before Merging

For the PR Assignees:

• Summarize commit messages into a comprehensive review of the PR

* Assign the one and two sigma band artists to the correct objects.
   - Amends PR https://github.com/scikit-hep/pyhf/pull/1377

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (970cd11) 98.28% compared to head (88bc5d5) 98.28%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2411   +/-   ##
=======================================
  Coverage   98.28%   98.28%           
=======================================
  Files          69       69           
  Lines        4539     4539           
  Branches      803      803           
=======================================
  Hits         4461     4461           
  Misses         45       45           
  Partials       33       33           

Flag	Coverage Δ
contrib	97.86% <100.00%> (ø)
doctest	60.71% <100.00%> (ø)
unittests-3.10	96.29% <0.00%> (ø)
unittests-3.11	96.29% <0.00%> (ø)
unittests-3.8	96.32% <0.00%> (ø)
unittests-3.9	96.34% <0.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[malin-horstmann]

Looks great.

[matthewfeickert]

Thanks @lukasheinrich and sorry for not noticing this until today (though with my laptop dead I would have had to review this on the phone anyway).

Here's an example that extends the docs example showing the bug

import numpy as np
import matplotlib.pyplot as plt
import pyhf
import pyhf.contrib.viz.brazil

pyhf.set_backend("numpy")
model  = pyhf.simplemodels.uncorrelated_background(
    signal=[12.0, 11.0], bkg=[50.0, 52.0], bkg_uncertainty=[3.0, 7.0]
)
observations = [51, 48]
data = observations + model.config.auxdata

test_pois = np.linspace(0, 5, 41)
results = [
    pyhf.infer.hypotest(test_poi, data, model, return_expected_set=True)
    for test_poi in test_pois
]
cls_obs = np.array([test[0] for test in results]).flatten()
cls_exp = [
    np.array([test[1][sigma_idx] for test in results]).flatten()
    for sigma_idx in range(5)
]
test_size = 0.05

fig, ax = plt.subplots()
artists = pyhf.contrib.viz.brazil.plot_brazil_band(
    test_pois, cls_obs, cls_exp, test_size, ax
)

one_sigma_band = artists[2]
one_sigma_band.set_color("blue")

ax.legend()

fig.savefig("pr_2411_bug.png")

[matthewfeickert]

Thanks @lukasheinrich. I'll go ahead and finish the dependabot updates, then merge this, and get a patch release out with this fix in it.

[matthewfeickert]

This fix is now out in pyhf v0.7.6.