arviz-devs / arviz

Exploratory analysis of Bayesian models with Python
https://python.arviz.org
Apache License 2.0
1.6k stars 397 forks source link

Minor fixes #2268

Closed OriolAbril closed 1 year ago

OriolAbril commented 1 year ago

Description

This PR fixes bad numba usage we had been doing which recently started causing errors and warnings. It also adds a numpy 2.0 cap to avoid issues once it is release (planned to happen around the end of this year already).

I think this should be included in a patch release for 0.16. However, it would be nice if someone could test locally and confirm that indeed, the numba warnings are gone. cc @guin0x @ahartikainen might also be related to #2244

Checklist


:books: Documentation preview :books:: https://arviz--2268.org.readthedocs.build/en/2268/

codecov[bot] commented 1 year ago

Codecov Report

Merging #2268 (f6bd1b8) into main (a1cefcc) will increase coverage by 0.03%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #2268      +/-   ##
==========================================
+ Coverage   87.87%   87.90%   +0.03%     
==========================================
  Files         122      122              
  Lines       12458    12459       +1     
==========================================
+ Hits        10947    10952       +5     
+ Misses       1511     1507       -4     
Impacted Files Coverage Δ
arviz/plots/backends/bokeh/forestplot.py 94.92% <100.00%> (ø)
arviz/plots/backends/bokeh/posteriorplot.py 98.36% <100.00%> (+0.01%) :arrow_up:
arviz/plots/backends/matplotlib/forestplot.py 97.86% <100.00%> (ø)
arviz/stats/density_utils.py 79.17% <100.00%> (ø)
arviz/stats/stats_utils.py 96.35% <100.00%> (ø)
arviz/utils.py 88.27% <100.00%> (ø)

... and 1 file with indirect coverage changes