Skip to main content

Replace == comparison with numpy.isnan()


ImportanceReview GuidanceRequires Scanning Tool
MediumMerge Without ReviewNo

Comparisons against numpy.nan always result in False. Thus comparing an expression directly against numpy.nan is always unintended. The correct way to compare a value for NaN is to use the numpy.isnan function.

Our changes look something like this:

import numpy as np

a = np.nan
-if a == np.nan:
+if np.isnan(a):

If you have feedback on this codemod, please let us know!


Why is this codemod marked as Merge Without Review?

We believe any use of == to compare with numpy.nan is unintended given that it is always False. Thus we consider this change safe.

Codemod Settings