New submission from Reed <rea...@gmail.com>:
If a float32 Numpy array is passed to statistics.variance(), an assertion
failure occurs. For example:
import statistics
import numpy as np
x = np.array([1, 2], dtype=np.float32)
statistics.variance(x)
The assertion error is:
assert T == U and count == count2
Even if you convert x to a list with `x = list(x)`, the issue still occurs. The
issue is caused by the following lines in statistics.py
(https://github.com/python/cpython/blob/ec007cb43faf5f33d06efbc28152c7fdcb2edb9c/Lib/statistics.py#L687-L691):
T, total, count = _sum((x-c)**2 for x in data)
# The following sum should mathematically equal zero, but due to rounding
# error may not.
U, total2, count2 = _sum((x-c) for x in data)
assert T == U and count == count2
When a float32 Numpy value is squared in the term (x-c)**2, it turns into a
float64 value, causing the `T == U` assertion to fail. I think the best way to
fix this would be to replace (x-c)**2 with (x-c)*(x-c). This fix would no
longer assume the input's ** operator returns the same type.
----------
components: Library (Lib)
messages: 359323
nosy: reed
priority: normal
severity: normal
status: open
title: Assertion failure when calling statistics.variance() on a float32 Numpy
array
type: behavior
versions: Python 3.8
_______________________________________
Python tracker <rep...@bugs.python.org>
<https://bugs.python.org/issue39218>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
