Numpy testing¶

Testing for array equality is very common when writing unit tests for code that heavily utilizes numpy. So, to further expand on my recent call to use smart python asserts here are a few great asserts customized just for numpy.

import numpy
experiment = numpy.arange(10, 20)
year = numpy.arange(1990, 2000)
numpy.testing.assert_array_equal(experiment, year)

AssertionError: Arrays are not equal

(mismatch 100.0%)
 x: array([10, 11, 12, 13, 14, 15, 16, 17, 18, 19])
 y: array([1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999])

This is so much more informative and easy than the naive assert method I used before:

import numpy
experiment = numpy.arange(10, 20)
year = numpy.arange(1990, 2000)

equal = year == experiment
self.assertTrue(equal.all(), "Mismatched arrays")

self.assertTrue(equal.all(), "Mismatched arrays")
AssertionError: Invalid return value