Confused about output with comparison operators is and == in this case

Asked Apr 19 '20 at 21:35

Active Apr 19 '20 at 21:35

Viewed 12 times

Say I have a list:

my_list = [0, 1, 2, 3, nan, 5, 6, 7, 8, 9]

Why are these 2 outputs different?

for x in my_list:
    if x == np.nan:
        continue
    print(x, end= " ")

0 1 2 3 nan 5 6 7 8 9

for x in my_list:
    if x is np.nan:
        continue
    print(x, end= " ")

0 1 2 3 5 6 7 8 9

asked Apr 19 '20 at 21:35

delvins78

Check [differences between `is and ==`](https://www.geeksforgeeks.org/difference-operator-python/) – DarrylG Apr 19 '20 at 21:38
It is because np.nan == np.nan yields false. It con not compare nan to nan – Vinod Kumar Apr 19 '20 at 21:39
1

A very good article to clear your doubts: https://towardsdatascience.com/navigating-the-hell-of-nans-in-python-71b12558895b – Rajat Mishra Apr 19 '20 at 21:39
1

Use `np.isnan(number)` – Patrick Artner Apr 19 '20 at 21:40
1

FWIW, the correct API for testing for `np.nan` is [np.isnan](https://numpy.org/doc/stable/reference/generated/numpy.isnan.html?highlight=nan#numpy.isnan). - edit: ah, Patrick beat me to it, leaving for the direct link :). – spectras Apr 19 '20 at 21:40

0 Answers0