Wrong values when iterating over numpy array

Question

Using the following code:

import numpy as np
k = .2
A = np.arange(.01, k, .01, dtype=np.float64)

# values
print(A)
print([A[i] for i in range(len(A))])

# comparison
print(A == [A[i] for i in range(len(A))])

# check for value
print(.06 in A)

The corresponding output is:

[0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1  0.11 0.12 0.13 0.14
 0.15 0.16 0.17 0.18 0.19]
[0.01, 0.02, 0.03, 0.04, 0.05, 0.060000000000000005, 0.06999999999999999, 0.08, 0.09, 0.09999999999999999, 0.11, 0.12, 0.13, 0.14, 0.15000000000000002, 0.16, 0.17, 0.18000000000000002, 0.19]
[ True  True  True  True  True  True  True  True  True  True  True  True
  True  True  True  True  True  True  True]
False

What sort of magic is happening here ? How am I supposed to iterate over the array keeping precision ?

EDIT: as precised in Is floating point math broken ? this lost of precision is a known issue, language agnostic. I would know what is the python way to handle this ? Using an epsilon seems to be a little hacky for me

One option is to check that the absolute value of the difference is less than some epsilon: print([abs(A[i] - 0.6) < 1e-10 for i in range(len(A))]). There are also infinite precision libraries available. It all depends on what you're trying to do. — pault
– pault, Commented Mar 28, 2018 at 14:08
@pault I understand my first question is duplicate but I would be interested in a response to the second one, more python specific — Loheek
– Loheek, Commented Mar 28, 2018 at 14:10
What is the desired output? To check the existence of the value in the array, you could do something like print(any(abs(A[i]-0.06) < 1e-10 for i in range(len(A)))) — pault
– pault, Commented Mar 28, 2018 at 14:14
Thank you, I will use this work around. Is there a more python way to handle this ? — Loheek
– Loheek, Commented Mar 28, 2018 at 14:17

Jay · Accepted Answer · 2018-03-28 14:29:58Z

0

If you just use a multi-line for loop for this the 'problem' goes away. It seems to be an 'issue' when using the generator.

for i in range (len(A)):
    print(A[i])

gives

0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
0.16
0.18
0.19

answered Mar 28, 2018 at 14:29

Jay

751 silver badge10 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

eapetcho Over a year ago

See help(np.arange). In fact, to be sure to obtain an exact representation of the solution, you need to use np.linspace() instead of np.arange() as np.arange() is more accurate with integer step

Collectives™ on Stack Overflow

Wrong values when iterating over numpy array

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related