Error when using the numpy.where function

Question

I want to use the numpy.where function to check whether an element in an array is a certain string, like for example coffee and then returning a certain vector in places where this is true, and a different one in places where this is not the case.

However, I keep getting the error message saying operands could not be broadcast together with shapes (4,) (1,3) (1,3).

Is there some other way I can do this without using for loops too much (the question explicitly says i should not use them)?

lst_1 = np.array(["dog", "dog1", "dog2", "dog3"])
a = np.where(lst_1 == "dog", [[1,0,0]], [[0,0,0]])
print(a)

The error message is explanatory, no? how are you going to index into (1,3) via (4, )? — Robin Gertenbach
– Robin Gertenbach, Commented Sep 19, 2021 at 13:47
Is your expected output [[1,0,0], [0,0,0], [0,0,0], [0,0,0]]? — not_speshal
– not_speshal, Commented Sep 19, 2021 at 13:47
Check out broadcasting rules here numpy.org/doc/stable/user/basics.broadcasting.html — caaax
– caaax, Commented Sep 19, 2021 at 13:53
Yes, the expected output is [[1,0,0], [0,0,0], [0,0,0], [0,0,0]]. — Amateur1
– Amateur1, Commented Sep 19, 2021 at 13:57

Robin Gertenbach · Accepted Answer · 2021-09-19 14:20:45Z

1

Can be done as a one-liner:

out = np.array([[0,0,0], [1,0,0]])
idx = lst_1 == dog
out[idx.astype(np.int32)]

Alternatively avoiding casting:

np.take([[0,0,0],[1,0,0]], lst_1 == "dog", axis=0)

edited Sep 19, 2021 at 14:20

answered Sep 19, 2021 at 14:08

Robin Gertenbach

10.9k3 gold badges30 silver badges38 bronze badges

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

1 Comment

Robin Gertenbach Over a year ago

If this solved your problem please accept the answer so people in the future can easily see that a solution was found if they have the same problem

caaax · Accepted Answer · 2021-09-19 14:15:16Z

0

If you want to do this without for loops, you can make use of lambda functions:

lst_1 = np.array(["dog", "dog1", "dog2", "dog3"])
a = list(map(lambda x: [1,0,0] if x=='dog' else [0,0,0], lst_1))

print(a)

> [[1, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0]]

edited Sep 19, 2021 at 14:15

answered Sep 19, 2021 at 14:08

caaax

4701 gold badge5 silver badges18 bronze badges

Collectives™ on Stack Overflow

Error when using the numpy.where function

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related