Assign value to subset of numpy array

Question

Given the following array:

x = np.zeros((1, 5))
x
# array([[0., 0., 0., 0., 0.]])

I'd like to be able to create array([[0., 3., 1., 0., 8.]]) using something like:

values = [3, 1, 8]
indices = [1, 2, 4]
x.iloc[indices] = values

I understand that this doesn't work - but I'm not sure what an idiomatic approach to this sort of thing in numpy would be.

The following works, but it doesn't seem like it's a sensible approach using numpy:

values = [3, 1, 8]
indices = [1, 2, 4]
for i, v in zip(indices, values):
    row[i] = v

Beside the point, but row is not defined. Looks like it's supposed to be row = x[0]. — wjandrea
– wjandrea, Commented Mar 22, 2022 at 21:32

oBObo2BI · Accepted Answer · 2022-03-22 21:43:08Z

4

You can use the put method from numpy to achieve this:

np.put(x, indices, values)

answered Mar 22, 2022 at 21:43

oBObo2BI

842 bronze badges

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

Comments

user17242583 · Accepted Answer · 2022-03-22 21:34:29Z

3

Just move the indexer to the second position (because you're modifying values in the second dimension):

x[:, indices] = values

Output:

>>> x
array([[0., 3., 1., 0., 8.]])

answered Mar 22, 2022 at 21:34

user17242583