Fill numpy array with other numpy array

Question

I have following numpy arrays:

 whole = np.array(
 [1, 0, 3, 0, 6]
 )
 sparse = np.array(
 [9, 8]
 )

Now I want to replace every zero in the whole array in chronological order with the items in the sparse array. In the example my desired array would look like:

 merged = np.array(
 [1, 9, 3, 8, 6]
 )

I could write a small algorithm by myself to fix this but if someone knows a time efficient way to solve this I would be very grateful for you help!

• willeM_ Van Onsem: Are you sure the number of 0s match with the sparse?
• FObersteiner: ...since you could just write whole[whole==0] = sparse then
• Adrian: Yes! Awesome, that was exactly what I was looking for. Thank you very much.

Answer 1

Do you assume that sparse has the same length as there is zeros in whole ?

If so, you can do:

import numpy as np
from copy import copy
whole = np.array([1, 0, 3, 0, 6])
sparse = np.array([9, 8])
merge = copy(whole)
merge[whole == 0] = sparse

if the lengths mismatch, you have to restrict to the correct length using len(...) and slicing.

• Nathan e: I was outdated by the comment above, but my method doesn't change 'whole' array in place.
• Kevin: Would it be inplace without the copy? I am a little confused about the purpose of the copy.
• user3483203: You should use numpy's copy as opposed to copy.copy
• Nathan e: To Kevin : if I don't use copy, when I update 'merge', the original array 'whole' is modified the same way'. So one has to copy the numpy array to keep the original data