Reshaping a numpy array in a function doesn't work

Question

... but changing the values of the numpy array works:

import numpy as np

def reshapeArray(arr):
    arr = arr.reshape((2, 2))
    arr /= 10
    print(arr) # prints [[0.1 0.3], [0.2 0.4]]

arr = np.array([1, 2, 3, 4], dtype=np.float32)
reshapeArray(arr)
print(arr) # prints [0.1 0.2 0.3 0.4]

The reshapeArray() function changed values of the array permanently, but changed the shape of the array temporarily. If I add a return line (return arr) to end of the function, and assign the output of the function to the array (arr = reshapeArray(arr)), this time it works. But I wonder why it didn't work without returning the array?

Itamar Mushkin · Accepted Answer · 2019-08-20 09:54:43Z

3

From the docs (numpy.reshape):

This will be a new view object if possible; otherwise, it will be a copy. Note there is no guarantee of the memory layout (C- or Fortran- contiguous) of the returned array.

As opposed to arr = arr / 10, which does makes a copy and reassigns it.

Apparently a view is lost when leaving scope...

answered Aug 20, 2019 at 9:54

Itamar Mushkin

2,9232 gold badges19 silver badges34 bronze badges

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

1 Comment

hpaulj Over a year ago

The arr=... breaks the back reference. That's basic python function variable scoping. The view is a new object.

Peter Nazarenko · Accepted Answer · 2019-08-20 10:05:50Z

0

Try to return an array from function and assign the returned value to the desired variable:

    return arr  # The last string of your function

arr = reshapeArray(arr)

answered Aug 20, 2019 at 10:05

Peter Nazarenko

4111 gold badge7 silver badges16 bronze badges

1 Comment

Itamar Mushkin Over a year ago

Note that OP explicitly said that's something they've tried (and know that it works) but would like to know why this works.

Collectives™ on Stack Overflow

Reshaping a numpy array in a function doesn't work

2 Answers 2

1 Comment

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related