numpy array inconsistency

Reporting a bug

• I have tried using the latest released version of Numba (most recent is visible in the change log (https://github.com/numba/numba/blob/main/CHANGE_LOG).
• I have included a self contained code sample to reproduce the problem. i.e. it’s possible to run as ‘python bug.py’.

The following function works fine outside of numba:

def myfun():
    y=np.zeros(10)
    y[0]=np.array([3])
    return y

However it fails with numba, as shown here:

@jit(nopython=True)
def myfun():
    y=np.zeros(10)
    y[0]=np.array([3])
    return y

But can be made to work using:

@jit(nopython=True)
def myfun():
    y=np.zeros(10)
    y[0]=np.array([3])[0]
    return y

This is awkward behaviour - would be nice if it could be solved. Thanks!

(I am using python 3.10 in conda, with numpy v 1.21.2 and numba v 0.55.0 which is the latest version available in conda). Use the following import statements:

import numpy as np
from numba import jit

[esc]

@amireson I have labelled it accordingly until such a time as a champion appears to fix it.

[stuartarchibald]

Further note. I think a way to write this that will work for any dimension single value array is using np.take:

from numba import njit
import numpy as np

@njit
def foo(x):
    y = np.zeros(4)
    y[0] = np.take(x, 0)
    return y

tmp = np.array(3)
for i in range(5):
    print(tmp)
    print(foo(tmp))
    tmp = np.expand_dims(tmp, 0)