Incrementing array location with rank 0 array

[aseyboldt]

On current main I can not increment a location of an array with a rank 0 value:

import numba

@numba.njit
def inc_array(x, val):
    x[0] += val

inc_array(np.zeros(5), np.array(2.))

TypingError: Failed in nopython mode pipeline (step: nopython frontend)
No implementation of function Function(<built-in function iadd>) found for signature:
 
 >>> iadd(float64, array(float64, 0d, C))
 
There are 18 candidate implementations:
      - Of which 16 did not match due to:
      Overload of function 'iadd': File: <numerous>: Line N[/A.](http://localhost:7890/A.)
        With argument(s): '(float64, array(float64, 0d, C))':
       No match.
      - Of which 2 did not match due to:
      Operator Overload in function 'iadd': File: unknown: Line unknown.
        With argument(s): '(float64, array(float64, 0d, C))':
       No match for registered cases:
        * (int64, int64) -> int64
        * (int64, uint64) -> int64
        * (uint64, int64) -> int64
        * (uint64, uint64) -> uint64
        * (float32, float32) -> float32
        * (float64, float64) -> float64
        * (complex64, complex64) -> complex64
        * (complex128, complex128) -> complex128

During: typing of intrinsic-call at [/tmp/ipykernel_359313/84858975.py](http://localhost:7890/tmp/ipykernel_359313/84858975.py) (5)

File "../../../../tmp/ipykernel_359313/84858975.py", line 5:
<source missing, REPL[/exec](http://localhost:7890/exec) in use?>

If I call it with a scalar value instead, it work:

inc_array(np.zeros(5), 2.)

[guilhermeleobas]

Thanks for the report @aseyboldt, can confirm the bug.

[guilhermeleobas]

Fix should be similar to what was done in #4516

[gmarkall]

As a workaround I think you can write

@numba.njit
def inc_array(x, val):
    x[()] += val

[aseyboldt]

Do you mean

@numba.njit
def inc_array(x, val):
    x[0] += val[()]

because otherwise this will increment all locations in x, not just the first.

[gmarkall]

Sorry, I think I mean something like that, yes - I'd come across this issue before of having to use variable[()] for the value of a rank 0 array, but I never realised this wasn't the way it was supposed to be.

[ricardoV94]

If it helps, this is only a problem when the lhs is a vector (and a number after indexing). It's fine when it's a matrix (and a vector after indexing)