No implementation of function Function getitem?

Numba Support: How do I do ...?
1

Hello,
I’m using version “0.55.1”.
I have a series of 3D points, that I need to digitize along the 2 first axes (X,Y) to produce 2D images as for each pixel one image is the sum over the Z and the other counts the number of points belonging to that pixel.

The regular numpy is ok. I had in mind to try to speed the for loop using numba although I am newby with numba so I may have done it wrong.

@nb.jit(nb.types.Tuple((nb.float32[:,:], nb.uint32[:,:]))(nb.float32[:,:,:],nb.float32[:,:], nb.uint32[:,:]))
def digitpts_and_updateimg(pts,img_sum,img_counts):
    pts_x_id = np.digitize(pts[0,:],binsx).astype(np.uint32)
    pts_y_id = np.digitize(pts[1,:],binsy).astype(np.uint32)
    pts_z = pts[2,:]
    for i,j,z in zip(pts_x_id,pts_y_id,pts_z):
        img_sum[j,i] += z
        img_counts[j,i] += 1
    return img_sum, img_counts

The error is the following:

 NumbaWarning: 
Compilation is falling back to object mode WITH looplifting enabled because Function "digitpts_and_updateimg" failed type inference due to: No implementation of function Function(<built-in function getitem>) found for signature:
 
 >>> getitem(array(float32, 2d, A), UniTuple(array(uint32, 1d, A) x 2))
 
There are 22 candidate implementations:
   - Of which 20 did not match due to:
   Overload of function 'getitem': File: <numerous>: Line N/A.
     With argument(s): '(array(float32, 2d, A), UniTuple(array(uint32, 1d, A) x 2))':
    No match.
   - Of which 2 did not match due to:
   Overload in function 'GetItemBuffer.generic': File: numba/core/typing/arraydecl.py: Line 166.
     With argument(s): '(array(float32, 2d, A), UniTuple(array(uint32, 1d, A) x 2))':
    Rejected as the implementation raised a specific error:
      NumbaNotImplementedError: only one advanced index supported
....

Any idea?

2

Is it related to this?

3

@jecampagne ,
binsx and binsy are not defined.
Do you have a working example?