Describe the issue:

np.version = 2.3.2

GE.shape = (2, 3, 2) <- 3rd (numeric) rank tensor
E = [False, True]

depending on indexing syntax produces inconsistent result
GE[0,:,E] = [[0 0 1]]
GE[0][:,E] = [[0]
[0]
[1]]

has shape error: GE[0,:,E].shape = (1, 3)
vs correct shape via GE[0][:,E].shape = (3, 1)

Reproduce the code example:

import numpy as np
print("np.__version__ = ", np.__version__)
GE = np.moveaxis( np.asarray([((0,0,0), (0,0,2)), ((0,0,1), (0,1,0)), ]), 0, 2 )
E  = [False, True]
print("GE.shape = ", GE.shape)
print("E  = ", E)
print()
print("GE[0,:,E] = ", GE[0,:,E])
print("GE[0][:,E] = ", GE[0][:,E])
print(f"has shape error: GE[0,:,E].shape = {GE[0,:,E].shape}  vs correct shape via  GE[0][:,E].shape = {GE[0][:,E].shape}")

Error message:

there is no error message, just the result is wrong(?)

Python and NumPy Versions:

2.3.2
3.12.3 (main, Jun 18 2025, 17:59:45) [GCC 13.3.0]

Runtime Environment:

ubuntu 24.04.2 LTS (<- cat /etc/os-release )
Python 3.12.3 (<- python --version )

Context for the issue:

obtain unexpected wrong shape of matrix in calculation