Closed
Description
Hiya!
Using:
Numpy version 1.10.2
Python 3.5
Mac OSX 10.10.3
This occurs when:
- First axis is sliced with a single index, and
- Any axis past the 2nd axis is indexed with a list
Self contained example:
A = np.zeros((2,3,4))
print(A[0, :, 0:2].shape) # (3,2) As one would expect
print(A[0, :, [0,1]].shape) # (2,3) UnexpectedI've inspected the values afterwards to confirm that it's a simple transposition that's occurring (no weird remapping of data). Also occurs for arrays with higher dimensions.
Metadata
Metadata
Assignees
Labels
No labels