Repository navigation
Possible conversion bug with record array #3351
Description
I got a weird output from the following script:
import numpy as np
U = np.zeros(1, dtype=[('x', np.float32, (4,4))])
# Problematic
U[0] = np.eye(4)
print U[0]
# Works as expected
U[0] = np.eye(4, dtype=np.float32)
U['x'][0] = np.eye(4)
Output:
([[0.0, 1.875, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 1.875], [0.0, 0.0, 0.0, 0.0]],)
Tested on numpy 1.7.1, confirmed on 1.6.0.
As noted by Chirs Barker on the mailing list, U is a rank-zero array, not a 4X4 array.
Activity
njsmith commentedon May 22, 2013
U[0] is a void scalar, and this scalar contains a 4x4 array, so that's the
explanation for that mystery.
The behaviour's still a bug though.
On Wed, May 22, 2013 at 7:28 PM, Nicolas Rougier
[email protected]:
seberg commentedon May 23, 2013
It goes down to direct buffers. To understand the result try:
np.eye(4).view('float32'). Directly interpreting buffers is probably wrong though when types are known...MAINT: improve struct assignment, allow multi fields views
MAINT: improve struct assignment, allow multi fields views