[Numpy-discussion] Possible conversion bug with record array

2013-05-22 Thread Nicolas Rougier 


Hi all,

I got a weird output from the following script:

import numpy as np

U = np.zeros(1, dtype=[('x', np.float32, (4,4))])

U[0] = np.eye(4)
print U[0]
# 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]],)

U[0] = np.eye(4, dtype=np.float32)
print U[0]
# output:  ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0], 
[0.0, 0.0, 0.0, 1.0]],)


The first output is obviously wrong. Can anyone confirm ?
(using numpy 1.7.1 on osx 10.8.3)


Nicolas
___
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Possible conversion bug with record array

2013-05-22 Thread Chris Barker - NOAA Federal 
On Wed, May 22, 2013 at 10:07 AM, Nicolas Rougier
 U = np.zeros(1, dtype=[('x', np.float32, (4,4))])

 U[0] = np.eye(4)
 print U[0]
 # 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]],)

I get the same thing. Note:

In [86]: U[0].shape
Out[86]: ()

it's a rank-zero array, not a 4X4 array -- some odd tings can happen there...

This works:

In [94]: U = np.zeros(1, dtype=[('x', np.float32, (4,4))])

In [95]: U['x'][0] = np.eye(4)

In [96]: U
Out[96]:
array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0,
0.0], [0.0, 0.0, 0.0, 1.0]],)],
  dtype=[('x', 'f4', (4, 4))])

Not that the first version isn't a bug!


-- 

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/ORR(206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115   (206) 526-6317   main reception

chris.bar...@noaa.gov
___
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Possible conversion bug with record array

2013-05-22 Thread eat 
Hi,

FWIW, apparently bug related to dtype of np.eye(.)


On Wed, May 22, 2013 at 8:07 PM, Nicolas Rougier
nicolas.roug...@inria.frwrote:



 Hi all,

 I got a weird output from the following script:

 import numpy as np

 U = np.zeros(1, dtype=[('x', np.float32, (4,4))])

 U[0] = np.eye(4)
 print U[0]
 # 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]],)

 U[0] = np.eye(4, dtype=np.float32)
 print U[0]
 # output:  ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0,
 0.0], [0.0, 0.0, 0.0, 1.0]],)


 The first output is obviously wrong. Can anyone confirm ?
 (using numpy 1.7.1 on osx 10.8.3)

In []: sys.version
Out[]: '2.7.2 (default, Jun 12 2011, 15:08:59) [MSC v.1500 32 bit (Intel)]'
In []: np.__version__
Out[]: '1.6.0'
In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])

In []: U[0]= np.eye(4)
In []: U
Out[]:
array([ ([[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]],)],
  dtype=[('x', 'f4', (4, 4))])

In []: U[0]= np.eye(4, dtype= np.float32)
In []: U
Out[]:
array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0],
[0.0, 0.0, 0.0, 1.0]],)],
  dtype=[('x', 'f4', (4, 4))])

and
In []: sys.version
Out[]: '3.3.0 (v3.3.0:bd8afb90ebf2, Sep 29 2012, 10:57:17) [MSC v.1600 64
bit (AMD64)]'
In []: np.__version__
Out[]: '1.7.0rc1'
In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])

In []: U[0]= np.eye(4)
In []: U
Out[17]:
array([ ([[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]],)],
  dtype=[('x', 'f4', (4, 4))])

U[0]= np.eye(4, dtype= np.float32)

In []: U
Out[]:
array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0],
[0.0, 0.0, 0.0, 1.0]],)],
  dtype=[('x', 'f4', (4, 4))])


My 2 cents,
-eat



 Nicolas
 ___
 NumPy-Discussion mailing list
 NumPy-Discussion@scipy.org
 http://mail.scipy.org/mailman/listinfo/numpy-discussion

___
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Possible conversion bug with record array

2013-05-22 Thread Nicolas Rougier 


Thanks, I filed a new issue on the bug tracker.

Nicolas




On May 22, 2013, at 8:15 PM, eat e.antero.ta...@gmail.com wrote:

 Hi,
 
 FWIW, apparently bug related to dtype of np.eye(.)
 
 
 On Wed, May 22, 2013 at 8:07 PM, Nicolas Rougier nicolas.roug...@inria.fr 
 wrote:
 
 
 Hi all,
 
 I got a weird output from the following script:
 
 import numpy as np
 
 U = np.zeros(1, dtype=[('x', np.float32, (4,4))])
 
 U[0] = np.eye(4)
 print U[0]
 # 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]],)
 
 U[0] = np.eye(4, dtype=np.float32)
 print U[0]
 # output:  ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 
 0.0], [0.0, 0.0, 0.0, 1.0]],)
 
 
 The first output is obviously wrong. Can anyone confirm ?
 (using numpy 1.7.1 on osx 10.8.3)
 In []: sys.version
 Out[]: '2.7.2 (default, Jun 12 2011, 15:08:59) [MSC v.1500 32 bit (Intel)]'
 In []: np.__version__
 Out[]: '1.6.0'
 In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])
 
 In []: U[0]= np.eye(4)
 In []: U
 Out[]: 
 array([ ([[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]],)], 
   dtype=[('x', 'f4', (4, 4))])
 
 In []: U[0]= np.eye(4, dtype= np.float32)
 In []: U
 Out[]: 
 array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0], 
 [0.0, 0.0, 0.0, 1.0]],)], 
   dtype=[('x', 'f4', (4, 4))])
 
 and 
 In []: sys.version
 Out[]: '3.3.0 (v3.3.0:bd8afb90ebf2, Sep 29 2012, 10:57:17) [MSC v.1600 64 bit 
 (AMD64)]'
 In []: np.__version__
 Out[]: '1.7.0rc1'
 In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])
 
 In []: U[0]= np.eye(4)
 In []: U
 Out[17]: 
 array([ ([[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]],)], 
   dtype=[('x', 'f4', (4, 4))])
 
 U[0]= np.eye(4, dtype= np.float32)
 
 In []: U
 Out[]: 
 array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0], 
 [0.0, 0.0, 0.0, 1.0]],)], 
   dtype=[('x', 'f4', (4, 4))]) 
 
 
 My 2 cents,
 -eat
 
 
 Nicolas
 ___
 NumPy-Discussion mailing list
 NumPy-Discussion@scipy.org
 http://mail.scipy.org/mailman/listinfo/numpy-discussion
 
 ___
 NumPy-Discussion mailing list
 NumPy-Discussion@scipy.org
 http://mail.scipy.org/mailman/listinfo/numpy-discussion

___
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Possible conversion bug with record array

2013-05-22 Thread Chris Barker - NOAA Federal 
On Wed, May 22, 2013 at 11:15 AM, eat e.antero.ta...@gmail.com wrote:

 FWIW, apparently bug related to dtype of np.eye(.)

sort of -- the issue shows up when assigning a float64 array (default
for eye()) to a rank-0 array with a custom dtype that has a single
object filed that is an arraynumpy should do the right thing
there. If you pull the arra yout of the custom dtype object, it does
do the right thing. See the example I posted earlier.

-Chris



 On Wed, May 22, 2013 at 8:07 PM, Nicolas Rougier nicolas.roug...@inria.fr
 wrote:



 Hi all,

 I got a weird output from the following script:

 import numpy as np

 U = np.zeros(1, dtype=[('x', np.float32, (4,4))])

 U[0] = np.eye(4)
 print U[0]
 # 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]],)

 U[0] = np.eye(4, dtype=np.float32)
 print U[0]
 # output:  ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0,
 0.0], [0.0, 0.0, 0.0, 1.0]],)


 The first output is obviously wrong. Can anyone confirm ?
 (using numpy 1.7.1 on osx 10.8.3)

 In []: sys.version
 Out[]: '2.7.2 (default, Jun 12 2011, 15:08:59) [MSC v.1500 32 bit (Intel)]'
 In []: np.__version__
 Out[]: '1.6.0'
 In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])

 In []: U[0]= np.eye(4)
 In []: U
 Out[]:
 array([ ([[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]],)],
   dtype=[('x', 'f4', (4, 4))])

 In []: U[0]= np.eye(4, dtype= np.float32)
 In []: U
 Out[]:
 array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0],
 [0.0, 0.0, 0.0, 1.0]],)],
   dtype=[('x', 'f4', (4, 4))])

 and
 In []: sys.version
 Out[]: '3.3.0 (v3.3.0:bd8afb90ebf2, Sep 29 2012, 10:57:17) [MSC v.1600 64
 bit (AMD64)]'
 In []: np.__version__
 Out[]: '1.7.0rc1'
 In []: U= np.zeros(1, dtype= [('x', np.float32, (4, 4))])

 In []: U[0]= np.eye(4)
 In []: U
 Out[17]:
 array([ ([[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]],)],
   dtype=[('x', 'f4', (4, 4))])

 U[0]= np.eye(4, dtype= np.float32)

 In []: U
 Out[]:
 array([ ([[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 0.0],
 [0.0, 0.0, 0.0, 1.0]],)],
   dtype=[('x', 'f4', (4, 4))])


 My 2 cents,
 -eat



 Nicolas
 ___
 NumPy-Discussion mailing list
 NumPy-Discussion@scipy.org
 http://mail.scipy.org/mailman/listinfo/numpy-discussion



 ___
 NumPy-Discussion mailing list
 NumPy-Discussion@scipy.org
 http://mail.scipy.org/mailman/listinfo/numpy-discussion




-- 

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/ORR(206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115   (206) 526-6317   main reception

chris.bar...@noaa.gov
___
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion