headers
Travis Oliphant | 1 Apr 05:54
Favicon

Release of NumPy 1.0.2 for Monday


I'm going to be tagging the tree for the NumPy 1.0.2 release tomorrow 
evening in preparation for the release on Monday.  I've closed several 
bugs.  Are there any show-stoppers remaining to be fixed?

-Travis
Permalink | Reply | 2 comments |
headers
Travis Vaught | 1 Apr 16:22

Re: organizational question

I was just about to respond that _substantial_ support by a non-profit best describes Travis Oliphant's many contributions through BYU.  I see now that BYU/SciPy/NumPy has been nominated ... excellent!

It might be appropriate for people from various constituencies to comment...

"Preference will be given to projects that benefit more and larger constituencies, demonstrate exceptional promise or performance, and meet particularly urgent needs."


Best,

Travis (Vaught)

On Mar 30, 2007, at 2:29 PM, Alan G Isaac wrote:

Is either NumPy or SciPy substantially supported
by an identifiable and actual non-profit organization?

I ask because I think both fit under
item 4.

Here is the announcement:

Note that universities are among the nominees:

Cheers,
Alan Isaac



_______________________________________________
Numpy-discussion mailing list


_______________________________________________
Numpy-discussion mailing list
Numpy-discussion <at> scipy.org
http://projects.scipy.org/mailman/listinfo/numpy-discussion
Permalink | Reply | 0 comments |
headers
Christopher Hanley | 1 Apr 19:58

test_multiarray.test_clip fails on Solaris 8 system

The following test fails on a Solaris 8 system:

======================================================================
FAIL: check_basic (numpy.core.tests.test_multiarray.test_clip)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/data/basil5/site-packages/lib/python/numpy/core/tests/test_multiarray.py", line 388, in check_basic
    self._clip_type('float',1024,-12.8,100.2, inplace=inplace)
  File "/data/basil5/site-packages/lib/python/numpy/core/tests/test_multiarray.py", line 382, in _clip_type
    assert_equal(byteorder,x.dtype.byteorder)
  File "/data/basil5/site-packages/lib/python/numpy/testing/utils.py", line 143, in assert_equal
    assert desired == actual, msg
AssertionError: 
Items are not equal:
 ACTUAL: '<'
 DESIRED: '='

----------------------------------------------------------------------
Ran 562 tests in 19.376s

FAILED (failures=1)
Permalink | Reply | 6 comments |
headers
Charles R Harris | 1 Apr 20:14
Picon

Re: test_multiarray.test_clip fails on Solaris 8 system



On 4/1/07, Christopher Hanley <chanley <at> stsci.edu> wrote:
The following test fails on a Solaris 8 system:

======================================================================
FAIL: check_basic (numpy.core.tests.test_multiarray.test_clip)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/data/basil5/site-packages/lib/python/numpy/core/tests/test_multiarray.py", line 388, in check_basic
    self._clip_type('float',1024,-12.8,100.2, inplace=inplace)
  File "/data/basil5/site-packages/lib/python/numpy/core/tests/test_multiarray.py", line 382, in _clip_type
    assert_equal(byteorder,x.dtype.byteorder)
  File "/data/basil5/site-packages/lib/python/numpy/testing/utils.py", line 143, in assert_equal
    assert desired == actual, msg
AssertionError:
Items are not equal:
ACTUAL: '<'
DESIRED: '='

----------------------------------------------------------------------
Ran 562 tests in 19.376s

FAILED (failures=1)


_

Hmm, Sun hardware is big endian, no? I wonder what happens on PPC? I don't see any problems here on Athlon64.

Chuck


_______________________________________________
Numpy-discussion mailing list
Numpy-discussion <at> scipy.org
http://projects.scipy.org/mailman/listinfo/numpy-discussion
Permalink | Reply | 5 comments |
headers
Christopher Hanley | 1 Apr 20:28

Re: test_multiarray.test_clip fails on Solaris 8 system

Sun hardware is big endian.  To be specific, this test was done on a Sun 
Ultra 10.  I don't have access to a PPC right now.  I can check tomorrow 
once I am in the office.

Chris

>
> Hmm, Sun hardware is big endian, no? I wonder what happens on PPC? I 
> don't see any problems here on Athlon64.
>
> Chuck
>
Permalink | Reply | 3 comments |
headers
Francesc Altet | 1 Apr 21:35

Re: Release of NumPy 1.0.2 for Monday

El ds 31 de 03 del 2007 a les 21:54 -0600, en/na Travis Oliphant va
escriure:
> I'm going to be tagging the tree for the NumPy 1.0.2 release tomorrow 
> evening in preparation for the release on Monday.  I've closed several 
> bugs.  Are there any show-stoppers remaining to be fixed?

Mmm... PyTables test suite segfaults with latest SVN. The problem is
described in #487 and I'm afraid that it is a clear stopper for us.
Sorry for running the PyTables test suite so late before the new NumPy
release.

Also, I've found a patch for #483. Although it is not critical for us,
it would be nice if it can be added to NumPy 1.0.2 final.

Good night!

--

-- 
Francesc Altet    |  Be careful about using the following code --
Carabos Coop. V.  |  I've only proven that it works, 
www.carabos.com   |  I haven't tested it. -- Donald Knuth
Permalink | Reply | 1 comment |
headers
Charles R Harris | 1 Apr 22:54
Picon

Violation of array scalar multiplication rules?

Just asking.

In [35]: type(array(1.0)*2)
Out[35]: <type 'numpy.float64'>

In [36]: type(array(1.0))
Out[36]: <type 'numpy.ndarray'>

Chuck

_______________________________________________
Numpy-discussion mailing list
Numpy-discussion <at> scipy.org
http://projects.scipy.org/mailman/listinfo/numpy-discussion
Permalink | Reply | 1 comment |
headers
Travis Oliphant | 1 Apr 23:16
Favicon

Re: Violation of array scalar multiplication rules?

Charles R Harris wrote:
> Just asking.
>
> In [35]: type(array(1.0)*2)
> Out[35]: <type 'numpy.float64'>
>
> In [36]: type(array(1.0))
> Out[36]: <type 'numpy.ndarray'>

No,  in ufuncs 0-d arrays are considered scalars, as are Python scalars 
and array scalars.

Also, ufuncs that result in scalars return NumPy scalars.

-Travis
Permalink | Reply | 0 comments |
headers
Stefan van der Walt | 1 Apr 23:22
Picon
Picon
Favicon

Re: test_multiarray.test_clip fails on Solaris 8 system

Hi Chris

Would you please run the following commands and show their output?

import sys
print sys.byteorder

import numpy as N
print N.array([1,2,3],N.dtype(N.int16).newbyteorder('<')).dtype.byteorder
print N.array([1,2,3],N.dtype(N.int16).newbyteorder('>')).dtype.byteorder
print N.array([1,2,3],N.dtype(N.int16).newbyteorder('=')).dtype.byteorder

Output on my little-endian system is

little
<
>
=

and I'd be curious to see if the output on a big-endian system follows
the same pattern.

I'd expect

big
<
>
=

Cheers
Stéfan
Permalink | Reply | 2 comments |
headers
Christopher Hanley | 2 Apr 01:03

Re: test_multiarray.test_clip fails on Solaris 8 system

Hi Stefan,

This is what I get:
 >>> import sys
 >>> print sys.byteorder
big
 >>> import numpy as N
 >>> print 
N.array([1,2,3],N.dtype(N.int16).newbyteorder('<')).dtype.byteorder
<
 >>> print 
N.array([1,2,3],N.dtype(N.int16).newbyteorder('>')).dtype.byteorder
 >
 >>> print 
N.array([1,2,3],N.dtype(N.int16).newbyteorder('=')).dtype.byteorder
=
 >>>

Stefan van der Walt wrote:
> Hi Chris
>
> Would you please run the following commands and show their output?
>
> import sys
> print sys.byteorder
>
> import numpy as N
> print N.array([1,2,3],N.dtype(N.int16).newbyteorder('<')).dtype.byteorder
> print N.array([1,2,3],N.dtype(N.int16).newbyteorder('>')).dtype.byteorder
> print N.array([1,2,3],N.dtype(N.int16).newbyteorder('=')).dtype.byteorder
>
> Output on my little-endian system is
>
> little
> <
>   
> =
>
> and I'd be curious to see if the output on a big-endian system follows
> the same pattern.
>
> I'd expect
>
> big
> <
>   
> =
>
> Cheers
> Stéfan
> _______________________________________________
> Numpy-discussion mailing list
> Numpy-discussion <at> scipy.org
> http://projects.scipy.org/mailman/listinfo/numpy-discussion
>
Permalink | Reply | 0 comments |

Gmane