Re: [Numpy-discussion] Intel MKL - was: parallel numpy - any info?

Hello On Jan 10, 2008 6:56 AM, David Cournapeau <[EMAIL PROTECTED]> wrote: > The one thing which I am not sure about is: say one MKL binary does not > work, and say I (or anyone outside your company) build numpy with the > MKL ro debug it, can I redistribute a new binary, even if it is just for >

Re: [Numpy-discussion] BzrMirror page on numpy wiki

David, Thanks, that's great. I am curious: how did you do the conversion? Did you work directly from the existing svn repo or did you make a local mirror first? Eric David Cournapeau wrote: > Hi, > > I started a page on how to use bzr, with a link to an import I have > done on the numpy

[Numpy-discussion] subclassing matrix

Hello folks, In the course of a project that involved heavy use of geometry and linear algebra, I found it useful to create a Vector subclass of numpy.matrix (represented as a column vector in my case). I'd like to hear comments about my use of this "class promotion" statement in __new__: ret

[Numpy-discussion] BzrMirror page on numpy wiki

Hi, I started a page on how to use bzr, with a link to an import I have done on the numpy trunk: http://scipy.org/scipy/numpy/wiki/BzrMirror I thought it would be useful for people not familiar with DVCS, and are curious to try. In perticular, since I have done the svn->bzr import, the di

Re: [Numpy-discussion] Intel MKL - was: parallel numpy - any info?

Ray Schumacher wrote: > At 10:38 PM 1/8/2008, Albert Strasheim wrote: > > >>> It is still unclear to me whether Python/numpy compiled with MKL >>> would be freely re-distributable, as the MSVC version is. >>> >> Read the License Agreement on Intel's site. My interpretation is that >> it w

Re: [Numpy-discussion] Does float16 exist?

On Jan 8, 2008 11:26 PM, David Cournapeau <[EMAIL PROTECTED]> wrote: If what you want to do is to be able to use simple templates (by simple, > I mean template function parametrized with the numerical type), C++ is > overkill IMHO. > So I pulled the template subsystem out of Django

Re: [Numpy-discussion] numpy.linalg.eigvals crashes whn calling lapack_lite.pyd

Charles R Harris gmail.com> writes: [snip] > > Which specific version of numpy for 2.5 did you use? Can you try one of the other versions on the sourceforge site?Chuck > > I have just tried: numpy-1.0.3.1.win32-py2.5.exe and it works! I was using: numpy-1.0.4.win32-py2.5.msi Have no idea

Re: [Numpy-discussion] Failing to understand vectorize behavior

otypes needs to be set for this to work. In [19]: vstrip = vectorize( string.strip, otypes=[object]) In [20]: s = [' aa ' , ' bb ', ' cc '] In [21]: vstrip(s) Out[21]: array([aa, bb, cc], dtype=object)

Re: [Numpy-discussion] Failing to understand vectorize behavior

Yes, 32 bits. On a 64 bits machine, I get 8 characters long strings like you. L. On 1/9/08, David Huard <[EMAIL PROTECTED]> wrote: > > Lorenzo, > > 2008/1/9, lorenzo bolla <[EMAIL PROTECTED]>: > > > > I don't think it's expected: mine are cropped to 4 characters! > > > > > I am on a 64 bit machin

Re: [Numpy-discussion] Does float16 exist?

On Jan 9, 2008, at 00:00 , Robert Kern wrote: > Charles R Harris wrote: > >> I see that there are already a number of parsers available for >> Python, >> SPARK, for instance is included in the 2.5.1 distribution. > > No, it isn't. It's used to generate the new parser in 2.5 (Parser/spark.py).

Re: [Numpy-discussion] Initial Hg experiences on Windows

David Cournapeau wrote: > > I think that at least wrt windows support, bzr is today better than > hg, if only for the fact that some bzr developers are primely windows > developers, with the "windows culture". bzr GUI - olive works great on windows Niki Spahiev _

[Numpy-discussion] Intel MKL - was: parallel numpy - any info?

At 10:38 PM 1/8/2008, Albert Strasheim wrote: > > It is still unclear to me whether Python/numpy compiled with MKL > > would be freely re-distributable, as the MSVC version is. > >Read the License Agreement on Intel's site. My interpretation is that >it would be redistributable. > >http://www.inte

Re: [Numpy-discussion] Failing to understand vectorize behavior

Lorenzo, 2008/1/9, lorenzo bolla <[EMAIL PROTECTED]>: > > I don't think it's expected: mine are cropped to 4 characters! > I am on a 64 bit machine. Are you on a 32 bit one ? In [101]: vstrip = vectorize( string.strip) > In [102]: s = [' aa ' , ' bb ', ' > c

Re: [Numpy-discussion] Failing to understand vectorize behavior

I don't think it's expected: mine are cropped to 4 characters! In [101]: vstrip = vectorize( string.strip) In [102]: s = [' aa ' , ' bb ', ' cc '] In [103]: vstrip(s) Out[103]: array(['', '', ''], dtype='|S4') You can obviously use "

[Numpy-discussion] Failing to understand vectorize behavior

Hi all, I'm having trouble understanding the behavior of vectorize on the following example: >>> import string >>> from numpy import vectorize >>> vstrip = vectorize(string.strip) >>> s = [' aa ' , ' bb ', ' cc '] >>> vstrip(s) array(['', 'b

Re: [Numpy-discussion] Does float16 exist?

On Jan 9, 2008 6:00 AM, Matthieu Brucher <[EMAIL PROTECTED]> wrote: > > One would naturally have to be very careful about which features to > > > use. In particular, don't use anything that throws, and I suppose one > > > would want to avoid having to link to the stdc++ library also. I don't > > >

Re: [Numpy-discussion] Does float16 exist?

> > > One would naturally have to be very careful about which features to > > use. In particular, don't use anything that throws, and I suppose one > > would want to avoid having to link to the stdc++ library also. I don't > > know if that is possible. > Avoiding linking to the stdc++ would be quit

Re: [Numpy-discussion] Does float16 exist?

Charles R Harris wrote: > > > On Jan 8, 2008 11:26 PM, David Cournapeau > <[EMAIL PROTECTED] > > wrote: > > Charles R Harris wrote: > > > > > > > > > > The C++ name mangling can be worked around. > name mangling is just the top of the iceberg. There

Re: [Numpy-discussion] numpy.linalg.eigvals crashes whn calling lapack_lite.pyd

On Jan 9, 2008 12:24 AM, Simon <[EMAIL PROTECTED]> wrote: > Charles R Harris gmail.com> writes: > > > > > > > On Jan 8, 2008 6:49 PM, Simon gmail.com> wrote: > > Newbie here. Trying to generate eigenvalues from a matrix using:print > numpy.linalg.eigvals(matrix)This works with small matrices, sa

Re: [Numpy-discussion] Does float16 exist?

On Jan 8, 2008 11:26 PM, David Cournapeau <[EMAIL PROTECTED]> wrote: > Charles R Harris wrote: > > > > > > > > The C++ name mangling can be worked around. > name mangling is just the top of the iceberg. There are problems wrt to > static initialization, exception, etc...; ABI compatibility is mu

Re: [Numpy-discussion] def of var of complex

Robert Kern wrote: > Travis E. Oliphant wrote: > >> Robert Kern wrote: >> >>> Neal Becker wrote: >>> >>> I noticed that if I generate complex rv i.i.d. with var=1, that numpy says: var () -> (close to 1.0) var () -> (close to 1.0) but var (c