Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread Bruce Southey
On Thu, Mar 26, 2009 at 9:32 AM, David Cournapeau wrote: > Bruce Southey wrote: >> Hi, >> Apparently not: >> >> Python 2.6.1 (r261:67517, Dec  4 2008, 16:51:00) [MSC v.1500 32 bit >> (Intel)] on win32 >> > > Well, installing 64 bits numpy on 32 bits python will not work very well :) > > I am surpr

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread David Cournapeau
Bruce Southey wrote: > Hi, > Apparently not: > > Python 2.6.1 (r261:67517, Dec 4 2008, 16:51:00) [MSC v.1500 32 bit > (Intel)] on win32 > Well, installing 64 bits numpy on 32 bits python will not work very well :) I am surprised the installation worked at all (I noticed msi were less robust t

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread Bruce Southey
On Thu, Mar 26, 2009 at 8:53 AM, David Cournapeau wrote: > Bruce Southey wrote: >> David Cournapeau wrote: >> >>> On Thu, Mar 26, 2009 at 7:48 PM, David Cournapeau >>> wrote: >>> >>> >>> It is built with an updated toolchain + a few patches to mingw I have yet submitted upstream,

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread David Cournapeau
Bruce Southey wrote: > David Cournapeau wrote: > >> On Thu, Mar 26, 2009 at 7:48 PM, David Cournapeau wrote: >> >> >> >>> It is built with an updated toolchain + a few patches to mingw I have >>> yet submitted upstream, >>> >>> >> I created a ticket as well to track this iss

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread Bruce Southey
David Cournapeau wrote: > On Thu, Mar 26, 2009 at 7:48 PM, David Cournapeau wrote: > > >> It is built with an updated toolchain + a few patches to mingw I have >> yet submitted upstream, >> > > I created a ticket as well to track this issue: > > I added my comments to it. Is there a wa

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread David Cournapeau
On Thu, Mar 26, 2009 at 7:48 PM, David Cournapeau wrote: > It is built with an updated toolchain + a few patches to mingw I have > yet submitted upstream, I created a ticket as well to track this issue: http://projects.scipy.org/numpy/ticket/1068 ___

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-26 Thread David Cournapeau
Hi Bruce On Fri, Mar 20, 2009 at 10:45 PM, Bruce Southey wrote: > I still have the same problem on my Intel vista 64 system (Intel > QX6700  CPUZ reports the instruction set as MMX, SSE, SSE2, SSE3, > SSSE3, EM64T) with McAfee. The binary is built with every optimization turned off, and no ATLA

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-20 Thread David E. Sallis
David Cournapeau said the following on 3/20/2009 6:03 AM: > On Fri, Mar 20, 2009 at 7:59 PM, David Cournapeau wrote: >> On Fri, Mar 20, 2009 at 2:50 AM, David E. Sallis >> wrote: >>> David Cournapeau said the following on 3/18/2009 9:43 PM: I am pleased to announce the release of the first

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-20 Thread Bruce Southey
On Fri, Mar 20, 2009 at 6:03 AM, David Cournapeau wrote: > On Fri, Mar 20, 2009 at 7:59 PM, David Cournapeau wrote: >> On Fri, Mar 20, 2009 at 2:50 AM, David E. Sallis >> wrote: >>> David Cournapeau said the following on 3/18/2009 9:43 PM: I am pleased to announce the release of the first

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-20 Thread David Cournapeau
On Fri, Mar 20, 2009 at 7:59 PM, David Cournapeau wrote: > On Fri, Mar 20, 2009 at 2:50 AM, David E. Sallis > wrote: >> David Cournapeau said the following on 3/18/2009 9:43 PM: >>> I am pleased to announce the release of the first beta for numpy 1.3.0. >> >> I would totally love to begin using

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-20 Thread David Cournapeau
On Fri, Mar 20, 2009 at 2:50 AM, David E. Sallis wrote: > David Cournapeau said the following on 3/18/2009 9:43 PM: >> I am pleased to announce the release of the first beta for numpy 1.3.0. > > I would totally love to begin using this.  Can I trouble you to include MD5 > (or PGP, or SHA) signatu

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Gael Varoquaux
On Fri, Mar 20, 2009 at 11:31:18AM +0900, David Cournapeau wrote: > > OK, let's put it this way: if you feel it makes your life easier for me > > to have SVN acces, I don't mind. But I won't use it for anything else > > than docs and text file (or maybe a patch after review, but my volume of > > pa

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David Cournapeau
On Thu, Mar 19, 2009 at 11:15 PM, Gael Varoquaux wrote: > On Thu, Mar 19, 2009 at 11:01:38PM +0900, David Cournapeau wrote: >> On Thu, Mar 19, 2009 at 11:00 PM, Gael Varoquaux >> >> > No. And I am not asking for them. I'd rather have a buffer between me and >> > numpy, because I don't feel I know

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David Cournapeau
On Thu, Mar 19, 2009 at 10:36 PM, Sturla Molden wrote: > How do we do that? > > E.g. I added an important update to memmap.py in ticket 1053, but it has > not even been reviewed. In Gael's case, I was just talking about mentioning his contributions in the release notes (those French are lazy).

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
On Thu, Mar 19, 2009 at 1:01 PM, Charles R Harris wrote: > > > 2009/3/19 Robert Pyle > >> >> On Mar 19, 2009, at 1:38 PM, Pauli Virtanen wrote: >> >> Thanks for tracking this! Can you check what your platform gives for: >> >> > import numpy as np >> > info = np.finfo(np.longcomplex) >> > print "

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
2009/3/19 Robert Pyle > > On Mar 19, 2009, at 1:38 PM, Pauli Virtanen wrote: > > Thanks for tracking this! Can you check what your platform gives for: > > > import numpy as np > > info = np.finfo(np.longcomplex) > > print "eps:", info.eps, info.eps.dtype > > print "tiny:", info.tiny, info.tiny.dt

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Stéfan van der Walt
Hi Sturla 2009/3/19 Sturla Molden : > You should really set up a better system for receiving and reviewing > contributions. Otherwise people will not care. The ticket was not set "ready for review" until Pauli did it this evening. You're supposed to have permission to do that, but if you don't i

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Jim Vickroy
Apologies for the spam, but I would suggest that the subject be changed to reflect the topic (i.e., Mac OS X problem with this release). Thanks, -- jv Robert Pyle wrote: On Mar 19, 2009, at 1:24 PM, Charles R Harris wrote: Yep, that's it. Can you see what info.tiny/eps is in this case. Als

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Robert Pyle
On Mar 19, 2009, at 1:38 PM, Pauli Virtanen wrote: Thanks for tracking this! Can you check what your platform gives for: > import numpy as np > info = np.finfo(np.longcomplex) > print "eps:", info.eps, info.eps.dtype > print "tiny:", info.tiny, info.tiny.dtype > print "log10:", np.log10(info.tin

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Robert Pyle
On Mar 19, 2009, at 1:24 PM, Charles R Harris wrote: > Yep, that's it. Can you see what info.tiny/eps is in this case. Also > info.tiny and eps separately. > > Chuck eps = 1.3817869701e-76 info.tiny = -1.08420217274e-19 info.tiny/eps = -7.84637716375e+56 Bob _

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David E. Sallis
David Cournapeau said the following on 3/18/2009 9:43 PM: > I am pleased to announce the release of the first beta for numpy 1.3.0. I would totally love to begin using this. Can I trouble you to include MD5 (or PGP, or SHA) signatures for your download files in your release notes as you have f

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Pauli Virtanen
Thu, 19 Mar 2009 13:19:18 -0400, Robert Pyle wrote: [clip] > When I run this, it says x_series is an array of 200 NaNs. That would > certainly explain why the assertion in test_umath.py failed! Thanks for tracking this! Can you check what your platform gives for: import numpy as np info = np.fin

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
On Thu, Mar 19, 2009 at 11:19 AM, Robert Pyle wrote: > Hi Chuck, > On Mar 19, 2009, at 1:01 PM, Charles R Harris wrote: > > Is that any help? > > > > Not yet ;) I think there is a problem with the range of values in x > > that might have their source in the finfo values. So it would help > > if yo

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Robert Pyle
Hi Chuck, On Mar 19, 2009, at 1:01 PM, Charles R Harris wrote: > Is that any help? > > Not yet ;) I think there is a problem with the range of values in x > that might have their source in the finfo values. So it would help > if you could pin down just where x goes wrong by printing it out. >

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
On Thu, Mar 19, 2009 at 10:21 AM, Robert Pyle wrote: > > On Mar 19, 2009, at 11:35 AM, Charles R Harris wrote: > > > > > > > On Thu, Mar 19, 2009 at 9:17 AM, Robert Pyle > > wrote: > > I'm getting one test failure with 1.3.0b1 --- > > > > FAIL: test_umath.TestComplexFunctions.test_loss_of_precisi

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Robert Pyle
On Mar 19, 2009, at 11:35 AM, Charles R Harris wrote: > > > On Thu, Mar 19, 2009 at 9:17 AM, Robert Pyle > wrote: > I'm getting one test failure with 1.3.0b1 --- > > FAIL: test_umath.TestComplexFunctions.test_loss_of_precision( 'numpy.complex256'>,) > --

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Alan G Isaac
I am really grateful to have NumPy on Python 2.6! Alan Isaac Python 2.6.1 (r261:67517, Dec 4 2008, 16:51:00) [MSC v.1500 32 bit (Intel)] on win32 Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> numpy.test() Running unit tests for numpy NumPy version

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
On Thu, Mar 19, 2009 at 9:17 AM, Robert Pyle wrote: > Hi, > > First of all, thanks to everyone for all the hard work. > > On Mar 18, 2009, at 10:43 PM, David Cournapeau wrote: > > > I am pleased to announce the release of the first beta for numpy > > 1.3.0. You can find source tarballs and instal

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Robert Pyle
Hi, First of all, thanks to everyone for all the hard work. On Mar 18, 2009, at 10:43 PM, David Cournapeau wrote: > I am pleased to announce the release of the first beta for numpy > 1.3.0. You can find source tarballs and installers for both Mac OS X I'm on a dual G5 Mac running OS X 10.5.6 an

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Charles R Harris
On Thu, Mar 19, 2009 at 5:02 AM, Peter < numpy-discuss...@maubp.freeserve.co.uk> wrote: > On Thu, Mar 19, 2009 at 2:43 AM, David Cournapeau > wrote: > > New defines > > ~~~ > > > > New public C defines are available for ARCH specific code through > > numpy/npy_cpu.h: > > > >* NPY_

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Gael Varoquaux
On Thu, Mar 19, 2009 at 11:01:38PM +0900, David Cournapeau wrote: > On Thu, Mar 19, 2009 at 11:00 PM, Gael Varoquaux > > > No. And I am not asking for them. I'd rather have a buffer between me and > > numpy, because I don't feel I know numpy well-enough to commit directly. > committing a text fil

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David Cournapeau
On Thu, Mar 19, 2009 at 11:00 PM, Gael Varoquaux > No. And I am not asking for them. I'd rather have a buffer between me and > numpy, because I don't feel I know numpy well-enough to commit directly. committing a text file should be easy enough, even for you ;) David

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Gael Varoquaux
On Thu, Mar 19, 2009 at 10:39:12PM +0900, David Cournapeau wrote: > On Thu, Mar 19, 2009 at 10:32 PM, Gael Varoquaux > wrote: > > On Thu, Mar 19, 2009 at 10:07:08PM +0900, David Cournapeau wrote: > >> Well, why not adding it yourself, then ? It is generally easier, > >> faster and more accurate fo

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David Cournapeau
On Thu, Mar 19, 2009 at 10:32 PM, Gael Varoquaux wrote: > On Thu, Mar 19, 2009 at 10:07:08PM +0900, David Cournapeau wrote: >> Well, why not adding it yourself, then ? It is generally easier, >> faster and more accurate for people related with the changes to add it >> themselves to the release not

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Sturla Molden
On 3/19/2009 2:07 PM, David Cournapeau wrote: > Well, why not adding it yourself, then ? It is generally easier, > faster and more accurate for people related with the changes to add it > themselves to the release notes :) How do we do that? E.g. I added an important update to memmap.py in ticke

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Gael Varoquaux
On Thu, Mar 19, 2009 at 10:07:08PM +0900, David Cournapeau wrote: > Well, why not adding it yourself, then ? It is generally easier, > faster and more accurate for people related with the changes to add it > themselves to the release notes :) I am being stupid. How do I do this? I am sure you alre

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread David Cournapeau
On Thu, Mar 19, 2009 at 6:12 PM, Gael Varoquaux wrote: > On Thu, Mar 19, 2009 at 11:43:29AM +0900, David Cournapeau wrote: >> I am pleased to announce the release of the first beta for numpy >> 1.3.0. You can find source tarballs and installers for both Mac OS X >> and Windows on the sourceforge p

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Peter
On Thu, Mar 19, 2009 at 2:43 AM, David Cournapeau wrote: > New defines > ~~~ > > New public C defines are available for ARCH specific code through > numpy/npy_cpu.h: > >        * NPY_CPU_X86: x86 arch (32 bits) >        * NPY_CPU_AMD64: amd64 arch (x86_64, NOT Itanium) >        * NPY_CPU_P

Re: [Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-19 Thread Gael Varoquaux
On Thu, Mar 19, 2009 at 11:43:29AM +0900, David Cournapeau wrote: > I am pleased to announce the release of the first beta for numpy > 1.3.0. You can find source tarballs and installers for both Mac OS X > and Windows on the sourceforge page: > https://sourceforge.net/projects/numpy/ > The releas

[Numpy-discussion] [Announce] Numpy 1.3.0b1

2009-03-18 Thread David Cournapeau
Hi, I am pleased to announce the release of the first beta for numpy 1.3.0. You can find source tarballs and installers for both Mac OS X and Windows on the sourceforge page: https://sourceforge.net/projects/numpy/ The release note for the 1.3.0 release are below, The Numpy developers