Re: [Numpy-discussion] NumPy 1.0.3 release next week

Hello all A quick update: we're down to 16 active tickets, with 38 tickets already closed for the 1.0.3 release. Thanks to all the bugfixers. Summary of tickets for 1.0.3: http://projects.scipy.org/scipy/numpy/milestone/1.0.3%20Release Active tickets for 1.0.3: http://projects.scipy.org/scipy

Re: [Numpy-discussion] pretty printing record array element - datetime

John Hunter wrote: > I have a numpy record array and I want to pretty print a single > element. I was trying to loop over the names in the element dtype and > use getattr to access the field value, but I got fouled up because > getattr is trying to access the dtype attribute of one of the python >

Re: [Numpy-discussion] umath.maximum.reduce

On Tuesday 15 May 2007 17:34:15 Charles R Harris wrote: > I see no one answered ;) I think it is traditional behaviour from Numeric, > so can't be changed without a bit of thought. > Why not file a ticket for the > 1.1 release and see how folks respond? I'll try to. > Is there any reason why

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Christopher Barker noaa.gov> writes: > Is your Python Universal? I suspect so. If you don't want a Universal > build, then you may have to go find the right flags to build Python for > PPC only, then distutils should work for you. If you do want a Universal > Python, then why not a Universal n

Re: [Numpy-discussion] pretty printing record array element - datetime

In addition to Matt's answer: Yes, we have a tmulti where you can store dates and other variables in a record-like TimeSeries. However, there must be simpler: Why are you using getattr to access the fields ? Your object tr is a record array, so you could use for n in tr.dtype.names: print n,

Re: [Numpy-discussion] umath.maximum.reduce

On 5/8/07, Pierre GM <[EMAIL PROTECTED]> wrote: All, On a 2D array, numpy.core.umath.maximum.reduce returns a 1D array (the default axis is 0). An axis=None is not recognized as a valid argument for numpy.core.umath.maximum, unfortunately... Is this the wanted behavior ? Thanks a lot in advance

Re: [Numpy-discussion] pretty printing record array element - datetime

> I have a numpy record array and I want to pretty print a single > element. I was trying to loop over the names in the element dtype and > use getattr to access the field value, but I got fouled up because > getattr is trying to access the dtype attribute of one of the python > objects (datetime.

[Numpy-discussion] pretty printing record array element - datetime

I have a numpy record array and I want to pretty print a single element. I was trying to loop over the names in the element dtype and use getattr to access the field value, but I got fouled up because getattr is trying to access the dtype attribute of one of the python objects (datetime.date) that

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: On Tue, 15 May 2007, Charles R Harris apparently wrote: > What I would like it to look like, except for the type > stuff, is attached. Ooof! Aside from obviously agreeing with the location of notes and examples, I have to say, the readability

Re: [Numpy-discussion] Documentation

On Tue, 15 May 2007, Charles R Harris apparently wrote: > What I would like it to look like, except for the type > stuff, is attached. Ooof! Aside from obviously agreeing with the location of notes and examples, I have to say, the readability of your example is greatly reduced from the epydoc

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: On Tue, 15 May 2007, Charles R Harris apparently wrote: > The item in a consolidated list is divided into three > parts: name, type, explanation. I see no way in the CSS to > control the display of those parts, although it might be > an undocum

Re: [Numpy-discussion] Documentation

On Tue, 15 May 2007, Charles R Harris apparently wrote: > The item in a consolidated list is divided into three > parts: name, type, explanation. I see no way in the CSS to > control the display of those parts, although it might be > an undocumented feature. I basically want it to look like >

Re: [Numpy-discussion] Documentation

On 5/15/07, Charles R Harris <[EMAIL PROTECTED]> wrote: On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: > > On Tue, 15 May 2007, Charles R Harris apparently wrote: > > If a CSS can set all the font styles, item spacing, and > > treatment of explanatory text, that would be the more > > flex

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Robert Kern gmail.com> writes: > > Ryan Welch wrote: > > > When I run a normal build, it always uses -arch ppc -arch i386 to compile C > > sources. I'd like to get it to do -arch ppc64 instead, and remove the > > unnecessary i386 if at all possible. > > Those parameters come from the BASECFLA

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Ryan Welch wrote: > When I run a normal build, it always uses -arch ppc -arch i386 to compile C > sources. I'd like to get it to do -arch ppc64 instead, and remove the > unnecessary i386 if at all possible. Is your Python Universal? I suspect so. If you don't want a Universal build, then you may

Re: [Numpy-discussion] Documentation

On Tue, 15 May 2007, Charles R Harris apparently wrote: > Looks like the font used for section headings can be > specified, but there is nothing in the epydoc.css file > that looks like it affects the consolidated lists or the > movement of a section heading to the top of the output. > I think

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Ryan Welch wrote: > When I run a normal build, it always uses -arch ppc -arch i386 to compile C > sources. I'd like to get it to do -arch ppc64 instead, and remove the > unnecessary i386 if at all possible. Those parameters come from the BASECFLAGS variable in $prefix/lib/python2.5/config/Makefi

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: On Tue, 15 May 2007, Charles R Harris apparently wrote: > If a CSS can set all the font styles, item spacing, and > treatment of explanatory text, that would be the more > flexible route. I'm not sure what "treatment" meams here, but if it is

Re: [Numpy-discussion] Documentation

On 5/15/07, Gael Varoquaux <[EMAIL PROTECTED]> wrote: On Tue, May 15, 2007 at 01:15:12PM -0400, Alan G Isaac wrote: > Bullets in consolidated lists: > if you don't like them, don't use them. > Use a definition list instead. > (I like this much better myself.) Can't this be fixed by

Re: [Numpy-discussion] Documentation

On Tue, 15 May 2007, Charles R Harris apparently wrote: > If a CSS can set all the font styles, item spacing, and > treatment of explanatory text, that would be the more > flexible route. I'm not sure what "treatment" meams here, but if it is a matter of font decisions or spacing, the answer sh

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Christopher Barker noaa.gov> writes: > What happens with a plain old: > > > python setup.py build > > distutils should just "do the right thing" > > If you use that python to run setup.py, and it's a PPC binary, you > should be all set. > > -CHB > When I run a normal build, it always uses

Re: [Numpy-discussion] 64-bit Numpy (ppc64)

Ryan Welch wrote: > I'm trying to figure out how to compile Numpy for -arch ppc64 on darwin, and I What happens with a plain old: > python setup.py build distutils should just "do the right thing" > Note: I have a 64-bit Python 2.5 binary already compiled, and I've confirmed > it > can alloc

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: > On Tue, May 15, 2007 at 01:15:12PM -0400, Alan G Isaac wrote: >> Bullets in consolidated lists: >> if you don't like them, don't use them. >> Use a definition list instead. >> (I like this much better myself.) On Tue, 15 May 200

Re: [Numpy-discussion] Documentation

> On Tue, May 15, 2007 at 01:15:12PM -0400, Alan G Isaac wrote: >> Bullets in consolidated lists: >> if you don't like them, don't use them. >> Use a definition list instead. >> (I like this much better myself.) On Tue, 15 May 2007, Gael Varoquaux apparently wrote: > Can't this

Re: [Numpy-discussion] Documentation

On Tue, May 15, 2007 at 01:15:12PM -0400, Alan G Isaac wrote: > Bullets in consolidated lists: > if you don't like them, don't use them. > Use a definition list instead. > (I like this much better myself.) Can't this be fixed by a CSS ? Gaël ___

Re: [Numpy-discussion] Documentation

On Tue, 15 May 2007, Charles R Harris apparently wrote: > It also causes the headings to be set in large bold type. Ugly. Well I agree. But again this is just a style issue. All font decisions (and *much* else) can be set in a CSS file. > I think reST is great, and properly used we get good >

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: > On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: >> Just to add a bit more specificity, I believe that the >> HOWTO_DOCUMENT.txt document at >> http://svn.scipy.org/svn/numpy/trunk/numpy/doc/HOWTO_DOCUMENT.txt >> summarizes a substantial d

Re: [Numpy-discussion] Documentation

> On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: >> Just to add a bit more specificity, I believe that the >> HOWTO_DOCUMENT.txt document at >> http://svn.scipy.org/svn/numpy/trunk/numpy/doc/HOWTO_DOCUMENT.txt >> summarizes a substantial discussion and broad agreement. >> I.e., it is cur

[Numpy-discussion] 64-bit Numpy (ppc64)

Hi guys, I'm trying to figure out how to compile Numpy for -arch ppc64 on darwin, and I can't seem to get it to work. Any pointers for how I can alter the build process? (I'm not too familiar with distutils, been doing some reading but so far no luck in figuring out how to pass in my own compiler

Re: [Numpy-discussion] Documentation

On 5/15/07, Charles R Harris <[EMAIL PROTECTED]> wrote: On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: > > > Gary Ruben wrote: > >> In you post, is your numpy/doc reference referring to the > >> file HOWTO_DOCUMENT.txt? > >> I notice that this is not the same as the one here > >> < http:/

Re: [Numpy-discussion] Documentation

On 5/15/07, Alan G Isaac <[EMAIL PROTECTED]> wrote: > Gary Ruben wrote: >> In you post, is your numpy/doc reference referring to the >> file HOWTO_DOCUMENT.txt? >> I notice that this is not the same as the one here >> >> which I think may be the preferred sta

Re: [Numpy-discussion] Documentation

> Gary Ruben wrote: >> In you post, is your numpy/doc reference referring to the >> file HOWTO_DOCUMENT.txt? >> I notice that this is not the same as the one here >> >> which I think may be the preferred standard. If someone can confirm >> this, it should

Re: [Numpy-discussion] Documentation

Gary Ruben wrote: > Hi Charles, > > In you post, is your numpy/doc reference referring to the file > HOWTO_DOCUMENT.txt? > > I notice that this is not the same as the one here > > which I think may be the preferred standard. If someone can confirm > this, it

[Numpy-discussion] call for chapeters in a book about freeware/opensource in chemometrics

Dear All, me and other two colleagues will be interested in writing a tutorial book about the use of opensource/freeware software in chemometrics (mainly python oriented). I've contacted the editor at Blackwell publishing that has told me that can be interested in it and sent me the module for s

Re: [Numpy-discussion] Documentation

On 5/15/07, Gary Ruben <[EMAIL PROTECTED]> wrote: Hi Charles, In you post, is your numpy/doc reference referring to the file HOWTO_DOCUMENT.txt? I notice that this is not the same as the one here which I think may be the preferred standard. If someone can c

Re: [Numpy-discussion] numpy array sharing between processes? (and ctypes)

Agreed. Could someone with wiki admin access please delete the Ctypes page and rename type Ctypes2 page to Ctypes? As far as I know, Ctypes2 is really what you want to look at (at least it was, last time I worked on it). Thanks. Cheers, Albert On Mon, 14 May 2007, Stefan van der Walt wrote:

Re: [Numpy-discussion] Documentation

Hi Charles, In you post, is your numpy/doc reference referring to the file HOWTO_DOCUMENT.txt? I notice that this is not the same as the one here which I think may be the preferred standard. If someone can confirm this, it should probably replace the HOWTO_D