matplotlib-users

[Matplotlib-users] Re: matplotlib and scipy_core

[Eric F. <ef...@ha...>]

> Chris Fonnesbeck wrote:
> 
>> I think there is more to it than that. There are all sorts of
>> dependencies on either Numeric or numarray that need to be addressed.
>> Try installing matplotlib without either Numeric or numarray -- I dont
>> think you will get very far.
>>  
>>
> Ah, good point.  It would be good to begin work on this, but IMHO the 
> new scipy_core has a bit to go before it is ready for prime time.  
> Someone just pointed out on the numpy list, for example, that it 
> presently does not include masked arrays, a capability I use often and 
> support for which was painstakingly added to Matplotlib over the last 
> year or so.
> 

No, it *does* include masked arrays.  Try:

import scipy.base.ma as ma

To use scipy as a complete replacement for Numeric or numarray will 
require work in numerix, in the setup.py build system, and possibly in 
cntr.c, which needs to build separate versions for numarray and Numeric. 
  (I expect little or no change will actually be needed in cntr.c.)

I haven't figured out yet exactly what is present and what is missing 
relative to old Numeric, and the things that numerix imports, so I am 
not sure whether scipy_core has everything needed for a one-to-one 
substitution; it looks like it is close, at least.

Eric

Re: [Matplotlib-users] Re: matplotlib and scipy_core

[Eric F. <ef...@ha...>]

John,

There may be a couple of related questions:

1) For how long do we maintain compatibility with older versions of 
Numeric and numarray?  Even if the array interface has been added, or is 
being added, to numarray, it won't be present in older versions or on 
classic Numeric.

2) Even if the array interface is the way to go for the future, it may 
be easier for now to let cntr.c continue to use the Numeric C API, which 
I think is maintained intact in scipy.  In fact, doing it this way looks 
very easy indeed.

It may be that I am misunderstanding something fundamental about how the 
array interface works and is implemented; I haven't looked at it closely.

Eric

John Hunter wrote:
>>>>>>"Eric" == Eric Firing <ef...@ha...> writes:
> 
> 
>     Eric> To use scipy as a complete replacement for Numeric or
>     Eric> numarray will require work in numerix, in the setup.py build
>     Eric> system, and possibly in cntr.c, which needs to build
>     Eric> separate versions for numarray and Numeric. (I expect little
>     Eric> or no change will actually be needed in cntr.c.)
> 
> I talked to Travis about this at scipy and he said that with the new
> array interface, we wouldn't need to compile separate extensions for
> each of Numeric, numarray and scipy (as we do now, eg with
> _nc_transforms, _na_transforms, etc).  This is good news because it
> implies shorter compile times and smaller binary distributions, but I
> can't quite understand how this would work yet, eg, to support the
> there packages with one binary *.so).  But this may be some of the
> black magic of the new array interface.  If correct, this approach
> would require *either* Numeric 24, numarray (something recent) or
> scipy_base.
> 
> Perhaps Todd has further insight, as the author of numerix and
> presumably the implementor of the array interface for numarray...
> 
> JDH

Re: [Matplotlib-users] matplotlib and scipy_core

[Chris F. <fon...@gm...>]

On 12/3/05, mat...@li... <matplotlib-user=
s-
> >
> >
> > Chris, are you using Daishi's patches?  I'm not quite sure if they're i=
n
> > CVS yet, but it's certain that you can NOT build mpl 0.85 + scipy_core
> > without them.
> >

Just tried it. One issue arises with the import syntax in
matplotlib/numerix/fft/__init__.py:

      7 elif which[0] =3D=3D "scipy":
----> 8     from scipy.fftpack import *
        global scipy.fftpack =3D undefined

I think the correct syntax is "from scipy import fftpack", etc.

--
Chris Fonnesbeck
Atlanta, GA

Re: [Matplotlib-users] matplotlib and scipy_core

[John H. <jdh...@ac...>]

>>>>> "Chris" == Chris Fonnesbeck <fon...@gm...> writes:
    Chris> I think the correct syntax is "from scipy import fftpack",
    Chris> etc.

I made some changes to support the new scipy in the numerix layer.
You may want to update from CVS and try again.  There still appears to
be a problem with the color mapping code, but this is the only one I
am aware of after light testing.

Checking in lib/matplotlib/numerix/mlab/__init__.py;
/cvsroot/matplotlib/matplotlib/lib/matplotlib/numerix/mlab/__init__.py,v
<--  __init__.py
new revision: 1.3; previous revision: 1.2

or later.

JDH

Re: [Matplotlib-users] Re: matplotlib and scipy_core

[John H. <jdh...@ac...>]

>>>>> "Eric" == Eric Firing <ef...@ha...> writes:

    Eric> To use scipy as a complete replacement for Numeric or
    Eric> numarray will require work in numerix, in the setup.py build
    Eric> system, and possibly in cntr.c, which needs to build
    Eric> separate versions for numarray and Numeric. (I expect little
    Eric> or no change will actually be needed in cntr.c.)

I talked to Travis about this at scipy and he said that with the new
array interface, we wouldn't need to compile separate extensions for
each of Numeric, numarray and scipy (as we do now, eg with
_nc_transforms, _na_transforms, etc).  This is good news because it
implies shorter compile times and smaller binary distributions, but I
can't quite understand how this would work yet, eg, to support the
there packages with one binary *.so).  But this may be some of the
black magic of the new array interface.  If correct, this approach
would require *either* Numeric 24, numarray (something recent) or
scipy_base.

Perhaps Todd has further insight, as the author of numerix and
presumably the implementor of the array interface for numarray...

JDH

Re: [Matplotlib-users] Re: matplotlib and scipy_core

[Stephen W. <ste...@cs...>]

Eric Firing wrote:

> No, [new scipy core] *does* include masked arrays.

Someone should tell Paul Dubois then :-) . See the bottom of this thread:

http://sourceforge.net/mailarchive/forum.php?thread_id=8355108&forum_id=4890