RE: [Numpy-discussion] Integrating cephes with numpy

[Paul F D. <pa...@pf...>]

FWIW, I am about to create an 'Extras' area in the Numeric cvs =
repository
where non-core, distutils-enabled addons can be put by the numpy =
developers.
These would be things that install as their own packages, not a part of
Numeric or Numarray.


> -----Original Message-----
> From: num...@li...=20
> [mailto:num...@li...] On=20
> Behalf Of Michiel Jan Laurens de Hoon
> Sent: Wednesday, March 05, 2003 9:25 PM
> To: num...@li...
> Subject: [Numpy-discussion] Integrating cephes with numpy
>=20
>=20
> Hi everybody,
>=20
> You probably all know Travis Oliphant's SpecialFuncs=20
> (previously called=20
> cephes) module, which contains a large number of special functions.=20
> Currently it is not easy to install this module, since SpecialFuncs=20
> became a part of SciPy, which is great in itself but cannot=20
> be installed=20
> easily as other parts of SciPy depend on a host of other packages. As=20
> Scipy will get extended in the future, this problem will only=20
> get worse.=20
> Recently I have been asked several times for such a module=20
> for special=20
> functions, but I don't know where I can refer people to=20
> without qualms,=20
> especially for newbies.
>=20
> Since the 'cephes' part of SpecialFuncs is basically an extension of=20
> umathmodule.c in numpy, it seems that numpy would be the=20
> natural place=20
> for cephes. So I would suggest to integrate cephes with=20
> numpy, either by=20
> adding cephes' special functions to umathmodule.c or as a separate=20
> module (similar to the LinearAlgebra or RandomArray parts of=20
> numpy). In=20
> the process, we can solve some installation problems in cephes which=20
> seem to be recurring frequently (see the numpy mailing list for some=20
> examples).
>=20
> For the moment, I slapped together a version of the cephes=20
> module that=20
> can be installed more easily; however, I would think it is=20
> better to do=20
> this the right way and to avoid multiple variations of basically the=20
> same package floating around in cyberspace.
>=20
> Any opinions on this? If this seems like a good idea, I'd be=20
> happy to do=20
> some additional coding if needed to set this up, though=20
> Travis Oliphant=20
> has basically done everything already so I wouldn't think=20
> much further=20
> coding is needed.
>=20
> --Michiel de Hoon, University of Tokyo.
>=20
> --=20
> Michiel de Hoon, Assistant Professor
> University of Tokyo, Institute of Medical Science
> Human Genome Center
> 4-6-1 Shirokane-dai, Minato-ku
> Tokyo 108-8639
> Japan
> http://bonsai.ims.u-tokyo.ac.jp/~mdehoon
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Etnus, makers of=20
> TotalView, The debugger=20
> for complex code. Debugging C/C++ programs can leave you=20
> feeling lost and=20
> disoriented. TotalView can help you find your way. Available=20
> on major UNIX=20
> and Linux platforms. Try it free. www.etnus.com=20
> _______________________________________________
> Numpy-discussion mailing list Num...@li...
> https://lists.sourceforge.net/lists/listinfo/numpy-discussion
>=20