[SciPy-dev] patch: allow arrayobject.h to compile in c++ compiler

[Charles R Harris]

>
> I'll make it up by working on numerical code I can actually do something
> useful with.  I'll try to contribute something for fitting, the included
> polyfit is nice but numerically less uniform in error and stable (due to
> condition issues) than a Chebyshev-based fit.  There's also the lingering
> issue of numerically stable evaluation of polyval()...  I'll find
> something to
> pay for my mistakes with :)
>
> Apologies...
>
> Cheers,



I've attached the Python routines I use for Chebychev stuff. It is also
possible to solve ODEs over an interval by sampling at the Chebychev points
and treating the derivative as a matrix. Of course, boundary conditions need
to be added so that the resulting matrix isn't singular.

Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20051123/81e5e671/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Chebychev.py
Type: application/octet-stream
Size: 20212 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20051123/81e5e671/attachment.obj>