[SciPy-dev] some scipy.interpolate functions appear to be inaccessible

[Travis E. Oliphant]

Tom Grydeland wrote:
> Hi developers,
>
> I tried figuring out how to use scipy.interpolate to create cubic
> splines with "natural" boundary conditions (i.e. y'' = 0 at the end
> points), but failed.  When I look in the source file for
> scipy/interpolate/interpolate.py it certainly looks like there is code
> to create the tridiagonal system to solve for such splines (
> _get_spline3_Bb()), but this function appears to be orphaned.  There
> are a number of functions _find_xxx() that do nothing except raise
> NotImplementedError, although these should be called from splmake(),
> from the looks of it.
>
>   
I'm the one who started the process of trying to add more options to the 
creation of cubic splines.   Much of the infrastructure was created and 
a few options were provided as well as some low-level operations.  
However, as you noticed not all of the details are worked out for all 
the possible options.  

Basically, the issue is that splines for order N have N-1 additional 
degrees of freedom and I wanted there to be some way to both provide for 
an arbitrary setting of these degrees of freedom as well as an easy 
interface to the "typical" ways the degrees of freedom are resolved.
> Does anyone know the intended structure of these functions?  I could
> try to make them work, given a roadmap.
>   
If you have time to work on the code, I could answer questions, but 
don't have a roadmap to share.    I'm teaching this week --- maybe I'll 
have time to work on the interpolation again --- but don't count on it.

Best regards,

-Travis

-- 

Travis Oliphant
Enthought, Inc.
(512) 536-1057 (office)
(512) 536-1059 (fax)
http://www.enthought.com
oliphant at enthought.com