I like it, but the function name should not say only "extrap", but should 
reflect the fact that it performs both interpolation and extrapolation.

Also, a flag can be set indicating if the operation was interpolation or 
extrapolation.


BTW, as we are on the subject: any plans for multidimensional interpolation?


--
jeremy

On Thursday 02 May 2013 23:54:08 Rhys Ulerich wrote:
> > I also miss the good old days when extrapolation was automatically
> > computed using the chosen interpolation scheme. Can this behavior be
> > added as a flag or as a new function?
> 
> I could see a gsl_extrap_eval() and friends that were identical in
> functionality but that did not perform this x < xmin || x > xmax
> check.  gsl_interp_eval() could then be implemented as this check
> followed by calling gsl_extrap_eval().  People wanting the old
> behavior could call the new function which would then self-document
> that extrapolation is a possibility.
> 
> Thoughts?
> 
> - Rhys

Reply via email to