[SciPy-User] Linear Interpolation

[Tom Harris]

Greetings.

I would like some help in linear interpolation. Not in doing the
interpolation, but in choosing the points for the interpolation (are these
called interpolants?).

I have experimental data that gives y for a few values of x, and while I
can get good approximations using scipy's spline functions, I need good
linear interpolation points so that I can perform an approximation in code
that is run on a tiny embedded system that can only do integer arithmetic.
I have been using a fairly inadequate method of choosing my linear
interpolation points: simply putting points on the splined curve and adding
points at the point of greatest error until the error reduces to a low
enough value. But there has got to be a better way. Do people simply not
bother with linear interpolation and go straight to splines?


-- 

Tom Harris <celephicus at gmail.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20120719/03bf5671/attachment.html>