On Fri, Jul 25, 2008 at 1:37 PM, Raymond Hettinger <report@bugs.python.org> wrote:
Raymond Hettinger <rhettinger@users.sourceforge.net> added the comment:

Can you also implement blending of approximations:  (1-t)*f1(x) + t*f2
(x)

Is this necessary?  Are the approximations significantly different near the transition points?

(Haven't had a chance to download the patch and try it myself as a I'm getting on a plane in the morning--sorry)
--
Daniel Stutzbach, Ph.D.
President, Stutzbach Enterprises, LLC