Personal tools
You are here: Home / CarnegiePython / Burns' Python Scripts / Modules / spline2


A python wrapper for spline2 (aka hyper-spline) by B. Thijsse et al. Uses the Durbin-Watson statistic rather than chi-square as a test for the number of knots in the approximating spline, which is supposed to be less sensitive to incorrect variances in the data. Requires Numeric python.

tar archive icon spline2-1.0.tgz — tar archive, 3.19 MB (3345149 bytes)