Last active
January 7, 2020 20:37
-
-
Save adamwlev/a05f9e4aa172cd0d6615c18eb1a6c219 to your computer and use it in GitHub Desktop.
Natural Cubic Splines
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Python implementation of Natural Cubic Splines from Elements of Statistical Learning (HTF).
Natural Cubic Splines have the restriction of being linear at the edges and therefore are better at extrapolation. This implementation uses internal Cross Validation during a fit to find the best number of degrees of freedom. Natural Cubic Splines do better than Cubic Splines especially when the number of observations is small and/or there is a significant amount of noise in the observed data.