Hacker News new | past | comments | ask | show | jobs | submit login

assuming any "kind" of spline before you have your data is jumping the gun; if you have 20 points, then anything under a 20th order polynomial is not guaranteed to actually fit your data, so: how lossy do you want the fit to be? You could use a poly-whatever, like a catmull-rom curve, or something based on B-spline



Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: