Fix spline plots to draw a better generalized line.
Current spline plots can be unintuitive when your dataset isn't quite a perfect curve. A dataset of 10,10,10,0 shows a huge spike in the line right when it really shouldn't exist.
anonymous
shared this idea
Version 2.1 released
-
Corey Haines
commented
It would be nice to have several options, as well as a method for replacing it with an external algorithm that we want.
For our graphs on MercuryApp.com, the cubic spline does not match what we are using the data for (dips and bumps are inappropriate). We need something much more smooth.
-
Spline is good but sometimes not acceptable. We need Akima interpolation algorithm. Look comparison on this chart:
http://www.alglib.net/interpolation/spline3.php
Your SPLINE behaves like CUBIC interpolation - Akima behaves differently.
-
Administrator
commented
Akima is better to presents values on charts because:
For example: data 150,150,150,170
Result (red arrow shows not acceptable part of draw on chart):
http://lh3.ggpht.com/_6UWC9mTMv5w/TGfcQKyHSxI/AAAAAAAAFdg/NGLWUfjJj3g/s800/akima.jpg
