[root]/branches/budaev/var/akima
nlb
(1 files, 1793 lines)
760
(0 files, 0 lines)
Drivers
(0 files, 0 lines)
Sp
(2 files, 498 lines)
Src
(0 files, 0 lines)
Sp
(1 files, 1234 lines)
761
(0 files, 0 lines)
Drivers
(0 files, 0 lines)
Sp
(2 files, 289 lines)
Src
(0 files, 0 lines)
Sp
(2 files, 8522 lines)
src
(3 files, 526 lines)
Author | Changes | Lines of Code | Lines per Change |
---|---|---|---|
sbu062 | 4 (100.0%) | 325 (100.0%) | 81.2 |
akima package from https://github.com/andrewning/akima
1-Dimensional Akima spline implementation [1] with derivatives not only of the function, but also with respect to the original data points. An Akima spline is a polynomial spline method that avoids overshooting issues common with many other splines (e.g., cubic splines), resulting in a more natural curve. I find Akima splines to be particularly useful in optimization applications for defining distributions of design variables at a small number of points. This type of spline more readily permits using bound constraints directly on the design variables, without worrying about excessive overshooting between them.
325 lines of code changed in 4 files: