Korenberg's fast orthogonal search

Started by December 20, 2017
Has anybody used this algorithm? It seems to be what I need;
fitting 2 or 3 sinusoids to sampled data, where those sinusoids have
arbitrary frequencies (not harmonically related as in discrete fourier
On Wednesday, December 20, 2017 at 5:42:05 AM UTC-6, pedr...@lycos.com wrote:
> Has anybody used this algorithm?
I have to admit, I'd never heard of it before. But skimming over one of his papers just now (http://www.applied-mathematics.net/identification/papersLARS/FOS_original.pdf), it looks very promising. The last paragraph in section 3.1 is especially interesting. If you try it, please keep us posted about your results. Greg