User talk:Starrynte/Remez

I need some help. Currently I'm using newton's method to find extrema of some error function, using the previous extrema as the initial value for newton's method. The trouble is it's not accurate enough and sometimes has seemingly random behavior...any suggestions for a more accurate a priori guess? --Starrynte 05:10, 17 May 2009 (UTC)RESOLVED, now using bisection method, which although slow, it always works