Talk:Alpha max plus beta min algorithm

Is there a way to polish the solution?

Is there a way to compute a new estimate of hypot from this starting condition like is done with typical sqrt algorithms? --Jaded-view (talk) 03:29, 5 June 2008 (UTC)[reply]

Is there one wrong parameter set?

The parameter set alpha=7/8 and beta=15/16 can't be right. I tried to plot that and the graph looks completely different. My guess is that this should have been something like beta=15/32. Most beta values are close to 1/2.

TomF —Preceding unsigned comment added by 77.191.238.62 (talk) 22:56, 26 December 2008 (UTC)[reply]

Since the error values are signed, using the root-mean-square-error is more meaningful than the mean-error. Here is a similar table with minimal rms errors. All but the first row can be computed with adds and shifts. α β max rms 0.947564607 0.392494792 5.24% 2.34% 243/256 25/64 5.26% 2.34% 121/128 25/64 5.53% 2.35% 61/64 25/64 4.98% 2.38% 31/32 3/8 4.98% 2.71% 15/16 3/8 7.19% 2.82% 1 3/8 6.80% 4.76% 1 1/4 11.61% 4.16% 1 1/2 11.80% 9.21%