I was slightly off with the initial condition. You need a fraction p/q where p^2=3 q^2 +/- 1. The simplest case is p=2, q=1. This puts you in the continued fraction expression for %:3. For these estimates, the absolute error is less than % (%:3)* *: q.
Best wishes, John ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
