On Thu, Aug 18, 2022 at 2:34 AM Elijah Stone <elro...@elronnd.net> wrote: > Have you ever wanted good performance out of a large array of > extended-precision integers (or, for that matter, rationals), where most of > the integers involved would not have fit in a machine integer (magnitude > approx. 2^63)?
Yes. Examples of this typically involve long (e.g. 1e7=#) sequences of algorithmically generated numbers (often involving primes and/or small offsets from powers of 2). That said, in many cases, these are intermediate results where the entire calculation could have been performed in a while loop. (And, yes, the current performance of J on these sorts of calculations has been pushing me towards personally implementing libgmp support as a replacement for J's 2x and 2r3 style numbers. I know other people are likely working on this, though, and it's a long haul.) Thanks, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
