All of these should run at about the same speed. Any differences are due to the vagaries of running benchmarks. Timing differences of less than 10% (say) are difficult to distinguish.
----- Original Message ----- From: bill lam <[EMAIL PROTECTED]> Date: Wednesday, December 12, 2007 7:22 Subject: Re: [Jprogramming] test for unique To: Programming forum <[email protected]> > I found that 0&e. could be marginally faster than *./ > Testing on x and y, both (0 [EMAIL PROTECTED] ~:) and ([: -. 0 e. ~:) are > faster than *./@~: > > Also (0 [EMAIL PROTECTED] ~:) is faster than ([: -. 0 e. ~:), is there any > particular reason > or it is just random. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
