You are right and if you want fast fib's, look at http://www.jsoftware.com/jwiki/Essays/Fibonacci_Sequence
But I wanted a solution as close as possible to those used in the other languages. Perhaps the conclusion should be: don't use J (as your first choice) for non memoized recurrent solutions. R.E. Boss > -----Oorspronkelijk bericht----- > Van: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Namens Viktor Cerovski > Verzonden: donderdag 3 januari 2008 15:37 > Aan: [email protected] > Onderwerp: Re: [Jgeneral] Performance comparison > > > > R.E. Boss wrote: > > > > In http://www.jsoftware.com/pipermail/chat/2008-January/000869.html the > > link http://cgi.cse.unsw.edu.au/~dons/blog is mentioned, at the end of > > which > > I found > > > > Language Time (N=36) > > Ruby (1.8.5) 64.26s > > Python (2.4) 25.16s > > Haskell (GHC 6.8) 0.48s > > Parallel Haskell (GHC 6.8) 0.42s > > > > This was for the most 'naive fibonacci algorithm': > > > > fib=: 3 : 0 > > if. y e. 0 1 do. 1 else. (fib y-1) + fib y-2 end. > > ) > > > > On my machine I got > > > > ts'fib 35' > > 213.43542 52736 > > > Directory entry for M. gives a very fast program: > > fib=: 3 : 0 M. > if. 1>:y do. y else. (fib y-1)+fib y-2 end. > ) > and that will speed up 'fib 35' about 10000 times, > which should be considerably faster than all other programs > from your list. > > Possibly even faster code, especially if you want all the > Fibonacci numbers below and including the 35-th, is: > > (,[EMAIL PROTECTED])^:34(1 0) > > -- > View this message in context: http://www.nabble.com/Performance- > comparison-tp14595315s24193p14597960.html > Sent from the J General mailing list archive at Nabble.com. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
