Probably. The Java BigInteger classes are not particularly fast, and do not seem to be a priority to Sun. Therefore Clojure is not competitive on large integer algorithms.
On Wed, Mar 11, 2009 at 2:21 AM, Tassilo Horn <tass...@member.fsf.org>wrote: > > Phil Hagelberg <p...@hagelb.org> writes: > > Hi Phil, > > >> If not, is there better way than inserting gazillions of printlns to > >> check why and where a function doesn't do the right thing? > > > > Most definitely! Break your functions up into smaller pieces, then > > write tests for them using test-is. If your functions are hard to > > test, it's probably because they need to be broken out differently. > > Yeah, might be. To get started with clojure I tried to translate the > Miller-Rabin pseudo prime algorithm from wikipedia. Basically it seems > to work correctly, but the preformance is really bad for numbers above a > certain limit. > > Investigated it a bit more (doing the algorithm step by step in the > repl) and now I know what's the culprit: The `expt' function from the > math contrib library is dead slow. > > For number theory you often need things like > > (mod (expt n exp) m) > > where the exponent may be very huge. Is that something which cannot be > done faster in Clojure because of the Java Integer/BigInteger stuff? > That would be a shame! I just fell in love with clojure cause the code > looks that concise and right! > > Bye, > Tassilo > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -~----------~----~----~----~------~----~------~--~---
