I would say that with good reducers, the reduce approach can eliminate seq creation and thus be at least a bit faster and less demanding for GC.
On Thu, Feb 13, 2014 at 4:16 PM, Michał Marczyk <michal.marc...@gmail.com>wrote: > str is much faster with apply than with reduce; in fact, with reduce > it is O(n^2), whereas with apply it is O(n). (The constants are better > with apply too.) > > In general, (apply foo ...) takes advantage of any optimizations the > author of foo might be aware of for speeding up its operation when > given multiple arguments. (reduce foo ...) is always a simple-minded > fold. Not that that's never useful; it is with reducers and with > anonymous functions. > > > On 12 February 2014 21:06, Vincent <vhenneb...@gmail.com> wrote: > > On Wednesday, February 12, 2014 2:47:07 PM UTC, Stuart Sierra wrote: > >> > >> On Wednesday, February 12, 2014 8:46:41 AM UTC-5, Vincent wrote: > >>> > >>> On a slightly different topic: why reduce and not apply? > >> > >> > >> The implementation of `+` with more than 2 arguments uses `reduce` > >> internally. So they amount to the same thing. > > > > > > Ah ok, interesting. I do seem to have, in the past, come across a > function > > that was significantly faster when used with apply rather than reduce, > but I > > can't remember which one. > > > > > >> > >> There isn't really a performance difference: > >> > >> user=> (dotimes [i 7] (time (reduce + (range 10000000)))) > >> "Elapsed time: 354.475 msecs" > >> "Elapsed time: 346.235 msecs" > >> "Elapsed time: 348.124 msecs" > >> "Elapsed time: 348.894 msecs" > >> "Elapsed time: 379.9 msecs" > >> "Elapsed time: 356.337 msecs" > >> "Elapsed time: 362.788 msecs" > >> nil > >> user=> (dotimes [i 7] (time (apply + (range 10000000)))) > >> "Elapsed time: 360.067 msecs" > >> "Elapsed time: 353.281 msecs" > >> "Elapsed time: 345.694 msecs" > >> "Elapsed time: 355.162 msecs" > >> "Elapsed time: 346.511 msecs" > >> "Elapsed time: 350.61 msecs" > >> "Elapsed time: 353.674 msecs" > >> nil > > > > > > Thanks, > > Vincent > > > > -- > > 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 > > Note that posts from new members are moderated - please be patient with > your > > first post. > > 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 > > --- > > You received this message because you are subscribed to the Google Groups > > "Clojure" group. > > To unsubscribe from this group and stop receiving emails from it, send an > > email to clojure+unsubscr...@googlegroups.com. > > For more options, visit https://groups.google.com/groups/opt_out. > > -- > 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 > Note that posts from new members are moderated - please be patient with > your first post. > 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 > --- > You received this message because you are subscribed to the Google Groups > "Clojure" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > -- 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 Note that posts from new members are moderated - please be patient with your first post. 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 --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
