From: artboreb@netscape.net
To: martin_jambon@emailuser.net
Cc: caml-list@inria.fr
Subject: Re: [Caml-list] What kind of sort do Array.sort ?
Date: Mon, 08 May 2006 19:52:32 -0400 [thread overview]
Message-ID: <8C841119BCFEEF0-1A64-9602@mblkn-m05.sysops.aol.com> (raw)
In-Reply-To: <Pine.LNX.4.63.0605081521120.20895@munge>
Hi Martin,
Thanks for your sugestion, I've been lazy and didn't check the
fast_sort,
the random numbers population take less than 10 secs, and the time
measurement is still high? Is fast_sort a quick-sort algorithm?
Ocaml fast-sort
real 0m51.602s
user 0m50.352s
sys 0m0.002s
Best regards
Arturo Borquez
-----Original Message-----
From: Martin Jambon <martin_jambon@emailuser.net>
To: artboreb@netscape.net
Cc: caml-list@inria.fr
Sent: Mon, 8 May 2006 15:23:56 -0700 (PDT)
Subject: Re: [Caml-list] What kind of sort do Array.sort ?
You can try Array.fast_sort instead of Array.sort (see the docs).
Also, check that the time is actually spent sorting the array, not
generating random numbers.
Martin
On Mon, 8 May 2006, artboreb@netscape.net wrote:
> Dear Calmers,
> Toying with a compiler (builded in Ocaml of course) of a language I
developed > which emits C code
> I've written some bemchmarks to test it. For my surprise the same
program > written in Ocaml lasts
> almost 6 times. It's about sorting an integer array, the timings
are:
>
> Ocaml (ocamlopt -unsafe unix.cmxa qc.ml -o qc)
> real 1m10.727s
> user 1m9.074s
> sys 0m0.001s
>
> Mine (gcc -I /usr/local/include -O3 -s x12.ao.c -o x12 >
/usr/local/lib/libgc.a -lpthread)
> real 0m18.536s
> user 0m18.018s
> sys 0m0.002s
>
> times to take sorting 1000000 an array of 100 random ints
> the Ocaml code is
>
> open Format
>
> let _ =
> if Array.length Sys.argv = 2 then
> begin
> Random.init (Unix.getpid ());
> let a = Array.make (int_of_string Sys.argv.(1)) 0 in
> for i = 0 to 1000000 do
> for j = 0 to Array.length a - 1 do a.(j) <- Random.bits () done;
> Array.sort (fun x y -> if x > y then 1 else if x = y then 0 else -1)
> a
> done;
> for i = 0 to 99 do printf "%010d\n" a.(i) done
> end
>
> I dont attach the C code emitted because is about 280 lines long,
> my language (for the courious)
>
> "ao.ao" INC
>
> [Array Int Int] # RAPIDO RECURSIVO > [(a l r) l r a l -> (hl hr p)
> [l r] [l r >=] [l]
> [ [r] [a r -> p >= l r < & !] [r] [r 1 -] TREC (r)
> l l r =/= [a a r -> l <<- 1 +] IFTE (l)
> [l] [a l -> p <= l r < & !] [l] [l 1 +] TREC (l) l
> r l r =/= [a a l -> r <<- 1 -] IFTE] TREC (l)
> a p l <- hl l < [hl l 1 - Qsort] IFTE hr l > [l 1 + hr Qsort] IFTE >
] Qsort FUN
>
> ARGC 2 =
> [ Getpid Srand [;Int] ARGV 1 -> Atoi MAKE Length 1 - (a n)
> [[(i) a Rand i <<-] 0 n Upto a 0 n Qsort X-] 1000000 Times
> a [?n] Fold
> #[] [Rand "%010d" FMT] Rnd_strnum FUN > ] IFTE
>
> Best regards,
>
> Arturo Borquez
> ___________________________________________________
> Try the New Netscape Mail Today!
> Virtually Spam-Free | More Storage | Import Your Contact List
> http://mail.netscape.com
>
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>
--
Martin Jambon, PhD
http://martin.jambon.free.fr
Edit http://wikiomics.org, bioinformatics wiki
___________________________________________________
Try the New Netscape Mail Today!
Virtually Spam-Free | More Storage | Import Your Contact List
http://mail.netscape.com
next prev parent reply other threads:[~2006-05-08 23:52 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-08 21:31 artboreb
2006-05-08 22:23 ` [Caml-list] " Martin Jambon
2006-05-08 23:52 ` artboreb [this message]
2006-05-09 0:22 ` Erick Tryzelaar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=8C841119BCFEEF0-1A64-9602@mblkn-m05.sysops.aol.com \
--to=artboreb@netscape.net \
--cc=caml-list@inria.fr \
--cc=martin_jambon@emailuser.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox