Best way to dispatch on two arguments

From: Markus Mottl (mottl@miss.wu-wien.ac.at)
Date: Sat Jan 23 1999 - 18:56:10 MET


From: Markus Mottl <mottl@miss.wu-wien.ac.at>
Message-Id: <199901231756.SAA28937@miss.wu-wien.ac.at>
Subject: Re: Okasaki's "Purely Functional Data Structures" translated to
To: bernardp@cli.di.unipi.it (Pierpaolo Bernardi)
Date: Sat, 23 Jan 1999 18:56:10 +0100 (MET)
In-Reply-To: <Pine.GSO.4.00.9901221221440.463-100000@carlotta.cli.di.unipi.it> from "Pierpaolo Bernardi" at Jan 22, 99 12:34:00 pm

Hello,

> > * A somewhat bigger change: the type of the list is not:
> >
> > type 'a ralist = Nil | Root of int * 'a tree * 'a ralist
> >
> > but now:
> >
> > type 'a t = (int * 'a tree) list
> >
> > I think that new users will understand details of implementation
> > easier with this change. It also comes closer to Okasaki's version.
>
> Okasaki in his paper on RALs presents both versions. The one I
> used is faster for Ocaml, so I suggest that you use the version I sent.
> Users should use the book or the paper to understand details.

I haven't read his paper, I just know his book - there is nothing about
the second possibility in it (maybe I have overlooked it).

Anyway, I have compared performance between the two (just for some
common operations). Your version was about 3% faster - not really much,
but I added it again to the repository and renamed the former version to
"sb_ralist2". This should satisfy both the speed-hungry and the advocats
of elegance...

By the way, chapter seven is now available, too. And I have made the
necessary changes so that "PhysicistsQueue" (chapter six) works.

As always:

  http://miss.wu-wien.ac.at/~mottl/ocaml_sources/intro.html

Best regards,
Markus

-- 
Markus Mottl, mottl@miss.wu-wien.ac.at, http://miss.wu-wien.ac.at/~mottl



This archive was generated by hypermail 2b29 : Sun Jan 02 2000 - 11:58:18 MET