English version
Accueil     À propos     Téléchargement     Ressources     Contactez-nous    

Ce site est rarement mis à jour. Pour les informations les plus récentes, rendez-vous sur le nouveau site OCaml à l'adresse ocaml.org.

Browse thread
[Caml-list] Caml Machine and GP
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2003-07-23 (19:05)
From: Lukasz Stafiniak <l_stafiniak@h...>
Subject: [Caml-list] Caml Machine and GP
Hi precious List!

I work on genetic programming with typed terms. I would like to use
OCaml bytecode-system to run my programs.

Evaluation scheme I've used so far: variables by de Brujin indices,
beta-reduction by substitution, recurrence by rational trees (with
mutable reference cells filled when first encountered in
evaluation). I wonder how much of a speedup could I get? -- I guess
MUCH. How big are the overheads? I guess for not computationally
involved tasks evaluation would be still better, due to the
overheads (but these dont give the kicks).

How should I do this? Look around in Toplevel, use the path:

Translcore.transl_exp --> Bytegen.compile_phrase -->
Emitcode.to_memory --> Meta.reify_bytecode?

Say, I put a function in Typedtree.expression at one side, and get the
result at the other side? How about treatment of types -- do I need to
supply them?  My terms are typed, but this would involve translating
types as well, unnecessarily. More -- how to call functions from
outside (perhaps use the environment argument -- what to put
there)... Best way to learn these things is reading the source, I
guess? Can I catch exceptions from code run this way -- I guess I can,
Toplevel does it probably.

I also welcome Hi! responses from people interested in genetic
programming. I plan to feature (type-based or some other) termination
of my programs, and I've developed some generalization algorithms to
use in "intelligent" recombination.

Best wishes,

Uwaga! Do koñca sierpnia przed³u¿yli¶my promocje, do pakietów
wielostanowiskowych dok³adamy PenDrive  Sprawd¼:

To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners