Re: MIU in Curry

From: Wolfgang Jeltsch <wolfgang_at_cs.ioc.ee>
Date: Fri, 26 Apr 2013 19:57:17 +0300

Am Freitag, den 26.04.2013, 18:38 +0200 schrieb Michael Hanus:
> On 04/26/2013 01:46 PM, Wolfgang Jeltsch wrote:
> > Anyway, I have now rewritten the applyRules function in a traditional
> > inductive style:
> >
> >> applyRules :: [Rule] -> Str -> [Str]
> >> applyRules [] _ = []
> >> applyRules (rule : rules) str = let
> >>
> >> str' = applyRule rule str
> >>
> >> in str' : applyRules rules str'
> >
> > Unfortunately, the program still doesn’t work.
>
> Since it was unclear to us why this program does not work,
> we made some experiments. It seems that this is actually
> an efficiency bug(?) of KiCS2 due to a complicated combination
> of functional patterns, unification, and search space explosion.
> If we try to separate these things, it works quite well,
> but the combination of all seems to cause a combinatorial explosion
> of the representation and traversal of the computation space.
> For instance, if we fix the number of rule applications,
> PAKCS (based on backtracking) computes the solutions much
> faster than KiCS2.
>
> Currently, we have no concrete idea of the reasons causing
> this problem, but, indeed, it is an interesting example
> to test many aspects of our implementation.

Thank you very much for this information. I already supposed that the
reason might be a pure efficiency issue.

Are there any other Curry implementations that support functional
patterns? Then I could try these.

Best wishes,
Wolfgang

_______________________________________________
curry mailing list
curry_at_lists.RWTH-Aachen.DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/curry
Received on Mon Apr 29 2013 - 09:49:05 CEST

This archive was generated by hypermail 2.3.0 : Thu Sep 19 2019 - 07:15:09 CEST