Re: clpfd constraints not taking effect

Hi Markus,

On 24 August 2013 19:47, Markus Triska <[email protected]> wrote:
> James Hogan <[email protected]> writes:
>
>> Ah, I think I get it, so in the apply_preargs example, SWI's shallow
>> look at apply_preargs sees the latter two functors as taking the same
>> argument types, so it has to remember it in case the first fails or
>> other solutions required.
>
> Yes, with your current version, the second argument of both clauses is a
> term of the form [_|_] (functor '.', arity 2), so SWI's current, shallow
> indexing cannot distinguish them, and choice-points are left behind.
> Relatedly, tail call optimization cannot be applied as a consequence.

Fixing get_event_time/3 had a massive (positive) effect on performance!

>
>>>    ?- person(enosh).
>>>    true ;
>>>    true ;
>>>    false.
>>
>> yes, I've been trying to get rid of these cases, especially the final
>> false (which now makes much more sense).
>
> It probably makes sense to precompute the definition of person/1 before
> your actual constraint solving: You can use setof/3 to collect
> everything you consider a person, using the various rules you defined,
> and then use assertz/1 to create a single clause for each person. This
> makes further uses of person/1 deterministic when the argument is bound.

very cunning, and seems to work well.

I've now managed to make process_db/2 fully deterministic :)

Thanks again
-- 
James Hogan