Re: what wrong with this setof

 
> Date: Sun, 16 Mar 2014 23:13:55 +1300
> Subject: Re: [SWIPL] what wrong with this setof
> From: [email protected]
> To: [email protected]
> 
> > Hello,
> >
> > I changed my clause to this :
> >
> > display_siblings(X,Y) :-
> >      setof(X-Y, (siblings(X,Y), \+X=Y), Sibs),
> >                member((X-Y,Y), Sibs),
> >                \+ (Y@<X, member((Y,X), Sibs)).
> 
> (1) People are not siblings of themselves.
>     If your definition of siblings/2 allows this, it is wrong.
>     If it doesn't, the \+(X=Y) test is pointless.
>     Probable state: you need X\==Y inside siblings/2 and NOT here.
> 
> (2) findall/3, bagof/3, and setof/3 agree that the third
>     argument will be bound to a list of instances (filled-in
>     copies) of the first argument.  So when setof/3 succeeds,
>     each member of Sibs is a pair X-Y where X and Y are
>     presumably names of people.  This means that
>     member((X-Y,Y), Sibs) cannot possibly find anything that
>     looks like (X-Y,Y).  Similarly, member((Y,X), Sibs)
>     cannot possibly succeed because the elements of Sibs
>     look like (X-Y) *not* like (X,Y).
> 
Thanks, 
I m playing with the way setof does the output.  According to another member of this list I could do this :  > You can make pairs by using any two-place functor such as the popular -/2:
> 
> ?- setof(X-Y, parents(X, Y), Parents).
> Parents = [adam-eve, alice-bob, harry-sally]. 
  		 	   		  
-------------- next part --------------
HTML attachment scrubbed and removed