It's been a while since I've programmed in Prolog. Today, I tried to make a simple program. It lists some facts of who belongs to the same family. If two people belong to the same family, they cannot give eachother gifts. I want to get all the people (or at least one person) to whom someone is allowed to give a gift.
family(john, jack).
family(matt, ann).
family(ann, jack).
family(jordan, michael).
family(michael, liz).
sameFamily(X, Y) :-
family(X, Y).
sameFamily(X, X) :-
false.
sameFamilySym(X, Y) :-
sameFamily(X, Y).
sameFamilySym(X, Y) :-
sameFamily(Y, X).
sameFamilyTrans(X, Z) :-
sameFamilySym(X, Y),
sameFamilySym(Y, Z).
gift(X, Y) :-
not(sameFamilyTrans(X, Y)).
Some queries if sameFamilyTrans/2 return false when they should in fact return true.
sameFamilyTrans/2 is obviously wrong. I think I need to keep a list of intermediate transitivities. Something like this:
sameFamilyTrans(X, Z, [Y|Ys]) :-
sameFamilySym(X, Y, []),
sameFamilyTrans(Y, Z, Ys).
But then I don't know how to call this.
P.S. I am using SWI-Prolog, if that makes any difference.
Yes, you were on the right track. The trick is to call the transitive closure with an empty accumulator, and check in each step whether a cycle is found (i.e., whether we have seen this member of the family before. As "false" has pointed out, the persons need to be instantiated already before going into the not, though.
So in sum, this works:
family(john, jack).
family(matt, ann).
family(ann, jack).
family(jordan, michael).
family(michael, liz).
sameFamily(X, Y) :-
family(X, Y).
sameFamilySym(X, Y) :-
sameFamily(X, Y).
sameFamilySym(X, Y) :-
sameFamily(Y, X).
sameFamilyTrans(X, Y, Acc) :-
sameFamilySym(X, Y),
not(member(Y,Acc)).
sameFamilyTrans(X, Z, Acc) :-
sameFamilySym(X, Y),
not(member(Y,Acc)),
sameFamilyTrans(Y, Z, [X|Acc]).
person(X) :- family(X, _).
person(X) :- family(_, X).
gift(X, Y) :-
person(X),
person(Y),
X \= Y,
not(sameFamilyTrans(X, Y, [])).
A bit of background: Transitive closure is not actually first-order definable (cf. https://en.wikipedia.org/wiki/Transitive_closure#In_logic_and_computational_complexity). So it can be expected that this would be a little tricky.
Negation is implemented in Prolog in a very rudimentary manner. You can essentially get a useful answer only if a negated query is sufficiently instantiated. To do this, define a relation person/1 that describes all persons you are considering. Then you can write:
gift(X,Y) :-
person(X),
person(Y),
\+ sameFamily(X,Y).
There is another issue with the definition of sameFamily/2.
Related
I am new to prolog and currently stuck trying to understand how to implement this.
I need a predicate to find the second and the second last elements of a list using recursion, so for example:
second_secondLast([1,2], X, Y). must return X=2, Y=1.
second_secondLast([1,2,3], X, Y). must return X=2, Y=2.
second_secondLast([1], X, Y). must print 'Error' and return false.
First, I have the error-checking clauses:
second_secondLast([], X, Y) :- print("Error"), !, fail.
second_secondLast([_], X, Y) :- print("Error"), !, fail.
Next, I tried something like this:
second_secondLast([Y,X],X,Y) :- !.
second_secondLast(L, X, Y) :-
second(L,X),
secondLast(L,Y).
second([_,S|_], X) :- X = S.
secondLast([P,_], Y) :- Y = P.
secondLast([F|R], Y) :- secondLast(R, Y).
However, the output using [1,2,3] is X=Y, Y=2.
I'm not sure if it is possible to force the output to be X=2 instead, or if there is a better method to do this.
First of all, the output X=Y, Y=2. has nothing to do with your program, it is an idiosyncracy of swipl (and maybe other interactive environments for Prolog implementations).
I think, your program looks fine, but you are asking for possible improvements.
second([_,S|_], S). is a more elegant version of your second([_,S|_], X) :- X = S..
Likewise, secondLast([P,_], P). is more elegant than your secondLast([P,_], Y) :- Y = P..
I would also prefer secondLast([_|R], Y) :- secondLast(R, Y). to your
secondLast([F|R], Y) :- secondLast(R, Y)..
Your error-checking clauses look fine to me.
You could also get rid of the predicate second and alter the definition of second_secondLast by using
second_secondLast([H,X|T], X, Y):-
secondLast([H,X|T], Y).
instead of your
second_secondLast(L, X, Y) :-
second(L,X),
secondLast(L,Y).
That change would also make it a bit more efficient.
Another possibility is to use
second_secondLast(L, X, Y):-
L= [_,X|_],
secondLast(L, Y).
Then you could also get rid of the predicate secondLast and alter the above clause to
second_secondLast(L, X, Y):-
L= [_,X|_],
append(_, [Y,_], L).
.
There is always a ton of possibilities...
I'm trying to rule to check if someone likes a friend other than the one stated in the argument. So for example,
likes(Alice,Bob).
likes(Bob,Alice).
likes(Alice,Jeff).
likes(Jeff,Alice).
I'm trying to create a rule friends(X,Y) that if both of them like each other, we look for another pair that either X or Y is linked to. Any help?
The first solution returns the first person found that one of the two people likes, in case these 2 people like each other.
likes(alice,bob).
likes(bob,alice).
likes(alice,jeff).
likes(jeff,alice).
friends_aux(X, Y, Z) :- likes(Y, Z), Z \= X.
friends_aux(X, Y, Z) :- likes(X, Z), Z \= Y.
friends(X, Y, Z) :- likes(X, Y), likes(Y, X), friends_aux(X, Y, Z), !.
And the second solution returns true if any of the two people have any other couple.
friends2_aux(X, Y) :- likes(Y, Z), Z \= X.
friends2_aux(X, Y) :- likes(X, Z), Z \= Y.
friends2(X, Y) :- likes(X, Y), likes(Y, X), friends2_aux(X, Y), !.
I attached an image of the results.
Results
Hopefully this is what you wanted.
I want to know why does the program goes in an infinite recursion in those cases:
?- love(kay, amanda).
and
?- love(rob, amanda).
And here is the code:
love(amanda, kay).
love(kay, geo).
love(geo, rob).
love(X, Y) :-
love(X, Z),
love(Z, Y).
love(X, Y) :-
love(Y, X).
First, your program always goes into an infinite loop. No matter what names you are using. Even ?- loves(amanda, kay). loops. To better see this rather ask ?- loves(amanda, kay), false. Why am I so sure that your program always loops?
To see this, we need some falsework. By adding goals false into your program, we get a new program requiring less (or equal) inferences.
love(amanda, kay) :- false.
love(kay, geo) :- false.
love(geo, rob) :- false.
love(X, Y) :-
love(X, Z), false,
love(Z, Y).
love(X, Y) :- false,
love(Y, X).
Because this fragment (called a failure-slice) does not terminate, your original program will not terminate. As you can see, all the persons have been removed. And thus actual names cannot influence the outcome.
If you want to fix commutativity, rather introduce a further predicate:
love2(X, Y) :- love1(X, Y).
love2(X, Y) :- love1(Y, X).
And to include the transitive closure, use closure/3:
love3(X, Y) :-
closure(love2, X, Y).
I'm currently trying to learn prolog. I hope you can help..
I have three rules:
reverse - retrieves the reverse of a list
startswith - checks if the second list is a prefix of the first list
suffix - checks if the first list is a suffix of the second list
reverse([H|T], Y) :- append(Z, [H], Y), reverse(T, Z).
reverse([], Y) :- Y = [].
startswith(_, []).
startswith([Xh|Xt], [Yh|Yt]) :- Xh=Yh, startswith(Xt, Yt).
suffix(X, Y) :- reverse(X, XR), reverse(Y, YR), startswith(YR,XR).
reverse and startswith seem to work as they should.
But suffix doesn't stop calculating. I cannot understand why?
In general, it's a bad idea to pass variables that have yet to be unified to predicates that you will unify later. Essentially, append(Z,[H],Y) is spiraling off into prolog never-never land because it's unifying Z over and over based on your version of Prolog's append.
Change
reverse([H|T], Y) :- append(Z, [H], Y), reverse(T, Z).
To
reverse([H|T], Y) :- reverse(T, Z), append(Z, [H], Y).
So that you unify Z before you pass it to append
Using recursion i need to find all blood relatives of any person in the family tree.
My attempt so far has failed.
Here is my code, with my attempt at the bottom
female(helen).
female(debbie).
female(louise).
female(yvonne).
female(belinda).
female(heather).
male(john).
male(andrew).
male(barry).
male(daniel).
male(charles).
parent(helen, debbie).
parent(helen, barry).
parent(helen, louise).
parent(john, debbie).
parent(john, barry).
parent(andrew, louise).
parent(debbie, yvonne).
parent(debbie, daniel).
parent(barry, charles).
parent(barry, belinda).
parent(louise, heather).
mother(X, Y) :-
female(X),
parent(X, Y).
father(X, Y) :-
male(X),
parent(X,Y).
child(X, Y) :-
parent(Y, X).
daughter(X, Y) :-
parent(Y, X),
female(X).
son(X, Y) :-
parent(Y,X),
male(X).
sister(X, Y) :-
female(X),
parent(Q,X),
parent(Q,Y).
brother(X, Y) :-
male(X),
parent(Q,X),
parent(Q,Y).
sibling(X, Y) :-
parent(Q,X),
parent(Q,Y),
X\=Y.
uncle(X, Y) :-
parent(P,Y),
brother(X,P).
aunt(X, Y) :-
parent(P,Y),
sister(X,P).
cousin(C, Cousin):-
parent(Parent,C),
sibling(Parent,AU),
child(Cousin,AU).
%Here is Relative
relative(An, Re):-
An\=Re,
parent(An, Re);
sibling(An, Re).
relative(An, Rela):-
parent(An, Child);
sibling(An, Rela),
relative(Child, Rela),
An\=Rela, C\=Rela.
Sort of works, but gets stuck in an infinite loop at the end.
Thanks.
not sure about 'relatives' (any person bound reachable in a parent/child relation ?), but your definition seems more complex than needed ( do you know what ; does ?).
I tried
relative(An, Re):-
parent(An, Re).
relative(An, Rela):-
parent(An, C),
relative(C, Rela).
that yields
16 ?- forall(relative(X,Y),writeln(X:Y)).
helen:debbie
helen:barry
helen:louise
john:debbie
john:barry
andrew:louise
debbie:yvonne
debbie:daniel
barry:charles
barry:belinda
louise:heather
helen:yvonne
helen:daniel
helen:charles
helen:belinda
helen:heather
john:yvonne
john:daniel
john:charles
john:belinda
andrew:heather
true.
edit I tried another relation, using a generalized parent/2, but still too permissive.
relative(Pers, Re):-
ancestor(Re, Pers) ; sibling(Pers, Re) ; cousin(Pers, Re) ; uncle(Re, Pers) ; aunt(Re, Pers).
ancestor(Anc, Pers) :- parent(Anc, Pers).
ancestor(Anc, Pers) :- parent(Anc, P), ancestor(P, Pers).
Maybe cousin/2 is too permissive also. Here is the graph
I guess that heather should have only luise,helen,andrew as relatives. It's this true ?
edit given latest comment, seems that the definition could be right. I get
24 ?- setln(X,relative(heather,X)).
andrew
barry
belinda
charles
daniel
debbie
helen
louise
yvonne
true.
that is everyone is related to heather apart john.
Here's one way that works, but it will sometimes produce duplicates. Using setof will give the unique collection. I avoided the miscellaneous relations and stuck with descendent or parent.
descendent(A, B) :-
parent(B, A).
descendent(A, B) :-
parent(C, A),
descendent(C, B).
relative(A, B) :-
descendent(B, A).
relative(A, B) :-
descendent(A, B).
relative(A, B) :-
descendent(A, C),
descendent(B, C),
A \= B.
setof(A, relative(heather, A), Relatives).
Relatives = [andrew,barry,belinda,charles,daniel,debbie,helen,louise,yvonne]
If you don't have setof, you can use the findall/3 and sort/2 ISO predicates:
findall(A, relative(heather, A), R), sort(R, Relatives).
Note that the solutions presented so far assume that all of the relatives have unique names. A general case of dealing with relatives with the same first name (and possibly the same last name) you would need to track and compare lineages for differences.