Answer Set Programming: Group into two sets so that those who like each other are in same set, and dislike = different set - answer-set-programming

I'm basically a beginner to Answer Set Programming (CLINGO), so I've been attempting this problem for hours now.
person(a;b;c;d;e;f).
likes(b,e; d,f).
dislikes(a,b; c,e).
People who like each other must be in the same set, and cannot be in the same set as someone they dislike.
So the output should be:
b,e | a, c, d,f
I know the logic behind it; partition it so that if an element is in both likes & dislikes, then it should be in its own set, and everything else in the other. But this is declarative programming, so I'm not sure how to tackle this. Any help would be appreciated.


Try this one, it should work for you:
person(a;b;c;d;e;f).
like(b,e; d,f).
dislike(a,b; c,e).
group(1..2).
% every person belongs to one group only.
1{in(S,G): group(G)}1 :- person(S).
% no two persons who do dislike each other are in the same group
:- in(X, G), in(Y, G), dislike(X,Y).
#show in/2.
The result you'll get is:
a & b are in different group.
and c & e are in different group.
The result you can get is like:

Related

How to create a variable in Matlab where certain subjects are coded as 1?

I want to create a variable called 'flag_artifact' where certain subjects from my dataset (for whom I know have bad quality images) are coded as e.g., 1. My dataset is stored in a table T with a certain number of rows and 'subject' is the 1st column in the table.
I managed to do it by creating a for loop. Surely there is a more efficient way to do this by perhaps directly creating the variable and using less lines of code? Could anyone give some advice?
Thank you very much! Here is what I have:
flag_artifact = {'T_300'}; %flagging subject number 300 for example
for i = 1:size(T,1)
if isequal(table2cell(T(i, 1)), flag_artifact)
T(i, 1) = {'1'};
end
end
However, when creating the variable flag_artifact = {'T_300'}, I would like it to include more than one subject. I tried using flag_artifact = {'T_300'}; {'T_301'}, as well as flag_artifact = {'T_300', 'T_301'} but it doesn't work because these subject identifiers do not get replaced with 1s.

How to replace choicerule to reduce "meaningless" answers that kill grounding process using asp (clingo)

I'm currently working on an answer set program to create a timetable for a school.
The rule base I use looks similar to this:
teacher(a). teacher(b). teacher(c). teacher(d). teacher(e). teacher(f).teacher(g).teacher(h).teacher(i).teacher(j).teacher(k).teacher().teacher(m).teacher(n).teacher(o).teacher(p).teacher(q).teacher(r).teache(s).teacher(t).teacher(u).
teaches(a,info). teaches(a,math). teaches(b,bio). teaches(b,nawi). teaches(c,ge). teaches(c,gewi). teaches(d,ge). teaches(d,grw). teaches(e,de). teaches(e,mu). teaches(f,de). teaches(f,ku). teaches(g,geo). teaches(g,eth). teaches(h,reli). teaches(h,spo). teaches(i,reli). teaches(i,ku). teaches(j,math). teaces(j,chem). teaches(k,math). teaches(k,chem). teaches(l,deu). teaches(l,grw). teaches(m,eng). teaches(m,mu). teachs(n,math). teaches(n,geo). teaches(o,spo). teaches(o,fremd). teaches(p,eng). teaches(p,fremd). teaches(q,deu). teaches(q,fremd). teaches(r,deu). teaches(r,eng). teaches(s,eng). teaches(s,spo). teaches(t,te). teaches(t,eng). teaches(u,bio). teaches(u,phy).
subject(X) :- teaches(_,X).
class(5,a). class(5,b). class(6,a). class(6,b). class(7,a). class(7,b). class(8,a). class(8,b). class(9,a). class(9,b). class(10,a). class(10,b).
%classes per week (for class 5 only at the moment)
classperweek(5,de,5). classperweek(5,info,0). classperweek(5,eng,5). classpereek(5,fremd,0). classperweek(5,math,4). classperweek(5,bio,2). classperweek(5,chem,0). classperweek(5,phy,0). classperweek(5,ge,1). classperweek(5,grw,0). cassperweek(5,geo,2). classperweek(5,spo,3). classperweek(5,eth,2). classperwek(5,ku,2). classperweek(5,mu,2). classperweek(5,tec,0). classperweek(5,nawi,0) .classperweek(5,gewi,0). classperweek(5,reli,2).
room(1..21).
%for monday to friday
weekday(1..5).
%for lesson 1 to 9
slot(1..9).
In order to creat a timetable I wanted to create every possible combination of all predicats I'm using and then filter all wrong answers.
This is how I created a timetable:
{timetable(W,S,T,A,B,J,R):class(A,B),teacher(T),subject(J),room(R)} :- weekday(W), slot(S).
Up to this point everything works, except that this solution is probably relatively inefficient.
To filter that no class uses the same room at the same time I formulated the following constraint.
:- timetable(A,B,C,D,E,F,G), timetable(H,I,J,K,L,M,N), A=H, B=I, G=N, class(D,E)!=class(K,L).
It looks like this makes to problem so big that the grounding fails, because I get the following error message
clingo version 5.4.0
Reading from timetable.asp
Killed
Therefore, I was looking for a way to create different instances of timetable without getting too many "meaningless" answers created by the choiserule.
One possibility I thought of is to use a negation cycle. So you could replace the choiserule
{a;b} with a :- not b. b :- not a. and exclude all cases where rooms are occupied twice.
Unfortunately I do not understand this kind of approach enough to apply it to my problem.
After a lot of trial and error (and online search), I have not found a solution to eliminate the choicerule and at the same time eliminate the duplication of rooms and teachers at the same time.
Therefore I wonder if I can use this approach for my problem or if there is another way to not create many pointless answersets at all.
edit: rule base will work now and updated the hours per lesson for class 5

I think you're looking for something like:
% For each teacher and each timeslot, pick at most one subject which they'll teach and a class and room for them.
{timetable(W,S,T,A,B,J,R):class(A,B),room(R),teaches(T,J)} <= 1 :- weekday(W);slot(S);teacher(T).
% Cardinality constraint enforcing that no room is occupied more than once in the same timeslot on the timetable.
:- #count{uses(T,A,B,J):timetable(W,S,T,A,B,J,R)} > 1; weekday(W); slot(S); room(R).
to replace your two rules.
Note that this way clingo won't generate spurious ground terms for teachers teaching a subject they don't know. Additionally by using a cardinality constraint as opposed to a binary clause, you get a big-O reduction in the grounded size (from O(n^2) in the number of rooms to O(n)).
Btw, you may be missing answers because of typos in the input. I would suggest phrasing it as:
teacher(a;b;c;d;e;f;g;h;i;j;k;l;m;n;o;p;q;r;s;t;u).
teaches(
a,info;
a,math;
b,bio;
b,nawi;
c,ge;
c,gewi;
d,ge;
d,grw;
e,de;
e,mu;
f,de;
f,ku;
g,geo;
g,eth;
h,reli;
h,spo;
i,reli;
i,ku;
j,math;
j,chem;
k,math;
k,chem;
l,deu;
l,grw;
m,eng;
m,mu;
n,math;
n,geo;
o,spo;
o,fremd;
p,eng;
p,fremd;
q,deu;
q,fremd;
r,deu;
r,eng;
s,eng;
s,spo;
t,te;
t,eng;
u,bio;
u,phy
).
subject(X) :- teaches(_,X).
class(
5..10,a;
5..10,b
).
%classes per week (for class 5 only at the moment)
classperweek(
5,de,5;
5,info,0;
5,eng,5;
5,fremd,0;
5,math,4;
5,bio,2;
5,chem,0;
5,phy,0;
5,ge,1;
5,grw,0;
5,geo,2;
5,spo,3;
5,eth,2;
5,ku,2;
5,mu,2;
5,tec,0;
5,nawi,0;
5,gewi,0;
5,reli,2
).
room(1..21).
%for monday to friday
weekday(1..5).
%for lesson 1 to 9
slot(1..9).

How do I generate a fixed sized list of facts (duplicates included)?

I'm new to ASP & Clingo and I need to work on a project for school. I thought about some basic music generator.
For now, I need to generate notes (I'm sticking with C major for now). I also want to generate them randomly and I don't know how to do that. How can I make the following code generate a random sequence of notes (duplicates too)?
note(c;d;e;f;g;a;b).
20 { play(X) : note(X)} 30.
#show play/1.
So far, the code won't allow for more than 7 as the upper bound, because it won't show duplicate notes.
Current output: play(b) play(g) play(e) play(c)
Wanted output: play(d) play(g) play(f) ...[20-30 randomly generated notes]
I want to be able to add constraints later (such as this note should not be followed by that note, and so on). I appreciate any tips since I know so little about this.

An answer set is a set. The atoms have no order and duplicates are not possible because it is a set.
You want to guess one note for each beat.
beat(1..8).
1 { play(N,B) : note(N) } 1 :- beat(B).

How should list be represented in ASP (Answer Set Programming)?

A processor 'a' takes care the header 'a' of a message 'a_b_c_d' and passes the payload 'b_c_d' to the another processor in the next level as following:
msg(a, b_c_d).
pro(a;b;c;d).
msg(b, c_d) :- pro(X), msg(X, b_c_d).
msg(c, d) :- pro(X), msg(X, c_d).
msg(d) :- pro(X), msg(X, d).
#hide. #show msg/2. #show msg/1.
How should I represent list 'a_b_c_d' in ASP, and change the above to general cases?

No, official way, but I think most people don't realize you can construct cons-cells in ASP.
For instance, here's how you can get items for all lists of length 5 from elements 1..6
element(1..6).
listLen(empty, 0).
listLen(cons(E, L), K + 1) :- element(E); listLen(L, K); K < 5.
is5List(L) :- listLen(L, 5).
#show is5List/1.
resulting in
is5List(cons(1,cons(1,cons(1,cons(1,cons(1,empty))))))
is5List(cons(1,cons(1,cons(1,cons(1,cons(2,empty))))))
is5List(cons(1,cons(1,cons(1,cons(1,cons(3,empty))))))
...

There is no 'official' way to handle lists in ASP as far as I know. But, DLV has built-in list handling similar to Prolog's.
The way you implement a list, the list itself cannot be used as a term and thus what if you want to bind between variables in the list and other elements of a rule? Perhaps you would like something such as p(t, [q(X), q(Y)]) :- X != Y.
You can try implementing a list as (a, b, c) and an append predicate but the problem is ASP requires grounding before computing answer-sets. Consequently a list defined in this way whilst more like lists in Prolog would mean the ground-program contains all ground-instances of all possible lists (explosion) regardless of whether they are used or not.
I therefore come back to my first point, try using DLV instead of Clingo if possible (for this task, at least).

By using index, I do have a way to walk a list, however, I do not know this is the official way to handle a list in ASP. Could someone has more experience in ASP give us a hand? Thanks.
index(3,a). index(2,b). index(1,c). index(0,d).
pro(a;b;c;d). msg(3,a).
msg(I-1,N) :- pro(P), msg(I,P), index(I,P), I>0, index(I-1,N).
#hide. #show msg/2.

You can use s(ASP) or s(CASP) ASP systems. Both of them support list operations like prolog. You might need to define the list built-in in ASP .

PigLatin - Removing A Tuple/Field From Relation

This is my first post on StackOverflow, so pardon me in advance if this is too lengthy and/or vague.
I have a Pig relation as follows:
my_relation = LOAD '/path/to/data' USING PigStorage(',')
AS (f1:chararray, f2:chararray, f3:chararray);
Now, I wanted to filter out the field 'f3' from the above relation 'my_relation'. I know I could do it like so:
my_new_relation = FOREACH my_relation GENERATE my_relation.f1, my_relation.f2;
The problem with this method comes when I have a large number of fields/tuples in the 'my_relation' relation. Say, my_relation had 900 fields/tuples, and I wanted all of them but for one. Now, with the above method, I'd have to list out 899 fields/tuples after my 'GENERATE' keyword!
My question: Is there an easy way to filter out a handful of fields/tuples from a relation in PigLatin?
Prior: My prior on Apache Pig and PigLatin in general is very weak (as can be told by the difficulty of this question). I'm still reading through the Pig documentation found here.
Thanks for reading this question! Any/all help is appreciated!

First of all, your syntax is not quite right. If you wanted to just save the fields f1 and f2, you would do it like this:
my_new_relation = FOREACH my_relation GENERATE f1, f2;
As to your question, you can use a project-range expression:
my_new_relation = FOREACH my_relation GENERATE f1 .. f345, f347 .. f900;

Also you can write your own loader where you can specify which all columns you want to load.