Download
/*
Set partition problem in Comet.
Problem formulation from
http://www.koalog.com/resources/samples/PartitionProblem.java.html
"""
This is a partition problem.
Given the set S = {1, 2, ..., n},
it consists in finding two sets A and B such that:
A U B = S,
|A| = |B|,
sum(A) = sum(B),
sum_squares(A) = sum_squares(B)
"""
This model uses a binary matrix to represent the sets.
Also, compare with other models which uses var sets:
* MiniZinc: http://www.hakank.org/minizinc/set_partition.mzn
* Gecode/R: http://www.hakank.org/gecode_r/set_partition.rb
This Comet model was created by Hakan Kjellerstrand (hakank@gmail.com)
Also, see my Comet page: http://www.hakank.org/comet
*/
// Licenced under CC-BY-4.0 : http://creativecommons.org/licenses/by/4.0/
import cotfd;
int t0 = System.getCPUTime();
int n = 16;
int num_sets = 2;
assert(n % num_sets == 0); // sanity: is the partition possible?
Solver<CP> m();
var<CP>{bool} a[1..num_sets, 1..n](m); // the set
Integer num_solutions(0);
exploreall<m> {
forall(i in 1..num_sets, j in i+1..num_sets) {
// same cardinality
m.post(sum(k in 1..n) a[i,k] == sum(k in 1..n) a[j,k]);
// same sum
m.post(sum(k in 1..n) k*a[i,k] == sum(k in 1..n) k*a[j,k]);
// same sum squared
m.post((sum(k in 1..n) (k*a[i,k])^2) == (sum(k in 1..n) (k*a[j,k])^2));
}
partition_sets(a);
// symmetry breaking (for num_sets == 2)
if (num_sets == 2)
m.post(a[1,1] == true);
} using {
labelFF(m);
num_solutions := num_solutions + 1;
cout << "sums: " << sum(j in 1..n) j*a[1,j] << endl;
cout << "sums squared: " << (sum(j in 1..n) (j*a[1,j])^2) << endl;
// show the partitions
forall(i in 1..num_sets) {
if ( sum(j in 1..n) a[i,j] > 0) {
cout << i << ": ";
forall(j in 1..n) {
if (a[i,j])
cout << j << " ";
}
cout << endl;
}
}
cout << endl;
}
cout << "\nnum_solutions: " << num_solutions << endl;
int t1 = System.getCPUTime();
cout << "time: " << (t1-t0) << endl;
cout << "#choices = " << m.getNChoice() << endl;
cout << "#fail = " << m.getNFail() << endl;
cout << "#propag = " << m.getNPropag() << endl;
//
// Partition the sets (binary matrix representation).
//
function void partition_sets(var<CP>{bool} [,] x) {
Solver<CP> m = x[1,1].getSolver();
range R1 = x.getRange(0);
range R2 = x.getRange(1);
forall(i in R1, j in R1 : i != j)
m.post(sum(k in R2) (x[i,k] && x[j,k]) == 0);
// ensure that all integers is in (exactly) one partition
m.post((sum(i in R1, j in R2) x[i,j]) == R2.getSize());
}