1. CSPlib
  2. Problems
  3. prob049
  4. Models
  5. set_partition_ortools.py

049: Number Partitioning

Download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
# Copyright 2010 Hakan Kjellerstrand hakank@gmail.com
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
 
"""
 
  Set partition problem in Google CP Solver.
 
  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
  * Comet: http://hakank.org/comet/set_partition.co
  * Gecode: http://hakank.org/gecode/set_partition.cpp
  * ECLiPSe: http://hakank.org/eclipse/set_partition.ecl
  * SICStus: http://hakank.org/sicstus/set_partition.pl
 
  This model was created by Hakan Kjellerstrand (hakank@gmail.com)
  Also see my other Google CP Solver models: http://www.hakank.org/google_or_tools/
"""
 
import sys
 
from constraint_solver import pywrapcp
 
 
#
# Partition the sets (binary matrix representation).
#
def partition_sets(x, num_sets, n):
    solver = x.values()[0].solver()
 
    for i in range(num_sets):
        for j in range(num_sets):
            if i != j:
                b = solver.Sum([x[i,k]*x[j,k] for k in range(n)])
                solver.Add(b == 0)
 
    # ensure that all integers is in
    # (exactly) one partition
    b = [x[i,j] for i in range(num_sets) for j in range(n) ]
    solver.Add(solver.Sum(b) == n)
 
 
def main(n=16,num_sets=2):
 
    # Create the solver.
    solver = pywrapcp.Solver('Set partition')
 
    #
    # data
    #
    print "n:", n
    print "num_sets:", num_sets
    print
 
    # Check sizes
    assert n % num_sets == 0, "Equal sets is not possible."
 
    #
    # variables
    #
 
    # the set
    a = {}
    for i in range(num_sets):
        for j in range(n):
            a[i,j] = solver.IntVar(0, 1, 'a[%i,%i]' % (i,j))
 
    a_flat = [a[i,j] for i in range(num_sets) for j in range(n)]
 
    #
    # constraints
    #
 
    # partition set
    partition_sets(a, num_sets, n)
 
    for i in range(num_sets):
        for j in range(i, num_sets):
 
            # same cardinality
            solver.Add(solver.Sum([a[i,k] for k in range(n)])
                       ==
                       solver.Sum([a[j,k] for k in range(n)]))
 
            # same sum
            solver.Add(solver.Sum([k*a[i,k] for k in range(n)])
                       ==
                       solver.Sum([k*a[j,k] for k in range(n)]))
 
 
            # same sum squared
            solver.Add(solver.Sum([(k*a[i,k])*(k*a[i,k])
                                   for k in range(n)])
                       ==
                       solver.Sum([(k*a[j,k])*(k*a[j,k])
                                   for k in range(n)]))
 
 
    # symmetry breaking for num_sets == 2
    if num_sets == 2:
        solver.Add(a[0,0] == 1)
 
 
    #
    # search and result
    #
    db = solver.Phase(a_flat,
                 solver.INT_VAR_DEFAULT,
                 solver.INT_VALUE_DEFAULT)
 
    solver.NewSearch(db)
 
 
    num_solutions = 0
    while solver.NextSolution():
        a_val = {}
        for i in range(num_sets):
            for j in range(n):
                a_val[i,j] = a[i,j].Value()
 
        sq = sum([(j+1)*a_val[0,j] for j in range(n)])
        print "sums:", sq
        sq2 = sum([((j+1)*a_val[0,j])**2 for j in range(n)])
        print "sums squared:", sq2
 
        for i in range(num_sets):
            if sum([a_val[i,j] for j in range(n)]):
                print i+1, ":",
                for j in range(n):
                    if a_val[i,j] == 1:
                        print j+1,
                print
 
        print
        num_solutions += 1
 
    solver.EndSearch()
 
    print
    print "num_solutions:", num_solutions
    print "failures:", solver.Failures()
    print "branches:", solver.Branches()
    print "WallTime:", solver.WallTime()
 
 
n = 16
num_sets = 2
if __name__ == '__main__':
    if len(sys.argv) > 1:
        n = int(sys.argv[1])
    if len(sys.argv) > 2:
        num_sets = int(sys.argv[2])
 
    main(n, num_sets)