examples/min01ks.mod
author Alpar Juttner <alpar@cs.elte.hu>
Sun, 05 Dec 2010 17:35:23 +0100
changeset 2 4c8956a7bdf4
permissions -rw-r--r--
Set up CMAKE build environment
alpar@1
     1
/* min01ks.mod - finding minimal equivalent 0-1 knapsack inequality */
alpar@1
     2
alpar@1
     3
/* Written in GNU MathProg by Andrew Makhorin <mao@gnu.org> */
alpar@1
     4
alpar@1
     5
/* It is obvious that for a given 0-1 knapsack inequality
alpar@1
     6
alpar@1
     7
      a[1] x[1] + ... + a[n] x[n] <= b,  x[j] in {0, 1}              (1)
alpar@1
     8
alpar@1
     9
   there exist infinitely many equivalent inequalities with exactly the
alpar@1
    10
   same feasible solutions.
alpar@1
    11
alpar@1
    12
   Given a[j]'s and b this model allows to find an inequality
alpar@1
    13
alpar@1
    14
      alfa[1] x[1] + ... + alfa[n] x[n] <= beta,  x[j] in {0, 1},    (2)
alpar@1
    15
alpar@1
    16
   which is equivalent to (1) and where alfa[j]'s and beta are smallest
alpar@1
    17
   non-negative integers.
alpar@1
    18
alpar@1
    19
   This model has the following formulation:
alpar@1
    20
alpar@1
    21
      minimize
alpar@1
    22
alpar@1
    23
         z = |alfa[1]| + ... + |alfa[n]| + |beta| =                  (3)
alpar@1
    24
alpar@1
    25
           = alfa[1] + ... + alfa[n] + beta
alpar@1
    26
alpar@1
    27
      subject to
alpar@1
    28
alpar@1
    29
         alfa[1] x[1] + ... + alfa[n] x[n] <= beta                   (4)
alpar@1
    30
alpar@1
    31
                              for all x satisfying to (1)
alpar@1
    32
alpar@1
    33
         alfa[1] x[1] + ... + alfa[n] x[n] >= beta + 1               (5)
alpar@1
    34
alpar@1
    35
                              for all x not satisfying to (1)
alpar@1
    36
alpar@1
    37
         alfa[1], ..., alfa[n], beta are non-negative integers.
alpar@1
    38
alpar@1
    39
   Note that this model has n+1 variables and 2^n constraints.
alpar@1
    40
alpar@1
    41
   It is interesting, as noticed in [1] and explained in [2], that
alpar@1
    42
   in most cases LP relaxation of the MIP formulation above has integer
alpar@1
    43
   optimal solution.
alpar@1
    44
alpar@1
    45
   References
alpar@1
    46
alpar@1
    47
   1. G.H.Bradley, P.L.Hammer, L.Wolsey, "Coefficient Reduction for
alpar@1
    48
      Inequalities in 0-1 Variables", Math.Prog.7 (1974), 263-282.
alpar@1
    49
alpar@1
    50
   2. G.J.Koehler, "A Study on Coefficient Reduction of Binary Knapsack
alpar@1
    51
      Inequalities", University of Florida, 2001. */
alpar@1
    52
alpar@1
    53
param n, integer, > 0;
alpar@1
    54
/* number of variables in the knapsack inequality */
alpar@1
    55
alpar@1
    56
set N := 1..n;
alpar@1
    57
/* set of knapsack items */
alpar@1
    58
alpar@1
    59
/* all binary n-vectors are numbered by 0, 1, ..., 2^n-1, where vector
alpar@1
    60
   0 is 00...00, vector 1 is 00...01, etc. */
alpar@1
    61
alpar@1
    62
set U := 0..2^n-1;
alpar@1
    63
/* set of numbers of all binary n-vectors */
alpar@1
    64
alpar@1
    65
param x{i in U, j in N}, binary, := (i div 2^(j-1)) mod 2;
alpar@1
    66
/* x[i,j] is j-th component of i-th binary n-vector */
alpar@1
    67
alpar@1
    68
param a{j in N}, >= 0;
alpar@1
    69
/* original coefficients */
alpar@1
    70
alpar@1
    71
param b, >= 0;
alpar@1
    72
/* original right-hand side */
alpar@1
    73
alpar@1
    74
set D := setof{i in U: sum{j in N} a[j] * x[i,j] <= b} i;
alpar@1
    75
/* set of numbers of binary n-vectors, which (vectors) are feasible,
alpar@1
    76
   i.e. satisfy to the original knapsack inequality (1) */
alpar@1
    77
alpar@1
    78
var alfa{j in N}, integer, >= 0;
alpar@1
    79
/* coefficients to be found */
alpar@1
    80
alpar@1
    81
var beta, integer, >= 0;
alpar@1
    82
/* right-hand side to be found */
alpar@1
    83
alpar@1
    84
minimize z: sum{j in N} alfa[j] + beta; /* (3) */
alpar@1
    85
alpar@1
    86
phi{i in D}: sum{j in N} alfa[j] * x[i,j] <= beta; /* (4) */
alpar@1
    87
alpar@1
    88
psi{i in U diff D}: sum{j in N} alfa[j] * x[i,j] >= beta + 1; /* (5) */
alpar@1
    89
alpar@1
    90
solve;
alpar@1
    91
alpar@1
    92
printf "\nOriginal 0-1 knapsack inequality:\n";
alpar@1
    93
for {j in 1..n} printf (if j = 1 then "" else " + ") & "%g x%d",
alpar@1
    94
   a[j], j;
alpar@1
    95
printf " <= %g\n", b;
alpar@1
    96
printf "\nMinimized equivalent inequality:\n";
alpar@1
    97
for {j in 1..n} printf (if j = 1 then "" else " + ") & "%g x%d",
alpar@1
    98
   alfa[j], j;
alpar@1
    99
printf " <= %g\n\n", beta;
alpar@1
   100
alpar@1
   101
data;
alpar@1
   102
alpar@1
   103
/* These data correspond to the very first example from [1]. */
alpar@1
   104
alpar@1
   105
param n := 8;
alpar@1
   106
alpar@1
   107
param a := [1]65, [2]64, [3]41, [4]22, [5]13, [6]12, [7]8, [8]2;
alpar@1
   108
alpar@1
   109
param b := 80;
alpar@1
   110
alpar@1
   111
end;