help-glpk
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Help-glpk] ILP problem , unexpected solution


From: satish raman
Subject: Re: [Help-glpk] ILP problem , unexpected solution
Date: Wed, 30 Dec 2009 00:43:01 +0530


On Tue, Dec 29, 2009 at 7:51 PM, Andrew Makhorin <address@hidden> wrote:
> I run your code with a call to glp_write_lp added.
> Below here is your instance passed to glp_simplex in cplex format:

> \* Problem: sample *\

> Maximize
>  obj: 0 x1

> Subject To
>  p: + x4 + x3 + x2 + x1 <= 5
>  q: - x2 + 10 x1 <= 5
>  r: - 2 x3 + x2 <= 5
>  o: - 0.5 x4 + 2 x3 = 1

> Bounds
>  0 <= x1 <= 1
>  0 <= x2 <= 1
>  0 <= x3 <= 1
>  0 <= x4 <= 1

> Generals
>  x1
>  x2
>  x3
>  x4

> End

Below here is the solution. (Note that in your program x1, x2,
x3, and x4 are declared as int.)

Problem:
Rows:       4
Columns:    4
Non-zeros:  10
Status:     OPTIMAL
Objective:  obj = 0 (MAXimum)

  No.   Row name   St   Activity     Lower bound   Upper bound
------ ------------ -- ------------- ------------- -------------
    1 p            B              1                           5
    2 q            NU             5                           5
    3 r            B             -1                           5
    4 o            NS             1             1             =

  No. Column name  St   Activity     Lower bound   Upper bound
------ ------------ -- ------------- ------------- -------------
    1 x1           B            0.5             0             1
    2 x4           NL             0             0             1
    3 x3           B            0.5             0             1
    4 x2           NL             0             0             1




Sorry for the botchup. I did some code juggling and posted incorrect code. I have posted the modified code at the end. I have upgraded to glpk 4.41. I actually want x1, x2 ,x3 ,x4 to take only binary values 0 and 1. I have used glp_set_col_kind function to set the column kind to  GLP_IV
  my foobar.lp(glp_write_lp function's output) is as follows

\* Problem: sample *\

Maximize
 obj: 0 x1

Subject To
 p: - x2 + 10 x1 <= 5
 q: - 2 x3 + x2 <= 5
 r: - 0.5 x4 + 2 x3 <= 5
 o: + x4 + x3 + x2 + x1 = 1

Bounds
 0 <= x1 <= 1
 0 <= x2 <= 1
 0 <= x3 <= 1
 0 <= x4 <= 1

Generals
 x1
 x2
 x3
 x4

End




the code
int main()
{
    glp_prob *lp;
    int ia[1+1000], ja[1+1000];
    double ar[1+1000];
    int  x1, x2, x3,x4;

   
    outfile = fopen("lpout.out","w");  
    float cur1, cur2, cur3, cur4;
    float bound_p, bound_q, bound_r;
   
    
    glp_iocp      parm;
    glp_smcp simplex_parm;
    int ret;
   
    
     /*10.00x1 + (-1.00)x2 < 5.00
             1.00x2 and +(-2.00)x3  < 5.00
             2.00x3 + (-0.50)x4  < 5.00
             */
     lp = glp_create_prob();
     glp_set_prob_name(lp, "sample");
     glp_set_obj_dir(lp, GLP_MAX);
   
    
     glp_add_rows(lp, 4);
    
     bound_p = 5 ;
     bound_q =  5;
     bound_r = 5 ;
    
     fprintf(outfile, "bound_p = %.2f bound_q %.2f bound_r =  %.2f \n ", bound_p, bound_q, bound_r);
    
     glp_set_row_name(lp, 1, "p");
     glp_set_row_bnds(lp, 1, GLP_UP, -DBL_MAX, bound_p);
     glp_set_row_name(lp, 2, "q");
     glp_set_row_bnds(lp, 2, GLP_UP, -DBL_MAX, bound_q);
     glp_set_row_name(lp, 3, "r");
     glp_set_row_bnds(lp, 3, GLP_UP, -DBL_MAX, bound_r);
   
     glp_set_row_name(lp, 4, "o");
     glp_set_row_bnds(lp, 4, GLP_FX, 1, 1);
                                    
     glp_add_cols(lp, 4);
     glp_set_col_name(lp, 1, "x1");
     glp_set_col_bnds(lp, 1, GLP_DB, 0, 1);
     glp_set_col_kind(lp, 1, GLP_IV);
     
     glp_set_col_name(lp, 2, "x2");
     glp_set_col_bnds(lp, 2, GLP_DB, 0, 1);
     glp_set_col_kind(lp, 2, GLP_IV);
    
     glp_set_col_name(lp, 3, "x3");
     glp_set_col_bnds(lp, 3, GLP_DB, 0, 1);
     glp_set_col_kind(lp, 3, GLP_IV);
    
     glp_set_col_name(lp, 4, "x4");
     glp_set_col_bnds(lp, 4, GLP_DB, 0, 1);
     glp_set_col_kind(lp, 4, GLP_IV);
            
    
    
     ia[1] = 1, ja[1] = 1, ar[1] = 10.0;
     ia[2] = 1, ja[2] = 2, ar[2] = -1.0;
     ia[3] = 1, ja[3] = 3, ar[3] = 0.0;
     ia[4] = 1, ja[4] = 4, ar[4] = 0.0;
    
    
  
    
     ia[5] = 2, ja[5] = 1, ar[5] = 0.0;
     ia[6] = 2, ja[6] = 2, ar[6] = 1.0;
     ia[7] = 2, ja[7] = 3, ar[7] = -2.0;
     ia[8] = 2, ja[8] = 4, ar[8] = 0.0;
    
    
     ia[9] = 3, ja[9] = 1, ar[9] = 0.0;
     ia[10] = 3, ja[10] = 2, ar[10] = 0.0;
     ia[11] = 3, ja[11] = 3, ar[11] = 2.0;
     ia[12] = 3, ja[12] = 4, ar[12] = -0.5;
    
     fprintf(outfile, "x1 + x2 +x3 +x4 = 1 \n");
     /*x1 + x2+ x3 + x4  = o */
     ia[13] = 4, ja[13] = 1, ar[13] = 1;
     ia[14] = 4, ja[14] = 2, ar[14] = 1;
     ia[15] = 4, ja[15] = 3, ar[15] = 1;
     ia[16] = 4, ja[16] = 4, ar[16] = 1;
         
     glp_load_matrix(lp, 16, ia, ja, ar);
    
     glp_write_lp(lp, NULL, "foobar.lp");
    
     glp_init_smcp(&simplex_parm);
     simplex_parm.presolve = GLP_OFF;
    
     if(glp_simplex(lp, &simplex_parm) != 0)
     {
         printf("failure of simplex\n");
         exit(-1);
     }


    
    x1 = glp_get_col_prim(lp, 1);
    x2 = glp_get_col_prim(lp, 2);
    x3 = glp_get_col_prim(lp, 3);
    x4 = glp_get_col_prim(lp, 4);
    
    fprintf(outfile,"\n x1 = %d; x2 = %d; x3 = %d x4 = %d \n", x1, x2, x3, x4);
  
    glp_delete_prob(lp);

    return 0;
}

I get x1 , x2, x3 and x4 all zero's
I think The right solution is x1 can't be one and one of other 3 variables should be 1. I don't know if i am doing something silly.
Is it okay to have x1, x2 , x3, x4 to be int type? Are there any special type that i should use.
thanks for replying so far.
thanks in advance for future help,
satish



reply via email to

[Prev in Thread] Current Thread [Next in Thread]