That is a good start.

GLPK can use this as a biasis for integer optimisation. The
output from simplex is a solution based on a continuous number
system. You want a discreet solution. It would be nice if just
rounding the continuous solution to the nearest integer produced
the answer. Usually it doesn't.

You must next call glp_intopt  using your simplex produced basis
as the parameter.

To obtain the answer you must then use glp_mip_col_val for
individual columns, or glp_print_mip for the entire solution. You
might like to also call glp_print_sol and compare how close the
actual values in glp_print_mip are to the values in glp_print_sol
if you just rounded them.

--
Nigel Galloway
[1][email protected]

On Sep 5 ’11, "name name" <[email protected]> wrote:

  I solve it with
    glp_simplex(lp, parm)  routine

On Mon, Sep 5, 2011 at 2:54 PM, Nigel Galloway
<[2][email protected]> wrote:

The question is how do you solve the problem, and how do you
obtain the solution.

When you make all these columns binary you have a Diophantine
Equation. GLPK can solve these, though there are other ways which
may be better.

GLPK will first solve it as a linear problem. It is then
necessary to perform Integer Optimization.

Having done this there are different routines to obtain the
integer optimal solution as opposed to the linear solution.
--
Nigel Galloway
[3][email protected]


On Mon, 05 Sep 2011 14:26 +0200, "name name"
<[4][email protected]> wrote:

On Mon, Sep 5, 2011 at 1:18 PM, Nigel Galloway
<[5][email protected]> wrote:

Do you obtain this value with GLPK.glp_get_mip_col_val ?

no with .glp_set_col_kind
Even if the type of variable is binary Lp give non integer value

--
Nigel Galloway
[6][email protected]

On Mon, 05 Sep 2011 09:52 +0200, "name name"
<[7][email protected]> wrote:


Hello,
I have a very strange problem
I solve un Lp problem but the and the value of my  variables are
not ineger I got x(0,1)=0.2... even thaught when I print the type
of my variables I got 3 coresponding to type binary( I got type
to binary for all variables with GLPK.glp_set_col_kind(lp, i,
GLPKConstants.GLP_BV)).
Can you tell my where the problem could be?
Thank you
Onor

_______________________________________________
Help-glpk mailing list

[8][email protected]
[9]https://lists.gnu.org/mailman/listinfo/help-glpk


--
[10]http://www.fastmail.fm - Email service worth paying for. Try it for fr
ee

  _______________________________________________
  Help-glpk mailing list
  [11][email protected]
  [12]https://lists.gnu.org/mailman/listinfo/help-glpk



--
[13]http://www.fastmail.fm - Does exactly what it says on the tin

References

1. mailto:[email protected]
2. mailto:[email protected]
3. mailto:[email protected]
4. mailto:[email protected]
5. mailto:[email protected]
6. mailto:[email protected]
7. mailto:[email protected]
8. mailto:[email protected]
9. https://lists.gnu.org/mailman/listinfo/help-glpk
  10. http://www.fastmail.fm/
  11. mailto:[email protected]
  12. https://lists.gnu.org/mailman/listinfo/help-glpk
  13. http://www.fastmail.fm/

-- 
http://www.fastmail.fm - mmm... Fastmail...

_______________________________________________
Help-glpk mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/help-glpk

Reply via email to