Subsets of variables

[Fabián Mancilla]

Hello everyone! I have a problem with sets and subsets definitions in a MIP formulation. I have 2 025; the firsts 225 are continue variables and the rest are binary variables. 

Because of this I write: 

k variables of the instance /k1*k2025/

l(k) binary variables(0 <= x <= 1) /k226*k2025/

And then I write:

VARIABLES    

z objetive

x(k) variables;

BINARY VARIABLES

x(l) variables binarias;

But gams give me: Error 184 - Domain list redefined

How I can fix this?

the complete code that I used is the following: 

$CALL GDXXRW 10teams.xlsx @teams.txt

$Title 10teams_ca

SETS

i filas matriz A /i1*i95/

j filas matriz M /j1*j135/

k  variables de la instania /k1*k2025/

l(k) subconjunto de variables binarys (0 <= x <= 1) /k226*k2025/

m1(k) subconjunto de variables continuas FX=1 /k1, k51, k105, k165, k225/

m2(k) subconjunto de variables continuas FX=0;

m2(k) = (not l(k)) - m1(k);

PARAMETERS

A(i,k) matriz A

M(j,k) matriz M

b(i)   vector b

g(j)   vector g

c(k)   vector de costos;

$GDXIN 10teams.gdx

$load A

$load M

$load b

$load g

$load c

$GDXIN

VARIABLES

z funcion objetivo

x(k) variables;

BINARY VARIABLES

x(l) variables binarias;

EQUATIONS

fobj funcion objetivo z

res1 restriccion 1

res2 restriccion 2;

fobj..    z =e= sum(k,c(k)*x(k));

res1(i).. sum(k,A(i,k)*x(k)) =l= b(i);

res2(j).. sum(k,M(j,k)*x(k)) =e= g(j);

x.fx(m1) = 1;

x.fx(m2) = 0;

Model Centro /all/;

option mip=coincbc,rmip=coincbc;

solve Centro USING MIP MINIMIZING z;

[Edson Cordeiro do Valle]

Fabi�n
Where are x(l) on your model?
You can substitute x(l) for another name: xx(l).
Why do you want the binaries a subset of the continuous variables?

Regards

Fabi�n Mancilla escreveu:

> --
> You received this message because you are subscribed to the Google
> Groups "gamsworld" group.
> To post to this group, send email to gams...@googlegroups.com.
> To unsubscribe from this group, send email to
> gamsworld+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/gamsworld?hl=en.

[babak saleckpay]

Hi
I think the easiest way is to seperate the variables :

k1     continious variables /1*225/
k2     binary variables       /225*2025/
;
Variables
x1(k1)
x2(k2)
;
BinaryVariables
x2
;

2010/12/22 Fabián Mancilla <mancill...@gmail.com>

--

You received this message because you are subscribed to the Google Groups "gamsworld" group.
To post to this group, send email to gams...@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

--
Babak Saleck Pay
Group of Industrial Engineering
M.Sc
Operations Research
University of Tehran
Tehran,Iran

[Fabián Mancilla]

Edson and Babak:

Well, I have a lot data of a LP problem in a GDX file. But I need consider some like binary and since the coeficients of the matrix "A" & "M" depend of the index "k" (System of the form Ax<=b; Mx=g), I define the subset "I" of "k" and I try to declaring "x(I)" binary variables. Thus the formulation of the equations is more easy (I think). However this seem not possible.

Regars and thanks