help-glpk
[Top][All Lists]

## [Help-glpk] how can I evaluate a previous solution within MathProg Model

 From: Rafael Torres Subject: [Help-glpk] how can I evaluate a previous solution within MathProg Model? Date: Thu, 14 Jun 2012 21:10:01 -0500

Hi help-glpk team:

I'm doing a research about a consensus and a got a question:

I have as parameter a matrix  A of size mxn like this

| y1  y2  y3
---- |------------------
x1  |  1   2   2
x2  |  1   2   1
x3  |  2   3   2
x4  |  3   2   3

where  every element in the matrix represent an label A=1 , B=2 and C=2

and when I resolve my model I get a vector solution Z of size mx1 like this:

|  Z[i]
----|--------
x1 |   2
x2 |   1
x3 |   2
x4 |   3

where every point represent the minimum distance with respect cell  aij in matrix A

my objective function is:
min sum{i in I, j in J } | aij - zi  |   (in my model, this objective function was already linearized )

now suppose that I have an number r o matrices A e.g. A1, A2, A3 ... At     and I get a solution z for every matrix like above.

then I need evaluate every solution z in every Matrix Ar , r = 1,...,t  except the matrix where the solution came from. i.e.  now I have matrix A and solution z and I don't need optimize only evaluate solutions found in previous matrices

Can you help me with this issue?

Best regards

--

RAFAEL TORRES ESCOBAR
Alumno Maestría en Ingeniería de Sistemas
Facultad de Ingeniería Mecánica y Eléctrica
México.

Programa de Posgrado en Ingeniería de Sistemas
F IME - U A N L
Cd. Universitaria
San Nicolás de los Garza, Nuevo León C.P. 66450
MÉXICO

http://pisis.fime.uanl.mx/  GBY