Error: all variables eliminated, but lower bound = 1 > 0

[Fernando Pineda]

Hi, i have a problem with my model.

I tried use couenne solver and it showed me following error:

Error: presolve, constraint maquina['mu3','jl4','ju10','mu4']:
Presolve eliminates 204 constraints and 64 variables.
Error: 	all variables eliminated, but lower bound = 1 > 0
Adjusted problem:
Error: presolve, constraint maquina['mu3','jl3','ju9','mu4']:
29 variables, all linear
Error: 	all variables eliminated, but lower bound = 1 > 0
256 constraints, all linear; 440 nonzeros
Error: presolve, constraint maquina['mu3','jl3','ju8','mu4']:
	36 equality constraints
Error: 	all variables eliminated, but lower bound = 1 > 0
	220 inequality constraints

What is that and What cant do to solve this problem?
Attached files if anyone can help me.

Thank you for the help.

[Stefano Coniglio]

It's a presolve message (returned by the presolve implemented in AMPL). In a nutshell, it tries to reduce the size of your model by removing unneeded variables and slack constraints. In your case, while performing such "model reduction" procedure, AMPL proved that the model you fed to it is infeasible ("lower bound = 1 > 0" means, roughly, that it managed to remove all the variables from the "maquina['mu3','jl3','ju9','mu4']:" constraints, obtaining some trivially false inequality with no variables, such as, e.g., 1 <= 0).

I'd suggest to doublecheck your model. A bug should pop up.

-S.

[Robert Fourer]

As a start on checking the model, I suggest picking one of the error messages, say,

 

   presolve, constraint maquina['mu3','jl4','ju12','mu4']:

      all variables eliminated, but lower bound = 1 > 0

 

and expanding the constraint mentioned in the message:

 

   ampl: expand maquina['mu3','jl4','ju12','mu4'];

 

   subject to maquina['mu3','jl4','ju12','mu4']:

      -XL['jl4','mu4'] = 0;

 

Next I would check the bounds on the one variable in this constraint:

 

   ampl: display XL['jl4','mu4'].lb,XL['jl4','mu4'].ub;

 

   XL['jl4','mu4'].lb = 1

   XL['jl4','mu4'].ub = 1

 

Here I see that AMPL's presolve phase has determined that this variable must equal one.  Thus the constraint becomes -1 = 0, which cannot possibly be satisfied.  Now you will have to make a further examination of your model to figure out why this contradiction is occurring.

 

Bob Fourer

am...@googlegroups.com

 

 

From: am...@googlegroups.com [mailto:am...@googlegroups.com]

On Behalf Of Stefano Coniglio
Sent: Thursday, October 10, 2013 11:58 AM
To: am...@googlegroups.com
Subject: [AMPL 7550] Re: Error: all variables eliminated, but lower bound = 1 > 0

[Marcelo Freitas da Silva]

Hello all, I need help in one issue of my model in AMPL.

I am trying to develop a model with CPLEX solver  visioning Optimal Transmission Switching (Power System) using a power flow based on IEEE 118 bus.

Regarding CPLEX being useful for linear programation, I set the equations of power flow linearized based on DCOPF (attachment "objective.jpg").

However, I am getting troubles in make the equation 5(attachment "objective.jpg") works. In my model i need to change the grid topology and analyse the costs involved.

For this i am trying to set the maximum number of lines ON (red equation) visioning getting the whole cost of system for different topologies, but this constraint is not working.

I am trying to sum binary variables and limitation to a maximum value algebric, but I think it is wrong...

I wouldlike to share my model for your comments...

Could you suggest anything to do? I really do not have any other idea.. I am trying to do many things, but the results is not as I expected.

I appreciate in advance you attention!

Thanks!