Re: [eclipse-clp-users] 'Problem is infeasible!' in eplex goes to stdout

From: Kish Shen <kisshen_at_...5...>
Date: Thu, 28 Aug 2014 18:59:56 +0100
Hi Sergii,

I should have updated on this earlier, but I was waiting to see if I got 
any responses from the Cbc mailing list to my post there about this 
problem, but as yet there is nothing...

On 20/08/2014 08:58, Sergii Dymchenko wrote:
> Hi,
> I've investigated the issue further and found that the message appears not
> for all infeasible instances, but only for some of them. The issue present
> in the version 6.1 #191 and not present in the version 6.1 #164 - so it was
> introduced sometime in between.

The problem was as I suspected: CbcSolver (the MIP solver used in this 
case) does print the 'Problem is infeasible!' directly to stdout 
(std::cout -- the code is written in C++) in several places, and there 
are also other messages that are printed to stdout. I am not sure why it 
is done this way, as most outputs are handled via their message handler, 
which is what the eplex output channels are connected to.

I updated the Cbc solver used with eplex in March this year, and I 
assume this is the reason for why you see the message only in the recent 
ECliPSe version.

The printing of 'Problem is infeasible!' message can be excluded from 
compiling Cbc, as the source for it are enclosed by a macro that allow 
the printing to be excluded from the compile. However, not all messages 
sent to std:cout in CbcSolver are enclosed by this macro, so unless the 
Cbc source is modified, messages can still be sent to stdout -- although 
these messages seems to be related to using CbcSolver as a stand-alone 
solver, so hopefully they will not be seen when CbcSolver is called from 
a program, as we do in eplex.

As I mentioned, I posted a message to the Cbc mailing list asking about 
these messages on 21 August, but have received no responses. I will try 
a follow-up post, and if I still don't hear anything, I will update the 
Cbc we use by compiling it with the macro to suppress the printing of 
these messages.


Received on Thu Aug 28 2014 - 18:35:08 CEST

This archive was generated by hypermail 2.2.0 : Mon Jul 09 2018 - 02:05:30 CEST