bug-glpk
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-glpk] Bug en glpk-4.56


From: Andrew Makhorin
Subject: Re: [Bug-glpk] Bug en glpk-4.56
Date: Mon, 09 Nov 2015 05:40:17 +0300

On Sun, 2015-11-08 at 20:20 -0500, Vicente Saldarriaga Yacila wrote:
> Ifound an error in the sensitivity report ofgklpk-4.56, when the range
> of activity restriction is determined.
> It is presented to solve a problem on a company that produces two
> types of TV sets, "Astro" and "Cosmo". For each there is a production
> line. The capacity of the production line is60 and 50, sets per
> day,respectively. The Astro set requires onemanhours of work, while
> Cosmo requires two. Currently there is a maximum of 120 manhours of
> work per day, which can be allocated to the production of such
> equipment. If the utility is $ 20 and $ 30 each, respectively, what
> should be the daily production?
> 
> I atach the following files: bugresumen, bug.mod, bug.sol, bug.sen and
> bug.dat.
> 

There is no bug. You need to note that constraint Linea[C] is inactive
in the optimal basic solution, i.e. its auxiliary variable is basic. 
In this case the sensitivity analysis concerns changing of objective
coefficients at that basic variable (while you try to change its
bounds). Since Linea[C] is an auxiliary variable, it always has zero
objective coefficient, and there is no way to change it to a non-zero.
But if this would be possible, then according to the sensitivity report
the basic solution would remain optimal if the objective coefficient at
Linea[C] would change in the range [-30, +10]. For more details please
consult the GLPK documentation (glpk/doc/glpk.pdf).






reply via email to

[Prev in Thread] Current Thread [Next in Thread]