bug-glpk
[Top][All Lists]
Advanced

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

Re: [Bug-glpk] glpsol ignores tmlim


From: Andrew Makhorin
Subject: Re: [Bug-glpk] glpsol ignores tmlim
Date: Fri, 4 Sep 2009 15:26:12 +0400

Hi Xypron,

> when starting glpsol for the model file below with
> glpsol -m test.mod --tmlim 100
> the solver does not terminate within the required time.

> A lot of lines like the following are output:
> | 19400: obj =  6.400000000e+001  infeas = 7.062e-014 (2)
> | 19600: obj =  6.400000000e+001  infeas = 1.024e-014 (2)
> | 19800: obj =  6.400000000e+001  infeas = 5.663e-014 (2)

Thank you for your bug report.

The solver does not stop, because the time limit is checked at the
beginning of the branch-and-bound loop, however, the dual simplex is
cycling and does not return.

BTW, your instance is pathological (currently I am trying to analyze
it). Though it is well scaled, in optimal solution to lp relaxation
some constraints have huge reduced costs that explains why the dual
simplex cannot terminate:

131 move[2,1,1]  B              0                           0
132 move[2,1,2]  NU             0                           0         < eps
133 move[2,1,3]  NU             0                           0   1.28218e+20
134 move[2,1,4]  NU             0                           0         < eps
135 move[2,1,5]  NU             0                           0   1.42691e+20
136 move[2,1,6]  NU             0                           0         < eps
137 move[2,1,7]  NU             0                           0   9.93974e+19
138 move[2,1,8]  NU             0                           0         < eps
139 move[2,2,1]  B              0                           0
140 move[2,2,2]  NU             0                           0   1.42098e+20
141 move[2,2,3]  B      -0.317426                           0
142 move[2,2,4]  NU             0                           0   2.09366e+20
143 move[2,2,5]  NU             0                           0         < eps
144 move[2,2,6]  NU             0                           0   1.90904e+20
145 move[2,2,7]  NU             0                           0         < eps
146 move[2,2,8]  NU             0                           0   9.94232e+19
147 move[2,3,1]  NU             0                           0   1.28218e+20
148 move[2,3,2]  B      -0.682574                           0
149 move[2,3,3]  NU             0                           0   2.55832e+20
150 move[2,3,4]  NU             0                           0         < eps


Andrew Makhorin





reply via email to

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