[AMPL 1857] column generation

[patrick-131]

The problem that I am working on has an incredible number of columns
(i.e. variables) so that there is simply no way to generate them all.
There about 1300 constraints.  My problem is that when I try to solve
this using column generation, it seems to work well for a while though
the reduced costs jump around a fair bit and also seem to level off.
After about 700 columns have been generated AMPL spits out a message
stating that all variables have been eliminated but lower bound = some
very small number in the order of 10^-8. I have tried setting the
tolerance at something bigger than 10^-8 but it seems to simply delay
the time until I get the same message.  Any suggestions?

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To post to this group, send email to ampl@...
To unsubscribe from this group, send email to ampl-unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ampl?hl=en
-~----------~----~----~----~------~----~------~--~---

[Robert Fourer-2]

Probably your linear program is being proved infeasible in AMPL's presolve
phase, without ever being sent to the solver.  Could you provide the complete
output from AMPL at the solve where you run into trouble, though?  (This would
include the exact message about all variables being eliminated.)

Also it might help to try a different test where you set "option presolve 0;"
at the beginning of the run.  That would force the solver to detect
infeasibility, and so it would provide a check on the correctness of AMPL's
presolve.

Bob Fourer
4er@...

> -----Original Message-----
> From: ampl@... [mailto:ampl@...]
> On Behalf Of patrick@...
> Sent: Tuesday, June 24, 2008 8:39 AM
> To: AMPL Modeling Language
> Subject: [AMPL 1857] column generation
>
> The problem that I am working on has an incredible number of columns
> (i.e. variables) so that there is simply no way to generate them all.
> There about 1300 constraints.  My problem is that when I try to solve
> this using column generation, it seems to work well for a while though
> the reduced costs jump around a fair bit and also seem to level off.
> After about 700 columns have been generated AMPL spits out a message
> stating that all variables have been eliminated but lower bound = some
> very small number in the order of 10^-8. I have tried setting the
> tolerance at something bigger than 10^-8 but it seems to simply delay
> the time until I get the same message.  Any suggestions?
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To post to this group, send email to ampl@...
To unsubscribe from this group, send email to ampl-unsubscribe@...
For more options, visit this group at http://groups.google.com/group/ampl?hl=en
-~----------~----~----~----~------~----~------~--~---