[Symphony] LP solver (Clp) won't stop after TM_time_limit & 2 newbie questions

[lebodicp at insa-rouen.fr]

Hello,

I have implemented an integer linear program thanks to the SYMPHONY user
callback functions. Few of them are redefined for now. I'm using the
symphony environment rather than the osi. I'm using an embedded parser
that reads custom data files rather than mps files. I'm handling large
problems (possibly 150000 variables and 100000 constraints). Finally, I
use Clp.
The program by itself works as expected: small problems are solved, large
problems are rarely solved.
I have encountered a few problems during further usage. I'm new to
SYMPHONY, so those probably aren't bugs.

-From the source code, setting TM_time_limit in the parameter file should
set LP time limit to the same value. As I'm handling large problems, it is
often the case when lp solving at root node is itself too long. However,
the LP solver never seems to stop after the time limit, or stops way after
it (way!). TM_time_limit affects TM behavior as expected though. Is this a
desired behavior? Am I doing something wrong? Is this a bug?

-I've tried using column generation, but couldn't get it right (for my
problem, variables are known from the start as required in SYMPHONY). I
followed the manual instructions but obviously I'm doing something wrong.
I can't find an example that uses column generation. Is there something of
that kind available?

-I didn't realize using sym environment rather than osi would prevent me
from using Cgl cuts. Now that the code is written, what would be the
easiest way to get Cgl cuts to work? Would it be possible to (1)export to
mps (2)read the mps file with another program that uses osi? What would be
the best way?
Thanks.

Best, Pierre