[Coin-symphony] Warm Start Problems

[David Goodlad]

Hi all

I am working on a solver that uses the warm start functionality to
analyze multiple objective functions for the same problem.  However,
there seems to be one case where it inexplicably fails.

In one problem, it happens that the optimal solution with the first
objective function is found in the root node (after about 500
iterations).  When I then set up the second objective function, load
the warm start data from the last run, and run sym_warm_solve, it
complains that the problem is infeasible!  I know, however, that if I
set the initial objective function to something else which causes more
than one node to be treated, the second objective function (same one
as before) works just fine.

I've had to setup a workaround for now testing for only 1 node being
treated, but this is ugly.  Is there a specific reason for this
behavior?

Thanks in advance,
Dave

-- 
Dave Goodlad
dgoodlad at gmail.com or dave at goodlad.ca
http://david.goodlad.ca/