Re: [Help-glpk] Strong branching

[Andrew Makhorin]

> I have used copy_prob for calculating LP relaxations of child nodes,
> however,
> when I find that one of the subproblems is infeasible, I want to fix
> the bound of that
> variable in the original subproblem at the current node itself. I
> basically want to implement concept of bound fixing or
> strong branching fixing.
> Can this be done?
> I have attached my subroutine along with this email.

You can fix variable(s) in response to GLP_IPREPRO (request for
preprocessing). This event precedes branching for the same subproblem.