Originally posted by: Doal90

Dear All,

I'm new with the use of Cplex library. I'm using Cplex with C in a computer with windows.
My programm modelling a MIP problem.
 

I'm interested to see if my instances generate trees with many nodes (if b&c has difficulty to find the optimal solution).

My problem is that the algorithms that analyze the root node find immediately the solution of the instance. So I get a solution like this:

Root node processing (before b&c):    
   Real time             =    0.20 sec. (30.32 ticks)    
Parallel b&c, 4 threads:       
   Real time             =    0.00 sec. (0.00 ticks)   
   Sync time (average)   =    0.00 sec.    
   Wait time (average)   =    0.00 sec.   
                            -------   
Total (root+branch&cut) =     0.20 sec. (30.32 ticks)

How can I change my programm to have Real time of Root node processing = 0.00 sec.(or something like that)? And to generate subproblems with primal simplex b&c after Cplex identifies a solution of the relaxed problem?

Best,

Alex

You may want to take a look at the list of parameters that are available to control CPLEX behavior. In order to reduce the work CPLEX performs at the root node, the following may be interesting:

• control presolve,
• control cutting planes,
• control heuristics

Note however that your model may still be too simple to produce a large tree.

Originally posted by: Doal90

Thank you Daniel, it was just what I needed. I posted only a piece of model.

Just another question, there is a system to get information of all nodes processed?

In the reference documentation you will find that class IloCplex has a function getNnodes() (or getNnodes64() for very large trees) which returns the number of nodes processed.