[Clp] infeasible, unbounded and popular decompositions

[Lijian Chen]

Dear All,

I am coding one of my research projects. Basically, I need to code the Dantzig-Wolfe decomposition or Bender's decomposition algorithm (which is essentially the same to the previous one) to solve a fair scale stochastic programming problem. I am not using the sample average approximation at this point because a precise solution is required. I used to apply CLP with C++ for other projects and had some experience. In the algorithm, we need to decide if some sub-problems are either infeasible or having positive reduced costs. 

Can CLP return some kind of result indicating infeasibility or positive reduced cost? Maybe I need to calculate the reduced costs myself. I do not have any clue about how to identify the infeasibility. Are there any subroutinaries having been developed for these decomposition method? Please help even it is a naive problem. Thanks! 


Sincerely Yours

Lijian Chen

Assistant Professor
Department of Industrial Engineering
J.B. Speed School of Engineering
University of Louisville
Louisville, KY 40292
(Tel) 502-852-2197
(Fax) 502-852-5633