Analyzing infeasibility: IIS

A general technique to analyze infeasibility is to find a small portion of the matrix that is itself infeasible. Xpress Optimizer does this by finding irreducible infeasible sets (IISs). An IIS is a minimal set of constraints and variable bounds which is infeasible, but becomes feasible if any constraint or bound in it is removed.

»Scenario 1 (Mosel)

Take a look at the Mosel model in file examples\getting_started\Mosel\folioiis.mos. This example retrieves the IIS sets for an LP-infeasible problem and displays their contents.

»Scenario 2 (BCL)

Similarly, you can use the IIS access functionality of BCL to retrieve and display IIS (see file folioiis.[c|cxx|java] in directory examples\bcl\[C|Cxx|Java]\UGExpl).

»Scenario 3 (Optimizer Python interface)

The Python script examples\python\example_infeasible.py showcases the usage of all functions for handling infeasibility in an optimization problem, specifically the functions that manage IIS.