Returns a vector proving that the Q matrix of QP is not positive semi-definite.
getNcols
where the variables of the active model are returned. getNcols
. By setting var[j]
to x[j]
for all variables, the quadratic term of the objective function evaluates to a negative value. When solving QPs, the quadratic terms of the objective function must form a positive semi-definite Q matrix (for a minimization problem, or negative semi-definite otherwise). If Cplex finds this not be true, it will discontinue the optimization. In such cases, this method can be used to compute assignments to all variables such that the quadratic term of the objective function evaluates to a negative value (x'Q x < 0
in matrix terms) to prove the indefiniteness.