The function fixes the values of all error vectors on their current values. It also removes their objective cost contribution.

The function is intended to support post optimization analysis, by removing any possible direct effect of the error vectors from the dual and reduced cost values.

The fixpenalties function will automatically reoptimize the linearization. However, as the XSLP convergence and infeasibility checks (regarding the original non-linear problem) will not be carried out, this function will not update the SLP solution itself. The updated values will be accessible using getlpsolution instead.