PSG Solution in MATLAB
Contents
The results of optimization problem are stored in solution report and solution structure with PSG Data Objects.
Solution report is a string with main information about optimization results:
Problem |
status of solution of the optimization problem: optimal | feasible | infeasible | unbounded | calculated (for calculation problems); |
Timing |
working time of solver including: data loading time, preprocessing time, solving time; |
Variables |
values of variables of optimization problem: |
Objective |
value of the objective function in optimal point; |
Constraint |
value of the constraint function in optimal point and it's maximal deviation from specified in problem bounds (in square brackets); |
Function |
value of functions used in optimization problem formulation in optimal point. |
Solution Structure includes optimal point, point with slacks values, point with dual constraints values and the other data stored as PSG Data Objects.
In PSG Toolbox GUI optimization results are available in Menu of the Main Toolbox Window-> Solution -> Report. All PSG Data Objects from solution structure are stored to Objects List.
In Programmatical PSG solution report and solution structure are outputs of PSG m-function tbpsg_run.
There are several PSG m-functions for convenient work with solution of optimization problem. See Extract Solution Results section.
Examples
This example considers CVaR minimization problem defined in the section Quick Start with PSG MATLAB. Run the following example from the folder ./Aorda/PSG/MATLAB/Examples/Problems/Example_CVaR_General.m to obtain the solution:
Full report for solution of the optimization problem is solution = output argument of tbpsg_run function:
solution_str =
Problem: solution_status = optimal
Timing: Data_loading_time = 0.00, Preprocessing_time = 0.00, Solving_time = 0.00
Variables: optimal_point = point_problem_cvar
Objective: objective_cvar = -4.350000000000
Constraint: constraint_budget = 1.000000000000e+000, [-4.440892098501e-016]
Constraint: constraint_return = 4.500000000000e+000, [0.000000000000e+000]
Function: avg_g(matrix_scenarios) = 4.500000000000e+000
Function: cvar_risk_example(9.500000E-01,matrix_scenarios) = -4.350000000000e+000
Function: linear(matrix_budget) = 1.000000000000e+000
The optimal point of the optimization problem is stored in the second output argument of tbpsg_run function outargstruc in PSG Point format:
outargstruc_arr =
object_name: 'point_problem_cvar'
variables: 'x1 x2 x3 x4'
type: 'point'
values: [0 0.6000 0.1900 0.2100]
See also