Performance Profile Summary (Objective Gap)
Results created on 01/18/09 17:30:47 using the PAVER Server/Performance tools available at http://www.gamsworld.eu/performance/paver
Performance profiles for different objective value tolerances. The profile utility compares solver outcomes of any number of given solvers (up to 8) over a model test set. The tool is described in detail at http://www.gamsworld.eu/performance/pprofile.htm.
A solver is considered feasible or optimal, if it has the proper model and solver return codes and the relative objective value error is within n % of the best possible solution. If the best found objective value is 0, then the absolute objective value error is used.
The following resource is used as measurement: from solver reported time
The following relative objective value tolerances are used:
- Only best objective value (i.e. 0 % of best solution found. Actual tolerance used is 1e-5 using relative error)
- No objective value tolerance, i.e. we do not check if obj is best found (within Inf%).
The following gap tolerances are used:
- Only closed gap (Actual tolerance used is 1e-5 using relative distance)
- No gap tolerance, i.e. we do not check if the gap is closed to a certain value (within Inf%).
- Relative gap is at most 10%.
Quality of Solution (within 0% wrt to best value, gap at most 0%)
Data File: profileg_0_0.txt
Black and White EPS file - profileg_0_0.eps
Quality of Solution (within 0% wrt to best value, gap at most 10%)
Data File: profileg_0_10.txt
Black and White EPS file - profileg_0_10.eps
Quality of Solution (within Inf% wrt to best value, gap at most 10%)
Data File: profileg_Inf_10.txt
Black and White EPS file - profileg_Inf_10.eps