Abstract
In this paper we describe PAVER 2.0, an environment (i.e. a process and a suite of tools supporting that process) for the automated performance analysis of benchmarking data. This new environment improves on its predecessor by addressing some of the shortcomings of the original PAVER (Bussieck et al. in Global optimization and constraint satisfaction, lecture notes in computer science, vol 2861, pp 223–238. Springer, Berlin, 2003. doi:10.1007/978-3-540-39901-8_17) and extending its capabilities. The changes serve to further the original goals of PAVER (automation of the visualization and summarization of benchmarking data) while making the environment more accessible for the use of and modification by the entire community of potential users. In particular, we have targeted the end-users of optimization software, as they are best able to make the many subjective choices necessary to produce impactful results when benchmarking optimization software. We illustrate with some sample analyses conducted via PAVER 2.0.
Similar content being viewed by others
Notes
While arithmetic means are sensitive to variations of data with relatively large range and insensitive to variations of data with relatively small range, geometric means are more sensitive to variations close to zero. As a compromise, PAVER can also compute shifted geometric means [1], which reduce the effect of data points close to zero in the geometric mean by shifting.
PAVER uses a gap tolerance of \(10^{-6}\) by default. However, we have run our solvers with a zero gap tolerance.
Note that our globallib.solu file contains known optimal values for only half of the GlobalLib instances, so the number of global optimal solutions found and the mean primal gap are computed w.r.t. these 214 instances only.
References
Achterberg, T.: Constraint Integer Programming. Ph.D. thesis, TU Berlin (2007). http://nbnresolving.de/urn:nbn:de:0297-zib-11129
Achterberg, T.: Benchmarking a MIP Solver. Talk in CPAIOR Master Class (2010). http://cpaior2010.ing.unibo.it/?q=node/10
Berthold, T.: Measuring the impact of primal heuristics. Oper. Res. Lett. 41(6), 611–614 (2013). doi:10.1016/j.orl.2013.08.007
Billups, S.C., Dirkse, S.P., Ferris, M.C.: A comparison of large scale mixed complementarity problem solvers. Comput. Optim. Appl. 7(1), 3–25 (1997). doi:10.1023/A:1008632215341
Bussieck, M.R., Drud, A.S., Meeraus, A.: MINLPLib—a collection of test models for mixed-integer nonlinear programming. INFORMS J. Comput. 15(1), 114–119 (2003). doi:10.1287/ijoc.15.1.114.15159
Bussieck, M.R., Drud, A.S., Meeraus, A., Pruessner, A.: Quality assurance and global optimization. In: Bliek, C., Jermann, C., Neumaier, A. (eds.) Global Optimization and Constraint Satisfaction, Lecture Notes in Computer Science, vol. 2861, pp. 223–238. Springer, Berlin (2003). doi:10.1007/978-3-540-39901-8_17
Crowder, H., Dembo, R.S., Mulvey, J.M.: On reporting computational experiments with mathematical software. ACM Trans. Math. Softw. 5(2), 193–203 (1979). doi:10.1145/355826.355833
Dolan, E.D., Moré, J.J.: Benchmarking optimization software with performance profiles. Math. Program. 91(2), 201–213 (2002). doi:10.1007/s101070100263
Dolan, E.D., Moré, J.J., Munson, T.S.: Benchmarking Optimization Software with COPS 3.0. Tech. Rep. ANL/MCS-273, Mathematics and Computer Science Division, Argonne National Laboratory (2004). http://www.mcs.anl.gov/more/cops
Dolan, E.D., Moré, J.J., Munson, T.S.: Optimality measures for performance profiles. SIAM J. Optim. 16(3), 891–909 (2006). doi:10.1137/040608015
Drud, A.S.: Testing and Tuning a New Solver Version Using Performance Tests. INFORMS 2002, San Jose, Session on ’Benchmarking and performance testing of optimization software’. http://www.gams.com/presentations/present_performance.pdf. Accessed 15 May 2013
Exler, O., Lehmann, T., Schittkowski, K.: A comparative study of SQP-type algorithms for nonlinear and nonconvex mixed-integer optimization. Math. Program. Comput. 4(4), 383–412 (2012). doi:10.1007/s12532-012-0045-0
GAMS Development: GAMS/Examiner, User’s Manual (2013). http://www.gams.com/dd/docs/solvers/examiner.pdf. Accessed 8 May 2013
Granlund, T.: The GMP Development Team: GNU MP: The GNU Multiple Precision Arithmetic Library (2012). http://gmplib.org
Hendel, G.: PyEvalGui - GUI Components to Facilitate Evaluation of SCIP and Other Solving Software (2013, in development)
Koch, T., Achterberg, T., Andersen, E., Bastert, O., Berthold, T., Bixby, R.E., Danna, E., Gamrath, G., Gleixner, A.M., Heinz, S., Lodi, A., Mittelmann, H., Ralphs, T., Salvagnin, D., Steffy, D.E., Wolter, K.: MIPLIB 2010—mixed integer programming library version 5. Math. Program. Comput. 3(2), 103–163 (2011). doi:10.1007/s12532-011-0025-9
Mahajan, A., Leyffer, S., Kirches, C.: Solving Mixed-Integer Nonlinear Programs by QP-Diving. Preprint ANL/MCS-P2071-0312, Argonne National Laboratory (2012). http://www.optimization-online.org/DB_HTML/2012/03/3409.html
Meeraus, A.: Globallib (2013). http://www.gamsworld.org/global/globallib.htm. Accessed 8 May 2013
Mittelmann, H.D.: An independent benchmarking of SDP and SOCP solvers. Math. Program. 95(2), 407–430 (2003). doi:10.1007/s10107-002-0355-5
Mittelmann, H.D.: DTOS—a service for the optimization community. SIAG/OPT Views-and-News 18, 17–20 (2007)
Mittelmann, H.D.: Decision Tree for Optimization Software (2013). http://plato.asu.edu/guide.html. Accessed 8 May 2013
Mittelmann, H.D., Pruessner, A.: A server for automated performance analysis of benchmarking data. Optim. Methods Softw. 21(1), 105–120 (2006). doi:10.1080/10556780500065366
SCIP Development Team: How to Run Automated Tests with SCIP. http://scip.zib.de/doc/html/TEST.shtml
Why open source? (2013). http://www.coin-or.org. Accessed 15 May 2013
Wikiquote: Winston churchill—wikiquote (2013). http://en.wikiquote.org/w/index.php?title=Winston_Churchill&oldid=1552921. Accessed 8 May 2013
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bussieck, M.R., Dirkse, S.P. & Vigerske, S. PAVER 2.0: an open source environment for automated performance analysis of benchmarking data. J Glob Optim 59, 259–275 (2014). https://doi.org/10.1007/s10898-013-0131-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10898-013-0131-5