| Tuning scenario | Training set of benchmark instances | Test set of benchmark instances | Instance/seed lists used for the repetitions of ParamILS (for the JAIR article, we used the first 25 for indepth and the first 10 for broad scenarios) |
Citation | Short description | Parameter configuration found in the FocusedILS repetition with best training performance, used for Figures 14 and 15 in our JAIR article |
| SAPS-SWGCP | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{GenHooProWal99, author = "I.~P. Gent and H.~H.~Hoos and P.~Prosser and T.~Walsh", title = "Morphing: Combining Structure and Randomness", booktitle = aaai99, pages = "654--660", year = "1999" } |
SAT-encoded graph colouring based on small world graphs. all instances satisfiable |
alpha=1.189, ps=0.033, rho=0.5, wp=0.05 in my format result comparison against default |
| Spear-SWGCP | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{GenHooProWal99, author = "I.~P. Gent and H.~H.~Hoos and P.~Prosser and T.~Walsh", title = "Morphing: Combining Structure and Randomness", booktitle = aaai99, pages = "654--660", year = "1999" } |
SAT-encoded graph colouring based on small world graphs. sat/unsat instances |
sp-clause-activity-inc=1,
sp-clause-decay=1.4, sp-clause-del-heur=1, sp-clause-inversion=0,
sp-first-restart=3200, sp-learned-clause-sort-heur=12,
sp-learned-clauses-inc=1.1, sp-learned-size-factor=0.2,
sp-max-res-lit-inc=0.5, sp-max-res-runs=8, sp-orig-clause-sort-heur=8,
sp-phase-dec-heur=1, sp-rand-phase-dec-freq=0.05,
sp-rand-phase-scaling=1.1, sp-rand-var-dec-freq=0.001,
sp-rand-var-dec-scaling=1.1, sp-res-cutoff-cls=16,
sp-res-cutoff-lits=1600, sp-res-order-heur=13, sp-resolution=2,
sp-restart-inc=1.9, sp-update-dec-queue=1, sp-use-pure-literal-rule=1,
sp-var-activity-inc=1, sp-var-dec-heur=0, sp-variable-decay=2.0 in my format result comparison against default |
| Old version of SAPS-QCP, used in ParamILS 2009 Tech report This scenario has a problem with the split into training and test set: the training set was systematically easier than the test set; I only keep it up here for completeness, but would advise against using it |
Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{GomSel97, author = "C.~P. Gomes and B. Selman", title = "Problem Structure in the Presence of Perturbations", booktitle = aaai97, year = "1997" } |
SAT-encoded quasigroup completion all instances satisfiable |
alpha=1.126, ps=0.2, rho=0, wp=0.05 in my format result comparison against default |
| New version of SAPS-QCP, used in JAIR paper | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{GomSel97, author = "C.~P. Gomes and B. Selman", title = "Problem Structure in the Presence of Perturbations", booktitle = aaai97, year = "1997" } |
SAT-encoded quasigroup completion all instances satisfiable |
alpha=1.126, ps=0.2, rho=0, wp=0.05 (same as for the old version of QCP) in my format result comparison against default |
| Spear-QCP | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{GomSel97, author = "C.~P. Gomes and B. Selman", title = "Problem Structure in the Presence of Perturbations", booktitle = aaai97, year = "1997" } |
SAT-encoded quasigroup completion sat/unsat instances |
sp-clause-activity-inc=0.5,
sp-clause-decay=1.1, sp-clause-del-heur=2, sp-clause-inversion=0,
sp-first-restart=1600, sp-learned-clause-sort-heur=13,
sp-learned-clauses-inc=1.4, sp-learned-size-factor=0.2,
sp-max-res-lit-inc=0.5, sp-max-res-runs=8, sp-orig-clause-sort-heur=13,
sp-phase-dec-heur=6, sp-rand-phase-dec-freq=0.001,
sp-rand-phase-scaling=1.1, sp-rand-var-dec-freq=0.0001,
sp-rand-var-dec-scaling=0.6, sp-res-cutoff-cls=8,
sp-res-cutoff-lits=200, sp-res-order-heur=14, sp-resolution=1,
sp-restart-inc=1.9, sp-update-dec-queue=1, sp-use-pure-literal-rule=0,
sp-var-activity-inc=1.5, sp-var-dec-heur=0, sp-variable-decay=1.1 in my format result comparison against default |
| CPLEX-regions100 | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@InProceedings{LeyPeaSho00, author = "K. Leyton-Brown and M. Pearson and Y. Shoham", title = "Towards a Universal Test Suite for Combinatorial Auction Algorithms", booktitle = "ACM Conference on Electronic Commerce (EC-00)", year={2000} } |
MIP-encoded combinatorial winner determination, 100 goods, 500 bids |
advance=1, barrier_algorithm=3,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=-1,
barrier_limits_corrections=-1, barrier_limits_growth=1e+12,
barrier_ordering=2, barrier_qcpconvergetol=1e-07, barrier_startalg=2,
emphasis_memory=yes, emphasis_mip=1, emphasis_numerical=no,
feasopt_mode=3, feasopt_tolerance=1e-04, lpmethod=1,
mip_cuts_cliques=1, mip_cuts_covers=1, mip_cuts_disjunctive=0,
mip_cuts_flowcovers=1, mip_cuts_gomory=-1, mip_cuts_gubcovers=-1,
mip_cuts_implied=-1, mip_cuts_mircut=2, mip_cuts_pathcut=1,
mip_limits_aggforcut=5, mip_limits_cutpasses=0,
mip_limits_cutsfactor=8, mip_limits_gomorycand=800,
mip_limits_gomorypass=0, mip_limits_polishtime=0,
mip_limits_probetime=1e+75, mip_limits_repairtries=0,
mip_limits_strongcand=5, mip_limits_strongit=0,
mip_limits_submipnodelim=2000, mip_ordertype=3,
mip_strategy_backtrack=0.9999, mip_strategy_bbinterval=4,
mip_strategy_branch=1, mip_strategy_dive=0, mip_strategy_file=1,
mip_strategy_heuristicfreq=-1, mip_strategy_lbheur=no,
mip_strategy_nodeselect=2, mip_strategy_order=no,
mip_strategy_presolvenode=-1, mip_strategy_probe=-1,
mip_strategy_rinsheur=-1, mip_strategy_startalgorithm=6,
mip_strategy_subalgorithm=2, mip_strategy_variableselect=4,
network_netfind=1, network_pricing=2, preprocessing_aggregator=-1,
preprocessing_boundstrength=0, preprocessing_coeffreduce=0,
preprocessing_dependency=3, preprocessing_dual=-1,
preprocessing_fill=20, preprocessing_numpass=-1,
preprocessing_presolve=yes, preprocessing_qpmakepsd=no,
preprocessing_reduce=1, preprocessing_relax=0,
preprocessing_repeatpresolve=0, preprocessing_symmetry=0, qpmethod=1,
read_scale=-1, sifting_algorithm=0, simplex_crash=-1,
simplex_dgradient=2, simplex_limits_perturbation=0,
simplex_limits_singularity=40, simplex_perturbation=no 1e-06,
simplex_pgradient=-1, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
| Spear-swv | Training set (302 instances) |
Test set (302 instances) |
Training instance/seed lists Test instance/seed list |
@inproceedings{babic07structural-abs, author = {Domagoj Babi\'c and Alan J. Hu}, title = {{Structural Abstraction of Software Verification Conditions}}, booktitle = {Computer Aided Verification: 19th International Conference, CAV 2007}, year = {2007}, pages={366--378} } |
SAT-encoded software verification | sp-clause-activity-inc=1,
sp-clause-decay=1.4, sp-clause-del-heur=2, sp-clause-inversion=0,
sp-first-restart=100, sp-learned-clause-sort-heur=16,
sp-learned-clauses-inc=1.4, sp-learned-size-factor=0.1,
sp-max-res-lit-inc=4, sp-max-res-runs=2, sp-orig-clause-sort-heur=12,
sp-phase-dec-heur=0, sp-rand-phase-dec-freq=0.0001,
sp-rand-phase-scaling=0.6, sp-rand-var-dec-freq=0.0001,
sp-rand-var-dec-scaling=0.9, sp-res-cutoff-cls=8,
sp-res-cutoff-lits=200, sp-res-order-heur=16, sp-resolution=1,
sp-restart-inc=1.3, sp-update-dec-queue=1, sp-use-pure-literal-rule=0,
sp-var-activity-inc=1, sp-var-dec-heur=6, sp-variable-decay=1.1 in my format result comparison against default |
| Spear-ibm | We cannot provide these instances online due to copyright issues. You can acquire them from the IBM Formal Verification Benchmarks Library. We used 40 random subsets of these instances. Here are the names of our 382 training instances and names of our 383 test instances. | Training instance/seed lists Test instance/seed list |
@inproceedings{zarpas05benchmarking, author = {Emanuel Zarpas}, title = {{Benchmarking SAT Solvers for Bounded Model Checking}}, booktitle = {SAT '05: Proc.~of the 8th International Conference on Theory and Applications of Satisfiability Testing}, year = {2005}, pages = {340--354} } |
SAT-encoded hardware verification (BMC) | sp-clause-activity-inc=1, sp-clause-decay=1.1, sp-clause-del-heur=0,
sp-clause-inversion=1, sp-first-restart=1600,
sp-learned-clause-sort-heur=1, sp-learned-clauses-inc=1.5,
sp-learned-size-factor=0.8, sp-max-res-lit-inc=1, sp-max-res-runs=2,
sp-orig-clause-sort-heur=10, sp-phase-dec-heur=0,
sp-rand-phase-dec-freq=0.0001, sp-rand-phase-scaling=1,
sp-rand-var-dec-freq=0.0001, sp-rand-var-dec-scaling=0.6,
sp-res-cutoff-cls=20, sp-res-cutoff-lits=200, sp-res-order-heur=13,
sp-resolution=0, sp-restart-inc=1.1, sp-update-dec-queue=0,
sp-use-pure-literal-rule=0, sp-var-activity-inc=0.5, sp-var-dec-heur=6,
sp-variable-decay=2.0 in my format result comparison against default |
|
| SAPS-random | Training set (363 instances) |
Test set (363 instances) |
Training instance/seed lists Test instance/seed list |
See, e.g. @inproceedings{LeBSim04, author = {{Le~Berre}, D. and Simon, L. }, title = {Fifty-five solvers in {Vancouver}: The {SAT} 2004 competition}, booktitle = sat04, year = {2004}, pages = {321--344} } |
All satisfiable instances in the RANDOM category from all SAT competitions until 2007 | alpha=1.256, ps=0.066, rho=0.333, wp=0.01 in my format result comparison against default |
| SAPS-crafted | Training set (189 instances) |
Test set (188 instances) |
Training instance/seed lists Test instance/seed list |
See, e.g. @inproceedings{LeBSim04, author = {{Le~Berre}, D. and Simon, L. }, title = {Fifty-five solvers in {Vancouver}: The {SAT} 2004 competition}, booktitle = sat04, year = {2004}, pages = {321--344} } |
All satisfiable instances in the CRAFTED category from all SAT competitions until 2007 | alpha=1.066, ps=0, rho=1, wp=0.02 in my format result comparison against default |
| CPLEX-regions200 | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
@InProceedings{LeyPeaSho00, author = "K. Leyton-Brown and M. Pearson and Y. Shoham", title = "Towards a Universal Test Suite for Combinatorial Auction Algorithms", booktitle = "ACM Conference on Electronic Commerce (EC-00)", year={2000} } |
MIP-encoded combinatorial winner determination, 200 goods, 1000 bids |
advance=1, barrier_algorithm=0,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=0,
barrier_limits_corrections=-1, barrier_limits_growth=1e+8,
barrier_ordering=0, barrier_qcpconvergetol=1e-07, barrier_startalg=4,
emphasis_memory=no, emphasis_mip=0, emphasis_numerical=no,
feasopt_mode=3, feasopt_tolerance=1e-06, lpmethod=0,
mip_cuts_cliques=2, mip_cuts_covers=-1, mip_cuts_disjunctive=0,
mip_cuts_flowcovers=0, mip_cuts_gomory=-1, mip_cuts_gubcovers=0,
mip_cuts_implied=0, mip_cuts_mircut=1, mip_cuts_pathcut=-1,
mip_limits_aggforcut=3, mip_limits_cutpasses=0,
mip_limits_cutsfactor=2, mip_limits_gomorycand=200,
mip_limits_gomorypass=0, mip_limits_polishtime=0,
mip_limits_probetime=2, mip_limits_repairtries=0,
mip_limits_strongcand=10, mip_limits_strongit=0,
mip_limits_submipnodelim=500, mip_ordertype=0,
mip_strategy_backtrack=0.9999, mip_strategy_bbinterval=7,
mip_strategy_branch=0, mip_strategy_dive=2, mip_strategy_file=1,
mip_strategy_heuristicfreq=-1, mip_strategy_lbheur=no,
mip_strategy_nodeselect=3, mip_strategy_order=yes,
mip_strategy_presolvenode=-1, mip_strategy_probe=0,
mip_strategy_rinsheur=0, mip_strategy_startalgorithm=3,
mip_strategy_subalgorithm=0, mip_strategy_variableselect=4,
network_netfind=2, network_pricing=0, preprocessing_aggregator=-1,
preprocessing_boundstrength=-1, preprocessing_coeffreduce=2,
preprocessing_dependency=2, preprocessing_dual=0,
preprocessing_fill=10, preprocessing_numpass=-1,
preprocessing_presolve=yes, preprocessing_qpmakepsd=yes,
preprocessing_reduce=3, preprocessing_relax=-1,
preprocessing_repeatpresolve=-1, preprocessing_symmetry=0, qpmethod=0,
read_scale=0, sifting_algorithm=3, simplex_crash=1,
simplex_dgradient=2, simplex_limits_perturbation=0,
simplex_limits_singularity=10, simplex_perturbation=no 1e-06,
simplex_pgradient=4, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
| CPLEX-conic.sch | We cannot provide these instances online due to copyright issues. You can acquire them from the Berkeley Computational Optimization Lab. Here are the names of our 172 training instances and names of our 171 test instances. |
Training instance/seed lists Test instance/seed list |
@TECHREPORT{AAG:csch:tr, author = {S. M. Akt{\"u}rk and A. Atamt{\"u}rk and S. G{\"u}rel}, title = {A Strong Conic Quadratic Reformulation for Machine-Job Assignment with Controllable Processing Times}, type = {Research Report}, number = {BCOL.07.01}, month = {April}, year = {2007}, institution = {University of California-Berkeley} } |
MIP-encoded Reformulation for Machine -Job Assignment | advance=1, barrier_algorithm=0,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=0,
barrier_limits_corrections=-1, barrier_limits_growth=1e+8,
barrier_ordering=1, barrier_qcpconvergetol=1e-07, barrier_startalg=1,
emphasis_memory=yes, emphasis_mip=1, emphasis_numerical=no,
feasopt_mode=0, feasopt_tolerance=1e-06, lpmethod=3,
mip_cuts_cliques=2, mip_cuts_covers=-1, mip_cuts_disjunctive=1,
mip_cuts_flowcovers=1, mip_cuts_gomory=1, mip_cuts_gubcovers=1,
mip_cuts_implied=0, mip_cuts_mircut=-1, mip_cuts_pathcut=1,
mip_limits_aggforcut=3, mip_limits_cutpasses=0,
mip_limits_cutsfactor=4, mip_limits_gomorycand=100,
mip_limits_gomorypass=0, mip_limits_polishtime=0,
mip_limits_probetime=5, mip_limits_repairtries=0,
mip_limits_strongcand=10, mip_limits_strongit=0,
mip_limits_submipnodelim=125, mip_ordertype=0,
mip_strategy_backtrack=0.999999, mip_strategy_bbinterval=7,
mip_strategy_branch=1, mip_strategy_dive=2, mip_strategy_file=1,
mip_strategy_heuristicfreq=0, mip_strategy_lbheur=yes,
mip_strategy_nodeselect=3, mip_strategy_order=yes,
mip_strategy_presolvenode=0, mip_strategy_probe=0,
mip_strategy_rinsheur=0, mip_strategy_startalgorithm=4,
mip_strategy_subalgorithm=0, mip_strategy_variableselect=2,
network_netfind=1, network_pricing=0, preprocessing_aggregator=-1,
preprocessing_boundstrength=0, preprocessing_coeffreduce=2,
preprocessing_dependency=-1, preprocessing_dual=1,
preprocessing_fill=40, preprocessing_numpass=-1,
preprocessing_presolve=yes, preprocessing_qpmakepsd=yes,
preprocessing_reduce=1, preprocessing_relax=0,
preprocessing_repeatpresolve=1, preprocessing_symmetry=-1, qpmethod=0,
read_scale=-1, sifting_algorithm=0, simplex_crash=0,
simplex_dgradient=1, simplex_limits_perturbation=0,
simplex_limits_singularity=20, simplex_perturbation=no 1e-06,
simplex_pgradient=1, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
|
| CPLEX-CLS | We cannot provide these instances online due to copyright issues. You can acquire them from the Berkeley Computational Optimization Lab. Here are the names of our 50 training instances and names of our 50 test instances. |
Training instance/seed lists Test instance/seed list |
@ARTICLE{AM:ls-poly,
|
MIP-encoded capacitated lot-sizing | advance=1, barrier_algorithm=2,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=-1,
barrier_limits_corrections=-1, barrier_limits_growth=1e+14,
barrier_ordering=0, barrier_qcpconvergetol=1e-07, barrier_startalg=1,
emphasis_memory=yes, emphasis_mip=0, emphasis_numerical=no,
feasopt_mode=0, feasopt_tolerance=1e-06, lpmethod=0,
mip_cuts_cliques=1, mip_cuts_covers=0, mip_cuts_disjunctive=3,
mip_cuts_flowcovers=2, mip_cuts_gomory=0, mip_cuts_gubcovers=0,
mip_cuts_implied=0, mip_cuts_mircut=1, mip_cuts_pathcut=0,
mip_limits_aggforcut=3, mip_limits_cutpasses=0,
mip_limits_cutsfactor=4, mip_limits_gomorycand=200,
mip_limits_gomorypass=0, mip_limits_polishtime=0,
mip_limits_probetime=1e+75, mip_limits_repairtries=0,
mip_limits_strongcand=10, mip_limits_strongit=0,
mip_limits_submipnodelim=250, mip_ordertype=2,
mip_strategy_backtrack=0.9999, mip_strategy_bbinterval=7,
mip_strategy_branch=0, mip_strategy_dive=1, mip_strategy_file=1,
mip_strategy_heuristicfreq=80, mip_strategy_lbheur=no,
mip_strategy_nodeselect=1, mip_strategy_order=yes,
mip_strategy_presolvenode=0, mip_strategy_probe=0,
mip_strategy_rinsheur=-1, mip_strategy_startalgorithm=0,
mip_strategy_subalgorithm=0, mip_strategy_variableselect=0,
network_netfind=2, network_pricing=0, preprocessing_aggregator=-1,
preprocessing_boundstrength=-1, preprocessing_coeffreduce=2,
preprocessing_dependency=3, preprocessing_dual=0,
preprocessing_fill=10, preprocessing_numpass=-1,
preprocessing_presolve=yes, preprocessing_qpmakepsd=no,
preprocessing_reduce=1, preprocessing_relax=-1,
preprocessing_repeatpresolve=2, preprocessing_symmetry=0, qpmethod=3,
read_scale=0, sifting_algorithm=2, simplex_crash=1,
simplex_dgradient=5, simplex_limits_perturbation=0,
simplex_limits_singularity=10, simplex_perturbation=no 1e-06,
simplex_pgradient=1, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
|
| CPLEX-MIK | We cannot provide these instances online due to copyright issues. You can acquire them from the Berkeley Computational Optimization Lab. Here are the names of our 60 training instances and names of our 60 test instances. |
Training instance/seed lists Test instance/seed list |
@ARTICLE{A:mip,
|
Mixed-integer knapsack | advance=1, barrier_algorithm=3,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=1,
barrier_limits_corrections=-1, barrier_limits_growth=1e+6,
barrier_ordering=2, barrier_qcpconvergetol=1e-07, barrier_startalg=1,
emphasis_memory=yes, emphasis_mip=2, emphasis_numerical=no,
feasopt_mode=1, feasopt_tolerance=1e-06, lpmethod=0,
mip_cuts_cliques=-1, mip_cuts_covers=-1, mip_cuts_disjunctive=0,
mip_cuts_flowcovers=-1, mip_cuts_gomory=0, mip_cuts_gubcovers=0,
mip_cuts_implied=0, mip_cuts_mircut=2, mip_cuts_pathcut=-1,
mip_limits_aggforcut=3, mip_limits_cutpasses=0,
mip_limits_cutsfactor=4, mip_limits_gomorycand=800,
mip_limits_gomorypass=0, mip_limits_polishtime=0,
mip_limits_probetime=10, mip_limits_repairtries=0,
mip_limits_strongcand=10, mip_limits_strongit=0,
mip_limits_submipnodelim=500, mip_ordertype=1,
mip_strategy_backtrack=0.99, mip_strategy_bbinterval=7,
mip_strategy_branch=-1, mip_strategy_dive=3, mip_strategy_file=1,
mip_strategy_heuristicfreq=80, mip_strategy_lbheur=no,
mip_strategy_nodeselect=2, mip_strategy_order=yes,
mip_strategy_presolvenode=-1, mip_strategy_probe=-1,
mip_strategy_rinsheur=-1, mip_strategy_startalgorithm=0,
mip_strategy_subalgorithm=0, mip_strategy_variableselect=1,
network_netfind=3, network_pricing=1, preprocessing_aggregator=-1,
preprocessing_boundstrength=-1, preprocessing_coeffreduce=1,
preprocessing_dependency=-1, preprocessing_dual=-1,
preprocessing_fill=20, preprocessing_numpass=-1,
preprocessing_presolve=yes, preprocessing_qpmakepsd=yes,
preprocessing_reduce=3, preprocessing_relax=-1,
preprocessing_repeatpresolve=1, preprocessing_symmetry=-1, qpmethod=2,
read_scale=-1, sifting_algorithm=4, simplex_crash=0,
simplex_dgradient=1, simplex_limits_perturbation=0,
simplex_limits_singularity=10, simplex_perturbation=no 1e-06,
simplex_pgradient=4, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
|
| CPLEX-QP | Training set (1000 instances) |
Test set (1000 instances) |
Training instance/seed lists Test instance/seed list |
To come | Quadratic programs from RNA energy parameter optimization | advance=1, barrier_algorithm=2,
barrier_colnonzeros=0, barrier_convergetol=1e-08, barrier_crossover=2,
barrier_limits_corrections=-1, barrier_limits_growth=1e+12,
barrier_ordering=0, barrier_qcpconvergetol=1e-07, barrier_startalg=3,
emphasis_memory=no, emphasis_mip=3, emphasis_numerical=no,
feasopt_mode=5, feasopt_tolerance=1e-04, lpmethod=6, network_netfind=1,
network_pricing=0, preprocessing_aggregator=-1,
preprocessing_boundstrength=0, preprocessing_coeffreduce=2,
preprocessing_dependency=-1, preprocessing_dual=0,
preprocessing_fill=2, preprocessing_numpass=-1,
preprocessing_presolve=no, preprocessing_qpmakepsd=no,
preprocessing_reduce=0, preprocessing_relax=0,
preprocessing_repeatpresolve=0, preprocessing_symmetry=2, qpmethod=2,
read_scale=1, sifting_algorithm=2, simplex_crash=0,
simplex_dgradient=4, simplex_limits_perturbation=0,
simplex_limits_singularity=10, simplex_perturbation=yes 1e-06,
simplex_pgradient=-1, simplex_pricing=0, simplex_refactor=0,
simplex_tolerances_feasibility=1e-06,
simplex_tolerances_markowitz=0.01, simplex_tolerances_optimality=1e-06 in my format result comparison against default |
Please send any questions, concerns or comments to Frank Hutter