| Option Name | MPL Name | Solver Param | ParamNr | Type | Default | Min | Max |
|---|---|---|---|---|---|---|---|
| Algorithm | OptimizeAlgorithm | SolveMethod | 1 | list | 1 | 0 | 5 |
| Presolve level | PresolveType | PresolveType | 2 | list | 0 | 0 | 2 |
| Scaling | Scaling | Scaling | 3 | list | 0 | 0 | 2 |
| Perturbation | Perturbation | Perturbation | 4 | list | 0 | -MAXREAL | MAXREAL |
| Log output level | LogLevel | LogLevel | 7 | int | 1 | 0 | 4 |
| Iteration limit | IterationLimit | MaxIter | 8 | int | 99999999 | 0 | MAXINT |
| Time limit | SolTimeLimit | MaxSeconds | 16 | real | -1 | -1 | MAXREAL |
| Use crash method | CrashInd | CrashInd | 9 | flag | 0 | 0 | 1 |
| Crash pivoting method | CrashPivot | CrashPivot | 10 | list | 0 | -1 | 3 |
| Crash gap | CrashGap | CrashGap | 11 | real | 1000.0 | 0 | MAXREAL |
| Primal objective limit | PrimalObjectLimit | PrimalObjLim | 12 | real | MAXREAL | 0.0 | MAXREAL |
| Dual objective limit | DualObjectLimit | DualObjLim | 13 | real | MAXREAL | 0.0 | MAXREAL |
| Primal tolerance | PrimalTolerance | PrimalTolerance | 14 | 1e-7 | 0 | MAXREAL | |
| Dual tolerance | DualTolerance | DualObjTol | 15 | 1e-7 | 0 | MAXREAL | |
| Node limit | MipNodeLimit | MIPMaxNodes | 17 | int | 9999999 | 0 | 9999999 |
| Maximum integer solutions | MipSolutionLimit | MIPMaxSol | 18 | int | 9999999 | 0 | 9999999 |
| MIP time limit | MipSolTimeLimit | MIPMaxSec | 19 | int | 9999999 | 0 | 9999999 |
| Use fathoming discipline | MipFathomDiscipline | MIPFathomDisc | 20 | flag | 0 | 0 | 1 |
| Minimum drop | MipMinimumDrop | MIPMinimumDrop | 22 | int | 1e-4 | -MAXINT | MAXINT |
| Maximum cut passes | MipMaxCutPasses | MIPMaxCutPass | 23 | int | 10 | 0 | MAXINT |
| Maximum root node cut passes | MIPMaxPassRoot | MipMaxPassRoot | 24 | int | 20 | 0 | MAXINT |
| Strong branching level | MipStrongBranching | MipStrongBranch | 25 | int | 5 | 0 | MAXINT |
| Scanning for global cuts | MipScanGlobalCuts | MIPScanGlobCuts | 26 | int | 1 | 0 | MAXINT |
| Integral feasibility tolerance | MipIntegerTol | MIPIntTol | 30 | real | 1e-6 | 0 | 1 |
| MIP Cut-off value | MipCutoffIncrement | MIPCutOff | 32 | real | 1e-5 | 0 | 1 |
| Absolute optimality gap | MipAbsoluteGap | MIPAbsGap | 33 | real | 1e-10 | 0 | MAXREAL |
| Use probing | MipCutProbing | MIPCutProbing | 110 | flag | 1 | 0 | 1 |
| Probing frequency | MipProbeFrequency | MIPProbeFreq | 111 | int | -1 | -MAXINT | MAXINT |
| Use objective in probe | MipProbeUseObjective | MIPProbeUseObj | 113 | flag | 1 | 0 | 1 |
| Maximum node passes for probing | MipProbeMaxPass | MIPProbeMaxPass | 114 | int | 3 | 0 | MAXINT |
| Probing variable limit | MipProbeMaxProbe | MIPProbeMaxProbe | 115 | int | 100 | 0 | MAXINT |
| Maximum single probing | MipProbeMaxLook | MIPProbeMaxLook | 116 | int | 50 | 0 | MAXINT |
| Row cuts | MipProbeRowCuts | MIPProbeRowCuts | 117 | list | 0 | 0 | 3 |
| Gomory cuts | MipCutGomory | MIPCutGomory | 120 | flag | 1 | 0 | 1 |
| Gomory frequency | MipGomoryFrequency | MIPGomoryFreq | 121 | int | -1 | -MAXINT | MAXINT |
| Gomory limit | MipGomoryLimit | MIPGomoryLimit | 122 | int | 50 | 1 | MAXINT |
| Gomory tolerance | MipGomoryAway | MIPGomoryAway | 123 | real | 0.05 | 0.0001 | 0.5 |
| Knapsack cuts | MipCutKnapsack | MIPCutKnapsack | 130 | flag | 1 | 0 | 1 |
| Knapsack frequency | MipKnapsackFrequency | MIPKnapsackFreq | 131 | int | -1 | -MAXINT | MAXINT |
| Knapsack limit | MipKnapsackMaxIn | MIPKnapsackMaxIn | 132 | int | 50 | 1 | MAXINT |
| Oddhole cuts | MipCutOddHole | MIPCutOddhole | 140 | flag | 1 | 0 | 1 |
| Oddhole frequency | MipOddHoleFrequency | MIPOddholeFreq | 141 | int | -1 | -MAXINT | MAXINT |
| Oddhole minimum violation | MipOddHoleFrequency | MIPOddholeMinViol | 142 | real | 0.001 | 1.01e-8 | 0.5 |
| Oddhole minimum violation entry | MipOddHoleMinViolPer | MIPOddholeMinViolPer | 143 | real | 0.0003 | 1.01e-8 | 0.25 |
| Oddhole limit | MipOddHoleMaxEntries | MIPOddholeMaxEntries | 144 | int | 100 | 3 | MAXINT |
| Clique cuts | MipCutClique | MIPCutClique | 150 | flag | 1 | 0 | 1 |
| Clique frequency | MipCliqueFrequency | MIPCliqueFrequency | 151 | int | -1 | -MAXINT | MAXINT |
| Clique set packing | MipCliquePacking | MIPCliquePacking | 152 | flag | 0 | 0 | 1 |
| Clique star | MipCliqueStar | MIPCliqueStar | 153 | flag | 1 | 0 | 1 |
| Clique star candidate threshold | MipCliqueStarMaxLen | MIPCliqueStarMaxLen | 155 | int | 12 | 0 | MAXINT |
| Clique star report | MipCliqueStarMaxLen | MIPCliqueStarMaxLen | 156 | flag | 1 | 0 | 1 |
| Clique row | MipCliqueRow | MIPCliqueRow | 157 | flag | 1 | 0 | 1 |
| Clique row candidate threshold | MipCliqueRowMaxLen | MIPCliqueRowMaxLen | 158 | int | 12 | 0 | MAXINT |
| Clique row report | MipCliqueRowReport | MIPCliqueRowReport | 159 | flag | 1 | 0 | 1 |
| Clique minimum violation | MipCliqueMinViolation | MipCliqueMinViolation | 160 | real | -1 | -MAXREAL | MAXREAL |
| Lift and project cuts | MipCutLiftAndProject | MIPCutLiftProject | 170 | flag | 1 | 0 | 1 |
| Lift and project cut frequency | MipLiftCoverFrequency | MIPLiftProFreq | 171 | int | -1 | -MAXINT | MAXINT |
| Lift and project normalization | MipLiftCoverBetaOne | MIPLiftProBetaOne | 172 | flag | 1 | 0 | 1 |
| Simple rounding cuts | MipCutSimpleRounding | MIPCutSimpRound | 180 | flag | 1 | 0 | 1 |
| Simple rounding cut frequency | MipSimpleRoundFrequency | MIPSimpRoundFreq | 180 | int | -1 | -MAXINT | MAXINT |
Indicates which algorithm type to use when solving LPs.
| Dual Simplex (0) | Use the Dual Simplex method. |
| Primal Simplex (1) | Use the Primal Simplex method. |
| Primal Simplex or Sprint (2) | Use the Primal Simplex or Sprint method. |
| Barrier (3) | Use the Barrier Interior point method. |
| Barrier with no crossover (4) | Use the Barrier method with no crossover. |
| Automatic (5) | Solver decides based on model characteristics. |
This refers if CoinMP's presolver will be invoked.
| None (0) | No presolve done. |
| Do presolve (1) | Uses the presolver. |
| Presovle passes (2) | Applies a number of presolve passes. |
This sets what type of scaling procedure will be employed. The possible values are:
| No scaling (0) | No scaling performed. |
| Equilibrium (1) | Uses equilibrium scaling. |
| Geometric (2) | Uses geometric scaling. |
| Automatic (3) | Solvers decides. |
| Dynamic (4) | Uses dynamic scaling. |
Selects what manor of perturbation is performed.
| None (0) | No perturbation is used. |
| Use perturbation (50) | Perturbation is used. |
| Automatic perturbation (100) | Automatic perturbation is performed. |
| Perturbation not repeated (102) | Perturbation will not be repeated. |
Setting that relays iterative information to the log, higher setting relays more information to the output log file.
This is a limit on the number of iterations the solver will perform before terminating. The default is 99999999.
This is a time limit in seconds for CoinMP to solve the optimization problem at hand. The default value of -1 imposes no time limit.
This is a flag indicating if an initial solution will be computed using crash routines.
Defines the pivoting steps undertaken during the crash procedure, possible setting are:
| No pivoting (-1) | No pivoting used. |
| Automatic (0) | Algorithm chooses. |
| Simple pivoting (1) | Uses simple pivoting eg. GUB. |
| Mini iterations (2) | Uses mini iterations. |
| Free variables in basis (3) | Just place all the free variables in the basis. |
Sets the gap limit between the bounds to move the corresponding variable to the basis.
Sets the maximum possible value the primal objective function can attain.
Sets the maximum possible value the dual objective function can attain.
An individual variable is deemed primal feasible if it is less than the primal tolerance referred to by these methods below its lower bound and less than it above its upper bound (for a minimization problem reverse is true for a maximization problem).
An individual variable is deemed dual feasible if its reduced cost is greater than minus the dual tolerance or its distance to the upper bound is less than primal tolerance and the reduced cost is less than plus the tolerance or the distance to lower bound is less than primal tolerance (for a minimization problem reverse is true for a maximization problem).
This sets the limit on the maximum number of possible nodes generated during the branch and bound process. The default value is 9999999.
This sets the limit on the maximum number of possible integer feasible solutions found during the MIP optimization process. The default value is 9999999.
This is a time limit in seconds for which CoinMP will search through the branch and bound tree. The default value is 1e100.
Boolean option that determines whether to use a strict fathoming discipline. If enabled, a small tolerance is added to the new cut-off. This can avoid problems due to roundoff when the target value is exact.
Set the minimum drop to continue cuts. The default value is 0.0001.
Specifies the maximum number of passes performed to generate the cuts. The default value is 10.
Specifies the maximum number of passes performed at the root node to generate the cuts. The default value is 20.
This sets the maximum number of candidates at a node to be evaluated for strong branching. The default is 5.
This option specifies the intensity in searching for cuts that can be applied globally to the model. The default value is 1, larger values means more time and effort will be used in finding global cuts which may or may not be an effective use of computational time.
An integer variable is considered integer feasible if the absolute difference from the nearest integer is smaller than this. The default value is 0.000001.
The incremental cut off value by in which an incumbent integer solution has to be be better than the best solution found thus far in the MIP search. The default value is 0.00001.
The absolute tolerance on the gap between the best bound and best known integer feasible solution.
Boolean option that determines whether to enable the probing facility within CoinMP. The default setting is enabled.
Sets the level of probing that is to be performed, higher the value can result increased computational time before starting the branch and bound search.
This is a boolean option indicating whether the objective function is used in probing.
This sets the maximum number of passes that can be done per node for probing purposes. The default value is 3.
Sets the maximum number of unsatisfied variables to look at. The default value is 100.
Sets the maximum number of variables to look at in one probe. The default is 50.
Sets the type of row cuts used, possible settings are:
| None (0) | No cuts generated. |
| Disaggregation (1) | Just cuts of dissagregation types. |
| Coefficient (2) | Just coefficient row cut types. |
| Both (3) | Does both coefficient and dissagregation type cuts. |
This flag controls if gomory cuts will be used during the MIP optimization process. The default setting is enabled.
This sets the frequency of the gomory cut generation.
Specifies the how dense the cuts can be generated, the default value of 50 means CoinMP will not generate cuts with more than 50 non-zero coefficients, to generate more dense cuts one needs to increase this limit.
This value specifies the tolerance, whereby the gomory cut generator only looks at basic integer variables whose current value is at least this value from being integer.
This flag controls if gomory cuts will be used during the MIP optimization process. The default setting is enabled.
This sets the frequency of the knapsack cut generation.
Specifies the how dense the cuts can be generated, the default value of 50 means CoinMP will not generate cuts with more than 50 non-zero coefficients, to generate more dense cuts one needs to increase this limit.
This flag controls if oddhole cuts will be used during the MIP optimization process. The default setting is enabled.
This sets the frequency of the oddhole cut generation.
Sets the minimum violations that are allowed for the oddhole cuts. The default is 0.001.
Sets the minimum violations that are allowed per entry for the oddhole cuts. The default is 0.0003.
This sets the maximum number of entries in a oddhole cut. The default value is 100.
This flag controls if clique cuts will be used during the MIP optimization process. The default setting is enabled.
This sets the frequency of the clique cut generation.
Boolean option that specifies whether the problem at hand is a set packing problem, if so CoinMP will generate specific cuts for the problem.
Boolean option that specifies if star cliques will be enabled, the default setting is on.
This controls the maximum number of candidates considered for the clique star generation of cuts. The default value is 12.
Boolean option that specifies if one creates a report on the clique star generation. The default is enabled.
Boolean option that specifies if row cliques will be enabled, the default setting is on.
This controls the maximum number of candidates considered for the clique row generation of cuts. The default value is 12.
Boolean option that specifies if one creates a report on the clique row generation. The default is enabled.
This determines the minimum violation allowed for the clique cut generation.
Boolean option that specifies if lift and project cuts will be generated, the default setting is enabled.
This sets the frequency of the lift and project cut generation.
Boolean option that specifies if normalization is done during cut generation, the default setting is enabled.
Boolean option that specifies if simple rounding cuts will be generated, the default setting is enabled.
This sets the frequency of the simple rounding cut generation.