Mosek - Computer Sciences Department
Mosek - Computer Sciences Department Mosek - Computer Sciences Department
MOSEK 8The first option specifies the maximum number of interior-point iterations, in this case 20. The seond optionindicates a scaling option of 1, which is no scaling.We remark that users can also use symbolic constants in place of numerical values. For example, for the scalingoption users could use MSK SCALING NONE in place of the value 1. For a complete list of applicable symbolicconstants, consult the MOSEK parameter list available online at www.mosek.com.3.1 Memory Considerations for Nonlinear ProblemsThe GAMS workfactor option can be used to increase the amount of memory available to MOSEK. The generalsyntax is(modelname).workfactor = (value)with a default value of 1. See the section on “Using Solver Specific Options” for details. If GAMS/MOSEK runsout of memory, an error message is printed out:*** GAMS/MOSEK interface error.The size estimate for Hessian of Lagrangian is too small.Try to increase workfactor option from 1 to a larger value.GAMS/MOSEK estimates the size of the Hessian as 5∗(number of nonlinear variables)∗(workfactor). Becauseof symmetry, the size of the Hessian is bounded byH d ∗ (number of nonlinear variables) 2 /2where H d detotes the density of the Hessian and H d ∈ [0, 1]. Therefore, one can choose the workfactor as:workfactor = H d ∗ (number of nonlinear variables) ∗ /(5 ∗ 2)Note that for a seperable model (see “Modeling Issues Involving Convex Programs”), the workfactor can in factbe reduced to 1/5.4 Summary of MOSEK Options4.1 General and Preprocessing OptionsMSK IPAR OPTIMIZERoptimizer selectionMSK DPAR OPTIMIZER MAX TIMEtime limitMSK IPAR PRESOLVE ELIMINATOR USEswitch for free variable eliminationMSK IPAR PRESOLVE ELIM FILLfill-in control during presolveMSK IPAR PRESOLVE LINDEP USElinear dependency checkMSK IPAR PRESOLVE LINDEP WORK LIMmaximum work for finding linear dependenciesMSK IPAR PRESOLVE USEswitch for presolveMSK IPAR CACHE SIZE L1L1 cache size usedMSK IPAR CACHE SIZE L2
MOSEK 9L2 cache size usedMSK IPAR CPU TYPEspecifies the CPU typeMSK SPAR PARAM READ FILE NAMEname of a secondary MOSEK option fileMSK IPAR CONCURRENT NUM OPTIMIZERSmaximum number of optimizers during concurrent runMSK IPAR CONCURRENT PRIORITY DUAL SIMPLEXpriority of dual simplex algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY FREE SIMPLEXpriority of free simplex algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY INTPNTpriority of interior point algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEXpriority of primal simplex algorithm in concurrent runMSK IPAR INFEAS REPORT AUTOswitch for infeasibility reportMSK IPAR INFEAS REPORT LEVELoutput level for infeasibility report4.2 Problem Data OptionsMSK DPAR DATA TOL AIJzero tolerance for matrix coefficientsMSK DPAR DATA TOL AIJ LARGEwarning for large coefficients in matrixMSK DPAR DATA TOL BOUND INFbound value for infinityMSK DPAR DATA TOL BOUND WRNwarning for large boundsMSK DPAR DATA TOL CJ LARGEwarning for large coefficients in objectiveMSK DPAR DATA TOL C HUGEerror for huge coefficients in objectiveMSK DPAR DATA TOL QIJzero tolerance for Q matrix coefficientsMSK DPAR DATA TOL Xtolerance for fixed variablesMSK IPAR CHECK CONVEXITYlevel of convexity check for quadratic problemsMSK DPAR LOWER OBJ CUTlower objective limitMSK DPAR LOWER OBJ CUT FINITE TRHupper objective limit threasholdMSK DPAR UPPER OBJ CUTupper objective limitMSK DPAR UPPER OBJ CUT FINITE TRHlower objective limit threashold4.3 Output Options
- Page 2 and 3: MOSEK 2• Interior-point optimizer
- Page 4 and 5: MOSEK 41.5 Nonlinear ProgramsMOSEK
- Page 6 and 7: MOSEK 6The original problem is:∑d
- Page 10 and 11: MOSEK 10MSK IPAR LOG BIoutput contr
- Page 12 and 13: MOSEK 12MSK IPAR SIM PRIMAL CRASHpr
- Page 14 and 15: MOSEK 14MSK IPAR PRESOLVE LINDEP US
- Page 16 and 17: MOSEK 16MSK DPAR DATA TOL QIJ (real
- Page 18 and 19: MOSEK 18MSK DPAR INTPNT CO TOL NEAR
- Page 20 and 21: MOSEK 20MSK IPAR INTPNT STARTING PO
- Page 22 and 23: MOSEK 22MSK IPAR SIM PRIMAL CRASH (
- Page 24 and 25: MOSEK 24Coefficient reduction +512G
- Page 26 and 27: MOSEK 268 The optimizer for nonconv
- Page 28 and 29: MOSEK 28Simplex - cpu time: 0.00Pri
- Page 30: MOSEK 30Presolve - time : 0.00Preso
MOSEK 9L2 cache size usedMSK IPAR CPU TYPEspecifies the CPU typeMSK SPAR PARAM READ FILE NAMEname of a secondary MOSEK option fileMSK IPAR CONCURRENT NUM OPTIMIZERSmaximum number of optimizers during concurrent runMSK IPAR CONCURRENT PRIORITY DUAL SIMPLEXpriority of dual simplex algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY FREE SIMPLEXpriority of free simplex algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY INTPNTpriority of interior point algorithm in concurrent runMSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEXpriority of primal simplex algorithm in concurrent runMSK IPAR INFEAS REPORT AUTOswitch for infeasibility reportMSK IPAR INFEAS REPORT LEVELoutput level for infeasibility report4.2 Problem Data OptionsMSK DPAR DATA TOL AIJzero tolerance for matrix coefficientsMSK DPAR DATA TOL AIJ LARGEwarning for large coefficients in matrixMSK DPAR DATA TOL BOUND INFbound value for infinityMSK DPAR DATA TOL BOUND WRNwarning for large boundsMSK DPAR DATA TOL CJ LARGEwarning for large coefficients in objectiveMSK DPAR DATA TOL C HUGEerror for huge coefficients in objectiveMSK DPAR DATA TOL QIJzero tolerance for Q matrix coefficientsMSK DPAR DATA TOL Xtolerance for fixed variablesMSK IPAR CHECK CONVEXITYlevel of convexity check for quadratic problemsMSK DPAR LOWER OBJ CUTlower objective limitMSK DPAR LOWER OBJ CUT FINITE TRHupper objective limit threasholdMSK DPAR UPPER OBJ CUTupper objective limitMSK DPAR UPPER OBJ CUT FINITE TRHlower objective limit threashold4.3 Output Options
Change language