Class SQPOption
java.lang.Object
org.hipparchus.optim.nonlinear.vector.constrained.SQPOption
- All Implemented Interfaces:
OptimizationData
Parameter for SQP Algorithm.
- Since:
- 3.1
-
Field Summary
FieldsModifier and TypeFieldDescriptionstatic final doubleDefault parameter for alpha min during line search.static final doubleDefault parameter for quadratic line search.static final intDeprecated.static final doubleDefault tolerance for convergence and active constraint(proportional to sqrt of function an constraints evaluation accuracy)static final doubleDefault tolerance for QP convergence(proportional to function and constraints evaluation accuracy)static final doubleDefault Gradient EPS (proportional to function and constraints evaluation accuracy).static final GradientModeDefault Gradient mode.static final intDefault parameter for non monotone line search history.static final intDefault Max Iterations.static final intDefault max iteration before reset hessian.static final doubleDefault parameter for evaluation of Armijo condition for descend direction.static final intDefault max iteration admitted for QP subproblem.static final doubleDefault weight for augmented QP subproblem.static final doubleDefault max value admitted for additional variable in QP subproblem.static final booleanDefault flag for using BFGS update formula. -
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptiondoubleGet parameter line search alpha min.doublegetB()Get parameter for quadratic line search.intDeprecated.as of 4.1, only used by deprecatedSQPOptimizerSandSQPOptimizerGM; will be removed in 5.0Get Convergence criteria FunctiondoublegetEps()Get tolerance for convergence and active constraint evaluation.doublegetEpsQP()Get tolerance for QP convergencedoubleGet Gradient Eps.Get Gradient Mode.intGet parameter fot non-monotone line search history.intGet max IterationintGet max Iteration for the line searchdoublegetMu()Get parameter for evaluation of Armijo condition for descend direction.intGet max iteration admitted for QP subproblem evaluation.doubleGet weight for augmented QP subproblem.doubleGet max value admitted for the solution of the additional variable in QP subproblem.voidsetAlphaMin(double alphaMin) Set parameter for line search alpha min.voidsetB(double b) Set parameter for quadratic line search.voidsetConvCriteria(int convCriteria) Deprecated.as of 4.1, only used by deprecatedSQPOptimizerSandSQPOptimizerGM; will be removed in 5.0voidSet custom convergence criteria functionvoidsetEps(double eps) Set tolerance for convergence Choose value proportional to sqrt of function and constraints evaluation accuracyvoidsetEpsQP(double epsQP) Set tolerance for QP convergence Choose value proportional to function and constraints evaluation accuracyvoidsetGradientEps(double gradientEps) Set Gradient Eps Choose value proportional to function and constraints evaluation accuracy)voidsetGradientMode(GradientMode gradientMode) Set Gradient modevoidsetHistory(int history) Set parameter for non-monotone line search history.voidsetMaxIteration(int maxIteration) Set max IterationvoidsetMaxLineSearchIteration(int maxLineSearchIteration) Set max Iteration for the line searchvoidsetMu(double mu) Set parameter for evaluation of Armijo condition for descend direction.voidsetQpMaxLoop(int qpMaxLoop) Set max iteration admitted for QP subproblem evaluation.voidsetRhoCons(double rhoCons) Set weight for augmented QP subproblem.voidsetSigmaMax(double sigmaMax) Set max value admitted for the solution of the additional variable in QP subproblem.voidsetUseFunHessian(boolean useFunHessian) Enable or Disable using direct the function Hessian.booleanCheck if using direct the function Hessian is enabled or disabled.
-
Field Details
-
DEFAULT_CONV_CRITERIA
Deprecated.as of 4.1, only used by deprecatedSQPOptimizerSandSQPOptimizerGM; will be removed in 5.0Default convergence criteria.- See Also:
-
DEFAULT_EPSILON
public static final double DEFAULT_EPSILONDefault tolerance for convergence and active constraint(proportional to sqrt of function an constraints evaluation accuracy)- See Also:
-
DEFAULT_EPSILON_QP
public static final double DEFAULT_EPSILON_QPDefault tolerance for QP convergence(proportional to function and constraints evaluation accuracy) -
DEFAULT_RHO
public static final double DEFAULT_RHODefault weight for augmented QP subproblem.- See Also:
-
DEFAULT_SIGMA_MAX
public static final double DEFAULT_SIGMA_MAXDefault max value admitted for additional variable in QP subproblem.- See Also:
-
DEFAULT_QP_MAX_LOOP
public static final int DEFAULT_QP_MAX_LOOPDefault max iteration admitted for QP subproblem.- See Also:
-
DEFAULT_MU
public static final double DEFAULT_MUDefault parameter for evaluation of Armijo condition for descend direction.- See Also:
-
DEFAULT_B
public static final double DEFAULT_BDefault parameter for quadratic line search.- See Also:
-
DEFAULT_ALPHA_MIN
public static final double DEFAULT_ALPHA_MINDefault parameter for alpha min during line search.- See Also:
-
DEFAULT_HISTORY
public static final int DEFAULT_HISTORYDefault parameter for non monotone line search history.- See Also:
-
DEFAULT_USE_FUNCTION_HESSIAN
public static final boolean DEFAULT_USE_FUNCTION_HESSIANDefault flag for using BFGS update formula.- See Also:
-
DEFAULT_MAX_LINE_SEARCH_ITERATION
public static final int DEFAULT_MAX_LINE_SEARCH_ITERATIONDefault max iteration before reset hessian.- See Also:
-
DEFAULT_GRADIENT_MODE
Default Gradient mode. -
DEFAULT_GRAD_EPS
public static final double DEFAULT_GRAD_EPSDefault Gradient EPS (proportional to function and constraints evaluation accuracy). -
DEFAULT_MAX_ITERATION
public static final int DEFAULT_MAX_ITERATIONDefault Max Iterations.- See Also:
-
-
Constructor Details
-
SQPOption
public SQPOption()Simple constructor.This constructor uses all defaults values.
-
-
Method Details
-
setGradientMode
Set Gradient mode- Parameters:
gradientMode- gradient mode
-
getGradientMode
Get Gradient Mode.- Returns:
- Gradient Mode
-
setGradientEps
public void setGradientEps(double gradientEps) Set Gradient Eps Choose value proportional to function and constraints evaluation accuracy)- Parameters:
gradientEps- gradient mode
-
getGradientEps
public double getGradientEps()Get Gradient Eps.- Returns:
- Gradient Eps
-
setConvCriteria
Deprecated.as of 4.1, only used by deprecatedSQPOptimizerSandSQPOptimizerGM; will be removed in 5.0Set convergence criteria.- Parameters:
convCriteria- convergence criteria
-
getConvCriteria
Deprecated.as of 4.1, only used by deprecatedSQPOptimizerSandSQPOptimizerGM; will be removed in 5.0Get convergence criteria.- Returns:
- convergence criteria
-
setEps
public void setEps(double eps) Set tolerance for convergence Choose value proportional to sqrt of function and constraints evaluation accuracy- Parameters:
eps- tolerance for convergence and active constraint evaluation
-
getEps
public double getEps()Get tolerance for convergence and active constraint evaluation.- Returns:
- tolerance for convergence and active constraint evaluation
-
setEpsQP
public void setEpsQP(double epsQP) Set tolerance for QP convergence Choose value proportional to function and constraints evaluation accuracy- Parameters:
epsQP- tolerance for convergence
-
getEpsQP
public double getEpsQP()Get tolerance for QP convergence- Returns:
- tolerance for QP convergence
-
setRhoCons
public void setRhoCons(double rhoCons) Set weight for augmented QP subproblem.- Parameters:
rhoCons- weight for augmented QP subproblem
-
getRhoCons
public double getRhoCons()Get weight for augmented QP subproblem.- Returns:
- weight for augmented QP subproblem
-
setSigmaMax
public void setSigmaMax(double sigmaMax) Set max value admitted for the solution of the additional variable in QP subproblem.- Parameters:
sigmaMax- max value admitted for the solution of the additional variable in QP subproblem
-
getSigmaMax
public double getSigmaMax()Get max value admitted for the solution of the additional variable in QP subproblem.- Returns:
- max value admitted for the solution of the additional variable in QP subproblem
-
setQpMaxLoop
public void setQpMaxLoop(int qpMaxLoop) Set max iteration admitted for QP subproblem evaluation.- Parameters:
qpMaxLoop- max iteration admitted for QP subproblem evaluation
-
getQpMaxLoop
public int getQpMaxLoop()Get max iteration admitted for QP subproblem evaluation.- Returns:
- max iteration admitted for QP subproblem evaluation
-
setMu
public void setMu(double mu) Set parameter for evaluation of Armijo condition for descend direction.- Parameters:
mu- parameter for evaluation of Armijo condition for descend direction
-
getMu
public double getMu()Get parameter for evaluation of Armijo condition for descend direction.- Returns:
- parameter for evaluation of Armijo condition for descend direction
-
setB
public void setB(double b) Set parameter for quadratic line search.- Parameters:
b- parameter for quadratic line search
-
getB
public double getB()Get parameter for quadratic line search.- Returns:
- parameter for quadratic line search
-
setAlphaMin
public void setAlphaMin(double alphaMin) Set parameter for line search alpha min.- Parameters:
alphaMin- parameter line search
-
getAlphaMin
public double getAlphaMin()Get parameter line search alpha min.- Returns:
- parameter for quadratic line search
-
setHistory
public void setHistory(int history) Set parameter for non-monotone line search history.- Parameters:
history- parameter line search
-
getHistory
public int getHistory()Get parameter fot non-monotone line search history.- Returns:
- parameter for non-monotone line search history
-
setMaxLineSearchIteration
public void setMaxLineSearchIteration(int maxLineSearchIteration) Set max Iteration for the line search- Parameters:
maxLineSearchIteration- max Iteration for the line search
-
getMaxLineSearchIteration
public int getMaxLineSearchIteration()Get max Iteration for the line search- Returns:
- max Iteration for the line search
-
setMaxIteration
public void setMaxIteration(int maxIteration) Set max Iteration- Parameters:
maxIteration- max Iteration
-
getMaxIteration
public int getMaxIteration()Get max Iteration- Returns:
- max Iteration
-
setUseFunHessian
public void setUseFunHessian(boolean useFunHessian) Enable or Disable using direct the function Hessian.- Parameters:
useFunHessian- enable or Disable using direct the function Hessian
-
useFunHessian
public boolean useFunHessian()Check if using direct the function Hessian is enabled or disabled.- Returns:
- true if using direct the function Hessian is enabled
-
setConvergenceFunction
Set custom convergence criteria function- Parameters:
fun- convergence criteria function
-
getConvergenceFunction
Get Convergence criteria Function- Returns:
- the function of the custom convergenc criteria
-
SQPOptimizerSandSQPOptimizerGM; will be removed in 5.0