This class represents the settings for Gradient Descent. More...
#include <DerivativeSettings.h>
Public Member Functions | |
| DerivativeSettings () | |
| Creates a new set of monotonicity checking settings. | |
| bool | isFeasibleInstantiationSearchSet () const |
| Retrieves whether a feasible instance should be found by Gradient Descent. | |
| boost::optional< std::string > | getDerivativeAtInstantiation () const |
| Retrieves whether an extremum should be found by Gradient Descent. | |
| double | getLearningRate () const |
| Retrieves the learning rate for the gradient descent. | |
| uint_fast64_t | getMiniBatchSize () const |
| Retrieves the mini batch size of the gradient descent. | |
| double | getAverageDecay () const |
| Retrieves the decay of the decaying step average of the ADAM algorithm. | |
| double | getSquaredAverageDecay () const |
| Retrieves the decay of the squared decaying step average of the ADAM algorithm. | |
| bool | isPrintJsonSet () const |
| Retrieves whether the GradientDescentInstantiationSearcher should print the run as json after finishing. | |
| boost::optional< derivative::GradientDescentConstraintMethod > | getConstraintMethod () const |
| Retrieves the gradient descent method constraint method. | |
| std::string | getConstraintMethodAsString () const |
| Retrieves the gradient descent method constraint method as a string. | |
| double | getTerminationEpsilon () const |
| Retrieves the termination epsilon. | |
| boost::optional< derivative::GradientDescentMethod > | getGradientDescentMethod () const |
| Retrieves the gradient descent method. | |
| std::string | getGradientDescentMethodAsString () const |
| Retrieves the gradient descent method as a string. | |
| bool | areInconsequentialParametersOmitted () const |
| Are inconsequential parameters omitted? | |
 Public Member Functions inherited from storm::settings::modules::ModuleSettings | |
| ModuleSettings (std::string const &moduleName) | |
| Constructs a new settings object. | |
| virtual | ~ModuleSettings () |
| virtual bool | check () const |
| Checks whether the settings are consistent. | |
| virtual void | finalize () |
| Prepares the modules for further usage, should be called at the end of the initialization, before checks are executed. | |
| std::unique_ptr< storm::settings::SettingMemento > | overrideOption (std::string const &name, bool requiredStatus) |
| Sets the option with the given name to the required status. | |
| std::string const & | getModuleName () const |
| Retrieves the name of the module to which these settings belong. | |
| std::vector< std::shared_ptr< Option > > const & | getOptions () const |
| Retrieves the options of this module. | |
| uint_fast64_t | getPrintLengthOfLongestOption (bool includeAdvanced) const |
| Retrieves the (print) length of the longest option. | |
| void | restoreDefaults () |
| Restores the default values for all arguments of all options. | |
Static Public Attributes | |
| static const std::string | moduleName = "derivative" |
Additional Inherited Members | |
| Protected Member Functions inherited from storm::settings::modules::ModuleSettings | |
| Option & | getOption (std::string const &longName) |
| Retrieves the option with the given long name. | |
| Option const & | getOption (std::string const &longName) const |
| Retrieves the option with the given long name. | |
| bool | isSet (std::string const &optionName) const |
| Retrieves whether the option with the given name was set. | |
| void | set (std::string const &name) |
| Sets the option with the specified name. | |
| void | unset (std::string const &name) |
| Unsets the option with the specified name. | |
| void | addOption (std::shared_ptr< Option > const &option) |
| Adds and registers the given option. | |
Detailed Description
This class represents the settings for Gradient Descent.
Definition at line 17 of file DerivativeSettings.h.
Constructor & Destructor Documentation
◆ DerivativeSettings()
| storm::settings::modules::DerivativeSettings::DerivativeSettings | ( | ) |
Creates a new set of monotonicity checking settings.
Definition at line 31 of file DerivativeSettings.cpp.
Member Function Documentation
◆ areInconsequentialParametersOmitted()
| bool storm::settings::modules::DerivativeSettings::areInconsequentialParametersOmitted | ( | ) | const |
Are inconsequential parameters omitted?
Definition at line 130 of file DerivativeSettings.cpp.
◆ getAverageDecay()
| double storm::settings::modules::DerivativeSettings::getAverageDecay | ( | ) | const |
Retrieves the decay of the decaying step average of the ADAM algorithm.
Definition at line 107 of file DerivativeSettings.cpp.
◆ getConstraintMethod()
| boost::optional< derivative::GradientDescentConstraintMethod > storm::settings::modules::DerivativeSettings::getConstraintMethod | ( | ) | const |
Retrieves the gradient descent method constraint method.
Definition at line 134 of file DerivativeSettings.cpp.
◆ getConstraintMethodAsString()
| std::string storm::settings::modules::DerivativeSettings::getConstraintMethodAsString | ( | ) | const |
Retrieves the gradient descent method constraint method as a string.
Definition at line 138 of file DerivativeSettings.cpp.
◆ getDerivativeAtInstantiation()
| boost::optional< std::string > storm::settings::modules::DerivativeSettings::getDerivativeAtInstantiation | ( | ) | const |
Retrieves whether an extremum should be found by Gradient Descent.
Definition at line 93 of file DerivativeSettings.cpp.
◆ getGradientDescentMethod()
| boost::optional< derivative::GradientDescentMethod > storm::settings::modules::DerivativeSettings::getGradientDescentMethod | ( | ) | const |
Retrieves the gradient descent method.
Definition at line 122 of file DerivativeSettings.cpp.
◆ getGradientDescentMethodAsString()
| std::string storm::settings::modules::DerivativeSettings::getGradientDescentMethodAsString | ( | ) | const |
Retrieves the gradient descent method as a string.
Definition at line 126 of file DerivativeSettings.cpp.
◆ getLearningRate()
| double storm::settings::modules::DerivativeSettings::getLearningRate | ( | ) | const |
Retrieves the learning rate for the gradient descent.
Definition at line 101 of file DerivativeSettings.cpp.
◆ getMiniBatchSize()
| uint_fast64_t storm::settings::modules::DerivativeSettings::getMiniBatchSize | ( | ) | const |
Retrieves the mini batch size of the gradient descent.
Definition at line 104 of file DerivativeSettings.cpp.
◆ getSquaredAverageDecay()
| double storm::settings::modules::DerivativeSettings::getSquaredAverageDecay | ( | ) | const |
Retrieves the decay of the squared decaying step average of the ADAM algorithm.
Definition at line 110 of file DerivativeSettings.cpp.
◆ getTerminationEpsilon()
| double storm::settings::modules::DerivativeSettings::getTerminationEpsilon | ( | ) | const |
Retrieves the termination epsilon.
Definition at line 118 of file DerivativeSettings.cpp.
◆ isFeasibleInstantiationSearchSet()
| bool storm::settings::modules::DerivativeSettings::isFeasibleInstantiationSearchSet | ( | ) | const |
Retrieves whether a feasible instance should be found by Gradient Descent.
Definition at line 89 of file DerivativeSettings.cpp.
◆ isPrintJsonSet()
| bool storm::settings::modules::DerivativeSettings::isPrintJsonSet | ( | ) | const |
Retrieves whether the GradientDescentInstantiationSearcher should print the run as json after finishing.
Definition at line 114 of file DerivativeSettings.cpp.
Member Data Documentation
◆ moduleName
|
static |
Definition at line 89 of file DerivativeSettings.h.
The documentation for this class was generated from the following files:
- src/storm-pars/settings/modules/DerivativeSettings.h
- src/storm-pars/settings/modules/DerivativeSettings.cpp
