autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT > Class Template Reference

#include <optimization_problem.hpp>

Inheritance diagram for autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >:

Collaboration diagram for autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >:

Public Types
using	DomainValue = DomainValueT
using	Value = autoware::common::types::float64_t
using	Jacobian = Eigen::Matrix< Value, NumVars, NumJacobianCols >
using	Hessian = Eigen::Matrix< Value, NumVars, NumVars >
using	JacobianRef = Eigen::Ref< Jacobian >
using	HessianRef = Eigen::Ref< Hessian >

Public Member Functions
Value	operator() (const DomainValue &x)
void	jacobian (const DomainValue &x, JacobianRef out)
void	hessian (const DomainValue &x, HessianRef out)
void	evaluate (const DomainValue &x, const ComputeMode &mode)

Static Public Attributes
static constexpr auto	NumJacobianCols = NumJacobianColsT
static constexpr auto	NumVars = NumVarsT

Additional Inherited Members
Protected Member Functions inherited from autoware::common::helper_functions::crtp< Derived >
const Derived &	impl () const
Derived &	impl ()

Detailed Description

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>
class autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >

CRTP base class for a mathematical expression.

Template Parameters

Derived	Implementation class. Must implement score_(..), jacobian_() and hessian(..).
DomainValueT	Parameter type.
NumJacobianColsT	Number of columns in the jacobian matrix.
NumVarsT	Number of variables.

Member Typedef Documentation

DomainValue

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::DomainValue = DomainValueT

Hessian

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::Hessian = Eigen::Matrix<Value, NumVars, NumVars>

HessianRef

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::HessianRef = Eigen::Ref<Hessian>

Jacobian

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::Jacobian = Eigen::Matrix<Value, NumVars, NumJacobianCols>

JacobianRef

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::JacobianRef = Eigen::Ref<Jacobian>

Value

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

using autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::Value = autoware::common::types::float64_t

Member Function Documentation

evaluate()

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

void autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::evaluate ( const DomainValue &  x, const ComputeMode &  mode  )

inline

Pre-compute and cache the score/jacobian/hessian values. Which terms are to be computed is specified with the ComputeMode argument. By default, this function does nothing. You can implement and hide this function in your implementation to allow for some caching behavior. Though it is encouraged to use CachedExpression which already handles a big portion of the cache state management.

Parameters

x	Parameter value.
mode	Computation mode. Which terms (score, jacobian, hessian) are to be computed can be set within this element.

hessian()

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

void autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::hessian ( const DomainValue &  x, HessianRef  out  )

inline

Get the hessian at a given parameter value.

Parameters

x	Parameter value.{
out	Evaluated hessian matrix.

jacobian()

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

void autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::jacobian ( const DomainValue &  x, JacobianRef  out  )

inline

Get the jacobian at a given parameter value.

Parameters

x	Parameter value.
out	Evaluated jacobian matrix.

operator()()

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

Value autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::operator() ( const DomainValue &  x )

inline

Get the result of an expression for a given parameter value.

Parameters

x	Parameter value

Returns

Evaluated score

Member Data Documentation

NumJacobianCols

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

constexpr auto autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::NumJacobianCols = NumJacobianColsT

static

NumVars

template<typename Derived, typename DomainValueT, int NumJacobianColsT, int NumVarsT>

constexpr auto autoware::common::optimization::Expression< Derived, DomainValueT, NumJacobianColsT, NumVarsT >::NumVars = NumVarsT

static

---

The documentation for this class was generated from the following file:

• optimization_problem.hpp