Namespaces | Classes | Typedefs | Functions | Variables
roboptim Namespace Reference

Meta-functions, functions and solvers related classes. More...

Namespaces

namespace  detail
namespace  fg
namespace  finiteDifferenceGradientPolicies
 

Contains finite difference gradients policies.


namespace  visualization
 

Graphic visualization.


Classes

class  ConstantFunction
 Constant function. More...
class  DerivableParametrizedFunction
 Parametrized function with parameter derivative available. More...
struct  derivativeSize< Function >
struct  derivativeSize< DifferentiableFunction >
struct  derivativeSize< TwiceDifferentiableFunction >
struct  derivativeSize< NTimesDerivableFunction< N > >
class  GenericDifferentiableFunction
 Define an abstract derivable function ( $C^1$). More...
class  CachedFunction
 Store previous function computation. More...
class  Split
class  BadGradient
 Exception thrown when a gradient check fail. More...
class  GenericFiniteDifferenceGradient
 Compute automatically a gradient with finite differences. More...
struct  GenericFunctionTraits
 GenericFunction traits. More...
class  GenericFunction
 Define an abstract mathematical function ( $C^0$). More...
struct  GenericFunctionTraits< EigenMatrixDense >
 Trait specializing GenericFunction for Eigen dense matrices. More...
struct  GenericFunctionTraits< EigenMatrixSparse >
 Trait specializing GenericFunction for Eigen sparse matrices. More...
class  NoSolution
 Tag a result if no solution has been found. More...
struct  EigenMatrixDense
 Tag type for functions using Eigen dense matrices. More...
struct  EigenMatrixSparse
 Tag type for functions using Eigen sparse matrices. More...
class  GenericSolver
 Abstract interface satisfied by all solvers. More...
class  IdentityFunction
 Identity function. More...
class  GenericLinearFunction
 Define an abstract linear function. More...
class  NTimesDerivableFunction< 2 >
 Explicit specialization for the stop case of NTimesDerivable class. More...
class  NTimesDerivableFunction
 Define a $\mathbb{R} \rightarrow \mathbb{R}^m$ function, derivable n times ( $n \geq 2$). More...
class  GenericNumericLinearFunction
 Build a linear function from a vector and a matrix. More...
class  NumericQuadraticFunction
 Build a quadratic function from a matrix and a vector. More...
class  ParametrizedFunction
 Define an abstract parametrized mathematical function ( $C^0$). More...
class  DummySolverLastState
 Dummy solver which always fails, but returns the last state of the solver. More...
class  DummySolver
 Dummy solver which always fails. More...
class  Problem< F, boost::mpl::vector<> >
class  Problem
class  GenericQuadraticFunction
 Define an abstract quadratic function. More...
class  ResultWithWarnings
 Represents the solution of an optimization problem when errors occurred during the solving process. More...
class  Result
 Represents the solution of an optimization problem. More...
class  SolverError
 Base exception class for solving errors. More...
class  SolverFactory
 Define a solver factory that instanciate the plug-ins. More...
class  SolverWarning
 Exception used for non-critical errors during optimization. More...
struct  Parameter
 Parameters type. More...
class  Solver
 Solver for a specific problem class. More...
class  SumOfC1Squares
 Sum of the squares of differentiable functions. More...
class  GenericTwiceDifferentiableFunction
 Define an abstract function which is derivable twice ( $C^2$). More...

Typedefs

typedef DifferentiableFunction DerivableFunction
 Legacy name of TwiceDifferentiableFunction.
typedef GenericFunction
< EigenMatrixDense > 		Function
 Dense function.
typedef GenericFunction
< EigenMatrixSparse > 		SparseFunction
 Sparse function.
typedef
GenericDifferentiableFunction
< EigenMatrixDense > 		DifferentiableFunction
 Dense differentiable function.
typedef
GenericDifferentiableFunction
< EigenMatrixSparse > 		DifferentiableSparseFunction
 Sparse differentiable function.
typedef
GenericNumericLinearFunction
< EigenMatrixDense > 		NumericLinearFunction
typedef
GenericTwiceDifferentiableFunction
< EigenMatrixDense > 		TwiceDifferentiableFunction
typedef
GenericTwiceDifferentiableFunction
< EigenMatrixSparse > 		TwiceDifferentiableSparseFunction
typedef GenericLinearFunction
< EigenMatrixDense > 		LinearFunction
typedef GenericLinearFunction
< EigenMatrixSparse > 		LinearSparseFunction
typedef
GenericQuadraticFunction
< EigenMatrixDense > 		QuadraticFunction
typedef
GenericQuadraticFunction
< EigenMatrixSparse > 		QuadraticSparseFunction
typedef TwiceDifferentiableFunction TwiceDerivableFunction
 Legacy name of DifferentiableFunction.

Functions

template<typename P , typename C >
void addNonScalarConstraint (P &problem, boost::shared_ptr< C > constraint, std::vector< Function::interval_t > interval, std::vector< Function::value_type > scale=std::vector< Function::value_type >()) throw (std::runtime_error)
template<typename T >
std::ostream & operator<< (std::ostream &o, const BadGradient< T > &f)
 Override operator<< to handle exception display.
template<typename T >
bool checkGradient (const GenericDifferentiableFunction< T > &function, typename GenericDifferentiableFunction< T >::size_type functionId, const typename GenericDifferentiableFunction< T >::vector_t &x, typename GenericDifferentiableFunction< T >::value_type threshold=finiteDifferenceThreshold) throw ()
 Check if a gradient is valid.
template<typename T >
void checkGradientAndThrow (const GenericDifferentiableFunction< T > &function, typename GenericDifferentiableFunction< T >::size_type functionId, const typename GenericDifferentiableFunction< T >::vector_t &x, typename GenericDifferentiableFunction< T >::value_type threshold=finiteDifferenceThreshold) throw (BadGradient<T>)
template<typename T >
std::ostream & operator<< (std::ostream &o, const GenericFunction< T > &f)
 Override operator<< to handle function display.
ROBOPTIM_DLLAPI std::ostream & operator<< (std::ostream &o, const GenericSolver &gs)
 Override operator<< to handle solver display.
ROBOPTIM_DLLAPI std::ostream & operator<< (std::ostream &o, const NoSolution &ns)
 Override operator<< to display ``no solution'' objects.
ROBOPTIM_DLLAPI long int & indent (std::ostream &o)
 The current indentation level for o.
ROBOPTIM_DLLAPI std::ostream & incindent (std::ostream &o)
 Increment the indentation.
ROBOPTIM_DLLAPI std::ostream & decindent (std::ostream &o)
 Decrement the indentation.
ROBOPTIM_DLLAPI std::ostream & resetindent (std::ostream &o)
 Reset the indentation.
ROBOPTIM_DLLAPI std::ostream & iendl (std::ostream &o)
 Print an end of line, then set the indentation.
ROBOPTIM_DLLAPI std::ostream & incendl (std::ostream &o)
 Increment the indentation, print an end of line, and set the indentation.
ROBOPTIM_DLLAPI std::ostream & decendl (std::ostream &o)
 Decrement the indentation, print an end of line, and set the indentation.
template<typename F >
std::ostream & operator<< (std::ostream &o, const ParametrizedFunction< F > &f)
 Override operator<< to handle function display.
template<typename F >
std::ostream & operator<< (std::ostream &o, const Problem< F, boost::mpl::vector<> > &pb)
 Override operator<< to handle problem display.
template<typename F , typename CLIST >
std::ostream & operator<< (std::ostream &o, const Problem< F, CLIST > &pb)
 Override operator<< to handle problem display.
ROBOPTIM_DLLAPI std::ostream & operator<< (std::ostream &o, const Result &r)
 Override operator<< to handle result display.
ROBOPTIM_DLLAPI std::ostream & operator<< (std::ostream &o, const SolverError &e)
 Override operator<< to handle error display.
template<typename T >
T * unionCast (void *ptr)
ROBOPTIM_DLLAPI std::ostream & operator<< (std::ostream &o, const Parameter &parameter)
 Override operator<< to display ``parameters'' objects.
template<typename T >
std::ostream & operator<< (std::ostream &, const std::vector< T > &)
 Display a vector.
template<typename T1 , typename T2 >
std::ostream & operator<< (std::ostream &, const std::pair< T1, T2 > &)
 Display a pair.
template<typename T >
std::ostream & operator<< (std::ostream &, const Eigen::MatrixBase< T > &)
 Display an Eigen object with the appropriate IOFormat.

Variables

static const double finiteDifferenceThreshold = 1e-4
 Default threshold for checkGradient.
static const double finiteDifferenceEpsilon = 1e-8
 Default epsilon for finite difference class.

Detailed Description

Meta-functions, functions and solvers related classes.

Typedef Documentation

typedef DifferentiableFunction roboptim::DerivableFunction

Legacy name of TwiceDifferentiableFunction.

typedef GenericDifferentiableFunction<EigenMatrixDense> roboptim::DifferentiableFunction

Dense differentiable function.

Examples:
problem-cc.cc.
typedef GenericDifferentiableFunction<EigenMatrixSparse> roboptim::DifferentiableSparseFunction

Sparse differentiable function.

typedef GenericFunction<EigenMatrixDense> roboptim::Function

Dense function.

Examples:
visualization-gnuplot-function.cc.
typedef GenericLinearFunction<EigenMatrixDense> roboptim::LinearFunction
typedef GenericLinearFunction<EigenMatrixSparse> roboptim::LinearSparseFunction
typedef GenericNumericLinearFunction<EigenMatrixDense> roboptim::NumericLinearFunction
typedef GenericQuadraticFunction<EigenMatrixDense> roboptim::QuadraticFunction
typedef GenericQuadraticFunction<EigenMatrixSparse> roboptim::QuadraticSparseFunction
typedef GenericFunction<EigenMatrixSparse> roboptim::SparseFunction

Sparse function.

typedef TwiceDifferentiableFunction roboptim::TwiceDerivableFunction

Legacy name of DifferentiableFunction.

typedef GenericTwiceDifferentiableFunction<EigenMatrixDense> roboptim::TwiceDifferentiableFunction
typedef GenericTwiceDifferentiableFunction<EigenMatrixSparse> roboptim::TwiceDifferentiableSparseFunction

Function Documentation

std::ostream & roboptim::decendl ( std::ostream &  o)

Decrement the indentation, print an end of line, and set the indentation.

References decindent(), and iendl().

std::ostream & roboptim::decindent ( std::ostream &  o)

Decrement the indentation.

References indent().

Referenced by decendl(), roboptim::detail::printConstraint< P >::operator()(), roboptim::ConstantFunction::print(), roboptim::IdentityFunction::print(), roboptim::BadGradient< T >::print(), roboptim::GenericNumericLinearFunction< T >::print(), roboptim::NumericQuadraticFunction::print(), roboptim::Result::print(), roboptim::ResultWithWarnings::print(), roboptim::Solver< F, C >::print(), roboptim::Problem< F, boost::mpl::vector<> >::print(), and roboptim::Problem< F, CLIST >::print().

std::ostream & roboptim::iendl ( std::ostream &  o)

Print an end of line, then set the indentation.

References indent().

Referenced by decendl(), incendl(), roboptim::detail::printConstraint< P >::operator()(), roboptim::ConstantFunction::print(), roboptim::IdentityFunction::print(), roboptim::BadGradient< T >::print(), roboptim::GenericNumericLinearFunction< T >::print(), roboptim::NumericQuadraticFunction::print(), roboptim::Result::print(), roboptim::ResultWithWarnings::print(), roboptim::Solver< F, C >::print(), roboptim::Problem< F, boost::mpl::vector<> >::print(), and roboptim::Problem< F, CLIST >::print().

std::ostream & roboptim::incendl ( std::ostream &  o)

Increment the indentation, print an end of line, and set the indentation.

References iendl(), and incindent().

Referenced by roboptim::Problem< F, boost::mpl::vector<> >::print(), and roboptim::Problem< F, CLIST >::print().

std::ostream & roboptim::incindent ( std::ostream &  o)

Increment the indentation.

References indent().

Referenced by incendl(), roboptim::detail::printConstraint< P >::operator()(), roboptim::ConstantFunction::print(), roboptim::IdentityFunction::print(), roboptim::BadGradient< T >::print(), roboptim::GenericNumericLinearFunction< T >::print(), roboptim::NumericQuadraticFunction::print(), roboptim::Result::print(), roboptim::ResultWithWarnings::print(), and roboptim::Solver< F, C >::print().

long int & roboptim::indent ( std::ostream &  o)

The current indentation level for o.

Referenced by decindent(), iendl(), incindent(), roboptim::GenericTwiceDifferentiableFunction< T >::print(), roboptim::GenericDifferentiableFunction< T >::print(), roboptim::GenericFunction< T >::print(), and resetindent().

template<typename T >
std::ostream & roboptim::operator<< ( std::ostream &  o,
const std::vector< T > &  vect 
)

Display a vector.

template<typename T1 , typename T2 >
std::ostream & roboptim::operator<< ( std::ostream &  o,
const std::pair< T1, T2 > &  p 
)

Display a pair.

template<typename T >
std::ostream & roboptim::operator<< ( std::ostream &  o,
const Eigen::MatrixBase< T > &  matrix 
)

Display an Eigen object with the appropriate IOFormat.

template<typename T >
std::ostream & roboptim::operator<< ( std::ostream &  o,
const BadGradient< T > &  f 
)

Override operator<< to handle exception display.

Parameters:
ooutput stream used for display
ffunction to be displayed
Returns:
output stream
std::ostream & roboptim::operator<< ( std::ostream &  o,
const Result &  r 
)

Override operator<< to handle result display.

Parameters:
ooutput stream used for display
rresult to be displayed
Returns:
output stream

References roboptim::Result::print().

std::ostream & roboptim::operator<< ( std::ostream &  o,
const SolverError &  e 
)

Override operator<< to handle error display.

Parameters:
ooutput stream used for display
eerror to be displayed
Returns:
output stream

References roboptim::SolverError::print().

template<typename F >
std::ostream & roboptim::operator<< ( std::ostream &  o,
const ParametrizedFunction< F > &  f 
)

Override operator<< to handle function display.

Parameters:
ooutput stream used for display
ffunction to be displayed
Returns:
output stream
std::ostream & roboptim::operator<< ( std::ostream &  o,
const GenericSolver &  gs 
)

Override operator<< to handle solver display.

Parameters:
ooutput stream used for display
gssolver to be displayed
Returns:
output stream

References roboptim::GenericSolver::print().

std::ostream & roboptim::operator<< ( std::ostream &  o,
const NoSolution &  ns 
)

Override operator<< to display ``no solution'' objects.

Parameters:
ooutput stream used for display
nsNoSolution object, ignored
Returns:
output stream
std::ostream & roboptim::resetindent ( std::ostream &  o)

Reset the indentation.

References indent().

template<typename T >
T* roboptim::unionCast ( void *  ptr)

Variable Documentation

const double roboptim::finiteDifferenceEpsilon = 1e-8 [static]

Default epsilon for finite difference class.

const double roboptim::finiteDifferenceThreshold = 1e-4 [static]

Default threshold for checkGradient.

 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines