|
T | apply (const Matrix &P, OptionalJacobian< -1, -1 > H={}) const |
| Manifold evaluation. More...
|
|
| ManifoldEvaluationFunctor () |
| For serialization. More...
|
|
| ManifoldEvaluationFunctor (size_t N, double x) |
| Default Constructor. More...
|
|
| ManifoldEvaluationFunctor (size_t N, double x, double a, double b) |
| Constructor, with interval [a,b]. More...
|
|
T | operator() (const Matrix &P, OptionalJacobian< -1, -1 > H={}) const |
| c++ sugar More...
|
|
Vector | apply (const Matrix &P, OptionalJacobian< -1, -1 > H={}) const |
| M-dimensional evaluation. More...
|
|
Vector | operator() (const Matrix &P, OptionalJacobian< -1, -1 > H={}) const |
| c++ sugar More...
|
|
EIGEN_MAKE_ALIGNED_OPERATOR_NEW | VectorEvaluationFunctor () |
| For serialization. More...
|
|
| VectorEvaluationFunctor (size_t M, size_t N, double x) |
| Default Constructor. More...
|
|
| VectorEvaluationFunctor (size_t M, size_t N, double x, double a, double b) |
| Constructor, with interval [a,b]. More...
|
|
|
using | Jacobian = Eigen::Matrix< double, -1, -1 > |
|
void | calculateJacobian () |
|
double | apply (const typename DERIVED::Parameters &p, OptionalJacobian<-1, -1 > H={}) const |
| Regular 1D evaluation. More...
|
|
| EvaluationFunctor () |
| For serialization. More...
|
|
| EvaluationFunctor (size_t N, double x) |
| Constructor with interval [a,b]. More...
|
|
| EvaluationFunctor (size_t N, double x, double a, double b) |
| Constructor with interval [a,b]. More...
|
|
double | operator() (const typename DERIVED::Parameters &p, OptionalJacobian<-1, -1 > H={}) const |
| c++ sugar More...
|
|
void | print (const std::string &s="") const |
|
Jacobian | H_ |
|
size_t | M_ |
|
Weights | weights_ |
|
template<typename DERIVED>
template<class T>
class gtsam::Basis< DERIVED >::ManifoldEvaluationFunctor< T >
Manifold EvaluationFunctor at a given x, applied to a parameter Matrix. This functor is used to evaluate a parameterized function at a given scalar value x. When given a specific M*N parameters, returns an M-vector the M corresponding functions at x, possibly with Jacobians wrpt the parameters.
The difference with the VectorEvaluationFunctor is that after computing the M*1 vector xi=F(x;P), with x a scalar and P the M*N parameter vector, we also retract xi back to the T manifold. For example, if T==Rot3, then we first compute a 3-vector xi using x and P, and then map that 3-vector xi back to the Rot3 manifold, yielding a valid 3D rotation.
Definition at line 293 of file Basis.h.