|
CMP++: Uncertainty Quantification & Bayesian Calibration
|
Epanechnikov (parabolic) density kernel. More...
#include <kernel++.h>


Public Member Functions | |
| Epanechnikov ()=default | |
| double | eval (const Eigen::VectorXd &z_diff) const override |
| Evaluates the unnormalized Epanechnikov kernel. | |
| double | normalizationConstant (const size_t &dim) const override |
| Returns the normalization constant for a given dimension. | |
| double | applyToGradient (const Eigen::VectorXd &z, const Eigen::VectorXd &grad_z_i) const override |
| Computes the gradient of the kernel evaluation. | |
Public Member Functions inherited from cmp::kernel::Kernel | |
| virtual | ~Kernel ()=default |
Static Public Member Functions | |
| static std::shared_ptr< Kernel > | make () |
| Factory method for creating an Epanechnikov kernel. | |
Epanechnikov (parabolic) density kernel.
Mathematical Formulation The multivariate Epanechnikov kernel:
\[ K(\mathbf{z}) \propto \begin{cases} 1 - \|\mathbf{z}\|_2^2 & \text{if } \|\mathbf{z}\|_2 \le 1 \\ 0 & \text{otherwise} \end{cases} \]
|
default |
|
overridevirtual |
Computes the gradient of the kernel evaluation.
Implements cmp::kernel::Kernel.
|
overridevirtual |
Evaluates the unnormalized Epanechnikov kernel.
Implements cmp::kernel::Kernel.
|
inlinestatic |
Factory method for creating an Epanechnikov kernel.
|
overridevirtual |
Returns the normalization constant for a given dimension.
Implements cmp::kernel::Kernel.