SG++-Doxygen-Documentation
|
Do transformation in all dimensions. More...
#include <OperationRosenblattTransformationKDE.hpp>
Public Member Functions | |
virtual void | doShuffledTransformation (base::DataMatrix &pointsCdf, base::DataMatrix &pointsUniform) |
virtual void | doTransformation (base::DataMatrix &pointsCdf, base::DataMatrix &pointsUniform) |
Rosenblatt Transformation with mixed starting dimension. More... | |
double | doTransformation1D (double x, base::DataVector &samples1d, double sigma, base::DataVector &kern) |
Rosenblatt transformation for one data point with given samples and and kernel evaluations, see doTransformation for details. More... | |
OperationRosenblattTransformationKDE (KernelDensityEstimator &kde, std::uint64_t seed=std::mt19937_64::default_seed) | |
virtual | ~OperationRosenblattTransformationKDE () |
Do transformation in all dimensions.
|
explicit |
|
virtual |
|
virtual |
References doTransformation1D(), sgpp::datadriven::Kernel::eval(), sgpp::base::DataVector::get(), sgpp::datadriven::KernelDensityEstimator::getDim(), sgpp::datadriven::KernelDensityEstimator::getKernel(), sgpp::base::DataMatrix::getNrows(), sgpp::base::DataMatrix::getRow(), sgpp::datadriven::KernelDensityEstimator::getSamples(), python.statsfileInfo::i, sgpp::base::DataVector::setAll(), and sgpp::base::DataMatrix::setRow().
|
virtual |
Rosenblatt Transformation with mixed starting dimension.
pointsCdf | Output base::DataMatrix (rows: # of samples, columns: # of dims) |
pointsUniform | data points to be transformed base::DataMatrix (rows: # of samples, columns: # of dims) |
References doTransformation1D(), sgpp::datadriven::Kernel::eval(), sgpp::base::DataVector::get(), sgpp::datadriven::KernelDensityEstimator::getKernel(), sgpp::base::DataMatrix::getNrows(), sgpp::base::DataMatrix::getRow(), sgpp::datadriven::KernelDensityEstimator::getSamples(), sgpp::base::DataVector::setAll(), and sgpp::base::DataMatrix::setRow().
double sgpp::datadriven::OperationRosenblattTransformationKDE::doTransformation1D | ( | double | x, |
base::DataVector & | samples1d, | ||
double | sigma, | ||
base::DataVector & | kern | ||
) |
Rosenblatt transformation for one data point with given samples and and kernel evaluations, see doTransformation for details.
x | data point |
samples1d | training samples in the dimension to be transformed |
sigma | bandwidth of the kernels in the current dimension |
kern | kernel evaluations |
References sgpp::datadriven::Kernel::cdf(), and sgpp::datadriven::KernelDensityEstimator::getKernel().
Referenced by doShuffledTransformation(), and doTransformation().