SG++-Doxygen-Documentation
|
A simple shuffling functor for data samples that performs a random shuffling of the data. More...
#include <DataShufflingFunctorRandom.hpp>
Public Member Functions | |
DataShufflingFunctor * | clone () const override |
Clone pattern. More... | |
DataShufflingFunctorRandom (int64_t seed=1337) | |
Standard constructor. More... | |
size_t | operator() (size_t idx, size_t numSamples) override |
Overload the function-call operator that maps indexes to indexes via a permutation of the entire index set. More... | |
Public Member Functions inherited from sgpp::datadriven::DataShufflingFunctor | |
DataShufflingFunctor () | |
DataShufflingFunctor (const DataShufflingFunctor &rhs)=default | |
DataShufflingFunctor (DataShufflingFunctor &&rhs)=default | |
DataShufflingFunctor & | operator= (const DataShufflingFunctor &rhs)=default |
DataShufflingFunctor & | operator= (DataShufflingFunctor &&rhs)=default |
virtual | ~DataShufflingFunctor ()=default |
A simple shuffling functor for data samples that performs a random shuffling of the data.
A Feistel network is used to compute the permutation of [0...N-1] without linear memory complexity.
|
explicit |
Standard constructor.
seed | the seed for the random shuffling (seed == -1 corresponds to a random seed) |
References python.statsfileInfo::i.
|
overridevirtual |
Clone pattern.
Implements sgpp::datadriven::DataShufflingFunctor.
References python.statsfileInfo::i.
|
overridevirtual |
Overload the function-call operator that maps indexes to indexes via a permutation of the entire index set.
The permutation used is the identity.
idx | the original index |
numSamples | the number of indexes to permute in total |
Implements sgpp::datadriven::DataShufflingFunctor.