Dakota
Version
Explore and Predict with Confidence
|
Base class/interface for the bootstrap sampler. More...
Public Member Functions | |
BootstrapSamplerBase (size_t data_size, Data orig_data) | |
Constructor for the bootstrap functor base. | |
virtual | ~BootstrapSamplerBase () |
Destructor. | |
virtual void | operator() (size_t num_samp, Data &bootstrapped_sample)=0 |
Generate and store a new bootstrapped sample into bootstrapped_sample. | |
virtual size_t | getDataSize () |
Obatin the number of samples used in the empirical distribution. | |
virtual void | operator() (Data &bootstrapped_sample) |
Generate and store an dataSize out of dataSize bootstrap sample. | |
virtual Data | operator() () |
Return bootstrapped sample. | |
Static Public Member Functions | |
static void | set_seed (size_t seed) |
Protected Attributes | |
boost::random::uniform_int_distribution | sampler |
Uniform distribution to provide samples from the empirical distribution. | |
const size_t | dataSize |
Size of the dataset defining the empirical distribution. | |
Data | origData |
Original data defining the empirical distribution TODO: Consider if it should be const (breaks Teuchos) | |
Static Protected Attributes | |
static boost::random::mt19937 | bootstrapRNG |
Random number generator to use for sampling. More... | |
Base class/interface for the bootstrap sampler.
BootstrapSamplerBase defines the minimum interface for a bootstrap sampler and handles initialization of the random variate generation used by the bootstrap. Functor is templated on the data type, but does not actually define a data member.
|
staticprotected |
Random number generator to use for sampling.
The boostrap random number generator.
Referenced by BootstrapSampler< Data >::operator()(), BootstrapSamplerWithGS< Data, Getter, Setter >::operator()(), and BootstrapSampler< Teuchos::SerialDenseMatrix< OrdinalType, ScalarType > >::operator()().