28 virtual void execute()
override;
Sampler & _sampler
The adaptive Monte Carlo sampler.
std::vector< Real > _prev_val_out
Storage for previously accepted output value.
virtual void finalize() override
A class used to perform Parallel Subset Simulation Sampling.
const bool _gp_used
Check if a GP is used.
A class used to perform Adaptive Importance Sampling using a Markov Chain Monte Carlo algorithm...
AdaptiveMonteCarloDecision(const InputParameters ¶meters)
const InputParameters & parameters() const
virtual void execute() override
void reinitChain()
This reinitializes the Markov chain to the starting value until the Gaussian process training is comp...
std::vector< std::vector< Real > > _inputs_sto
Storage for the previously accepted sample inputs across all the subsets.
std::vector< std::vector< Real > > _inputs_sorted
Store the sorted input samples according to their corresponding outputs.
std::vector< std::vector< Real > > _prev_val
Storage for previously accepted input values. This helps in making decision on the next proposed inpu...
const int *const _gp_training_samples
Store the GP training samples.
libMesh::Parallel::Communicator & _local_comm
Communicator that was split based on samples that have rows.
const std::vector< Real > & _output_value
Model output value from SubApp.
const ParallelSubsetSimulation *const _pss
Parallel Subset Simulation sampler.
std::vector< Real > _output_sorted
Store the sorted output sample values.
virtual void initialize() override
AdaptiveMonteCarloDecision will help make sample accept/reject decisions in adaptive Monte Carlo sche...
int _check_step
Ensure that the MCMC algorithm proceeds in a sequential fashion.
static InputParameters validParams()
std::vector< Real > _outputs_sto
Storage for previously accepted sample outputs across all the subsets.
DIE A HORRIBLE DEATH HERE typedef LIBMESH_DEFAULT_SCALAR_TYPE Real
std::vector< Real > & _output_required
Modified value of model output by this reporter class.
const AdaptiveImportanceSampler *const _ais
Adaptive Importance Sampler.
std::vector< std::vector< Real > > & _inputs
Model input data that is uncertain.
Real _output_limit
Store the intermediate ouput failure thresholds.