Inheritance diagram for sampler_stan_adapt_dense_e_nuts_MCMC:

Collaboration diagram for sampler_stan_adapt_dense_e_nuts_MCMC:

Public Member Functions
	sampler_stan_adapt_dense_e_nuts_MCMC (int seed, likelihood &L, std::vector< std::string > var_names, size_t dimension)

void	set_initial_location (std::vector< double > initial_parameters)

void	run_sampler (int nsteps, int thin, int refresh, int verbosity=1)

void	set_adaptation_parameters (int num_adapt, bool save_adapt)
	sets the adaption stuff with the applicable window stuff for stan. More...

void	set_stepsize (double step_size)

void	set_max_depth (int max_depth)

void	set_delta (double delta)

void	set_gamma (double gamma)
	Sets the adaptation regularization scale which must be greater than zero [default is 0.05].

void	set_kappa (double kappa)

void	set_t0 (double t0)

void	set_window_parameters (unsigned int init_buffer, unsigned int term_buffer, unsigned int window)
	Sets the window parameters for STAN, these rarely need to be changed.

void	set_initial_inverse_metric (Eigen::MatrixXd inverse_metric)
	Sets the initial metric, default is the identity metric.

void	get_chain_state (std::vector< double > &parameters, std::vector< double > &sample_state)

void	set_chain_state (std::vector< double > model_values, std::vector< double > state_values)

void	write_checkpoint (std::ostream &)

void	read_checkpoint (std::string ckpt_file)
	Restarts the sampler from the state specified in the ckpt_file. More...

void	read_checkpoint (std::istream &ckpt_in)

virtual double	get_rand ()

virtual void	reset_rng_seed (int seed)

Public Member Functions inherited from sampler_MCMC_base
	sampler_MCMC_base (int seed, likelihood &L, std::vector< std::string > var_names, size_t dimension)

virtual void	set_output_stream (std::string chain_name, std::string state_name, std::string sampler_name, int output_precision=8)
	Sets the output names for the chains and sampler states. This must be implemented in any classes inheriting sampler_exploration_mcmc_base. More...

virtual std::vector< double >	find_best_fit (std::string chain_file, std::string state_file) const
	Finds the best fit within provided chain file and returns it. More...

virtual std::vector< double >	find_best_fit () const
	Finds the best fit from the chain and state file saved in set_output_stream. More...

void	set_checkpoint (int ckpt_stride, std::string ckpt_file)
	Sets the checkpoint/restart functionality. More...

virtual void	reset_sampler_step ()

void	reset_likelihood_sum ()

double	get_likelihood_sum ()

void	set_mpi_communicator (MPI_Comm comm)
	Sets the MPI_Comm to be used for the likelihood evaluation. More...

void	close_streams ()
	Helper function that closes the streams from a sampler.

Private Member Functions
void	write_state ()

void	write_chain_header ()

void	write_state_header ()

void	write_sampler_header ()

void	generate_transition (int step, int start, int finish, int nthin, int refresh, bool adaptation, bool save, int verbosity)

Private Attributes
themistan_model	_stan_model

boost::ecuyer1988	_stan_rng

stan::mcmc::adapt_dense_e_nuts< themistan_model, boost::ecuyer1988 >	_stan_sampler

unsigned int	_init_buffer

unsigned int	_term_buffer

unsigned int	_window

Eigen::MatrixXd	_inverse_metric

std::vector< double >	_initial

stan::mcmc::sample	_current_sample

stan::callbacks::stream_logger	_logger

bool	_adaptation

bool	_save_adaptation

int	_nadapt

int	_start_step
	restart flag.

bool	_restart

Additional Inherited Members
Protected Attributes inherited from sampler_MCMC_base
int	_seed

Ran2RNG	_rng

std::vector< std::string >	_var_names

size_t	_dimension

size_t	_step_count

likelihood *	_L

std::string	_chain_file

std::string	_state_file

std::string	_sampler_file

std::string	_ckpt_file

std::ofstream	_chain_out

std::ofstream	_state_out

std::ofstream	_sampler_out

int	_output_precision

int	_ckpt_stride

double	_sum_lklhd

MPI_Comm	_comm

int	_rank

Member Function Documentation

void read_checkpoint ( std::string ckpt_file )

inlinevirtual

Parameters

ckpt_file The file where the checkpoint information was saved.

Reimplemented from sampler_MCMC_base.

Here is the call graph for this function:

void set_adaptation_parameters	(	int	num_adapt,
		bool	save_adapt
	)

Parameters

num_warmup	The number of adaptation steps to take. Default is 1000.
save_adapt	Whether to save the adaptation steps.

Note: Be aggressive with adaptation. If HMC is poorly tuned the performance tanks.

void set_delta ( double delta )

Sets the target acceptance rate for nuts [default of 0.8 which is the theoretical optimal result in high dim]

void set_kappa ( double kappa )

Sets the adaptation relaxation scale which must be greater than zero because it should damp. [default is 0.75]

void set_max_depth ( int max_depth )

Sets the max tree depth (max number of leapfrog steps is 2^tree_depth) for the sampler [default 10]

void set_stepsize ( double step_size )

Set the initial step size for the hmc algorithm. [default is 0.1]

void set_t0 ( double t0 )

Sets the adaptation iteration offset, which must be greater than 0 [default is 10]

The documentation for this class was generated from the following files:

sampling/sampler_stan_adapt_dense_e_nuts_MCMC.h
sampling/sampler_stan_adapt_dense_e_nuts_MCMC.cpp

Public Member Functions

Private Member Functions

Private Attributes

Additional Inherited Members

Member Function Documentation