hmc_nuts_dense_e.hpp
Go to the documentation of this file.
1 #ifndef STAN_SERVICES_SAMPLE_HMC_NUTS_DENSE_E_HPP
2 #define STAN_SERVICES_SAMPLE_HMC_NUTS_DENSE_E_HPP
3 
4 #include <stan/callbacks/interrupt.hpp>
5 #include <stan/callbacks/logger.hpp>
6 #include <stan/callbacks/writer.hpp>
7 #include <stan/math/prim/mat/fun/Eigen.hpp>
8 #include <stan/math/prim/mat.hpp>
9 #include <stan/mcmc/hmc/nuts/dense_e_nuts.hpp>
10 #include <stan/services/error_codes.hpp>
11 #include <stan/services/util/run_sampler.hpp>
12 #include <stan/services/util/create_rng.hpp>
13 #include <stan/services/util/initialize.hpp>
14 #include <stan/services/util/inv_metric.hpp>
15 #include <vector>
16 
17 namespace stan {
18  namespace services {
19  namespace sample {
20 
21  /**
22  * Runs HMC with NUTS without adaptation using dense Euclidean metric
23  * with a pre-specified Euclidean metric.
24  *
25  * @tparam Model Model class
26  * @param[in] model Input model to test (with data already instantiated)
27  * @param[in] init var context for initialization
28  * @param[in] init_inv_metric var context exposing an initial dense
29  inverse Euclidean metric (must be positive definite)
30  * @param[in] random_seed random seed for the random number generator
31  * @param[in] chain chain id to advance the pseudo random number generator
32  * @param[in] init_radius radius to initialize
33  * @param[in] num_warmup Number of warmup samples
34  * @param[in] num_samples Number of samples
35  * @param[in] num_thin Number to thin the samples
36  * @param[in] save_warmup Indicates whether to save the warmup iterations
37  * @param[in] refresh Controls the output
38  * @param[in] stepsize initial stepsize for discrete evolution
39  * @param[in] stepsize_jitter uniform random jitter of stepsize
40  * @param[in] max_depth Maximum tree depth
41  * @param[in,out] interrupt Callback for interrupts
42  * @param[in,out] logger Logger for messages
43  * @param[in,out] init_writer Writer callback for unconstrained inits
44  * @param[in,out] sample_writer Writer for draws
45  * @param[in,out] diagnostic_writer Writer for diagnostic information
46  * @return error_codes::OK if successful
47  */
48  template <class Model>
49  int hmc_nuts_dense_e(Model& model, stan::io::var_context& init,
50  stan::io::var_context& init_inv_metric,
51  unsigned int random_seed, unsigned int chain,
52  double init_radius, int num_warmup, int num_samples,
53  int num_thin, bool save_warmup, int refresh,
54  double stepsize, double stepsize_jitter,
55  int max_depth,
56  callbacks::interrupt& interrupt,
57  callbacks::logger& logger,
58  callbacks::writer& init_writer,
59  callbacks::writer& sample_writer,
60  callbacks::writer& diagnostic_writer) {
61  boost::ecuyer1988 rng = util::create_rng(random_seed, chain);
62 
63  std::vector<int> disc_vector;
64  std::vector<double> cont_vector
65  = util::initialize(model, init, rng, init_radius, true,
66  logger, init_writer);
67 
68  Eigen::MatrixXd inv_metric;
69  try {
70  inv_metric =
71  util::read_dense_inv_metric(init_inv_metric, model.num_params_r(),
72  logger);
73  util::validate_dense_inv_metric(inv_metric, logger);
74  } catch (const std::domain_error& e) {
75  return error_codes::CONFIG;
76  }
77 
78  stan::mcmc::dense_e_nuts<Model, boost::ecuyer1988>
79  sampler(model, rng);
80 
81  sampler.set_metric(inv_metric);
82 
83  sampler.set_nominal_stepsize(stepsize);
84  sampler.set_stepsize_jitter(stepsize_jitter);
85  sampler.set_max_depth(max_depth);
86 
87  util::run_sampler(sampler, model, cont_vector, num_warmup, num_samples,
88  num_thin, refresh, save_warmup, rng, interrupt,
89  logger,
90  sample_writer, diagnostic_writer);
91  return error_codes::OK;
92  }
93 
94  /**
95  * Runs HMC with NUTS without adaptation using dense Euclidean metric,
96  * with identity matrix as initial inv_metric.
97  *
98  * @tparam Model Model class
99  * @param[in] model Input model to test (with data already instantiated)
100  * @param[in] init var context for initialization
101  * @param[in] random_seed random seed for the random number generator
102  * @param[in] chain chain id to advance the pseudo random number generator
103  * @param[in] init_radius radius to initialize
104  * @param[in] num_warmup Number of warmup samples
105  * @param[in] num_samples Number of samples
106  * @param[in] num_thin Number to thin the samples
107  * @param[in] save_warmup Indicates whether to save the warmup iterations
108  * @param[in] refresh Controls the output
109  * @param[in] stepsize initial stepsize for discrete evolution
110  * @param[in] stepsize_jitter uniform random jitter of stepsize
111  * @param[in] max_depth Maximum tree depth
112  * @param[in,out] interrupt Callback for interrupts
113  * @param[in,out] logger Logger for messages
114  * @param[in,out] init_writer Writer callback for unconstrained inits
115  * @param[in,out] sample_writer Writer for draws
116  * @param[in,out] diagnostic_writer Writer for diagnostic information
117  * @return error_codes::OK if successful
118  *
119  */
120  template <class Model>
121  int hmc_nuts_dense_e(Model& model, stan::io::var_context& init,
122  unsigned int random_seed, unsigned int chain,
123  double init_radius, int num_warmup, int num_samples,
124  int num_thin, bool save_warmup, int refresh,
125  double stepsize, double stepsize_jitter,
126  int max_depth,
127  callbacks::interrupt& interrupt,
128  callbacks::logger& logger,
129  callbacks::writer& init_writer,
130  callbacks::writer& sample_writer,
131  callbacks::writer& diagnostic_writer) {
132  stan::io::dump dmp =
133  util::create_unit_e_dense_inv_metric(model.num_params_r());
134  stan::io::var_context& unit_e_metric = dmp;
135 
136  return hmc_nuts_dense_e(model, init, unit_e_metric,
137  random_seed, chain, init_radius, num_warmup,
138  num_samples, num_thin, save_warmup, refresh,
139  stepsize, stepsize_jitter, max_depth,
140  interrupt, logger,
141  init_writer, sample_writer, diagnostic_writer);
142  }
143 
144  }
145  }
146 }
147 #endif
stan::services::sample::num_warmup
Definition: defaults.hpp:51
stan::services::sample::num_samples
Definition: defaults.hpp:14
stan::mcmc::base_hmc< Model, dense_e_metric, expl_leapfrog, BaseRNG >::set_nominal_stepsize
virtual void set_nominal_stepsize(double e)
Definition: base_hmc.hpp:152
stan::services::util::validate_dense_inv_metric
void validate_dense_inv_metric(const Eigen::MatrixXd &inv_metric, callbacks::logger &logger)
Definition: validate_dense_inv_metric.hpp:20
stan::services::util::create_unit_e_dense_inv_metric
stan::io::dump create_unit_e_dense_inv_metric(size_t num_params)
Definition: create_unit_e_dense_inv_metric.hpp:21
stan::services::sample::save_warmup
Definition: defaults.hpp:88
dense_e_nuts.hpp
Model
rosenbrock_model_namespace::rosenbrock_model Model
Definition: bfgs_minimizer_test.cpp:6
error_codes.hpp
stan::mcmc::base_hmc< Model, dense_e_metric, expl_leapfrog, BaseRNG >::set_stepsize_jitter
virtual void set_stepsize_jitter(double j)
Definition: base_hmc.hpp:165
stan::services::sample::hmc_nuts_dense_e
int hmc_nuts_dense_e(Model &model, stan::io::var_context &init, stan::io::var_context &init_inv_metric, unsigned int random_seed, unsigned int chain, double init_radius, int num_warmup, int num_samples, int num_thin, bool save_warmup, int refresh, double stepsize, double stepsize_jitter, int max_depth, callbacks::interrupt &interrupt, callbacks::logger &logger, callbacks::writer &init_writer, callbacks::writer &sample_writer, callbacks::writer &diagnostic_writer)
Definition: hmc_nuts_dense_e.hpp:49
stan::services::error_codes::CONFIG
Definition: error_codes.hpp:16
stan::services::util::read_dense_inv_metric
Eigen::MatrixXd read_dense_inv_metric(stan::io::var_context &init_context, size_t num_params, callbacks::logger &logger)
Definition: read_dense_inv_metric.hpp:27
stan::io::var_context
Definition: var_context.hpp:29
stan::mcmc::base_nuts< Model, dense_e_metric, expl_leapfrog, BaseRNG >::set_metric
void set_metric(const Eigen::MatrixXd &inv_e_metric)
Definition: base_nuts.hpp:54
productionTest.chain
chain
Check that an output directory exists.
Definition: productionTest.py:60
stan::services::sample::max_depth
Definition: defaults.hpp:414
inv_metric.hpp
stan::services::util::initialize
std::vector< double > initialize(Model &model, stan::io::var_context &init, RNG &rng, double init_radius, bool print_timing, stan::callbacks::logger &logger, stan::callbacks::writer &init_writer)
Definition: initialize.hpp:68
stan::math::domain_error
void domain_error(const char *function, const char *name, const T &y, const char *msg1, const char *msg2)
Definition: domain_error.hpp:29
stan::mcmc::base_nuts< Model, dense_e_metric, expl_leapfrog, BaseRNG >::set_max_depth
void set_max_depth(int d)
Definition: base_nuts.hpp:62
stan::services::sample::stepsize_jitter
Definition: defaults.hpp:482
stan::mcmc::dense_e_nuts
Definition: dense_e_nuts.hpp:16
stan::services::util::run_sampler
void run_sampler(stan::mcmc::base_mcmc &sampler, Model &model, std::vector< double > &cont_vector, int num_warmup, int num_samples, int num_thin, int refresh, bool save_warmup, RNG &rng, callbacks::interrupt &interrupt, callbacks::logger &logger, callbacks::writer &sample_writer, callbacks::writer &diagnostic_writer)
Definition: run_sampler.hpp:36
e
Float_t e
Definition: plot.C:35
create_rng.hpp
stan::services::error_codes::OK
Definition: error_codes.hpp:11
model
const XML_Char XML_Content * model
Definition: expat.h:151
stan::services::util::create_rng
boost::ecuyer1988 create_rng(unsigned int seed, unsigned int chain)
Definition: create_rng.hpp:25
refresh
void refresh()
Definition: show_event.C:21
run_sampler.hpp
Generated on Mon Mar 1 2021 14:15:25 for NOvA by   doxygen 1.8.11