BayesOpt
bayesopt::Parameters Class Reference
+ Collaboration diagram for bayesopt::Parameters:

Public Member Functions

 Parameters (bopt_params c_params)
 
bopt_params generate_bopt_params ()
 
void set_learning (std::string name)
 
std::string get_learning ()
 
void set_score (std::string name)
 
std::string get_score ()
 

Public Attributes

size_t n_iterations
 Maximum BayesOpt evaluations (budget)
 
size_t n_inner_iterations
 Maximum inner optimizer evaluations.
 
size_t n_init_samples
 Number of samples before optimization.
 
size_t n_iter_relearn
 Number of samples before relearn kernel.
 
size_t init_method
 Sampling method for initial set 1-LHS, 2-Sobol (if available), other value-uniformly distributed.
 
int random_seed
 >=0 -> Fixed seed, <0 -> Time based (variable). More...
 
int verbose_level
 Neg-Error,0-Warning,1-Info,2-Debug -> stdout 3-Error,4-Warning,5-Info,>5-Debug -> logfile.
 
std::string log_filename
 Log file path (if applicable)
 
size_t load_save_flag
 1-Load data,2-Save data, 3-Load and save data. More...
 
std::string load_filename
 Init data file path (if applicable)
 
std::string save_filename
 Sava data file path (if applicable)
 
std::string surr_name
 Name of the surrogate function.
 
double sigma_s
 Signal variance (if known). More...
 
double noise
 Variance of observation noise (and nugget)
 
double alpha
 Inverse Gamma prior for signal var. More...
 
double beta
 Inverse Gamma prior for signal var. More...
 
score_type sc_type
 Score type for kernel hyperparameters (ML,MAP,etc)
 
learning_type l_type
 Type of learning for the kernel params.
 
bool l_all
 Learn all hyperparameters or only kernel.
 
double epsilon
 For epsilon-greedy exploration.
 
size_t force_jump
 If >0, and the difference between two consecutive observations is pure noise, for n consecutive steps, force a random jump. More...
 
KernelParameters kernel
 Kernel parameters.
 
MeanParameters mean
 Mean (parametric function) parameters.
 
std::string crit_name
 Name of the criterion.
 
vectord crit_params
 Criterion hyperparameters (if needed)
 

Private Member Functions

void init_default ()
 
void bostrdup (char *d, const char *s)
 

Detailed Description

Definition at line 61 of file parameters.hpp.

Member Data Documentation

◆ alpha

double bayesopt::Parameters::alpha

Inverse Gamma prior for signal var.

Used in StudentTProcessNIG

Definition at line 90 of file parameters.hpp.

Referenced by bayesopt::BOptState::loadFromFile().

◆ beta

double bayesopt::Parameters::beta

Inverse Gamma prior for signal var.

Used in StudentTProcessNIG

Definition at line 92 of file parameters.hpp.

Referenced by bayesopt::BOptState::loadFromFile().

◆ force_jump

size_t bayesopt::Parameters::force_jump

If >0, and the difference between two consecutive observations is pure noise, for n consecutive steps, force a random jump.

Avoid getting stuck if model is bad and there is few data, however, it might reduce the accuracy.

Definition at line 100 of file parameters.hpp.

Referenced by bayesopt::BayesOptBase::stepOptimization().

◆ load_save_flag

size_t bayesopt::Parameters::load_save_flag

1-Load data,2-Save data, 3-Load and save data.

Definition at line 80 of file parameters.hpp.

Referenced by bayesopt::BOptState::loadFromFile(), bayesopt::BayesOptBase::optimize(), and bayesopt::BayesOptBase::saveInitialSamples().

◆ random_seed

int bayesopt::Parameters::random_seed

>=0 -> Fixed seed, <0 -> Time based (variable).

Definition at line 74 of file parameters.hpp.

Referenced by bayesopt::BayesOptBase::BayesOptBase(), and bayesopt::BOptState::loadFromFile().

◆ sigma_s

double bayesopt::Parameters::sigma_s

Signal variance (if known).

Used in GaussianProcess and GaussianProcessNormal

Definition at line 86 of file parameters.hpp.

Referenced by bayesopt::BOptState::loadFromFile().


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