StochTree::OrdinalSampler Class Reference

Sampler for ordinal model hyperparameters. More...

#include <ordinal_sampler.h>

Public Member Functions
void	UpdateLatentVariables (ForestDataset &dataset, Eigen::VectorXd &outcome, std::mt19937 &gen)
	Update truncated exponential latent variables (Z)
void	UpdateGammaParams (ForestDataset &dataset, Eigen::VectorXd &outcome, double alpha_gamma, double beta_gamma, double gamma_0, std::mt19937 &gen)
	Update gamma cutpoint parameters.
void	UpdateCumulativeExpSums (ForestDataset &dataset)
	Update cumulative exponential sums (seg)

Static Public Member Functions
static double	SampleTruncatedExponential (std::mt19937 &gen, double rate, double low=0.0, double high=1.0)
	Sample from truncated exponential distribution.

Detailed Description

Sampler for ordinal model hyperparameters.

This class handles MCMC sampling for ordinal-specific parameters:

• Truncated exponential latent variables (Z)
• Cutpoint parameters (gamma)
• Cumulative sum of exp(gamma) (seg) [derived parameter]

Member Function Documentation

SampleTruncatedExponential()

static double StochTree::OrdinalSampler::SampleTruncatedExponential ( std::mt19937 &  gen, double  rate, double  low = 0.0, double  high = 1.0  )

static

Sample from truncated exponential distribution.

Samples from exponential distribution truncated to [low,high]

Parameters

gen	Random number generator
rate	Rate parameter for exponential distribution
low	Lower truncation bound
high	Upper truncation bound

Returns

Sampled value from truncated exponential

UpdateLatentVariables()

void StochTree::OrdinalSampler::UpdateLatentVariables	(	ForestDataset &	dataset,
		Eigen::VectorXd &	outcome,
		std::mt19937 &	gen
	)

Update truncated exponential latent variables (Z)

Parameters

dataset	Forest dataset containing training data (covariates) and auxiliary data needed for sampling
outcome	Vector of outcome values
gen	Random number generator

UpdateGammaParams()

void StochTree::OrdinalSampler::UpdateGammaParams	(	ForestDataset &	dataset,
		Eigen::VectorXd &	outcome,
		double	alpha_gamma,
		double	beta_gamma,
		double	gamma_0,
		std::mt19937 &	gen
	)

Update gamma cutpoint parameters.

Parameters

dataset	Forest dataset containing training data (covariates) and auxiliary data needed for sampling
outcome	Vector of outcome values
alpha_gamma	Shape parameter for log-gamma prior on cutpoints gamma
beta_gamma	Rate parameter for log-gamma prior on cutpoints gamma
gamma_0	Fixed value for first cutpoint parameter (for identifiability)
gen	Random number generator

UpdateCumulativeExpSums()

void StochTree::OrdinalSampler::UpdateCumulativeExpSums

(

ForestDataset &

dataset

)

Update cumulative exponential sums (seg)

Parameters

dataset

Forest dataset containing training data (covariates) and auxiliary data needed for sampling

---

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

• include/stochtree/ordinal_sampler.h