sampleSLGP: Draw posterior predictive samples from a SLGP model

In SLGP: Spatial Logistic Gaussian Process for Field Density Estimation

Draw posterior predictive samples from a SLGP model

Description

Samples from the predictive distributions modeled by a SLGP at new covariate inputs. This method uses inverse transform sampling on the estimated posterior CDFs.

Usage

sampleSLGP(
  SLGPmodel,
  newX,
  n,
  interpolateBasisFun = "WNN",
  nIntegral = 101,
  nDiscret = 101,
  seed = NULL
)

Arguments

SLGPmodel	A trained SLGP model object (SLGP-class).
newX	A data frame of new covariate values at which to draw samples.
n	Integer or integer vector specifying how many samples to draw at each input point.
interpolateBasisFun	Character string specifying interpolation scheme for basis evaluation. One of "nothing", "NN", or "WNN" (default).
nIntegral	Integer; number of quadrature points for density approximation.
nDiscret	Integer; discretization step for the response axis.
seed	Optional integer to set a random seed for reproducibility.

Value

A data frame containing sampled responses from the SLGP model, with covariate columns from newX and one response column named after SLGPmodel@responseName.

Examples

# Load Boston housing dataset
library(MASS)
data("Boston")
# Set input and output ranges manually (you can also use range(Boston$age), etc.)
range_x <- c(0, 100)
range_response <- c(0, 50)

# Train an SLGP model using Laplace estimation and RFF basis
modelMAP <- slgp(medv ~ age,        # Use a formula to specify response and covariates
                 data = Boston,     # Use the original Boston housing data
                 method = "MAP",    # Train using Maximum A Posteriori estimation
                 basisFunctionsUsed = "RFF",         # Random Fourier Features
                 sigmaEstimationMethod = "heuristic",  # Auto-tune sigma2 (more stable)
                 predictorsLower = range_x[1],         # Lower bound for 'age'
                 predictorsUpper = range_x[2],         # Upper bound for 'age'
                 responseRange = range_response,       # Range for 'medv'
                 opts_BasisFun = list(nFreq = 200,     # Use 200 Fourier features
                                      MatParam = 5/2), # Matern 5/2 kernel
                 seed = 1)                             # Reproducibility

# Let's draw new sample points from the SLGP

newDataPoints <- sampleSLGP(modelMAP,
                            newX = data.frame(age=c(0, 25, 95)),
                            n = c(10, 1000, 1), # how many samples to draw at each new x
                            interpolateBasisFun = "WNN")

SLGP documentation built on Sept. 9, 2025, 5:25 p.m.