This repository provides R code implementing Gibbs sampler for Bayesian spatial quantile smoothing, as proposed by the following paper.

Onizuka, T., Hashimoto. S. and Sugasawa, S. (2022), Locally Adaptive Spatial Quantile Smoothing: Application to Monitoring Crime Density in Tokyo. arXiv:2202.09534.

The repository includes the following files.

• BSQS-function.R: Implementation of Gibbs sampling for Bayesian spatial quantile smoothing methos: BQTF, SAR, and GP.

• BSQS-example-plot.R: One-shot example of fitting Bayesian spatial quantile smoothing.

• TrueSignal-function.R: The true signal functions such as two block structure and exponential function.

• NoiseDistribution-function.R: The data-generating functions such as (I) Homogeneous, (II) Block heterogeneous, and (III) Smooth heterogeneous.