bayesian-qr

Bayesian quantile regression project completed for STAT 525 at Rice University. The final report is included as report.pdf. Additionally, we have included an associated slideshow presentation in slides.pdf. This report specifically discusses standard Bayesian quantile regression and longitudinal Bayesian quantile regression. We have also included code that implements the methods discussed in the report.

Repository Contents

• OBQR_syntheticdata.R: Ordinary Bayesian quantile regression applied to synthetic data set. This code generates Figures 1-4 and 9-12.
• OBQR_realdata.R: Ordinary Bayesian quantile regression applied to real data set. This code generates Figures 5.
• BQRcode.R: Ordinary Bayesian quantile regression applied on a sythentic data generated from a linear mixed-effects model (uses the bayesQR R package).
• BQRGScode.R: Bayesian quantile regression Gibbs sampler for data generated from a linear mixed-effects model with 2 fixed effects and 2 random effects.
• BQRMHcode.R: Bayesian quantile regression Metropolis-Hastings sampler for data generated from a linear mixed-effects model with 2 fixed effects and 2 random effects.