Illustrate Bivariate Normal Gibbs Sampler
rbiNormGibbs implements a Gibbs Sampler for the bivariate normal distribution. Intermediate moves are plotted and the output is contrasted with the iid sampler. This function is designed for illustrative/teaching purposes.
rbiNormGibbs(initx=2, inity=-2, rho, burnin=100, R=500)
initx |
initial value of parameter on x axis (def: 2) |
inity |
initial value of parameter on y axis (def: -2) |
rho |
correlation for bivariate normals |
burnin |
burn-in number of draws (def: 100) |
R |
number of MCMC draws (def: 500) |
(θ_1, θ_2) ~ N( (0,0), Σ) with Σ = matrix(c(1,rho,rho,1),ncol=2)
R x 2 matrix of draws
Peter Rossi, Anderson School, UCLA, perossichi@gmail.com.
For further discussion, see Chapters 2 and 3, Bayesian Statistics and Marketing by Rossi, Allenby, and McCulloch.
http://www.perossi.org/home/bsm-1
## Not run: out=rbiNormGibbs(rho=0.95)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.