This is a c++ implementation of Gibbs sampling SUR model with power prior
sur_sample_gibbs_cpp(
Sigma,
M,
X,
X0,
XtX,
X0tX0,
Y,
Y0,
y,
y0,
a0,
pvec,
burnin,
thin
)
initial value for covariance matrix
number of samples
design matrix for current data
design matrix for historical data
matrix that is crossprod(cbind(X1, ..., XJ))
matrix that is crossprod(cbind(X01, ..., X0J))
future response as matrix (Y1, ..., YJ)
historical response as matrix (Y01, ..., Y0J)
future response as vector
historical response as vector
power prior parameter
vector
giving number of covariates per endpoint
Burn-in parameter
Thin parameter
sampled covariance matrix