Skip to contents

Predictions from the structural regression only: the copula terms never enter predict() or fitted() because they are endogeneity controls, not part of the causal model being predicted.

Usage

# S3 method for class 'copregbayes'
predict(object, newdata = NULL, ...)

Arguments

object

An object of class "copregbayes", as returned by CopRegBAYES.

newdata

An optional data.frame of new predictor values. If NULL (the default), the fitted values of the original data are returned.

...

Not used; present for S3 method consistency.

Value

A numeric vector of predictions, one per row of newdata (or of the original data if newdata is NULL), computed at the posterior mean coefficients.

References

Haschka, R. E. (2025). Bayesian inference for joint estimation models using copulas to handle endogenous regressors. Oxford Bulletin of Economics and Statistics. doi:10.1111/obes.70023

Examples

set.seed(1)
n <- 60
x <- rnorm(n); z <- x + rnorm(n); y <- 1 + z + x + rnorm(n)
dat <- data.frame(y = y, z = z, x = x)
fit <- CopRegBAYES(y ~ z | x, data = dat,
                    iterations = 200, burnin = 50, thin = 5, verbose = FALSE)
predict(fit, newdata = dat[1:5, ])
#> [1] 3.7202734 1.1811879 0.2712549 4.3285814 0.2603980