Trivial accessors for a fitted "copreg" object, following the usual
lm()-style generics. coef() returns the coefficients of the
augmented regression (structural regressors and copula terms together),
vcov() their bootstrap covariance matrix, nobs() the number
of observations used, and formula() the two- or three-part
Formula the model was fitted with.
Arguments
- object, x
a fitted model of class
"copreg", as returned by one of the cross-sectional estimator functions (e.g.CopRegPG(),CopReg2sCOPE()).- ...
currently unused; present for method consistency.
Value
coef returns a named numeric vector of the augmented
regression coefficients (the structural regressors' coefficients
followed by the copula terms'). vcov returns the corresponding
bootstrap covariance matrix, with matching row and column names.
nobs returns the number of observations used in the fit, as a
single integer. formula returns the two- or three-part
Formula object (endogenous | exogenous | discrete) the model was
fitted with.
References
Qian, Y., A. Koschmann, and H. Xie (2025). A practical guide to endogeneity correction using copulas. Journal of Marketing.
Examples
set.seed(1)
n <- 150
w <- rnorm(n)
p <- 0.4 * w + rt(n, df = 3)
xi <- 0.5 * p + rnorm(n)
y <- 1 + 2 * p + 1.5 * w + xi
dat <- data.frame(y = y, p = p, w = w)
fit <- endogCopula:::.copreg_fit(
formula = y ~ p | w, data = dat,
ctor = endogCopula:::.ctor_twostage(TRUE),
method = "2sCOPE", cdf = "rank.n", ties = "max",
nboots = 25, verbose = FALSE)
coef(fit)
#> (Intercept) p w p_cop
#> 0.8776032 2.8387045 1.4293344 -0.5092000
vcov(fit)
#> (Intercept) p w p_cop
#> (Intercept) 0.0083106040 -0.0002842064 0.0009440729 0.003448208
#> p -0.0002842064 0.0517789378 -0.0121223049 -0.079621463
#> w 0.0009440729 -0.0121223049 0.0134089215 0.017620101
#> p_cop 0.0034482078 -0.0796214629 0.0176201005 0.132615426
nobs(fit)
#> [1] 150
formula(fit)
#> y ~ p | w
#> <environment: 0x55e6efbc3640>