Displaying 1 result from an estimated 1 matches for "reml_sigma_squared".
2013 Apr 20
0
Calculate confidence intervals in mgcv for unconditional on the, smoothing parameters
...de(it`s a bit lengthy and has
to be tidied up but I wanted to to some manual calculations on the
objects afterwards)
###
# Simulate 100000 coefficient vectors from the posterior for ß and
calculate the critical value
###
cov_beta_diff <- vcov(gam_fit)
basis_matrix <- model.matrix(gam_fit)
reml_sigma_squared <- gam.vcomp(gam_fit)
sigma_2_epsilon <- reml_sigma_squared[2,1]^2
smooth_matrix <-
basis_matrix%*%vcov(gam_fit,dispersion=1)%*%t(basis_matrix)
stdv_f_x <- sqrt(diag(sigma_2_epsilon * smooth_matrix))
set.seed(123)
R_sim <- 100000
mean_val <- rep(0, nrow(cov_beta_diff))
coef_diff...