search

timcdlucas / INLAutils

R package providing utilities for INLA
Other
26 stars 10 forks source link

posterior p values are wrong #63

Open timcdlucas opened 1 year ago

timcdlucas commented 1 year ago 
x <- rnorm(100)
y <- x + rnorm(100, sd = 0.2)
d <- data.frame(x, y)

result1 = inla(y ~ x, family="normal", data = d, 
               control.predictor = list(compute = TRUE, link = 1),
               control.compute = list(return.marginals.predictor=TRUE))

plot_inla_residuals(result1, y)

looks horrible even with correct model

timcdlucas commented 1 year ago

Maybe I should just switch to PIT values.