Dear Prof Harrell,
I've noticed that the HRs from a loglog orm do not fall within their corresponding CIs (kindly refer to the minimal example below). Thank you for looking into this and for creating the rms package!
Yonghao
## create a datadist object
mtcars.dd <- datadist(mtcars)
options(datadist="mtcars.dd")
mod.loglog <- orm (mpg ~ drat + vs , family =loglog, data = mtcars) # loglog CPM
## HR ratios are not within their corresponding CIs
## HRs for the predictors are probably exp(beta)
summary (mod.loglog)
Dear Prof Harrell, I've noticed that the HRs from a loglog orm do not fall within their corresponding CIs (kindly refer to the minimal example below). Thank you for looking into this and for creating the rms package! Yonghao