Hi all,
I would like to shrink a flexible parametric survival model that I have generated using the stpm2 package using a boostrap-estimated shrinkage coefficient. Could someone guide me as to the process required? I would ideally like to save the shrunk model into a .ster file. I have achieved this for a logistic model by multiplying the coefficients by the shrinkage coefficient and then recalculating the intercept but there are a few more variables in a stpm2 model like the restricted cubic splines of the baseline hazard and their derivatives that I am unsure how to account for.
We can use the following code as an example, where gender is a binary variable and age is a continuous variable:
Many thanks in advance,
Markos
I would like to shrink a flexible parametric survival model that I have generated using the stpm2 package using a boostrap-estimated shrinkage coefficient. Could someone guide me as to the process required? I would ideally like to save the shrunk model into a .ster file. I have achieved this for a logistic model by multiplying the coefficients by the shrinkage coefficient and then recalculating the intercept but there are a few more variables in a stpm2 model like the restricted cubic splines of the baseline hazard and their derivatives that I am unsure how to account for.
We can use the following code as an example, where gender is a binary variable and age is a continuous variable:
Code:
stpm2 age i.gender, scale(h) df(2)
Markos
