Dear All:
I have a data set that I want to analyse with Coxph that consist in ordered
repeated events (same type of events) from individuals that are clustered in
several units (farms). Therefore, I have 2 types of heterogeneities: one
intra-subject -related to repeated measurements and other due to lack of
independence of individuals from different clusters.
I structured the data in the counting process form described in the book of
Therneau-Grambsh and for the one type of heterogeneity I apply one of the
models that I was interested.
Then, I try to add the frailty function to the previous with the aim to add the
second type of heterogeneity, but I get error message that the dimmensions of a
given variable has less colums that it should have.
Any clue what could happens or what could be done?. Or simply I do it wrongly?
I have s-plus 2000 on windows.
Thanks in advance;
Best regards;
Gustavo Monti
|