CODEWITHSUNDEEP
rcox-regression
E.D.D
E.D.D

Reputation: 11

Can you use the cluster function twice in a cox regression in R?

So I have a cox query - so I am trying to do a cox regression and I wonder if I can cluster the data twice:

cox1<-coxph(formula = Surv(time.to.arrive..trap, Arrived.1.or.0) ~     
            Is.feeder.control.or.stimulus + cluster(id.location) + cluster(New.ID), 
            data = all.data)
cox1

I am trying to work out if there is a difference of time to arrive at the control or stimulus feeder - however I want to take into account that birds can appear multiple times in my data - (at control and stimulus, either or both at each of the five locations). Does it make sense to use the cluster function twice? In the above formula I clustered it by ID of the bird and by location.Id.

This may seem like an obvious question - but has been causing me no end of trouble! Thank you very much indeed in advance for any light you can shed on the matter!!

Kind regards

Upvotes: 1

Views: 546

Answers (1)

Whitebeard
Whitebeard

Reputation: 6203

I don't know the specific answer to your question regarding clustering two variables within coxph, but perhaps you can create an interaction factor between id.location and New.ID and then cluster that variable (and calling cluster only once)

For example, 

all.data$newvar <- interaction(all.data$id.location, all.data$New.ID)
cox1<-coxph(formula = Surv(time.to.arrive..trap, Arrived.1.or.0) ~     
        Is.feeder.control.or.stimulus + cluster(newvar), data = all.data)

Upvotes: 0

Related Questions