Raw Message

Message-ID: <200812291432.mBTEWIa17794@hsrnfs-101.mayo.edu>
Date: 2008-12-29T14:32:17Z
From: Terry Therneau
Subject: cox regression warning/error messages

1. I haven't seen that particular message before, so it is difficult to comment.

2. To go further, you will need to provide a reproducable example, i.e., some 
code/data that I could run to get the same error.

3. One note
 cox.NV <- coxph(Surv(recidivism$intDaysUntilFNVPO, 
	recidivism$Event_nv) ~ recidivism$strGender + 
	recidivism$intAgeAtMHCIndex + recidivism$PRE.nv + group + 
	recidivism$MHC.nv + recidivism$SNFP , data = recidivism)

would be much simpler to type as
 cox.NV <- coxph(Surv(intDaysUntilFNVPO, Event_nv) ~ strGender + 
	                intAgeAtMHCIndex + PRE.nv + group + MHC.nv + SNFP , 
	             data = recidivism)

(That is one purpose of the data= argument, to make it simpler.)

	Terry Therneau