error message

Hello Nonmem Users, We are getting an error message: 0PROGRAM TERMINATED BY FNLETA ERROR IN CELS WITH INDIVIDUAL 2 ID= 1.10050000000000E+01 0VAR-COV OF SOME OBS ESTIMATED TO BE 0 MESSAGE ISSUED FROM TABLE STEP The code runs fine, but does not create output tables. What can it be? Thanks, Pavel

Pavel, This indicates that the empirical Bayes (EBE aka 'posthoc') estimates of the parameters for ID=2 do not make sensible predictions of the residual error because the residual error is predicted to be zero. 0VAR-COV OF SOME OBS ESTIMATED TO BE 0 A common cause of this kind of error is to have a proportional residual error with a model prediction of zero e.g. Y=F*(1+EPS(1)) will fail if F=0 You may be able to workaround the problem by adding a small additive component to the model. Y=F*(1+EPS(1)) + EPS(2) You don't have to estimate SIGMA(2) but just fix it to something small. This may then allow the table step to complete and you can check the EBEs and perhaps understand why the prediction is causing an error. Nick