Proc Reliability

[sasprogram]

Hi,
I have a problem with the RELIABILITY Procedure.
I could not estimate the parameters in a recurrent events data model when exact event times are known and we have interval-censored data. When I define the left and right censoring time in the format of MODEL (variable1 variable2) I receive the following error:

“You must specify the numbers of recurrences in each interval with a FREQ statement, and the number of units observed in each interval with an NENTER statement”

However, these two statements are used when the event exact recurrence times are not known, but they are known exactly in my case.

[SteveDenham]

What happens when you use the extrema for the interval based on the known times?  Does that help?  If not, you may get a more focused answer by posting the question to the optimization, discrete-simulation and OR community, as RELIABILITY is a procedure in SAS/QC which that group is focused on.

SteveDenham