The parametric procedure that we are going to use to analyze the same business problem is PROC LIFEREG. This is a procedure that isn't as popular a procedure as LIFETEST or PHREG but still is an important one. As we discussed earlier, the Cox methodology cannot handle left and interval censored data. We also highlighted that the time-dependent covariates cannot be handled by PROC PHREG. In such cases, we will end up with multiple rows per subject when the covariates change values across time. If the survival function is known, then PROC LIFEREG can produce more accurate estimates. Let's try and explore the procedure using variations of the distribution types.
Here is the code for the PROC LIFEREG Weibull distribution: ...
