Replicating observation and compiling mean

[infinity255]

Hi-

There's a longitudinal data set I'm working with which I need to compare baseline weight to mean weight of patients from third week of study to end of study/time of censoring.

Each patient has different number of record for weight in the dataset (i.e. patientgp will be the groping variable.)

I thought about creating a variable "baselinewt" and replicating baseline weight within each patient (by "patientgp") so that I can compare mean weight of each patient to "baselinewt" after removing first two weeks. To do this, I would appreciate help replicating observations and compling mean within a group.

If there's easier way to solve this problem, I would welcome that as well. Thanks.