are sex and drug essential in modeling primary biliary cirrhosis data?

We analyze primary clinical biliary cirrhosis sequential (pbcseq) data in an r-package from a datascience angle in this talk. the problem under focus is when commonly, practitioners register the sexand drug variables; and benet from the descriptive analysis. the question that arises here is whetherwe need these critical variables in modeling. by regression modeling, we talk about the pros and cons.we conduct a penalized estimation for the longitudinal data and eciently select important variablesin the multivariate regression modeling. we found drugs and sex can be eliminated from the pbcseqanalysis when both serum bilirubin and serum albumin are measured as responses, and over time thedisease progresses. this study shows the critical and signicant role of statistics in data science forcorrect decision-making.