Hi everyone,
Do i have to repeat my lasso code multiple times even if i put an rseed and average the AUC and C-slopes (i noticed that lasso is selecting different lambda sometimes even though i am running the same code which is resulting in slightly different AUC and C-slope)
My dataset consists of 1,409 participants out of which 240 have the outcome, i want to get the predictors, AUC and Calibration slope after running lasso (cv).
My concern is if i should rerun this code multiple times and pool the results
the code i am using is:
lasso logit olump age gender sem1 lbs pa2 chab02 fh3 i.(motherbmi), selection(cv) rseed(1234)
Thank you
Do i have to repeat my lasso code multiple times even if i put an rseed and average the AUC and C-slopes (i noticed that lasso is selecting different lambda sometimes even though i am running the same code which is resulting in slightly different AUC and C-slope)
My dataset consists of 1,409 participants out of which 240 have the outcome, i want to get the predictors, AUC and Calibration slope after running lasso (cv).
My concern is if i should rerun this code multiple times and pool the results
the code i am using is:
lasso logit olump age gender sem1 lbs pa2 chab02 fh3 i.(motherbmi), selection(cv) rseed(1234)
Thank you
