BookmarkSubscribeRSS Feed
mahim
Calcite | Level 5
I am trying to:
1. Sample repeated measures data based on subject IDs for cross validation of a GENMOD model,
2. Bootstrap a dataset because of a highly skewed outcome measure recorded in an unbalanced long format (e.g a patient may may have 5 observations for t1, 0 for t2, 3 for t3, and so on, where 't's are time points).
 
Is there a well explained macro/guide available for resampling or bootstrapping repeated measures data based on subject IDs? I want to resample by ID so that all observations per subject are captured by the procedure in the train/test sets. Please do correct me if I am wrong; I believe simple random sampling by replacement, which may be done using proc surveyselect, would not be an efficient method for repeated measures data. How can I modify proc surveyselect to incorporate subject IDs?
2 REPLIES 2
PGStats
Opal | Level 21

@Rick_SAS has posted many great entries on Bootstrap. Look here

 

https://blogs.sas.com/content/?s=bootstrap

 

To make the subject ID the (re-)sampling unit, go for cluster sampling, where the subject is the cluster.

PG
fbaumer
Calcite | Level 5

Hi - I have this exact same question. I've been looking through the recommended link and have not found a clear answer in the documentation. Did you ever find a way to do this? I have done a year of masters' level statistics so do rely on very specific instructions if I try one of these models (rather than extrapolating from related things like proc mixed).

 

Thank you!

SAS Innovate 2025: Register Now

Registration is now open for SAS Innovate 2025 , our biggest and most exciting global event of the year! Join us in Orlando, FL, May 6-9.
Sign up by Dec. 31 to get the 2024 rate of just $495.
Register now!

Mastering the WHERE Clause in PROC SQL

SAS' Charu Shankar shares her PROC SQL expertise by showing you how to master the WHERE clause using real winter weather data.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 2 replies
  • 2088 views
  • 2 likes
  • 3 in conversation