topic Re: How to subset the longitudinal data into two separate datasets? in SAS Procedures

How to subset the longitudinal data into two separate datasets?

mantubiradar19 — Wed, 15 Jun 2016 04:21:28 GMT

I have a longituidnal data (Called as SBP & DBP which are contineous variables)- which is collected on different dates and the date variable looks like 10SEP2010. I have sample ID for each patient.

I would like to classify first measurement of SBP, DBP of each observation into one dataset and last measurement of SBP, DBP of each observation into another dataset.

How can I do this?

Re: How to subset the longitudinal data into two separate datasets?

zzone — Wed, 15 Jun 2016 05:18:55 GMT

Browse source dataset by patient_ID and sampling_date. Output first record of a group to one dataset and the last to another. something like code below.

data destination_dataset1 destination_dataset2;

set source_dataset;

by patient_id sampling_date;

if first.patient_id then output destination_dataset1;

if last.patient_id then output destination_dataset2;

run;

Re: How to subset the longitudinal data into two separate datasets?

mantubiradar19 — Wed, 15 Jun 2016 05:26:04 GMT

Thank you very much Zzone! Let me try this!

Re: How to subset the longitudinal data into two separate datasets?

mantubiradar19 — Thu, 16 Jun 2016 06:57:30 GMT

I'm trying to computete the difference between first and last observations. Can someone tell me why I have the following note?

Thanks

Re: How to subset the longitudinal data into two separate datasets?

zzone — Thu, 16 Jun 2016 07:54:45 GMT

"first" / "last" refer to a subgroup which means ordering. Your code does not browse the data set in an ordered manner (you should have a "by ... " clause in it.

I'd suggest you should read a little bit the basics to understand the mechanisms behind:

http://support.sas.com/documentation/cdl/en/basess/58133/HTML/default/viewer.htm#a001302699.htm

Re: How to subset the longitudinal data into two separate datasets?

mantubiradar19 — Thu, 16 Jun 2016 08:32:28 GMT

Thanks for the reference!