Prev: Creating new variables
Next: Estimation of marginal willingness to pay (MWTP) using discrete
From: William Whitworth on 3 Mar 2010 14:29
I have data in the following long format (4 obs per subject, with automan
and accession as unique identifiers):

data long;
input Subj_ID $ AUTOMAN $ ACCESSION $ TESTRESULT;

datalines;
24C-004 auto C07-1021 10
24C-004 manual C07-1021 22
24C-004 auto C07-1022 13
24C-004 manual C07-1022 20
24C-006 auto C07-1033 17
24C-006 manual C07-1033 25
24C-006 auto C07-1034 18
24C-006 manual C07-1034 11
;;;;
Proc Print; run;

I'd like to get the data into the following structure:
(The lowest accession is for the first two tests, the higher is for the
second two tests).

Subj_ID AUTO1 AUTO2 MANUAL1 MANUAL2
24C-004 10 13 22 20
24C-006 17 18 25 11

I've tried a few things (datastep array, proc transpose, SQL) but nothing
seems to work.
Can anybody offer any suggestions? Thanks!
 | 
Pages: 1
Prev: Creating new variables
Next: Estimation of marginal willingness to pay (MWTP) using discrete