Combining several continuous longitudinal variables into one

[MoZes]

Hello Team,

Am in the process of running a multivariate model with several continuous longitudinal responses. Am envisaging a wide format data structure with a number of responses, but currently working with two responses (S and D) as follows;

ID	S1	S2	S3	S4	D1	D2	D3	D4
1	124	125	136	140	77	79	85	90
2	150	145	140	145	82	83	80	82
3	140	135	142	138	80	79	82	78

The 'S' and 'D' measurements are taken per individual at four time points (1, 2, 3, 4). The sample data set is from three individuals uniquely identified by 'ID'. Am able to restructure this to a long format  as follows;

ID	T	S	D
1	1	124	77
1	2	125	79
1	3	136	85
1	4	140	90
2	1	150	82
2	2	145	83
2	3	140	80
2	4	145	82
3	1	140	80
3	2	135	79
3	3	142	82
3	4	138	78

I now wish to combine the two sequences into one sequence (SDcomb) by a stack up procedure such that the resulting data set will appear as follows;

ID	T	SDcomb
1	1	124
1	1	77
1	2	125
1	2	79
1	3	136
1	3	85
1	4	140
1	4	90
2	1	150
2	1	82
2	2	145
2	2	83
2	3	140
2	3	80
2	4	145
2	4	82
3	1	140
3	1	80
3	2	135
3	2	79
3	3	142
3	3	82
3	4	138
3	4	78

How do I achieve this? Thank you.

Kind Regards,

MoZes

[Rick_SAS]

data have;
input ID	T	S	D;
datalines;
1	1	124	77
1	2	125	79
1	3	136	85
1	4	140	90
2	1	150	82
2	2	145	83
2	3	140	80
2	4	145	82
3	1	140	80
3	2	135	79
3	3	142	82
3	4	138	78
;

data want;
set have;
SDComb = S; output;
SDComb = D; output;
drop S D;
run;

[MoZes · Accepted solution]

Hi Rick,

Thank you so much for helping me out. Have been cracking my head all day, finally sorted. Thanks Bro!

Regrads,

MoZes