how to use proc transpose transfer A to B?

[walterwang]

A is:

count	pct	year	y
32751	8.6	2012	0
35200	8.4	2013	1
34422	7.8	2014	2
36195	7.6	2015	3
39035	7.5	2016	4
43060	7.6	2017	5
360673	6.2	2012	0
357294	5.9	2013	1
341996	5.3	2014	2
341963	5.1	2015	3
351231	5.2	2016	4
366421	5.3	2017	5
25432	6.6	2012	0
25219	6.0	2013	1
23947	5.4	2014	2
25239	5.3	2015	3
27019	5.2	2016	4
29103	5.1	2017	5
274908	4.7	2012	0
257168	4.2	2013	1
235824	3.6	2014	2
234645	3.5	2015	3
237780	3.5	2016	4
247493	3.6	2017	5
7319	1.9	2012	0
9981	2.4	2013	1
10475	2.4	2014	2
10956	2.3	2015	3
12016	2.3	2016	4
13957	2.5	2017	5
85765	1.5	2012	0
100126	1.7	2013	1
106172	1.6	2014	2
107318	1.6	2015	3
113451	1.7	2016	4
118928	1.7	2017	5

B is :

32751	8.6	360673	6.2	25432	6.6	274908	4.7	7319	1.9	85765	1.5
35200	8.4	357294	5.9	25219	6.0	257168	4.2	9981	2.4	100126	1.7
34422	7.8	341996	5.3	23947	5.4	235824	3.6	10475	2.4	106172	1.6
36195	7.6	341963	5.1	25239	5.3	234645	3.5	10956	2.3	107318	1.6
39035	7.5	351231	5.2	27019	5.2	237780	3.5	12016	2.3	113451	1.7
43060	7.6	366421	5.3	29103	5.1	247493	3.6	13957	2.5	118928	1.7

[VDD]

what are you transposing on?

you have no column names in B.

[walterwang]

That is

count pct count pct count pct

[VDD · Accepted solution]

this is a nice link that explains proc transpose.

Example 3 should answer your needs.

https://stats.idre.ucla.edu/sas/modules/how-to-reshape-data-long-to-wide-using-proc-transpose/

reshaping data with 2 variables

[novinosrin]

See if this helps

data have;
input vcount	pct	year	y;
cards;
32751	8.6	2012	0
35200	8.4	2013	1
34422	7.8	2014	2
36195	7.6	2015	3
39035	7.5	2016	4
43060	7.6	2017	5
360673	6.2	2012	0
357294	5.9	2013	1
341996	5.3	2014	2
341963	5.1	2015	3
351231	5.2	2016	4
366421	5.3	2017	5
25432	6.6	2012	0
25219	6.0	2013	1
23947	5.4	2014	2
25239	5.3	2015	3
27019	5.2	2016	4
29103	5.1	2017	5
274908	4.7	2012	0
257168	4.2	2013	1
235824	3.6	2014	2
234645	3.5	2015	3
237780	3.5	2016	4
247493	3.6	2017	5
7319	1.9	2012	0
9981	2.4	2013	1
10475	2.4	2014	2
10956	2.3	2015	3
12016	2.3	2016	4
13957	2.5	2017	5
85765	1.5	2012	0
100126	1.7	2013	1
106172	1.6	2014	2
107318	1.6	2015	3
113451	1.7	2016	4
118928	1.7	2017	5
;
proc transpose data=have out=w;
by  year notsorted;
var vcount pct;
run;

proc sort data=w out=ww(keep=year col1);
by year;
run;

proc transpose data=ww out=want(drop=_name_);
by year;
var col1;
run;

[Tom]

Why?

[Tom]

That looks like a report instead of a dataset. 

Easy to do in PROC REPORT, but your source data has no grouping variable.

Looks like a new group starts whenever Y=0.

data have;
  input count pct year y @@;
  group+(y=0);
cards;
32751 8.6 2012 0 35200 8.4 2013 1 34422 7.8 2014 2 36195 7.6 2015 3 39035 7.5 2016 4 43060 7.6 2017 5
360673 6.2 2012 0 357294 5.9 2013 1 341996 5.3 2014 2 341963 5.1 2015 3 351231 5.2 2016 4 366421 5.3 2017 5
25432 6.6 2012 0 25219 6.0 2013 1 23947 5.4 2014 2 25239 5.3 2015 3 27019 5.2 2016 4 29103 5.1 2017 5
;

proc report data=have ;
 column group year,(count pct) ;
 define group / group ;
 define year / across ;
run;

[walterwang]

I need further process.

[Tom]

---

@walterwang wrote:

I need further process.

---

Then keep it in the form you have where it is usable.

[walterwang]

How to save the ourtput from tabulate as a data set? Any idea, thanks