Standard deviation of quarterly sales over the preceding 16 quarters

[joon1]

Hi Madam/Sir,

I want to construct the standard deviation of quarterly sales over the the preceding 16 quarters using the attached file ("songdata"). 

GVKEY	DATADATE	FYEARQ	FQTR	saleq
1004	19850228	1984	3	50.939
1004	19850531	1984	4	60.128
1004	19850831	1985	1	53.836
1004	19851130	1985	2	61.04
1004	19860228	1985	3	64.541
1004	19860531	1985	4	68.595
1004	19860831	1986	1	66.376
1004	19861130	1986	2	75.907
1004	19870228	1986	3	71.83
1004	19870531	1986	4	84.079
1004	19870831	1987	1	74.543
1004	19871130	1987	2	81.31
1004	19880229	1987	3	92.951
1004	19880531	1987	4	100.858
1004	19880831	1988	1	90.007
1004	19881130	1988	2	100.645

I try to find the relevant code, but it is hard to revise this code. Any help will be highly appreciated.

data a2; set a1;
by gvkey;
array t {16} _temporary_;
if first.gvkey then call missing(of t{*});
do fy=sum(lag(fyearq),1) to fyearq-1;
t{1+mod(fy,16)}=.;
stdsale=std(of t{*});
run;

Thank you

Joon1

[Reeza · Accepted solution]

data want;
array p{0:15} _temporary_;
set have; 
by gvkey;

if first.gvkey then call missing(of p{*});

p{mod(_n_,16)} = saleq;

stdsales = std( of p(*));

run;

For this approach to work you must have continuous time periods. If your time periods are not continuous and you're missing quarters this approach will not work. 

https://gist.github.com/statgeek/27e23c015eae7953eff2

[joon1]

Thank you so much. I greatly appreciate for your help.

Joon1