Data manipulation using proq sql in sas - sas

Question

I have the below table
enter image description here
I calculate the std deviation of y and z till from 2001 q4 to 2003 q2 .
I have to create a new table which should look like below
enter image description here
I tried using the case statement inside proc sql, but it did not work. Any assistance will be appreciated.
proc sql;
create table tablenew as
select
Date,
X,
case when Date >= "2003Q3" then (y+std(y)) else y end as y,
case when Date >= "2003Q3" then (z-std(z)) else z end as z
from have
;
quit;
But here, the std deviation of the entire column is calculated, i only want the standard deviation of columns y and z , from 2001 q3 to 2003 q2

Answer 1

Only using SQL is just going to make this problem more difficult than it needs to be, but the issue is you need to wrap the value you send to the STD() function in a CASE statement also. Here I have saved those trimmed values as their own variables so you can see what is happening.
create table tablenew as
select
Date
,X
,case when date >= '2003Q3' then y else . end as y_subset
,case when date >= '2003Q3' then z else . end as z_subset
,case when Date >= "2003Q3" then (y+std(calculated y_subset)) else y end as y
,case when Date >= "2003Q3" then (z-std(calculated z_subset)) else z end as z
from have
;
Much easier to break it down into logical steps and just use normal SAS code instead of SQL.
proc summary data=have ;
where date >= '2003Q3' ;
var y z ;
output out=std std=y_std z_std ;
run;
data want ;
set have ;
if _n_=1 then set std ;
if date >= '2003Q3' then do;
y=y+y_std;
z=z-z_std;
end;
run;