Issue while retaining the value

[Ishaan]

I have 2 dataset x and y. X is my master dataset and has id and date and Y is quarterly dataset having balance and bal.

Y table is getting created every Mar, Jun, Sep, and Dec while X is refreshed Monthly.

I am retaining the value populated in Mar to Apr and May. Once the data available in jun. I am retaining the Jun data to Jul and Aug. Till we get the data in next qtr. If any Id doesn't get the Balance and Bal in next qtr I am trying to keep the missing value for that Id.

Currently previous qtrs non missing value is populated in my final table where i want missing value.

Table X
id	date
1	01-12-2017
2	01-12-2017
3	01-12-2017
4	01-12-2017
5	01-12-2017
6	01-12-2017
7	01-12-2017
1	01-01-2018
2	01-01-2018
3	01-01-2018
4	01-01-2018
5	01-01-2018
6	01-01-2018
7	01-01-2018
1	01-02-2018
2	01-02-2018
3	01-02-2018
4	01-02-2018
5	01-02-2018
6	01-02-2018
7	01-02-2018
1	01-03-2018
2	01-03-2018
3	01-03-2018
4	01-03-2018
5	01-03-2018
6	01-03-2018
7	01-03-2018

Table Y
Id	date	Balance	Bal
1	01-12-2017	200	A
2	01-12-2017	500	B
3	01-12-2017	700	C
4	01-12-2017	800	C
1	01-03-2018	200	A
2	01-03-2018	500	B
5	01-03-2018	700	C
6	01-03-2018	800	C

Final table(Expected)
id	date	balance	bal
1	01-12-2017	200	A
2	01-12-2017	500	B
3	01-12-2017	700	C
4	01-12-2017	800	C
5	01-12-2017
6	01-12-2017
7	01-12-2017
1	01-01-2018	200	A
2	01-01-2018	500	B
3	01-01-2018	700	C
4	01-01-2018	800	C
5	01-01-2018
6	01-01-2018
7	01-01-2018
1	01-02-2018	200	A
2	01-02-2018	500	B
3	01-02-2018	700	C
4	01-02-2018	800	C
5	01-02-2018
6	01-02-2018
7	01-02-2018
1	01-03-2018	200	A
2	01-03-2018	500	B
3	01-03-2018
4	01-03-2018
5	01-03-2018	700	C
6	01-03-2018	800	C
7	01-03-2018

Code: 

proc sql;
create table final1 as select a.*, b.balance, b.bal
from x a left join y b
on a.id=b.id
and a.date=b.date;
quit;

proc sort data=final1;
by id date;
run;

data final;
set final1;
retain balance1 bal2;
by id date;
if first.id then do;
balance1=balance;
bal2=bal
end;
if month(date) in (3,6,9,12) then do;
balance1=balance;
if bal ne '' then bal2=bal;
end;
drop bal balance;
run;

[novinosrin]

Hi @Ishaan   Can you please check your table Y  dates  that has

1	01-03-2017	200	A
2	01-03-2017	500	B
5	01-03-2017	700	C
6	01-03-2017	800	C

as opposed to the final

1	01-03-2018	200	A
2	01-03-2018	500	B
3	01-03-2018
4	01-03-2018
5	01-03-2018	700	C
6	01-03-2018	800	C

Assuming I understand , my attempt:

data x;
input id	date :ddmmyy10.;
format date ddmmyy10.;
cards;
1	01-12-2017
2	01-12-2017
3	01-12-2017
4	01-12-2017
5	01-12-2017
6	01-12-2017
7	01-12-2017
1	01-01-2018
2	01-01-2018
3	01-01-2018
4	01-01-2018
5	01-01-2018
6	01-01-2018
7	01-01-2018
1	01-02-2018
2	01-02-2018
3	01-02-2018
4	01-02-2018
5	01-02-2018
6	01-02-2018
7	01-02-2018
1	01-03-2018
2	01-03-2018
3	01-03-2018
4	01-03-2018
5	01-03-2018
6	01-03-2018
7	01-03-2018
;


data y;
infile cards truncover;
input Id	date :ddmmyy10.	Balance	Bal $;
format date ddmmyy10.;
cards;
1	01-12-2017	200	A
2	01-12-2017	500	B
3	01-12-2017	700	C
4	01-12-2017	800	C
1	01-03-2017	200	A
2	01-03-2017	500	B
5	01-03-2017	700	C
6	01-03-2017	800
;

data want;
if _n_=1 then do;
if 0 then set x y(rename=date=_date);
  dcl hash H (dataset:'y(rename=date=_date)') ;
   h.definekey  ("id","_date") ;
   h.definedata ("_date","balance", "bal") ;
   h.definedone () ;
end;
set x;
rc=h.find(key:id,key:date);
if rc ne 0 then do;
call missing(balance, bal);
if intck('month',_Date,date)<=2 then rc1=h.find();
end;
drop rc: _:;
run;

[novinosrin]

So assuming your Y table is in sequence 

data y;
infile cards truncover;
input Id	date :ddmmyy10.	Balance	Bal $;
format date ddmmyy10.;
cards;
1	01-12-2017	200	A
2	01-12-2017	500	B
3	01-12-2017	700	C
4	01-12-2017	800	C
1	01-03-2018	200	A
2	01-03-2018	500	B
5	01-03-2018	700	C
6	01-03-2018	800  C
;

where March,2018 would follow Dec2017 i.e of Dec2017--Mar2018--Jun2018 and beyond

The above code will match your expected result. 

[ChrisNZ]

This does what you want though I am unsure why you'd want the 2018 entries to take the 2017 data.

proc sql; 
  select a.ID, b.DATE, Y.BALANCE, Y.BAL      
  from (select unique ID   from X) a
         cross join 
       (select unique DATE from X) b
         left join 
       Y
         on Y.ID=a.ID and intck('qtr.3',Y.DATE,b.DATE) in(0,4)  
  order 2,1;
quit;

ID	DATE	BALANCE	BAL
1	01/12/2017	200	A
2	01/12/2017	500	B
3	01/12/2017	700	C
4	01/12/2017	800	C
5	01/12/2017	.
6	01/12/2017	.
7	01/12/2017	.
1	01/01/2018	200	A
2	01/01/2018	500	B
3	01/01/2018	700	C
4	01/01/2018	800	C
5	01/01/2018	.
6	01/01/2018	.
7	01/01/2018	.
1	01/02/2018	200	A
2	01/02/2018	500	B
3	01/02/2018	700	C
4	01/02/2018	800	C
5	01/02/2018	.
6	01/02/2018	.
7	01/02/2018	.
1	01/03/2018	200	A
2	01/03/2018	500	B
3	01/03/2018	.
4	01/03/2018	.
5	01/03/2018	700	C
6	01/03/2018	800	C
7	01/03/2018	.