How to have sum in each subsequent observation

[Ramya2]

Hi All,

Can you please help me to add up each observation value with its previous var value.

Suppose below is my data,

Sub      var1

001       12

002       5

003         .

004       3

005       .

006        6       

output1:

I want a results as below,

sub     var1      var2

001       12        12

002       5          17

003         .         17

004       3           20

005       .            20

006        6          26

output2:

I want a results as below,

sub     var1      var2

001       12        12

002       5          17

003         .         .

004       3           20

005       .            .

006        6          26

[Jagadishkatam · Accepted solution]

Please try

data have;
input Sub var1;
cards;
001 12
002 5
003   .
004 3
005 .
006  6 
;
 
data want1(keep=sub var1 var3 rename=(var3=var2)) want2(keep=sub var1 var2);
set have ;
retain var3;
if var1 ne . then var3=sum(var1,var3);
if var1 ne . then var2=var3;
run;

[yabwon]

Just for fun with shorter syntax:

data have;
input Sub var1;
cards;
001 12
002 5
003 .
004 3
005 .
006 6
;
 
data 
  want1(drop = var2 rename=(var3=var2)) 
  want2(drop = var3)
;
  set have ;
  var2 + var1;
  var3 = ifn(var1,var2,var1);
run;

All the best

Bart

[Ramya2]

Thank you so much.

[ed_sas_member]

Hi @Ramya2 

Here is another approach:

data want1;
	set have;
	var2 + var1;
run;

data want2;
	set have;
	_var + var1;
	if not missing (var1) then var2 = _var;
	drop _var;
run;

Best,

[sustagens]

Try

data output1(drop=var2 rename=(cum_var1=var2)) output2(drop=cum_var1);
  set have;
  cum_var1+var1;
  if var1 eq . then var2=.;
  else var2=cum_var1;
run;