Count changes in column

PatrykSAS — Thu, 17 Jun 2021 09:19:22 GMT

Hello

I would like to mark value changes in a column like below:

ID	VALUE	CHANGE
1	10	0
1	10	0
1	12	1
1	12	0
1	15	1

In this example ID and VALUE is my data and I want to make CHANGE column equal 1 each time the value change in given ID

I've tried this code but it aplies 0 each row. Maybe lag function could be helpful but I'm not sure how to use it

data want;
set have;
by id value notsorted;
count=first.value;
if first.id then count=0;
run;

Can you help me with solution?

Re: Count changes in column

Kurt_Bremser — Thu, 17 Jun 2021 09:35:52 GMT

ID	VALUE	change
1	10	0
1	10	0
1	12	1
1	12	0
1	15	1

Works for me:

data have;
input ID VALUE;
datalines;
1 10
1 10
1 12
1 12
1 15
;

data want;
set have;
by id value notsorted;
change = first.value;
if first.id then change = 0;
run;

proc print data=want noobs;
run;

Result:

ID	VALUE	change
1	10	0
1	10	0
1	12	1
1	12	0
1	15	1

topic Count changes in column in SAS Programming

Count changes in column

Re: Count changes in column