Flag First Instance

[mikemangini]

Hi, I have a data set that tracks unique person identifier, and number of activities they had on a website and call center per YearMonth. An example below, I have the first 4 variables. What I would like to do is flag the YearMonth where the FIRST event occurred and populate the variable with 1 and 0 otherwise. Also my actual data set has multiple years of data.

Have				Want
PersonID	YearMonth	Logons	Calls	FirstLogon	FirstCall
A	201701	0	5	0	1
A	201702	0	0	0	0
A	201703	0	0	0	0
A	201704	1	0	1	0
A	201705	5	1	0	0
A	201706	0	0	0	0
A	201707	0	0	0	0
A	201708	2	0	0	0
A	201709	1	0	0	0
A	201710	0	2	0	0
A	201711	2	0	0	0
A	201712	1	0	0	0

[novinosrin · Accepted solution]

data have;
input PersonID $	YearMonth Logons	Calls;
datalines;
A	201701	0	5
A	201702	0	0
A	201703	0	0
A	201704	1	0
A	201705	5	1
A	201706	0	0
A	201707	0	0
A	201708	2	0
A	201709	1	0
A	201710	0	2
A	201711	2	0
A	201712	1	0
;


data want;
set have;
by PersonID;
retain _temp1	_temp2  ;
if first.personid then do;_temp1=0;	_temp2=0;end;
if Logons and not(_temp1)   then do; _temp1=1;FirstLogon=1;end;else FirstLogon=0;
if calls and not(_temp2)   then do; _temp2=1;Firstcall=1;end;else firstcall=0;
drop _:;
run;

[mikemangini]

worked perfectly. Thank you for your help!

Mike