Generating new variables

[sks521]

Hi folks,

I am trying to do the following;

My data looks like below. For each ID there can be multiple Epi_number. I want to count a record having same ID and Start_date and Dis_date as one admission regardless of Epi_number. For example ID 02 has three Epi_number and I want to take these three rows as one admission. Can someone suggest what to do please?

TaData Admissions_qtr;
input ID Epi_number Start_date Dis_date;
datalines;
01 1 01Jan2012 03Jan2012
02 1 17Mar2014 19Mar2014
02 2 17Mar2014 19Mar2014
02 3 17Mar2014 19Mar2014
03 1 19May2015 25May2015
03 2 19May2015 25May2015
03 3 19May2015 25May2015
03 4 19May2015 25May2015
;
run;

[sks521]

Just to say, there are 77650 rows in my data set

[novinosrin]

Not sure, how to imagine your expected output and guess. Can you post the expected sample too plz for the posted input

[sks521]

That's exactly what I am not sure about but I am assuming it will look like this;

Also I would like to sum up all 'admissions' for each month so I am assuming that it should give the desired output

TaData Admissions_qtr;
input ID Epi_number Start_date Dis_date Admissions;
datalines;
01 1 01Jan2012 03Jan2012 1
02 1 17Mar2014 19Mar2014 1
02 2 17Mar2014 19Mar2014 0
02 3 17Mar2014 19Mar2014 0
03 1 19May2015 25May2015 1
03 2 19May2015 25May2015 0
03 3 19May2015 25May2015 0
03 4 19May2015 25May2015 0
;
run;

[sks521]

'Admissions' is the new variable I have created here (assumption!)

[sks521]

And the challenge is some IDs have episode number up to 10.

[novinosrin · Accepted solution]

Doctor isn't sure what to order 🙂 ? and 

For what it's worth, i am posting this

Data Admissions_qtr;
infile cards expandtabs truncover; 
input ID $ Epi_number (Start_date Dis_date) (:date9.);
format Start_date Dis_date date9.;
datalines;
01 1 01Jan2012 03Jan2012
02 1 17Mar2014 19Mar2014
02 2 17Mar2014 19Mar2014
02 3 17Mar2014 19Mar2014
03 1 19May2015 25May2015
03 2 19May2015 25May2015
03 3 19May2015 25May2015
03 4 19May2015 25May2015
;
run;

data want;
 set Admissions_qtr;
 by id Start_date Dis_date;
 group_var=first.Dis_date;
run;