RETAIN Statement in SAS (with Examples)

This tutorial explains how to use the RETAIN statement in SAS, with examples. In SAS, it's a very easy and useful way to retain values with RETAIN statement.

Create Sample Data

The following program creates sample data for demonstration -

data abcd;
input x y;
cards;
1 25
1 28
1 27
2 23
2 35
2 34
3 25
3 29
;
run;

Uses of RETAIN Statement

The RETAIN statement simply copies retaining values by telling the SAS not to reset the variables to missing at the beginning of each iteration of the DATA step. If you would not use retain statement then SAS would return missing at the beginning of each iteration.

The retain statement keeps the value once assigned.

Generate Serial Number

Suppose you need to generate a serial number (or row index number) with data step.

data aaa;
set abcd;
retain z 0;
z = z + 1;
run;

Output Dataset

Output Data Set

The above SAS code initializes a variable "z" to 0 and increments it by 1 for each observation in the "aaa" data set. The result is a new data set with an additional variable "z" that has row numbers.

We can retain implicitly by using the +1 notation.

data aaa;
set abcd;
z + 1;
run;

z + 1; is a simplified way to increment the variable "z" by 1 for each observation. It returns row numbers starting from 1.

Cumulative Score

Suppose you need to calculate cumulative score. In financial data, we generally need to calculate cumulative score year to date.

data aaa;
set abcd;
retain z 0;
z = z + y;
run;

Output Data Set

Generate Serial Number by Group

Suppose you have a grouping variable say "region" and you need to generate a row index number by region.

proc sort data = abcd;
by x;
run;

data aaa;
set abcd;
retain z;
if first.x then z = 1;
else z = z + 1;
by x;
run;

SAS : Retain Statement

Cumulative Score by Group

Suppose you need to calculate cumulative sale by product categories.

data aaa1;
set aaa;
retain z1;
if first.x then z1 = y;
else z1 = z1 + y;
by x;
run;

SAS : Cumulative Score

The variable "z1" constitutes cumulative values of variable y by grouping variable x.

Number of Unique Observations

The number of unique rows by a group can easily be calculated with PROC FREQ and PROC MEANS. The following program explains how we can calculate number of observations in a categorical variable with Data Step.

data aaa2;
set abcd (drop = y);
retain z;
if first.x then z = 1;
else z = z + 1;
by x;
if last.x then output;
run;

Unique Count

Suppose you have more than 1 grouping variable

In the dataset below, we have two grouping (categorical) variables : "ID" and "ID1".

data temp;
input ID ID1 Score;
cards;
1 1 25
1 1 26
1 2 27
1 2 29
2 1 28
2 1 29
2 2 31
;
run;

When you have more than 1 grouping variable, we can use multiple FIRST. statements with OR operator to generate serial numbers.

data temp2;
set temp;
by ID ID1;
if first.ID or first.ID1 then N = 1;
else N+1;
proc print;
run;

Related Posts

SAS Tutorials : Top 100 SAS Tutorials

Spread the Word!
Share Share Tweet

About Author:

Deepanshu founded ListenData with a simple objective - Make analytics easy to understand and follow. He has over 10 years of experience in data science. During his tenure, he worked with global clients in various domains like Banking, Insurance, Private Equity, Telecom and HR.

While I love having friends who agree, I only learn from those who don't
Let's Get Connected Email LinkedIn