Previous examples of DATA steps used DO loops to generate one or more observations from one iteration of the DATA step. It is also possible to write a DATA step that reads a data set to compute the value of a new variable.

The SAS data set Finance.CDRates contains interest rates for certificates of deposit (CDs) that are available from several institutions.

Suppose you want to compare how much each CD earns at maturity with an investment of $5,000. The DATA step below creates a new data set, Work.Compare, that contains the added variable, Investment.

data work.compare(drop=i); 
   set finance.cdrates; 
   Investment=5000; 
   do i=1 to years; 
      investment+rate*investment; 
   end; 
run;
proc print data=work.compare noobs;
run;

The index variable is used only to execute the DO loop, so it is dropped from the new data set. Notice that the data set variable Years is used as the stop value in the iterative DO statement. As a result, the DO loop executes the number of times specified by the current value of Years.

Here is what happens during each iteration of the DATA step:

At the bottom of the DATA step, the first observation is written to the Work.Compare data set. Control returns to the top of the DATA step, and the next observation is read from Finance.CDRates. These steps are repeated for each observation in Finance.CDRates. The resulting data set contains the computed values of Investment for all observations that have been read from Finance.CDRates.