Reading the Same Number of Repeating Fields (1)

Creating Multiple Observations from a Single Record

Reading the Same Number of Repeating Fields

So far, you have created multiple observations from a single record by executing the DATA step once for each block of data in a record.

Now, look at another file that is organized differently.

Each record in the file Data97 contains a sales representative's ID number, followed by four repeating fields that represent his or her quarterly sales totals for 1997.

You want to pair each employee ID number with one quarterly sales total to produce a single observation. That way, four observations can be derived from one record.

Raw Data File Data97

1---+----10---+----20---+----30---+----40

0734 1,323.34 2,472.85 3,276.65 5,345.52

0943 1,908.34 2,560.38 3,472.09 5,290.86

1009 2,934.12 3,308.41 4,176.18 7,581.81

1043 1,295.38 5,980.28 8,876.84 6,345.94

1190 2,189.84 5,023.57 2,794.67 4,243.35

1382 3,456.34 2,065.83 3,139.08 6,503.49

1734 2,345.83 3,423.32 1,034.43 1,942.28

ID Quarter Sales

0734          1 1323.34

0734          2 2472.85

0734          3 3276.65

0734          4 5345.52

0943 1 1908.34

0943 2 2560.38

0943 3 3472.09

0943 4 5290.86

ID	Quarter	Sales
0734	1	1323.34
0734	2	2472.85
0734	3	3276.65
0734	4	5345.52
0943	1	1908.34
0943	2	2560.38
0943	3	3472.09
0943	4	5290.86

To accomplish this, you must execute the DATA step once for each record, repetitively reading and writing values in one iteration.

This means that a DATA step must

read the value for ID and hold the current record
create a new variable named Quarter to identify the fiscal quarter for each sales figure
read a new value for Sales and write the values to the data set as an observation
continue reading a new value for Sales and writing values to the data set three more times.