Reading Repeating Blocks of Data (1)

Creating Multiple Observations from a Single Record

Reading Repeating Blocks of Data

Each record in the file Tempdata contains three blocks of data. Each block contains a date followed by the day's high temperature in a small city located in the southern United States.

Raw Data File Tempdata

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

01APR90 68 02APR90 67 03APR90 70

04APR90 74 05APR90 72 06APR90 73

07APR90 71 08APR90 75 09APR90 76

10APR90 78 11APR90 70 12APR90 69

13APR90 71 14APR90 72 15APR90 74

16APR90 73 17APR90 71 18APR90 75

19APR90 75 20APR90 73 21APR90 75

22APR90 77 23APR90 78 24APR90 80

25APR90 78 26APR90 77 27APR90 79

28APR90 81 29APR90 81 30APR90 84

You could write a DATA step that reads each record and creates three different Date and Temp variables.

SAS Data Set

Date1 Temp1 Date2 Temp2 Date3 Temp3

11048 68 11049 67 11050 70

Date1	Temp1	Date2	Temp2	Date3	Temp3
11048	68	11049	67	11050	70

But if you create a separate observation for each block of data in a record, you can later use several statistical procedures to analyze the data for each day.

SAS Data Set

Date HighTemp

11048 68

11049 67

11050 70

Date	HighTemp
11048	68
11049	67
11050	70