6.2 - Identifying Observations

In the previous section, we learned how to suppress the observation column that appears in default reports. Alternatively, we can use the ID statement to replace the observation column with one or more variables.

Example 6.4 Section

Using the ID statement, we can emphasize one or more key variables. The ID statement, which automatically suppresses the printing of the observation number, tells SAS to print the variable(s) specified in the ID statement as the first column(s) of your output. Thus, the ID statement allows you to use the values of the variables to identify observations, rather than the (usually meaningless) observation number. The following SAS code illustrates the use of the ID statement option:

PROC PRINT data = basic;
   id name;
   var gender expense;
RUN;

name	gender	expense
Alice Smith	1	1001.98
Maryann White	1	2999.34
Thomas Jones	2	3904.89
Benedict Arnold	2	1450.23
Felicia Ho	1	1209.94
John Smith	2	1763.09
Jane Smiley	1	3567.00

Launch and run the SAS program, and review the output to convince yourself that name, having replaced the observation number, appears in the first column.

Example 6.5 Section

It is particularly useful to use the ID statement when observations are so long that SAS can't print them on one line. In that case, SAS breaks up the observations and prints them on two (or more lines). When that happens, it is helpful to use an ID variable (or more) so that you can keep track of the observations. The following SAS code illustrates such a situation:

OPTIONS LS = 64 PS = 58 NODATE;

PROC PRINT data = basic;
   id name;
   var subj name clinic gender 
       subj no_vis type_vis expense;
RUN;

name	subj	name	clinic	gender
Alice Smith	1024	Alice Smith	LEWN	1
Maryann White	1167	Maryann White	LEWN	1
Thomas Jones	1168	Thomas Jones	ALTO	2
Benedict Arnold	1201	Benedict Arnold	ALTO	2
Felicia Ho	1302	Felicia Ho	MNMC	1
John Smith	1471	John Smith	MNMC	2
Jane Smiley	1980	Jane Smiley	MNMC	1

name	subj	no_vis	type_vis	expense
Alice Smith	1024	7	101	1001.98
Maryann White	1167	2	101	2999.34
Thomas Jones	1168	10	190	3904.89
Benedict Arnold	1201	1	190	1450.23
Felicia Ho	1302	7	190	1209.94
John Smith	1471	6	187	1763.09
Jane Smiley	1980	5	190	3567.00

Let's take note of a couple of things about the code. Given that we are working with a small data set with just seven variables, I used the OPTIONS statement to intentionally shrink the line size to 64 so that SAS is unable to fit the requested variables on one line. Also note that the variable name appears not only in the ID statement but also in the VAR statement. When that happens, SAS displays the variable twice in the output. Finally, note that subj is included twice in the VAR statement again only to make sure that the requested variables do not fit on one line.

Launch and run the SAS program, and review the output to convince yourself that SAS behaves as described.