Sometimes, rather than just combining data from two data sets, you might want to update the data in one data set with data that is stored in another data set. That is, you might want to update a master data set by overwriting certain values with values that are stored in a transaction data set.

For example, suppose the data set Mylib.Empmaster contains data that is outdated. The current data is stored in another data set named Mylib.Empchanges. Mylib.Empmaster contains 148 observations, and Mylib.Empchanges contains six observations. The variable EmpID contains unique values in both data sets.

A partial listing of Mylib.Empmaster and the full listing of Mylib.Empchanges is shown below. Notice that there is one observation in each data set with a value of 1065 for EmpID. The values of JobCode and Salary are different in these observations.

If you could see the full listing of Mylib.Empmaster, you would see that each of the observations in Mylib.Empchanges has a matching observation in Mylib.Empmaster based on the values of EmpID. There are also many observations in Mylib.Empmaster that do not have a matching observation in Mylib.Empchanges. To update Mylib.Empmaster, you want to find all of the matching observations and change their values for JobCode and Salary to the new values from Mylib.Empchanges. You can use the UPDATE statement to make these changes.

You use the UPDATE statement to update a master data set with a transaction data set. The UPDATE statement can perform the following tasks:

The UPDATE statement replaces values in the master data set with values from the transaction data set for each observation with a matching value of the BY variable. Any observations in either the master data set or the transactional data set that have nonmatching values for the BY variable are included in the output data set. Also, by default, SAS does not replace existing values in the master data set with missing values if those values are coded as periods (for numeric variables) or blanks (for character variables) in the transaction data set.

When you use the UPDATE statement, keep in mind the following restrictions.

Remember that you want to update the master data set Mylib.Empmaster with the transactional data set Mylib.Empchanges. You can use the UPDATE statement to accomplish this task, as shown in the program below. Remember, both data sets must be sorted by or indexed on the BY variable.

proc sort data=mylib.empmaster;
   by empid;
run;

proc sort data=mylib.empchanges;
   by empid;
run;

data mylib.empmaster;
   update mylib.empmaster mylib.empchanges;
   by empid;
run;

The first eight observations of the updated Mylib.Empmaster data set are shown below. Notice that the observation that has a value of 1065 for EmpID now contains the updated values for JobCode and Salary.

proc print data=mylib.empmaster (obs=8) noobs;
run;