The HPDMDB Procedure

The HPDMDB procedure summarizes data. The following example uses the Sampsio.Hmeq data set, which includes information about 5,960 fictitious mortgages. Each case represents an applicant for a home equity loan, and all applicants have an existing mortgage. The binary target variable BAD indicates whether an applicant eventually defaulted or was ever seriously delinquent. There are 10 numeric input variables and three classification input variables.

 proc hpdmdb data=Sampsio.Hmeq 
     classout=cout varout=vout;

     var loan derog mortdue value yoj delinq 
         clage ninq clno debtinc;
     class bad(desc) reason(ascending) job;
 run;

 proc print data=cout;run;
 proc print data=vout;run;

The data set cout (shown in Figure 5.1) contains the class summary table with levels sorted according to the sort option in the CLASS statement of PROC HPDMDB. You can see that the levels for BAD are in descending order and the levels for REASON are in ascending order. The levels for JOB are in the default ascending order.

Numeric summaries are in the data set vout, shown in Figure 5.2.