The HPCDM Procedure(Experimental)

You can use the BY statement in the HPCDM procedure to process the input data set in groups of observations defined by the BY variables.

If you specify the BY statement, then PROC HPCDM expects the input data set to be sorted in the order of the BY variables unless you specify the NOTSORTED option.

The BY statement is always supported in the single-machine mode of execution. For the distributed mode, it is supported only when the DATA= data set resides on the client machine. In other words, the BY statement is supported only in the client-data (or local-data) mode of the distributed computing model and not for any of the alongside modes, such as the alongside-the-database or alongside HDFS mode.

If you specify the COUNTSTORE= option, then the BY group processing is not supported.