Null Value Processing Modes

The processing behavior of a null value depends on the mode, which is determined by how you connect to the DataFlux Federation Server.
  • A client application that connects to the DataFlux Federation Server with a client-side driver processes data by default by using ANSI SQL null value behavior.
  • When you use the SAS LIBNAME engine for DataFlux Federation Server, null values are processed by default with SAS missing value behavior.
In some ways a SAS missing value is analogous to an ANSI SQL null value; however, the processing behavior can be different. Therefore, if an application is processing data with SAS missing value behavior rather than ANSI SQL null behavior, then you need to be aware of processing differences.
  • You can sort a SAS missing value and evaluate it with standard comparison operators.
  • You cannot sort an ANSI SQL null value or evaluate it with standard comparison operators, because there is no data on which to operate. For example, to test for a null value, you cannot use arithmetic comparison operators such as = or <.
  • SAS missing values in a SAS data set are translated to ANSI SQL null values when the data is copied to a data source that processes in the ANSI SQL null mode (for example, an Oracle database).
  • Many relational databases, such as Oracle and DB2, implement ANSI SQL null values. Therefore, the concept of ANSI SQL null values with the DataFlux Federation Server languages is the same as with the Oracle SQL language.
Note: Because the SAS data set does not physically store null indicators, the DataFlux Federation Server languages emulate ANSI SQL null values for the data source.