Solving Operational Business Intelligence with InfoSphere Warehouse Advanced Edition

Chapter 10. Techniques for data mining in an operational warehouse 387

2. Publish Phase

Use the SAS server software to publish the scoring model to InfoSphere

Warehouse (DB2)

3. Populate Phase

Register the SAS scoring model with the SAS embedded process in DB2,

making it available natively to InfoSphere Warehouse applications.

Using the new embedded process allows for high performance execution of the

scoring model in an operational warehouse environment with InfoSphere

Warehouse 10.1.

10.3.3 Access the scoring model

The SAS embedded process was developed and installed by SAS, and runs as a

fenced process in DB2 (db2sasep). The process is started and stopped using

new db2ida DDL commands or can be configured to start and stop automatically

with the DB2 instance. Examples of starting and stopping manually are shown in

Example 10-1.

Example 10-1 Starting and stopping the SAS embedded process in DB2

db2ida -provider sas -start

db2ida -provider sas -stop

db2ida -provider sas -stopforce

After the process is running, scoring models can be invoked using the

ANALYZE_TABLE function embedded in a SQL expression as shown in

Example 10-2.

Example 10-2 Invoking the SAS scoring model

SELECT * FROM T1 ANALYZE_TABLE (IMPLEMENTATION ‘PROVIDER=SAS;

ROUTINE_SOURCE_TABLE=USER1.SOURCE_TABLE;

ROUTINE_SOURCE_NAME=REGRESSION;’);

Note the following:

򐂰 Scoring input (ROUTINE_SOURCE_TABLE) can be a table, view, or

common table expression.

򐂰 The ROUTINE_SOURCE_NAME refers to the name of the model, which is

REGRESSION in this example.