Creating a new data source

  1. To access the Data Source Editor, select Add a data source under either the Project Data Sources or My Data Sources area.
  2. Type a name for the data source.
  3. Select the source type. Note that the list of available databases is from those available on the IBM® SPSS® Modeler machine.
    • File. Many file types are supported, such as text (*.txt, *.csv, etc.), Excel, the *.sav and *.zsav formats used by IBM SPSS Statistics, and SAS files. See the topic File source for more information.
    • Database. Any database that supports ODBC, including Microsoft SQL Server, DB2, Oracle, and others. See the topic Database source for more information. To connect to a relational database, you must first define an ODBC datasource for that database on the IBM SPSS Modeler Server.
    • Data View. Creates a connection to be used with IBM Operational Decision Management rules. For complex rules that use a hierarchical data model, note that administrators may restrict use of such rules because they can degrade performance. See the topic Data View source for more information.
    • IBM Cognos BI server. Any database that supports IBM Cognos BI server. See the topic IBM Cognos BI source for more information.
    • Analytic Server. You cannot select this source type when creating a new data source. In IBM SPSS Modeler 16, IBM SPSS Analytic Server has been introduced, allowing the use of Hadoop Distributed File System (HDFS) as a data source. IBM SPSS Modeler users can create streams using the HDFS Analytic Server source node and build models based on the Hadoop big data system. IBM Analytical Decision Management users can open these model streams in IBM SPSS Modeler Advantage for viewing (they cannot create new projects using Analytic Server data sources).
  4. Specify options as appropriate for the source type.

    You can click the No. of records icon to display the number of records in the current data source, or click Preview to display the first ten records of the data source.