
Source data is often structured and formatted to satisfy the requirements of operational application programs, and is not well suited for users of data warehouses. Data must often undergo considerable transformation before it becomes part of the data warehouse. DB2 Universal Database and the DB2 Warehouse Manager include over 100 built-in transformers, statistical transformers, and user-defined functions.
The SQL functions assist you in performing various field-level and table-level transformations. The field-level transformations include string manipulation, data type conversion, arithmetic calculations, and Boolean operations. Table-level transformations include average, median, minimum, maximum, rollup along a dimension, cube calculations, and more. You use warehouse transformers to perform basic data transformations, including cleaning data, generating key columns, generating period tables, inverting data, and pivoting data. The statistical transformers perform the following functions: calculation of basic statistics, calculation of subtotals, calculation of moving averages, chi-square test, chi-square goodness-of-fit test, correlation analysis, regression, and analysis of variance (ANOVA). User-defined functions give you the ability to easily define and reuse your own transformers.