The original data available to a data scientist/analyst is only the starting point of the discovery process. Building additional fields, both to address issues in the data and to incorporate subject matter expertise by creating derived variables, is central to predictive analytics. SPSS provides a range of tools to facilitate the development of new data fields, using a very comprehensive set of mathematical and statistical functions. Essential tasks, such as revising or grouping codes can be handled using specially designed transformation commands that simplify the process. Transformations that require more complex rules can be performed using conditional logic and the nesting of the appropriate commands.
In this chapter, the most commonly used transformation commands--RECODE, COMPUTE and IF--were explored in detail. The more complex DO IF/END IF structure was introduced...