Introduction

IBM SPSS Modeler is the core product in an incredibly powerful and scalable predictive analytics platform. The visual, point-n-click Modeler interface makes it almost immediately accessible to business analysts or anyone with a quantitative background. However, having a tool that is easy to use but cannot be automated severely limits the usefulness and scalability of that tool. That's why CLEM scripting was added to the Modeler platform.

CLEM scripting is an important pillar of the Modeler platform scalability. It serves a number of functions that cover data preparation, modeling, evaluation, and deployment. First, scripting allows tedious, repetitive processes within data preparation to be automated. ETL and other data preparation processes can involve many disconnected steps. Scripting can connect these processes into a single connected workflow. Second, model building often involves much experimentation with variable combinations, row samples, time windows, and so...