Automating Data Preparation

The Quadstone System includes a suite of “data-build commands” that perform standard operations to create customer-level analysis datasets, deploy models or scores resulting from analysis, or generate reports. In this module, you will learn how to apply these commands manually at the command line, and then how to create automated scripts to carry out more complex operations.

Duration: ½ day

Audience:

• Data architects
• Systems integrators

and other individuals who want access to data-build and reporting functionality from a command line or in scripts, for example for

• Automating update processes
• Deploying models into operational systems
• Creating a web portal

Prerequisites

• Prior attendance at one of
• Introduction to Data Preparation (Quadstone System Version 5)
• Data Preparation with Decisionhouse (Quadstone System Version 4)
• Prior attendance at one of
• Advanced Data Preparation (Quadstone System Version 5)
• Data Preparation with Transaction Measurement Language (Quadstone System Version 4)
• Failing this, you should have some experience with importing flat files and deriving fields in the Quadstone System and basic experience of writing any programming or scripting language

Topics covered

• Overview of data-build commands
• Transforming data
• Import from and export to flat file and database
• Sorting and deriving
• Transaction measurement (aggregation or data rollup)
• Combining datasets
• Enhancing datasets by adding metadata
• Reporting
• Automating and parametrizing data builds
• Data-build design considerations

Back