Baseball Hacks

Slice and Dice Teams with Cubes

Use Perl to create your very own local OLAP multidimensional cubes that you can explore using Microsoft’s Office Web Components (OWC).

Many hacks in this book show you where to get data and how to store it in a database. Other hacks show you how to answer specific questions using this data. But what if you don’t know exactly what question you want to ask? Oftentimes, so much data is available that you don’t know where to start. Other times, you just want to kind of poke around and let one question lead you on a journey to five more, as if you were on some kind of data safari.

You can do this with a database, writing multiple SQL queries to fetch different subsets of data summarized in different ways. However, this quickly becomes tedious, especially if you ask the same types of questions repeatedly. A data cube is a popular tool for working with summarized, aggregated data that lets you easily slice and dice the data. Logically, these structures are called cubes because they’re seen as 3D objects (or, more technically, as nD objects), where the qualitative things (such as teams, height, and season) that you typically like to group by are called dimensions, and the pre-aggregated data and any calculated expressions from them (such as hits and home runs), are seen as measures. This hack shows you how to use a cool tool from Microsoft to work with data cubes.

Well, let me back up. The Microsoft SQL Server product comes bundled with Microsoft SQL Server ...