Bizarro Ball - Make a Hash of it

5 Likes

If it sounds bizarre - that's because it is! Bizarro was a comic book super-villian, a 'negative' clone of Superman who lived on the planet Htrae (Earth, backwards). On Htrae, everything is backwards - wrong is right, good is bad.

So what does this have to do with Baseball? Or even SAS, for that matter?!

Well - as it turns out - authors @DonH and @hashman have previously published (alongside their book), a set of SAS programs that will auto-generate random data for an entire baseball season. This came to be known as Bizarro Ball.

After becoming aware of this project, and recognising that the code / data is so obviously useful for training exercises, product demos, testing etc - with the author's permission, I loaded it onto github.

The reasons for this were severalfold:

Ease of Use

The original package involved downloading, unzipping, and saving / configuring the code on a file system. On github, there is now a build script which concatenates all the files into a single program - meaning that SAS Studio / University Edition users can simply click here and copy paste the code directly into their editor. If you have internet access from SAS, it's even easier - simply run:

filename bizarro url 
  "https://raw.githubusercontent.com/allanbowe/BizarroBall/master/bizarroball.sas";
%inc bizarro;

Documentation

Each SAS file was marked up with appropriate Doxygen header tags, enabling a documentation site to be generated. The great thing about the doxygen approach is that code becomes 'self documenting' - in the sense that the documentation for each file, lives inside the file itself. An example can be found here.

Audit Trail

As is normal for a git repository, every change to the project is visible in the commit history (ie what changed, by whom, and when).

Collaboration

Perhaps the greatest thing about having the project on github is that the entire world can contribute to make it better! Got ideas or suggestions? Raise an issue, or even a pull request (see CONTRIBUTIONS.md).

It Needed a New Home Anyway

The sascommunity.org site is being decommissioned, so it had to move at some point.

How is this different from SASHELP?

Apart from the randomness, and the Baseball domain interest, this data can also be modified (unlike SASHELP) - which can be useful in situations like class exercises. It is also interesting to follow and understand the hashing techniques used to create the data, described in more detail in this paper.

That's great! Where can I get more information?

* The book (obviously)

* The aforementioned paper

* The DMHashBook tag

* The DMHashBookData tag

Now that all this code / data is documented and so easily accessible, what will YOU build with it? Do let us know!