Tableau’s new data prep tool, Maestro, is now in public beta, and I’m loving it!  While the first beta doesn’t necessarily include all the features that might soon make it to a production release, what blows me away is the overall paradigm.  It brings the intuitiveness and transparency that Tableau brought to visual analytics and actually makes data prep fun!

(BTW, if you want to see a preview of Maestro in action, please check out this upcoming webinar!)

So to celebrate all the hard work the developers have put into Maestro and Tableau, I thought I’d have some fun and make a Christmas card using only 3 records of data plus Maestro and Tableau:

So, how did I use Maestro to create a fractal?

Well, I started with only 3 records of data:

These are the corners of the Sierpinski Triangle.  To generate more points, I self-join the data source to itself on the Join field creating a cross-join (meaning every record gets joined to every other record)

In the step after the join, I calculate new X and Y coordinates.  And I do this by adding together the X values from both sides of the join and dividing by 2 and doing the same for the Y values.  This gives me a point half-way between the two joined points.

And then I repeat this process, joining in the three records again and again and calculating the half-way points for each new record.

And each time I join in the 3 original points, my data set grows times 3! Here, for example, is the result of the 4th join, going from 81 records to 243 (and I love the join summary in Maestro!):

After 8 iterations of joins, I have nearly 20,000 records of data — more than enough points to draw the Sierpinski Triangle).  And with one final calculation in Maestro for the trunk of the fractal Christmas tree…

…I have the right data to output to a .tde and build the view in Tableau:

I’m looking forward to a great new year filled with Tableau and Maestro magic.  Merry Christmas everyone!

Click here if you’d like to download the data files, Maestro flow, and Tableau workbook