The first bit of my audio file is simply the audio from the "Consumption" time series. I added a little more range to the frequency changes (using “p.freq.target(Std.mtof(80.0 + w*22.0));” instead of p.freq.target(Std.mtof(80.0 + w*20.0));I also changed the wave to a saw wave. For the second voicing, I created my own data file with random numbers. The file starts with a bunch of zeroes, allowing the first voicing to have a solo section at the beginning. This second voicing only lasts a short period of time.
For my time series, I used the “Classification of Individual Consumption by purpose” in the United States from 1998-2013, as seen at: http://datamarket.com/topic/230/united-states