Making this cookbook reproducible

[brian-rose]

Related to my pre-publication review (#5), I'm noting here that in its current state this cookbook does not reproduce itself. Currently we have execute_notebooks: off in the config file, so the cookbook is deploying a pre-executed notebook which has been checked into the repository with output.

Now that the environment file is fixed up (#7), I am able to execute the notebook and reproduce the results locally on a laptop. I was also able to run the notebook manually on the Pythia binder.

I opened #9 to see if we can automate the execution on the binder.

[brian-rose]

My impression is that the execution speed of the notebook is limited by network bandwidth, not computing power. It looks like a good candidate for accelerating with a dask-based parallelization of the data subsetting.