Sacha Chua has a nice post that describes something that I really like: taking some data that she’s collected and analysing it to extract useful information. For some reason I really like that sort of thing and am always doing it myself with my own data.
In Chua’s case, she wanted to know how many Emacs News entries she’s written and plot them over time. It’s a little more difficult than you might think. First of all, she didn’t have the complete Emacs News index so she had to recreate the missing parts by converting the HTML entries into Org mode. She did that with a standard Unix pipeline using Pandoc.
Once she had everything in Org mode, she wrote some Elisp to extract the needed information and put it in a table. From there, it was easy to produce a graph showing how may entries she wrote by year. You can even see the years her daughter was a toddler because the number of entries went down for a couple of years. As Chua says, “Babies sleep a lot; toddlers, not so much.”
Take a look at Chua’s post to see all the code and the final graph. The data is mostly useful only for Chua just as mine is useful mostly for me but the way she goes about analyzing the data is useful to anyone that needs to take some data and massage it.