The May 2019 SWD challenge asks us to curate a dataset of our own, so that we know all its vagaries and can add information about context.

I chose to download my twitter archive for analysis... the initial impetus was that I thought that I have been primarily retweeting recently, and I wondered what my patterns of use have been over the years that I have been using twitter.

Questions I considered:

What proportion of my tweets are RTs and what proportion are original?

How many include links? How many include pictures? How does tweet frequency vary over time? WRT month of the year? Day of the week?

As in all things, time is limited, and I have a dozen other things that I'm also working on, so I wound up with more modest results... how has my twitter use varied over time, and are there obvious features in the graph?

The most interesting thing I saw was that since I tend to spend the summer outside, and often on the road, my August tweets are only about 2% of the total. I'm basically offline for the month of August, and this has apparently been consistent for the last 9 years.