I track several habits - sleep, running, github pushes, mastodon tweets Most...
I track several habits
- sleep, running, github pushes, mastodon tweets
Most interesting stat - time between habit. Some habits are Gamma distribution shaped, which is generated by the product of several random variables, i.e. several things have to line up for the habit to happen. The raw data doesn't say what those hidden variables are.
Self-replies
The other pattern is habit data can have really high autocorrelations, where if you're doing your habit it is because you did it yesterday and if you are not doing your habit, you are highly unlikely to start,
e.g. a Hawkes Process, where there is a baseline probability you'll do the thing, the probability goes up with the sum of how often you did it recently and a decay function where the impact of recent past dissipates, e.g. reading habits, if you were reading yesterday, you'll probably keep reading the book and then the novelty wears off and you go back to your baseline probability of reading.