Not being a heavy Git user, I was surprised to learn that to gain a deeper understanding and a solid mental model of how Git works, DAGs - well known from the world of causal inference, where I feel a bit more at home - are very handy. They make it much easier to deal with more complex operations like rebasing or fixing merge conflicts - at its core, it’s just graph manipulation.
Maybe I’m not the only one who was ignorant of this, so I’m sharing this small but useful learning of mine 🤓
If interested, check out, for example, the following video, which, imo, does a decent job of explaining Git through a DAG lens.
Related notes
- A plausible model of data-generating process eats ML algorithms for breakfast
- Beyond prediction: Exploiting organizational events for causal inference in people analytics
- A graph approach to reporting structures: stop wrestling recursion
- Causal insights with no code?
- Causal Inference with Bayesian Networks
📄 Read the original post with full outputs on my blog.