When data science projects get beyond a few hundred lines of code they can become hard to modify without breaking the whole thing. Visualising and understanding the pipeline is key and the #rstats package {targets} is brilliant for this, it can even be used as an orchestrator for Python code!
