Here’s a repo which explains the basics of scikit-learn pipelines and why they can both save time and prevent data leakage.
It introduces the sklearn.Pipeline class, how to instantiate a pipeline, how to modify an existing pipeline and how to apply them to different datasets.
Here is the repo. Rather than go through the code here, I’ve included a link to run through the tutorial on a remote environment and see the code for yourself. Click the myBinder badge below to get started.
