Real data¶
16 October to 14 November 4 projects
Clean data was the training wheels. Everything here is about data that arrives in the wrong shape, from a source you do not control, with problems nobody warned you about.
-
Web Scraping
Pull structured data out of pages built for humans, and work out when you are allowed to.
-
APIs
Ask a service for data the way it expects to be asked, and keep your keys out of your repository.
-
Data Preparation and Visualization
What you do in the first hour with a dataset you have never seen before.
-
Data Preprocessing
Turning a clean table into something a model can learn from, without leaking the answers.
Preparation and preprocessing are not the same thing
Preparation makes your data correct. Preprocessing makes it model-ready. You can do the first perfectly and still ruin a project with the second.