Data Cleaning

Data cleaning is the process of fixing problems in raw data so it can be trusted for analysis: correcting wrong types, standardizing inconsistent labels (“NY,” “N.Y.,” “New York”), handling missing values, removing duplicates, and catching impossible entries such as an age of 200.

It is famously the most time-consuming part of data work — practitioners often say the majority of a project is spent here. The cardinal rule is to clean with code, not by hand-editing cells, so every change is recorded, repeatable, and reviewable rather than a silent, unrepeatable edit.