SigmaAnalytics & Tech
C5PythonIntermediateAbout 120 min to complete

Data Cleaning

A repeatable cleaning pipeline that reports every change it makes.

Notebook

About 120 minutes to complete

Enrol to read the notebook and download it to run.

What this module covers

A repeatable cleaning pipeline that reports every change it makes.

By the end of it

  • Profile a dirty dataset systematically before touching it
  • Detect and remove duplicates, including near-duplicates on a subset of columns
  • Normalise inconsistent categories, whitespace, and casing
  • Correct dtypes safely, counting every coercion
  • Detect outliers with IQR and robust z-scores, and decide what to do about them
  • Validate rows against business rules
  • Build a cleaning pipeline that produces a full audit trail
  • Distinguish errors from legitimate extreme values