Real-world data is rarely ready to use. It arrives with duplicate rows, missing values, inconsistent spellings, and dates in three different formats. Analysts often spend more time cleaning data than analysing it, and any mistakes left in will flow straight into your reports.
In this module you'll learn to profile a messy dataset, spot what is wrong with it, and fix it. You'll do this first with SQL, using functions like TRIM, COALESCE and ROW_NUMBER, and then with Python and pandas, the tool analysts reach for when a dataset is too messy or too large for a spreadsheet.
Clean data is the foundation for what comes next. In the next module, Data Modeling & Warehousing, you'll see how clean data is organised so it can be analysed at scale.
A practical walkthrough of cleaning a messy e-commerce dataset in SQL, the way it is done on real projects.
A complete project from import to recap, cleaning the Nashville Housing dataset with date conversion, string functions, self-joins, and duplicate removal using ROW_NUMBER.
How to import a CSV, find and remove duplicates, and handle missing values and inconsistencies using pandas.