Unit 4: Python, Pandas, and Data Preparation

Real datasets arrive messy. Missing values, duplicate records, inconsistent labels, wrong types, and suspicious outliers are not side issues—they can change the conclusion.

Source lessons

1. Working with Data: Python and the Pandas Library
2. Working with Data: Data Preparation

Use the first lesson to build the dataframe mental model. Use the second to think about what must happen before analysis is trustworthy.

What you should be able to prove

By the end of the Unit, you should be able to:

• explain rows, columns, indexes, and data types in a dataframe;
• identify common data-quality failures;
• choose a transformation for a specific defect;
• explain what information a cleaning step could accidentally destroy; and
• preserve an audit trail of important transformations.

Evidence check

Finish with Evidence Check — Build a Data-Cleaning Plan.

Runtime boundary

The pinned Microsoft notebooks and code remain available as source resources. The Robotnix browser Python/Pandas workbench is the next implementation layer and is not yet claimed as executable here. The current evidence check can be completed with reasoning and pseudocode.