Python is a popular language with applications covering a variety of domains, such as web applications, automation scripts, and especially machine learning and data analysis. Program analysis techniques provide the means to validate correctness, analyze performance, and check security properties. With Python being a dynamic language, performing program analysis at runtime is an obvious choice, as it allows for precisely observing the actual runtime behavior. However, implementing dynamic analyses with traditional approaches requires significant engineering effort. DynaPyt allows for implementing analyses with minimal effort, and provides the ability to not only observe but to also modify executions. For example, DynaPyt can be used to implement a dynamic taint analysis, for runtime verification of API protocols, to build a dynamic call graph, or to detect possibly incorrect program behavior. This tutorial will provide a hands-on introduction into dynamically analyzing Python programs with DynaPyt. We will guide participants through setting up the tool and implementing several program analyses. After the tutorial, participants will be able to build on DynaPyt for their own future research and tool development.