While we all love the robustness of Jupyter Notebooks, sometimes their extensive usage can facilitate bad coding practices and incorrect execution flow. First, I will briefly discuss the major advantages and drawbacks of Jupyter Notebooks. I will demonstrate useful features, tricks and tools that can improve both our code and the experience from using interactive environments. The second part of the talk will be focused on two of the biggest issues in research code – code quality and reproducibility. I will offer advice and processes that can be adopted, to improve the quality of the code and ultimately, the reliability of the conducted experiments.

Presentation: https://docs.google.com/presentation/d/1WUAHsmzFdwqMu8vZOfTW53Pf3hhYvTigFBHQO59sY5Q/edit?usp=sharing
Code: https://github.com/gkaradzhov/research_code_jupyter/tree/master/notebooks_good_bad_and_ugly

A version of this presentation focused on research code: https://gkaradzhov.com/research-code-with-jupyter-notebooks/

Back to Resources