Write Perfect* Code First Time, Every Time

The traditional method of writing programs has been:

• Write some code
• Execute it
• Print variables out to the console to verify the code is working as expected. However, Python, being an interpreted language, has the benefit of being able to execute code on the fly.

Taking advantage of this, we explore a workflow, using Visual Studio Code’s (VScode) Python Extension, to try-before-you-buy each line of code before we write it in a script. That is, one can execute and verify code in VScode’s debug console without having to write-run-rewrite. This allows us to explore several options fast!

This workflow not only cuts debugging time significantly but also provides visibility in libraries lacking comprehensive documentation (as often is the case).

The basic layout of the talk will be:

• Brief comparison of interpreted and dynamic languages
• Walkthrough of what a large percentage of workflows look like. Type code in editor, go to terminal, execute script, observe error, fix error, print variable, see that it’s incorrect, fix logic error, and finally verify again.
• Walkthrough–Visual Studio Code workflow:
  • Everything in one screen
  • Integration of REPL, editor and debugger into a workflow
  • Ability to execute different code snippets while preserving the session, thus removing the need to re-execute the script.
  • Ability to inspect the attributes and methods of objects especially helpful when dealing with large libraries e.g. scikit-learn.
• Personal stories about how this workflow helped my students and friends solidify certain concepts about the Python language.