Effective Machine Learning Teams

Chapter 7. Supercharging Your Code Editor with Simple Techniques

There are no shortcuts in life, but there are many in coding.

Juntao Qiu, developer and author

Most ML practitioners that we know love to code. Code is our vehicle for transporting ideas in our head (“will this feature engineering technique improve the model’s performance?”) into reality. In fact, many bemoan the fact that there are too many distractions, too many meetings, and not enough time to write code. However, when we do get time to code, it’s common to find ourselves wasting valuable time on tedious tasks such as manual testing (the focus of Chapters 4 and 5), reading convoluted code (the focus of the next chapter), and unproductive coding practices (the focus of this chapter).

In this chapter, we will detail how you and your team can spend less time getting stuck or even lost in the weeds with unproductive coding practices. We’ll demonstrate how your integrated development environment (IDE) can help you to read and write code more effectively. An IDE is an application designed to aid developers, data scientists, and engineers to write, run, test, and debug code productively. It provides development tools such as a source code editor, integrated terminal, version control support, virtual environment management, and code suggestions. It can also assist you with refactoring toward a better designed and more readable codebase (see Figure 7-1).

To that end, this chapter will cover:

• How to configure your IDE ...