Successfully integrating data science into an organization can be a challenge, and requires data scientists to align well with business needs. The ability to test ideas quickly and show the value from data can be key, and a lot of things must happen before a model can reach the production environment. Following best practices from software engineering is essential to create and change any code, which can speed up the development of any data product, and bring any predictive model to production earlier. Actionable insights from data will be available sooner, innovative ideas can be tested faster, and the data product can reach the desired maturity earlier.
During the course, you will learn how to apply best practices from software development to your data science project, enabling you to create a code base that is maintainable, extensible and understandable. You will learn to recognize common problems with existing data science code bases, how to structure your code, and how to make you code not just readable, but beautiful as well.
We will explore more advanced Python features, such as decorators, and how you can benefit from an IDE like PyCharm. You will learn to set up a development stack that facilitates a typical data science workflow in Python, with the tooling to automate the boring bits. You will also gain hands-on experience in building a CI/CD pipeline and automated testing using Travis CI.