Software Design by Example

Most data scientists have taught themselves most of what they know about programming. As a result, many have gaps in their knowledge: they may be experts in some areas, but don't even know what they don't know about others.

One of those other areas is software design. A large program is not just a dozen short programs stacked on top of each other: doubling the size of a program more than doubles its complexity. Since our brains can only hold a small number of things at once, making large programs comprehensible, testable, shareable, and maintainable requires more than using functions and sensible variable names: it requires design.

The best way to learn design in any field is to study examples. These lessons therefore build small versions of tools that programmers use every day to show how experienced software designers think. Along the way, they introduce some fundamental ideas in computer science that most data scientists haven't encountered. Finally, we hope that if you know how programming tools work, you'll be more likely to use them and better able to use them well.