In this chapter, we TDDed a simple and more complex algorithm, and then built a larger test to run the comparative benchmarking. The final comparative test ran slowly, which is less than ideal, but it could be improved by optimizing the algorithms for efficiency. You were also introduced to viewing histograms as probability distributions that can be viewed for feedback on how well our algorithm is doing. Last but not least, you learned how to create a bootstrapping approach to multi-armed bandit problems.

In the next chapter, we will dive into linear regression, and we will begin to integrate third-party libraries such as scipy into our TDD practice.