Summary

In this chapter, we introduced the basic principles of optimization and applied those principles to a test application. When optimizing an application, the first thing to do is test and identify the bottlenecks in the application. We saw how to write and time a benchmark using the time Unix command, the Python timeit module, and the full-fledged pytest-benchmark package. We learned how to profile our application using cProfile, line_profiler, and memory_profiler, and how to analyze and navigate the profiling data graphically with KCachegrind.

Speed is undoubtedly an important component of any modern software. The techniques we have learned in this chapter will allow you to systematically tackle the problem of making your Python programs more efficient from different angles. Further, we have seen that these tasks can take advantage of Python built-in/native packages and do not require any special external tools.

In the next chapter, we will explore how to improve performance using algorithms and data structures available in the Python standard library. We will cover scaling and sample usage of several data structures, and learn techniques such as caching and memorization. We will also introduce Big O notation, which is a common computer science tool to analyze the running time of algorithms and data structures.