Python Profiling your Applications
On our last market notes:
we covered concepts in computer organization and architecture, as well as some Python specifics and quirks in the Cpython reference implementations.
In this market notes, we continue to talk about Python optimizations. We talk in high level about different approaches to code optimization, write concrete examples in using timing modules as well as through Python decorators.
We discuss useful libraries and utilities in the IPython kernel, LINUX/UNIX utilities, concepts such as page faults, context switching and so on.
We discuss function profiling using the built-in Python module known as cProfile, a deterministic profiler (as opposed to statistical ones), running through a pandas working example to diagnose and fix performance issues.
The modifications start from page 829. Next set of notes - we do more granular profiling, using line profilers, and then talk about actual optimizations and programming constructs.
The discount duration for the publication has been closed - welcome to the family. If you enjoy our content, I would really appreciate if you share my publication to your fellow peers, quants and developers. The only marketing we have so far is word-of-mouth, and I would really appreciate your honest recommendations.
Notes: (paid readers)