High Performance Python: Practical Performant Programming for Humans

This book is a hidden gem. It explains in-depth methods for how to profile your Python programs, how to use different compilers for python code to gain speedups, how your code gets interpreted and how it interacts with the underlying computer architecture, how to write async code and parallel code in python, how to conserve RAM and many more topics and concepts. Each one of the concepts that this book teaches had a lot of work and effort put into by the authors. Wether it's the text or the code examples, the quality is high. The book is written in a way that you can read it in a linear fashion but it can also serve as a reference once you know where everything is. The authors of this book are very talented and experienced python developers with background in data science. In addition to showing many advanced methods for making your code fast, they put a strong emphasis on when/where to use each method and the importance of profiling and benchmarking.This book is probably a better read for people with an intermediate understanding of Python, operating systems and parallel programming. It uses very interesting and high quality problems as examples (warning, they are not always easy to understand). I highly recommend this book to anyone who has been using Python for a few years and wants to take their understanding of the language to the next level.

Good read

A very comprehensive book that covers the memory and performance of Python. Even though the book's title contains "performance," it also has a chapter on memory management. Book has excellent and up to the point examples.

The cover should be a Reticulated Python. They are the highest performance python.

I’m learning a lot and half way through the book. I’ve written some python before but primarily an expert C++ programmer. Ironically I ordered a more fundamental python book and it’s stuck in the mail but still finding this book very understandable.

Good: The book does a nice job covering multiple aspects of speeding up Python code, as well as reducing memory usage, etc. Good examples with test results of different approaches.Not so good: The book is probably twice as long as it should be. There's a ton of filler in here, like the odd opinion notes about hard it is moving code to a cluster, or using multiprocessing. The final section appears to be random thoughts about Python code, and has nothing to do with performance.Also, the writing style is just simply weird. I was reading along about some topic, and all of a sudden the text says, "Ian doesn't like to have his laptop set up so that ...". I thought, "Who is Ian?, and why would I care about his opinion?" Well, it turns out he's one of the two authors. This approach is used multiple times, and boy is it jarring - such a weird way of writing, especially in a technical book.That said, if you can wade through the strange writing and ignore all the useless fluff, there is some good content in this.

Thank you

It's a great book with useful information. It will take your understanding of Python's internal workings, memory allocation and how to use better performance libraries. The only down side to this book is I found some of the examples to verbose and unnecessarily complex for the point being proven. Someone who isn't into data science as much may find the example arduous.

and this book was suggested. numba had maybe half a chapter (out of 10) worth of narration plus a bonus writeup from the tech lead of the library itself. useful yet I felt we had been gently led away from proper weeds. I would say 20% of the book overall is necessary trivia and ceremonies. the structure and delivery are good, each technique or class of approaches is showcased with gradual improvement, so some chapters even have a bit of character arch.

Kwik word of advice. Do not expect this to be a easy read. Buy a notebook and some pens because you will be taking notes. If you are a beginner in Python programming wait a year or two before reading this.

Sono ancora sotto le 100 pagine di lettura ma gia' dall' inizio si percepisce una qualita' superiore rispetto ad altri manuali sul python. Utilissimo il sito github per scaricare i sorgenti e fare prove in locale. Pienamente soddisfatto.

There was a time when this publisher's books were considered the most authoritative ones. And now for the past 5-6 years I am watching that worst and most superficial books are being released from this publisher. The have become sort of guaranteed of worst books on the topic where they write a lot from here there, try to cover lots of topic with almost zero depth. And end up suggesting you to buy a few more OReilly books, from where also you hardly get anything.This book was supposed to be on High performance computing with python, but they have written a chapter of Multiprocessing module, and without explaining it, they job explaining joblib module! You will see the chapter of Multiprocessing long enough, but it hardly gives any proper introduction and detailing. They don't even explain their codes point by point.Also the writing style looks incoherent. Some chapters look a bit better written while you feel that some chapters are written by some kid. No depth at all. They did not even cover properly or say anything on important topics like Futures.Entire book looks trying to impress you with bringing names of modules only with hardly any meaningful detailing. I am almost on the brink of stopping buying SPD and Oreilly books. They r just behind money and publishing junks, that too overpriced.