Book Image

Advanced Python Programming

By : Dr. Gabriele Lanaro, Quan Nguyen, Sakis Kasampalis
Book Image

Advanced Python Programming

By: Dr. Gabriele Lanaro, Quan Nguyen, Sakis Kasampalis

Overview of this book

This Learning Path shows you how to leverage the power of both native and third-party Python libraries for building robust and responsive applications. You will learn about profilers and reactive programming, concurrency and parallelism, as well as tools for making your apps quick and efficient. You will discover how to write code for parallel architectures using TensorFlow and Theano, and use a cluster of computers for large-scale computations using technologies such as Dask and PySpark. With the knowledge of how Python design patterns work, you will be able to clone objects, secure interfaces, dynamically choose algorithms, and accomplish much more in high performance computing. By the end of this Learning Path, you will have the skills and confidence to build engaging models that quickly offer efficient solutions to your problems. This Learning Path includes content from the following Packt products: • Python High Performance - Second Edition by Gabriele Lanaro • Mastering Concurrency in Python by Quan Nguyen • Mastering Python Design Patterns by Sakis Kasampalis
Table of Contents (41 chapters)
Title Page
Copyright
About Packt
Contributors
Preface
Index

How to work with the GIL


There are a few ways to deal with the GIL in your Python applications, which will be addressed as follows.

Implementing multiprocessing, rather than multithreading 

This is perhaps the most popular and easiest method to circumvent the GIL and achieve optimal speed in a concurrent program. As the GIL only prevents multiple threads from executing CPU-bound tasks simultaneously, processes executing over multiple cores of a system, each having its own memory space, are completely immune to the GIL.

Specifically, considering the preceding countdown example, let's compare the performance of that CPU-bound program when it is sequential, multithreading, and multiprocessing. Navigate to the Chapter22/example3.py file; the first part of the program is identical to what we saw earlier, but at the end we add in an implementation of a multiprocessing solution for the problem of counting down from 50,000,000, using two separate processes:

# Chapter22/example3.py

import time
import...