Fast Native Data Structures - C-C++ from Python

Explore advanced techniques for optimizing Python performance using C/C++ in this 46-minute EuroPython 2021 conference talk by Stefan Behnel and David Woods. Learn how to leverage Cython to compile Python code and utilize fast, memory-efficient native data types and structures. Discover methods for accessing Python data structure internals, processing NumPy array data, and implementing C++ STL standard library data structures as alternatives to high-level Python collections. Gain insights into creating high-level Python interfaces that enable rapid data processing while maintaining seamless integration with Python features and libraries. The talk covers topics such as static type declarations, memory allocation, C vectors, iterator protocols, C arrays, and advanced features, concluding with a Q&A session.

Intro
Static type declarations
Memory allocation
Python loop
C vector
Iterator protocol
Exposure to Python
C arrays
Python dict
C median
Advanced features
Questions