Are Python 3.6 Dictionaries Ordered, and How Does This Affect Their Performance?

Are Dictionaries Ordered in Python 3.6 and How They Perform Better

Introduction

Python dictionaries have gained insertion order functionality since version 3.6, allowing elements to be accessed based on the sequence in which they were added. This article explores this aspect, including the implementation details and performance enhancements it brings.

Insertion Order in Dictionaries

Dictionaries in Python 3.6 and later support insertion order, meaning that the elements are stored and retrieved in the order they were originally inserted. This differs from previous versions where dictionaries followed a random iteration order.

Implementation Details

The new dictionary implementation employs two arrays:

• dk_entries: Holds the actual entries in insertion order.
• dk_indices: Contains indices for dk_entries, acting as a hash table.

The indices array allows for faster lookups by reducing the memory overhead associated with sparse arrays.

Performance Enhancements

The new implementation outperforms its predecessor in terms of memory usage. By storing only the necessary entries and using a sparse array for indices, it reduces memory requirements compared to the previous version.

The performance difference in terms of speed is not as significant, with some operations, like key lookups, potentially experiencing slight regressions. However, other operations, such as iteration and resizing, see noticeable improvements due to the improved memory efficiency.

Conclusion

Dictionaries in Python 3.6 have adopted an insertion order feature, providing a more predictable and consistent access pattern. The implementation achieves this while also improving memory usage, resulting in better performance in many common use cases.

The above is the detailed content of Are Python 3.6 Dictionaries Ordered, and How Does This Affect Their Performance?. For more information, please follow other related articles on the PHP Chinese website!