-
Notifications
You must be signed in to change notification settings - Fork 312
Open
Description
Implement Fusion Tree for Fast Integer Key Search
Description
Fusion Trees are advanced search trees optimized for integer keys, achieving O(log log n) search time using word-level parallelism. Unlike traditional balanced trees, Fusion Trees use bitwise operations, sketching, and parallel comparisons to speed up searches. Implementing Fusion Trees in pydatastructs
would provide a highly efficient alternative to AVL and Red-Black Trees for large datasets with integer keys.
Tasks
- Implement a Fusion Tree with insert, search, and delete operations.
- Optimize the tree using bitwise sketching for compressed key comparisons.
- Implement multiplication-based fingerprinting for parallel search.
- Add unit tests to verify correctness and performance.
References
- Fusion Tree - Wikipedia
- Fredman & Willard (1990): “Fusion Trees”
Metadata
Metadata
Assignees
Labels
No labels