radix-heap 0.1.0

Monotone priority queue useful for shortest paths and more; wraps Rust's radix-heap crate

Radix Heap for Python

This module provides a Python implementation of Radix Heap-like data structures, designed to function as monotone priority queues. The heavy-lifting for the heap operations is powered by the radix-heap Rust crate from mpdn -- they did 99% of the hard work by implementing the data structure.

Monotone priority queues are useful for pathfinding in general (e.g., A*, Dijkstra's algorithm). A monotone priority queue must maintain the invariant where the extracted elements form a monotone sequence (e.g., for a max-heap, the extracted elements are in non-increasing order).

Overview

A Radix Heap is a type of priority queue where the extracted elements form a monotone sequence. This module includes implementations for both maximum and minimum heaps for integer and float keys.

Classes

• RadixHeapLike[K, V]: A generic base class for Radix Heap-like structures.
• RadixMaxHeapInt[V]: A radix max heap for integers.
• RadixMinHeapInt[V]: A radix min heap for integers.
• RadixMaxHeap[V]: A radix max heap for floats (or integers).
• RadixMinHeap[V]: A radix min heap for floats (or integers).

Methods

• push(key: K, value: V) -> None: Adds a key-value pair to the heap.
• pop() -> V: Removes and returns the minimum value from the heap.
• pop_with_key() -> tuple[K, V]: Removes and returns the minimum key-value pair from the heap.
• top() -> V | None: Returns the minimum value from the heap without removing it.
• clear() -> None: Removes all elements from the heap.
• __len__() -> int: Returns the number of elements in the heap.

Usage Examples

Radix Max Heap for Integers

from radix_heap import RadixMaxHeapInt

heap = RadixMaxHeapInt()
heap.push(10, "A")
heap.push(5, "B")
heap.push(8, "C")

while heap:
    print(heap.pop())
# Output: "A", "C", "B"

Installation

To install the module, you can use pip:

pip install radix-heap

Contributing

Contributions are welcome! Please check the repository and feel free to open issues or submit pull requests.

License

This project is licensed under the MIT License. See the LICENSE file for details.