I've made an early version of a new graph data science and analytics library for Python named PyGraphina. It's written in Rust and at the moment includes implementations for a large collection of popular graph algorithms, including:

- Centrality metrics: PageRank, betweenness centrality, etc.

- Community detection algorithms like connected components, Louvain, etc.

- Heuristics for hard graph algorithms, such as Max clique finding.

- Link prediction algorithms like Jaccard coefficients, Adamic-Adar index, etc.

The aim of the project is to make PyGraphina as feature-rich as NetworkX, with the performance benefits of Rust.

Project's GitHub repo: https://github.com/habedi/graphina/tree/main/pygraphina

PyGraphina's documentation: https://habedi.github.io/graphina/python