News for hackers. Stuff that matters.

nerdfaktor42

Re: on Thursday March 02, 2023 @05:01AM (#34993962)
Thanks for the pointers, I'll look into them. I read your blog post series yesterday and found it very well written and interesting. Gonna read the CIDR paper, too. Fyi, I'm the author of https://github.com/s1ck/graph where we use a read-only CSR and my day job is https://github.com/neo4j/graph-data-science which is built on top of a read-only, compressed CSR. Maybe we can have a chat at some point :)
Re: on Wednesday March 01, 2023 @07:42AM (#34981722)
Do you know of any research papers wrt "more update-friendly disk-based CSR-based indices"?
Looking forward to your talk tomorrow :)
Re:Who ranks better? Memgraph vs. NetworkX PageRank on Wednesday November 23, 2022 @11:49AM (#33721571)
This is a very interesting benchmark! I'd like to add a library to the comparison, which has recently been released.

It's called graph-mate (https://pypi.org/project/graph-mate/) and it's a wrapper around a Rust implementation for parallel graph algorithms (https://crates.io/crates/graph) that a friend and I work on.

We created a Jupyter notebook that loads the same Wikipedia articles and runs page rank with the same parameters: https://github.com/s1ck/graph/blob/d0a45116b1891daa39f493647... We converted the memgraph dataset to an edge list: https://gist.github.com/s1ck/97e23af14b2e117fa47c713addef751.... To reproduce, put the dataset next to the notebook.

On my machine (Ryzen 3950x, 64GB RAM) ..

  memgraph takes 1285 ms
  graph_mate takes 3.347 ms
I have not looked into the C++ implementation of page rank in memgraph. The Rust implementation is a pull-based Page Rank which is not uncommon in parallel graph libraries. See https://github.com/s1ck/graph/blob/036cda61e1bbf2d9ccb2b0e46... for details.

graph-mate and the Rust graph project are pretty young, we don't support that many algorithms, yet. But if people are interested in contributing, feel free to open a PR / issue :)