Our great sponsors
NetworkX | aoc2022 | |
---|---|---|
61 | 22 | |
14,178 | 2 | |
1.6% | - | |
9.6 | 10.0 | |
5 days ago | 8 months ago | |
Python | C | |
GNU General Public License v3.0 or later | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
NetworkX
-
Routes to LANL from 186 sites on the Internet
Built from this data... https://github.com/networkx/networkx/blob/main/examples/grap...
-
The Hunt for the Missing Data Type
I think one of the elements that author is missing here is that graphs are sparse matrices, and thus can be expressed with Linear Algebra. They mention adjacency matrices, but not sparse adjacency matrices, or incidence matrices (which can express muti and hypergraphs).
Linear Algebra is how almost all academic graph theory is expressed, and large chunks of machine learning and AI research are expressed in this language as well. There was recent thread here about PageRank and how it's really an eigenvector problem over a matrix, and the reality is, all graphs are matrices, they're typically sparse ones.
One question you might ask is, why would I do this? Why not just write my graph algorithms as a function that traverses nodes and edges? And one of the big answers is, parallelism. How are you going to do it? Fork a thread at each edge? Use a thread pool? What if you want to do it on CUDA too? Now you have many problems. How do you know how to efficiently schedule work? By treating graph traversal as a matrix multiplication, you just say Ax = b, and let the library figure it out on the specific hardware you want to target.
Here for example is a recent question on the NetworkX repo for how to find the boundary of a triangular mesh, it's one single line of GraphBLAS if you consider the graph as a matrix:
https://github.com/networkx/networkx/discussions/7326
This brings a very powerful language to the table, Linear Algebra. A language spoken by every scientist, engineer, mathematician and researcher on the planet. By treating graphs like matrices graph algorithms become expressible as mathematical formulas. For example, neural networks are graphs of adjacent layers, and the operation used to traverse from layer to layer is matrix multiplication. This generalizes to all matrices.
There is a lot of very new and powerful research and development going on around sparse graphs with linear algebra in the GraphBLAS API standard, and it's best reference implementation, SuiteSparse:GraphBLAS:
https://github.com/DrTimothyAldenDavis/GraphBLAS
SuiteSparse provides a highly optimized, parallel and CPU/GPU supported sparse Matrix Multiplication. This is relevant because traversing graph edges IS matrix multiplication when you realize that graphs are matrices.
Recently NetworkX has grown the ability to have different "graph engine" backends, and one of the first to be developed uses the python-graphblas library that binds to SuiteSparse. I'm not a directly contributor to that particular work but as I understand it there has been great results.
-
Build the dependency graph of your BigQuery pipelines at no cost: a Python implementation
In the project we used Python lib networkx and a DiGraph object (Direct Graph). To detect a table reference in a Query, we use sqlglot, a SQL parser (among other things) that works well with Bigquery.
- NetworkX โ Network Analysis in Python
-
Custom libraries and utility tools for challenges
If you program in Python, can use NetworkX for that. But it's probably a good idea to implement the basic algorithms yourself at least one time.
-
Google open-sources their graph mining library
For those wanting to play with graphs and ML I was browsing the arangodb docs recently and I saw that it includes integrations to various graph libraries and machine learning frameworks [1]. I also saw a few jupyter notebooks dealing with machine learning from graphs [2].
Integrations include:
* NetworkX -- https://networkx.org/
* DeepGraphLibrary -- https://www.dgl.ai/
* cuGraph (Rapids.ai Graph) -- https://docs.rapids.ai/api/cugraph/stable/
* PyG (PyTorch Geometric) -- https://pytorch-geometric.readthedocs.io/en/latest/
--
1: https://docs.arangodb.com/3.11/data-science/adapters/
2: https://github.com/arangodb/interactive_tutorials#machine-le...
-
org-roam-pygraph: Build a graph of your org-roam collection for use in Python
org-roam-ui is a great interactive visualization tool, but its main use is visualization. The hope of this library is that it could be part of a larger graph analysis pipeline. The demo provides an example graph visualization, but what you choose to do with the resulting graph certainly isn't limited to that. See for example networkx.
aoc2022
-
[2022 Day 9] "Rope Bridge". A particularly efficient implementation idea (for people who understand C++, but applicable to C/Java/Go/Rust as well)
It may be a nifty hash table (no, it is!:)) but the whole program runs about 12x slower on an M1 than my version where I first determine the max dimensions and simply allocate a grid of booleans... So I wonder how much the 8x8 bitset breakup could improve. But bitsets are a C++ feature. In C, like NRK: https://github.com/ednl/aoc2022/blob/main/09.c (my "startstoptimer.c" and .h are in the same repo)
-
[2022 Day 15 (Part 2)] [Python] I wrote a really fast solution for day 15 part 2 (less than 1ms). What do you think of the algorithm I came up with?
I also checked lines but only after doing a rotation by 45 degrees, so the lines are straight. Compiled in C, fastest run time on M1 was 26 ยตs: https://github.com/ednl/aoc2022/blob/main/15.c
-
-๐- 2022 Day 15 Solutions -๐-
Same code but with preprocessed input to make it all fit into memory, runs in 7 ms on an Arduino Uno! https://github.com/ednl/aoc2022/blob/main/aoc22-15/aoc22-15.ino
-
-๐- 2022 Day 13 Solutions -๐-
Complete program runs in 463 ยตs on Apple M1, 2.61 ms on Pi 4. See comments at the top of the source file for how I measured. My comparison function:
- -๐- 2022 Day 12 Solutions -๐-
-
[2022 day 11][C] Benching Monkeys
Not 100% sure this is Upping-The-Ante, maybe just Other. I wanted to share some benchmark results of my solution for today, day 11 with the 10,000 monkeys, and how I got there. I think the easiest way to compare performance is to use the same hardware, and nowadays fairly common & standardised hardware might be the Raspberry Pi 4. Although, you can't buy any for years now... Best score I got when running my solution on my Pi 4 home server is 15.6 ms.
-
-๐- 2022 Day 11 Solutions -๐-
I quickly saw that I could do "item = item modulo (product of all div-test numbers)" but the implementation took me a while in C without queues or circular buffers. But that's all part of the fun for me! I didn't look for further clever optimisations because the compiled program runs in 20 ms on a Raspberry Pi 4. That was fast enough for today, I thought. Source code: https://github.com/ednl/aoc2022/blob/main/11.c
-
-๐- 2022 Day 10 Solutions -๐-
Yay, embedded software engineering!! :) Short, fast & almost no memory needed in C: https://github.com/ednl/aoc2022/blob/main/10.c or the relevant bits:
-
-๐- 2022 Day 9 Solutions -๐-
That's great, but on what hardware? My solution in C runs in 0.8 ms average (0.6 ms minimum) using hyperfine to measure 1000 runs in a Mac Mini M1. Same on a Pi 4 in performance mode: 3.6 - 3.8 ms.
-
-๐- 2022 Day 8 Solutions -๐-
Well, it took me a while to realise that in part 1 you always have to check the whole row or column because a higher tree can come at any point ... And except for skipping the borders, I couldn't come up with any sort of clever optimisation that would help reduce the O(N^2) complexity. It still runs in under 1 ms on a Mac Mini M1 according to hyperfine. Full code 52 lines without space/comments: https://github.com/ednl/aoc2022/blob/main/08.c
What are some alternatives?
Numba - NumPy aware dynamic Python compiler using LLVM
rust-mos - Empowering everyone to build reliable and efficient software.
Dask - Parallel computing with task scheduling
AdventOfCode2022
julia - The Julia Programming Language
AdventOfCode - My solutions to Advent of Code
RDKit - The official sources for the RDKit library
aoc - KlongPy Advent of Code (AoC) solutions
snap - Stanford Network Analysis Platform (SNAP) is a general purpose network analysis and graph mining library.
AOC2022 - Advent of Code 2022, solved in Haskell
SymPy - A computer algebra system written in pure Python
aoc-go - A Golang tool for generating code for Advent of Code