Graph and network flow algorithms in GoLang

[Meir]

I am considering switching from C to GoLang in order to write more generic code faster for my research in AI. However, I noticed that the Go standard library is missing algorithms for working with graphs and networks and I am not eager to write my own code for graph partitioning or minimum cost flow -- that would undermine my goal of writing code faster!

I am guessing that the right thing to do could be (is it?) to link against existing C libraries, but I am afraid of the compatibility issues and the associated headache. So, I would appreciate if GoLang programmers could share their experience on this topic. Thanks!

[Gyu-Ho Lee]

I am glad to find your interest in graph. I don't know much about graph theory but have a huge interest in this field. I have just started building this for the sake of learning. And plan to learn network flow and minimum cost problem over the spring break. (It seems challenging for me, so I might need more reading in this topic).

https://github.com/gyuho/goraph

[Oleku Konko]

+1  nice

[David Skinner]

I am not an experienced go programmer but I do have a lot of c and c++ experience. I found the cgo documentation to be sparse but after reading a few of the sample files I am doing quite well with using go to make c calls to c++ libraries and making calls from c++ to go func. It cost me about a week and I may not be proficient but I do okay. Now that I am using it, I would never go back. I like that I can edit my cpp file and run go test and know quickly that all of my code is working as desired. It really is well integrated.

[Vincenzo Prignano]

I really like your project. Well done, will contribute where I can :)

On Saturday, March 22, 2014 6:43:12 PM UTC+1, Gyu-Ho Lee wrote:

[Dan Kortschak]

Jeff Juozapaitis and Sonya Keyes have been putting in a fair amount of
work to the gonum project's graph packages.

http://godoc.org/github.com/gonum/graph

There may also be other packages in that project you may interested in
for your research.

[Gyu-Ho Lee]

Thanks a lot! I will update more so that I can also visualize the algorithm too.