Hacker News new | past | comments | ask | show | jobs | submit login
Kite: An interactive visualization tool for graph theory (erkal.github.io)
189 points by homarp 14 days ago | hide | past | favorite | 14 comments



Wow, this looks really nice and reminds me alot of a tool I built myself 2 years ago. [0] The overlap in ideas is really astounding. I really like the way the nodes can be grouped in bags and how each step of running topo-sort or dijkstra is saved as a separate graph.

The way nodes are created by clicking on an existing edge feels almost exactly as in my project and I remember how excited I was coming up with the idea on my own. But obviously this project predates my project by about 2 years.

Really nice work.

[0] https://algorithms.laszlokorte.de


The editor is very slick. Where does the 'graph theory' part come into this though? As so far, all I've done is create some vertices and group them by 'bags'. Note that I'm a math theory novice


On the left side there are some tools for running algorithms on the graph. some are not implemented yet. But for example the "play" icons allows you to run djikstra or topological sort algorithm.


Hey! I tried to model an elementary "reverse linked list" (as a sort of Hello World), and couldn't figure out how to:

- Label a node once it is placed - Get everything to stop bouncing around - output/export a .dot file


Click the Force button (Darth Vader icon, bottom right) to make things stop bouncing around.

Am I the only one who really hates it when people name their projects after actual words?

Just makes it incredibly hard to search for later.


Also the same name as a very dubious IDE plug-in:

https://www.reddit.com/r/Python/comments/pn49pa/psa_kite_is_...


Also a very explicit 1998 anime.

And a operator in the lambda calculus. (Kite is just a function that returns the identity, aka as false or λx . λy . y)

Maybe Kite could consider introducing the unique label kitegraph, a bit like Go being searchable as golang

what edge and node file formats will it support? there is a big gap for graphviz excited for algorithms. you could ultimately do GNNs.

Looking forward to playing Planarity on this.


See you on IEEE VIS next month?

Interesting



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: