Twingly Summer of Code

So, today I’ve finally finished the graph producing algorithm, which is awesome! And what more is that it’s actually looking really good. Obviously there will be some need for tweaks, but other than that it’s looking really good!

Here is picture to show you how nice it looks, it’s still quite dotty, but I need some extra info from the database before I can do something about that, but as I said there is tweaking to do. Anyhow, would be nice to hear what you think about it, is there anything you think in particular is important to get into the picture to make it more easily understandable and easy to navigate? Please give me a shout and I’ll try to include that in the algorithm. It could be colouring, labels on the dot’s and so on….

For you who don’t know what we are doing, take a look at our first episode.

Over and out

This post was written by Jonas Kac

Cuz we will!


Fri Jun 13 15:50:50 2008
Now doing file: swedish_7000.gml
__________________________
Reading generated gmlfile...
Setting layout...
Finding communities...
Adding coords to the graph...
Setting colors...
Saving to GmlZ file...
Making picture...
Saving File: swedish_7000.png
Finished at: Fri Jun 13 16:07:11 2008

__________________________

Fri Jun 13 16:07:11 2008
Now doing file: swedish_2500.gml
__________________________
Reading generated gmlfile…
Setting layout…

THE TEAM

This post was written by Jonas Kac

I’ve been working quite a lot the past week whit iGraph and trying to get the graphs to render properly, and then add some extra color to them.

I’ve finally gotten everything working, and realized that iGraph is not that bad after all. There is one more thing that needs to be done in the graph department, and that is the algorithm that I told you about earlier. iGraph actually has it’s built in algorithm for making the exact graph that I want to do, but there is a big drawback, and that is that every time you make a new graph you get a totally new graph. Meaning that nodes aren’t even remotely close to where they were earlier, I guess you can be lucky and get them “quite” close, but that is not good enough. We want the graph to portray a familiarity every time you take a look at it, things will change, new nodes (blogs that is) will show get included, and some may choose to start writing about something else and then the node should fit somewhere else. But this is supposed to happen over time not instantaneously.

So that’s why there is a need to make a new algorithm that is able to handle this problem.

Fortunately there are good algorithms out there already.

But I’ll do that later, now I need to parse some XML so that you can actually see some progress soon!

Over and out for now

This post was written by Jonas Kac