Updates becoming a rarity? :o

I've noticed that my blog posts are now becoming weekly, which isn't necessarily bad, but I feel I should leave an update more often to ensure that I'm doing the work and to show the world my pretty pictures.

This week I've been writing up my work into some sort of essay-thing, which is now taking shape. On top of this I have begun collecting subjective evidence for the paper; running the application with the values given by the quantitative results and selecting "good" drawings to include. This is for both algorithms to show fairness.

Funny enough, although the results suggest no large change in quality when the force-strength-parameter (yet to be given a name) is below 1; the drawings appeal to my subjective taste when the value is low (0.0005) as opposed to higher values (0.5 - 1.0). This is no real surprise and was to be expected.

3D results had been on hold (what with problem solving and such) but will be next on my list.

But with progress comes reality to kick me back down, I have been comparing my runtime's against those given in other people's papers. ACE really does ruin it for me, with its millions of vertices in seconds (though I can live with this as it is spectral drawing as opposed to FDP - or so wikipedia tells me). Worse still, Yifan Hu's results are slightly better than those given by my implementation of his work and my own work, and in order to compete, I must spend more time improving (and finishing) my algorithm.

Now for some images:

This is a comparison of the graph fe_body, on the left is the output from my own algorithm, and on the left, that given on Yifan Hu's gallery. Although globally different, the structures are very similar (if not identical). A very badly draw picture below shows the similarities.

Admittedly, his looks more car-like... but his is a finished algorithm and not in comparison_mode.