I fear the universe may have a leak as I am becoming more aware of the age old saying that theres not enough time in the day. It's finally starting to hit me that 2 years are over and I have just over 11 months left; so its time to get rocking I guess.
I am still on the literature side of things this week (and for the past few weeks - hence the lack of updates), making my way in to Dynamic Graph Data Structures, a lengthy subject on its own. Having found what to aim for in drawings (aesthetics wise), decided on what drawing algorithm to use (FDP - Spring Embedder) and what viewing technique will be best (Animation), I am now looking at how previous authors control the flow of data and the complexity of operations associated with dynamic graphs.
Although I have a lot more to read, there is no definitive system for dealing with multilevel graphs - specifically updating the locations of vertices in coarser representations. In our static graph drawing work, we used a brute force post-order search of the approximation tree to update vertex positions, using the mean of its children - computationally expensive but also used in other work such as the Octree. This requires FDP be applied to all levels of the graph before the updates can be made - something which would be extremely inefficient if applied to dynamic graphs).
Further reading will hopefully identify a more useful method, or at least give me some ideas how to overcome this (NB: Veldhuizens work isn't clear on this subject but I've yet to read it in this context). That's all for this week folks, I will update again shortly - hopefully getting back into the swing of updating this. See you next time.
No comments:
Post a Comment