1. What were my goals for this week?
- Implement a couple different types of clustering, naive and hopefully something a little smarter
- Do a search through existing bibliographic tools and see what's out there and if we need to build our own
- Work on a good way to pick out "critical" points to represent motion without showing paths from every point
2. What did I accomplish this week?
- Looked through potential texts for fall graphics class
- Put list of bibliographic tools (along with quick descriptions of features) on the wiki
- Wrote rough Matlab program to pick "critical" points using local maxima
- Looked through k-means clustering algorithms to prepare for implementing in C++
- Did a C++ refresher read...
3. Why are numbers 1 and 2 different?
My coding skillz were a lot rustier than I had imagined they were - I've spent too long coding Matlab and GAMS and not real languages, so jumping into C++ was a lot more difficult than I'd thought it would be. The "implement this in C++" task on Thursday turned into "let's remember how to code C++ first". Thanks to that, the implementation of k-means still isn't done, and thus I haven't moved on to other types of clustering.
4. What are my goals for next week?
- Implement k-means clustering
- Find other potential clustering algorithms, read and implement
- Try some of them out on real data!
- Discuss possible better critical point selections with Aaron
5. How does this fit in my bigger picture?
Focusing on clustering will hopefully help with doing the motion representation more efficiently or at least more effectively - we'll try to cluster by vector direction and location, and compare the results with what the critical point selection shows.
6. What did I read this week?
- Suhre, Karsten and Yves-Henri Sanejouand, On the potential of normal-mode analysis for solving difficult molecular-replacement problems.
- Krebs, W.G. et al, Normal Mode Analysis of Macromolecular Motions in a Database Framework: Developing MOde Concentration as a Useful Classifying Statistic.
::