Last Week [18.03] - Holiday! I was away for a week to help with my parents project, nestled quietly in the middle of nowhere.
Two weeks ago [11.03] - Technical Report! Having had comments from my supervisors, making those changes, and then some further changes of my own. Hopefully the end is near for the technical report. In addition, meetings with supervisors to discuss these and my intentions when I return from holiday.
Four weeks ago [25.02] - Working on implementing the primitives functionality and moving to dynamic data structures (which are so slow that I will be providing support for both Collections and Arrays 5sec to 13sec for 4elt). As for this, all the primitive layout schemes appear to work and different primitive coarsening schemes - such as leaf and chain coarsening - appear to work (and work in combination with one another).
Since the last blog post [18.02] - I don't really remember this far back, but thanks to my notes, it appears I was working on some errors with random coarsening (over the linear method used previously). This has since been fixed and was due to a efficiency mechanism dependant on the order of vertices.
So this week, what are my intentions?
- Renumbering of test graphs and ensuring results do not differ (the algorithm should treat identical graphs in the same way).
- Collect runtime and quality results for the various preprocessing tasks and multi-matching tasks assuming previous problems with the matching are fixed
- Dynamic graph time slices (frames of some dynamic graph after changes occur) with some quality analysis for each (and/or additionally - mental/movement map stuffs)
- Collect results for the dynamic matching framework/stuff
Obviously not going to be completed in a week, especially when its the slow start after a break away, but these are my aims for the coming weeks. Time is running out, so we shall see how truly competent I am.
P.S. finally had my MPhil to PhD viva accepted by the Research Degrees Committee!
No comments:
Post a Comment