Weekly Report for week ending 1 June 2012
Got the event markers in the flot graphs working, fetching on the fly and
caching as required. Built up a few webpages to try to get a feel for the
types, numbers, timing and locations of events and have each one linked
back to these graph of AMP data with event markers which makes exploring
the results of the event detection a bit easier.
Spent some time looking over the events in the graphs and adjusting the
algorithms slightly to try to exclude ones that aren't really events. Also
started to think about how events could be grouped to show those that
likely have a similar cause, and how it could be determined whether or not
to alert anyone about them.