…I’ll blog about it!
In the interest of fully pointless disclosure: I’m trying to find a signal in this data for my research and utterly failing. Here’s a hint of what I’m up to:
Needless to say, I’m stuck, so I figured I’d blog about some other things that have been going on: namely, the fact that The Lady and I have officially pieced together our training plan for the Philly Marathon. It’s pretty freaking awesome: we combined two separate training plans, added our own twist (mainly to work around the insane September we’re going to have, which involves no fewer than 4 distinct racing events), and came up with something workable. At its peak, we’ll be cranking out a 42-mile week and not one, but two 20-mile runs.
Along those lines, I finally figured out a reasonable way to plot my running progress since purchasing my Garmin GPS watch, in order to see if there’s been any improvement over that time in my overall pace. The key was to tap into the watch’s autolap function: the way in which the watch stores each “lap” as its own entity, which for running is always (or 95% of the time) 1 mile. Treat each mile as independent (too strong an assumption, but over all 800+ miles I think it’s a reasonable one), calculate an average pace, and voila!
I was pretty excited to see a clear decrease in the regression value: a strong sign that I am, in fact, getting faster. I still want to do an update later which involves the kernel regression I previously tried to use but failed, as this new approach doesn’t even require that I hack the kernel to make it work. But for now, feel free to enjoy the updated code.
Other things I’m working on, for those who are curious:
- A new lab website. Mainly adding content. It takes a surprising amount of time before folks are satisfied with it.
- Getting a paper up and ready involving the previous mysterious graph. I already mentioned this? Yeah well, this is the current main thrust, given…
- Preparing to go out of town for a week. Going to Colorado (visit The Lady’s family) and Florida (a friend’s wedding!).
- Working on some outstanding Mahout issues. Currently have the Ronon Virtual Cluster up and running, crunching some data with 4.2+ million nodes.
- Still settling into my new apartment. And scrambling to get my ducks in order after I had to close my credit account due to suspicious activity (another story for another time).
So to all that, I say this: