Jul. 7th, 2017

rebeccmeister: (1x)
Another overnight in the lab last night. But I got data back from efforts on Monday, Tuesday, and Wednesday, and the data look really good, which is what usually happens with pilot experiments with too-small sample sizes. Still, it's a start.

I managed to stay out of the lab until 12:30 pm, by heading to the BAP at 10 am to row. I'd originally intended to meet up with the Serious Double at 6:30 am, but stayed up too late reading a trashy fantasy novel, and really needed the sleep anyway.

It wound up being a good thing that I'd waited. The weeds have gotten so bad that the only open water is a narrow channel barely wide enough for one boat. It's like rowing on a river that requires constant steering and vigilance.

Algae at Berkeley Aquatic Park

Algae at Berkeley Aquatic Park

So it was way less stressful to be out there by myself and not have to worry about dodging other paddlers or rowers.

I should have put on sunscreen.

I also met up with [personal profile] scrottie for dinner at a nice restaurant on Gilman Street, Lalime's. S had poked around on the internet and discovered that they included the Tell-Tale Tart from Boulevard Brewing on their bottled beer list. I was delighted to have a chance to dine there because I bike past the restaurant on my way home from work every day and it had piqued my curiosity. It was an extravagant dinner by our standards, but we relished it. I don't dip into the foodie culture out here very often, but who doesn't appreciate the occasional beautiful and well-cooked cuisine?

Last night's overnight trials were a wash. One of my minions was very very eager to help out with the late-night timepoints, so it was his trial-by-fire time, which meant several screw-ups and the stress that accompanies that. I cannot simultaneously concentrate on high-precision tasks and think and verbalize in the face of mistakes, you'll be surprised to learn. The net result was we achieved the same amount of data that I would have gathered had I done everything by myself, but with less sleep and more stress to go along with it. If I'd done everything by myself, I would have just split up the crickets across two timepoints, which would have made for a really busy but productive evening. It all just reinforced my notion that for projects like the circadian experiment, adding extra people ("help") is counterproductive. On the other hand, my minion now has a much greater appreciation for everything that's involved in running a highly precise and time-sensitive experiment. Now he'll be back to a less time-sensitive project and I hope he will have more patience for it. Many aspects of science are tedious, so one must get used to it.

One other surprise happened: in the spring, I interviewed another potential undergraduate researcher, and was disappointed to discover he was already doing research in another lab and looking to piggyback even more research experience. I decided I had to draw a line - involvement in two labs at the same time is too much and can lead to a lot of awkwardness. I encouraged him to stick with his existing research, but said to come back if he decided to transition out of that project and was still interested in working with us. And yesterday, out of the blue, he showed up! During our initial interview he struck me as someone who has really great potential but hasn't had access to great mentoring, which is something I'll do my darndest to supply.
rebeccmeister: (cricket)
My brain is complete mush today.

But I needed to squeeze in a Skype meeting this morning with my long-distance undergrad, who has been determinedly plugging away at learning the basics with R. We have two main tasks left with the current stage of data analysis, which we basically need to translate from English into syntax. The first is figuring out how to read in a bunch of files from a bunch of directories. This should be straightforward.

The second task is one I'd been scratching my head over. When recording cricket activity, I set up the timelapse video to record 4 crickets at a time. Our tracking software assigns individual numbers to individuals, but for various reasons individual crickets wind up having between 1 and 10 different numbers assigned to them. So, how to separate out data for each individual cricket? I'd been thinking of coming up with a method to subset the track files by the assigned ID numbers. This would require figuring out how to import a ragged data file, then figuring out a kind of complicated reference scheme.

But then today C pointed out that because the crickets are spatially separated from each other, it would probably be a whole lot simpler to just subset based on each cricket's xy coordinates. We don't expect cricket 1 to ever show up in the area occupied by cricket 2, unless something went seriously wrong with the setup (nothing did).

Foreheadslap to self. Good job, undergraduate! You rock.

This startup is a VERY good idea. I need to pay a visit.


rebeccmeister: (Default)

September 2017

3 4 56 789
10 11 1213 141516
17 1819 20 21 2223
24 252627282930

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 26th, 2017 06:11 pm
Powered by Dreamwidth Studios