?

Log in

No account? Create an account
Workshops End - The Mad Schemes of Dr. Tectonic [entries|archive|friends|userinfo]
Beemer

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

Workshops End [Aug. 5th, 2014|10:52 pm]
Beemer
Oh man. I am a bit worn out.

So July has at last come to an end, and after tomorrow, so will all of the workshops that I've been involved with. They were all for work, and all AT work (except for one, which was down the hill at NOAA), and some of them were a lot OF work, too.

I already posted about the first half of the month. The last two and a half weeks have been all about the colloquium that my boss was putting on (which means it was a bit more important than the other ones). For this one, I was not a participant, but I'm my boss's security blanket on the "technical things running properly" front, so after the first day or two (when we determined there was no need for me to help wrangle powerpoint files or anything), I got tagged to make sure that everything was going to work for this decision-making exercise at the end of the second week (that is, last Friday).

The idea is, you've got a predator-prey model that represents a fishery, and you're harvesting the prey species. And each team in the contest has to come up with a strategy that maximizes the amount of their harvest but is also sustainable. And climate change makes the predator reproduce faster as time goes on. (So imagine that the prey species is tuna, and the predator is some parasite that reproduces better when the water's warm.) And you know the dynamics of the system, but you don't know the values of the parameters that control it. So we give them some code (R scripts) that will simulate the system and let them explore the different possibilities and they have to come up with a function that describes how much they'll harvest each year for 100 years.

So I got a copy of the code (written by a colleague of RL, the guy running the exercise), and I was able to get it to run on my computer, but it was... it was pretty terrible. It made me itch. And it made me curse involuntarily when I was reading it. But y'know, it wasn't mine, and it functioned. We could use it for the exercise. So I left it alone.

And then near the end of the first week, RL asks me if we could get it to also do this thing and that thing, and I say, well, I could, but I'd probably have to do some, ahem, 'significant refactoring'. And RL says to me: "The code is yours now. It belongs to you."

Which is how I ended up going in to the office for six hours on a Saturday, and working extra in the evenings for several days, because I was able to burn all the terrible to the ground and write shiny new non-terrible, non-itchy code to replace it that did all the same things literally a hundred times faster. Like, sweeping over the parameters took more than 8 minutes originally, and after I was done it took less than five seconds. (Did I mention that the original code was terrible? Because it was.)

And I got it all done and added new stuff and on Friday of last week it worked, hooray! And then I was DONE with things I had to do for the workshop.

Yesterday I played catch-up with people requesting things over email, and today I got my AGU abstract submitted, and tomorrow I'm going to see all the colloquium participants' presentations on their group projects, and after that I am going to take. Some time. OFF.
LinkReply

Comments:
[User Picture]From: quirkstreet
2014-08-06 02:10 pm (UTC)
Oh, the entrapment power of bad code. I seem to be immune to such things, but they are indeed time consuming for those who hear the siren call. Some time off sounds like just the ticket!
(Reply) (Thread)
[User Picture]From: jofish22
2014-08-07 05:33 pm (UTC)
Huh! Sweet. I find myself intrigued.

But the last thing I should be doing right now is looking at other peoples code for projects I have no engagement in...
(Reply) (Thread)