Log in

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

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

Time [Oct. 31st, 2015|11:09 am]
Okay, how is it almost November? It really feels to me like it should be, like, late September.

More alarmingly, in six weeks and half an hour, roughly, I'll be heading off to San Francisco for AGU, where I need to give a talk on how much of a difference it makes if you throw fancy math at bias-correcting extreme precipitation in climate model output, and I haven't actually done the research yet!

I have written quite a lot of code -- I'm actually developing an R package for all this stuff -- and I'm hopeful that it's going a lot like building a new office building seems to, where they spend weeks and weeks digging a big hole in the ground, but once the foundation has been poured, the whole thing goes up in a matter of days. When I look at my to-do list, that does seem to be the case, but the fact that I haven't done an data analysis yet is a smidge nervous-making. Especially given that there's Thanksgiving and a trip to visit Jerry's mom coming up, so it's actually only five weeks to AGU.

We don't have anything in particular on deck for Halloween. We have D&D this afternoon, but we're not sure if we'll do anything afterwards. I did remember to pick up candy last weekend while I was shopping for other non-perishables. And last night I made a cute little bear jack-o-lantern with tiny vampire fangs at Sarah's pumpkin carving party. (Poor Monkey has been frequently working late this week.)

But right now, I have a kitty snoozing in my lap, so taking a shower and getting dressed will have to wait just a little longer.

[User Picture]From: geometrician
2015-11-01 08:36 pm (UTC)
R is driving me crazy. I'm working on structural equation modeling. I am coming to believe that in order to really comprehend what I am doing, I have to write the code myself. And really, I feel like I am only just beginning to be able to think like R, and definitely not consistently. I'll keep plugging away at it, and eventually it will click.
(Reply) (Thread)
[User Picture]From: dr_tectonic
2015-11-01 10:09 pm (UTC)
It has a pretty steep learning curve, but now that I've gotten the hang of it, I really like R.

I think the main thing is learning to think semi-functionally with vectors instead of procedurally with loops. Once you make that shift, things start to hang together more clearly.

My usual metaphor is that in other languages, I often feel like I'm spending lots of time and effort riding herd on a bunch of arrays, keeping them all synced up and marching in time with one another. Whereas coding in R feels more like I spend a long time thinking hard about how to do data origami and fold everything up into the right structures, and then I write five lines of code and magic happens.
(Reply) (Parent) (Thread)