One talk survived, only three more to go...

[rebeccmeister]

I gave an hour-long presentation about R to a group of engineering students today at lunchtime. They gave me a certificate of appreciation, a Department of Industrial and Systems Engineering notepad, and a slim laser pointer. I hope they got what they wanted from the talk. It was half Keynote presentation, half me-bumbling-through-scripts-and-websites. An hour isn't really long enough to dig in and play with R. I would put it in the column of "not my best talk, ever," but I hope it was sufficient to intrigue some of these students into poking at R on their own. The questions at the end were difficult to answer. I've never used MATLAB, so I don't know how R compares to MATLAB (apples and oranges?). I don't know what factors contribute to R's stability vs. instability or major causes of crashes. I don't think that R is the best choice for analyzing a super-complex dataset involving weather conditions, crop types, and crop yields. But then again, I don't know what program might be better for that purpose, either.

Oh well. The students were quite nice, and quite a homogenous non-white population (but not as male-dominated as you might think!). Interesting to see how racial and gender divides work at this university.

Up next, a talk on leafcutter ant nutrition at my brother's university next Thursday. I hope I don't completely embarrass myself in front of his entire department. At least they won't ask me any questions about MATLAB. On the other hand, for the talk after that one I'll be back among a math-oriented audience, presenting my non-mathy data. So I'm not completely out of the woods yet.

But before any of that, tonight, a bike ride to downtown Bryan, then my friend A's birthday party, then up early to volunteer at a bike ride tomorrow morning until 2 pm. Then a pause, then a contemporary dance performance. I might sleep on Sunday.

Comments

[altamira16.livejournal.com]

The weather people still work in Fortran.

[twoeleven.livejournal.com]

A lot of people do, Babbage have mercy on their souls.

[rebeccmeister.livejournal.com]

Oh geez. Figuring out what language to use for a given task can be a tricky project, in and of itself. I just hope this particular student saw enough to be able to poke around a bit and think over his questions some more.

[altamira16.livejournal.com]

It's actually a little stranger than I wrote. At my last job, people still worked in Fortran. In attempting to learn a little Fortran, I realized that some aerospace people and some atmospheric people still use Fortran. It is the language that people worked in when they wanted mathematically accurate results before Matlab and R were available, and some of the people who worked with Fortran as their mathematical language are still in the work place.

Someone that I met last week said she applied for an internship at National Center for Atmospheric Research, and Fortran was one of the skills that was needed for the position.

[twoeleven.livejournal.com]

I've never used MATLAB, so I don't know how R compares to MATLAB (apples and oranges?).
Pretty much.

I don't think that R is the best choice for analyzing a super-complex dataset involving weather conditions, crop types, and crop yields. But then again, I don't know what program might be better for that purpose, either.
Depending on what's meant by "super-complex", some R libraries might work. But if you mean what I think you mean, probably a dedicated data-mining program.

Anyway, I'm glad you survived your talk on how to do statistics like a pirate. :)

[randomdreams.livejournal.com]

I'm with "pretty much" having spent a year programming in matlab and at least some time on R. Kind of different.

[rebeccmeister.livejournal.com]

Okay, reassuring to know I wasn't way off-base! I've mostly just ever seen MATLAB output, but I can also say with some certainty that it's much more finicky to put together complex equations in R.

[rebeccmeister.livejournal.com]

Re: MATLAB, it's good to know I wasn't too off-base. :-) It's interesting, being on the fringes of this stuff - among programmers, I'm a biologist, but among biologists, I'm a programmer.