Reminder - Meeting tomorrow

[madelaine]

Okay, one last reminder - tomorrow, 2pm, Plaza branch, Brian, ggplot2.

Plus Earl might share something interesting about parallel computing
with R.

Madelaine

[Chris Seidel]

Thanks for the presentation Brian. By the way Earl, the free chapter
of the ggplot2 book that explains all about qplot with good examples
of plot types is at: http://had.co.nz/ggplot2/book/

-Chris

[Earl F. Glynn]

Thanks to Madelaine for setting up the meeting.

Ditto to thanks to Brian for his ggplot talk. That was a good reminder to
re-visit the ggplot2 package. Better looking graphics would be nice on many
projects.

Thanks, Chris, for the link to the online book. I'll take a look.

FWIW, here's the link to "Innocentive" web site that I briefly described:
http://www.innocentive.com/. I'm not so interested in the money they might
pay as just something that is technically challenging.

Some of the challenges are a bit like the CAMDA competition, and some are on
bioinformatics topics such as "Build an Efficient Pipeline to Find the Most
Powerful Predictors," https://www.innocentive.com/ar/challenge/9932794
That might be fun, but is a bit too big to explore with limited time. [Sadly
over time I'm losing much of my bioinformatics knowledge anyway.]

I've explored several of the Innocentive competitions but I've only
submitted one solution. The review of the submissions takes longer than the
time spent working on them. I'm still waiting to hear about a submission
from July. If only I had created ggplot2 graphics to make them look better
... <g>.

The project that needed multiple cores and a lot of computation in R (I was
trying some wavelet approaches) was here: Materials Identification Based on
Measurements of Passively Emitted Electromagnetic Radiation
http://tunedit.org/challenge/material-classification

I've worked a bit on this from time to time since May. Much of the data
look almost random. Finding "signal" is quite difficult.

efg

[Madelaine Gogol]

Here's the article about Kindergarten teachers that used the STAR data set, I thought it was interesting.

http://www.nytimes.com/2010/07/28/business/economy/28leonhardt.html

Thanks for the talk, Brian, I'll give it a try. We may try meeting at UMKC next meeting, I'll talk to Bob about it and let everyone know. Tentatively December 3rd at 2pm.

Madelaine