Revolutions blog recently posted a link to R code by Joshua Reich with self-contained examples of using machine learning techniques in R, including various clustering methods (k-means, nearest neighbor, and kernel), recursive partitioning (CART), principle components analysis, linear discriminant analysis, and support vector machines.  This post also links to some slides that go over the basics of machine learning.

I wish I would have discovered this long ago.  Loading data into R or MySQL requires you to specify the full path to the file.  If you do this on a Windows machine there are two annoyances.  First, if you save something to your desktop the path to your desktop is really long.  Second, windows by default uses backslashes "\" in the file path, while R or other software requires forward slashes "/".

It took me a while to figure this out, so I thought I'd share. I have a dataframe with millions of observations in it, and I want to estimate a density distribution, which is a memory intensive process. Running my kde2d function on the full dataframe throws and error -- R tries to allocate a vector that is gigabytes in size. A reasonable alternative is to run the function on a smaller subset of the data.

As far as I know there are 3 different systems for producing figures in R: (1) base graphics, included with R, (2) the lattice package, and (3) ggplot2, one of the newer plotting systems which is, according to the creator Hadley Wickham, "based on the grammar of graphics, which tries to take the good parts of base and lattice graphics and none of the bad parts.

There's no doubt that by now you've noticed that we're big fans of R around here.

For readers at Vanderbilt: At yesterday's R course I found out that Theresa Scott in the Biostatistics department holds a weekly R clinic and encourages new R users who want to learn more to bring any questions about R, or even your own code and data. The R clinic is held weekly on Thursday from 2:00-3:00 in MCN.

Writing from the previously mentioned intro to R course at the Kennedy Center. If you couldn't make it you can download all the course materials from Theresa Scott's website, under the "Current Teaching Material" heading. Here is a direct link to the PDF for the overview materials that we're going over today, along with the R code from the examples in this guide.

Hierarchical clustering is a technique for grouping samples/data points into categories and subcategories based on a similarity measure. Being the powerful statistical package it is, R has several routines for doing hierarchical clustering. The basic command for doing HC is Nearly all clustering approaches use a concept of distance.

The New York Times published this interesting article on how the ability to design and perform computer simulations is a highly marketable skill for careers across many disciplines.In methodology development we use simulation nearly every day.

I've linked to UCLA's stat computing resources once before on a previous post about choosing the right analysis for the questions your asking and the data types you have.