University of Minnesota, Twin Cities     School of Statistics     Stat 5101     Rweb

Stat 5101 (Geyer) Computing

R

R is a general purpose computing language designed especially for statistics. That is, you can do any computational problem in R, but what it is really good at is statistics.

R is free software (both free as in "free beer" and free as in "free speech"). If you want to use it at home, you can download from CRAN (the Comprehensive R Archive Network). It is available for Microsoft Windows and all versions of UNIX (including Linux). Unfortunately, the Macintosh port seems moribund. You do not have to download it if you don't want to. R is installed in the computers in the lab (Room B53, Ford Hall), and R can be used over the web on any computer that runs a web browser (see the section on Rweb).

Rweb Example

The Rweb "general" interface looks something like the form below, only there are more boxes and buttons, the boxes are different shapes, and the boxes are empty. You have to fill them in to get any work done.

For this example, we've filled in the box so that Rweb will look up the value of the standard normal cumulative distribution function at x = 4.156. Clicking the "Submit" button runs R and produces the result.

Go ahead and try it. Use the "Back" button on your browser to return to this page.

If the example doesn't make any sense, you will just have to wait until we cover the material: cumulative distribution functions (Section 3.2 in Lindgren) and normal distributions (Section 6.8 in Lindgren).

The old-fashioned, pencil-and-paper way to do this is to look up the result in a table like Table I (pp. 576--577 in Lindgren). In this case the table isn't much help because it doesn't go far enough. The argument we want (4.156) is off the table.

Rweb

To use Rweb on other datasets go to the Rweb home page. There you will be asked to choose one of Rweb's two modes of operation:

You should probably stick to the "general" version that looks like the example above until you are familiar with R and Rweb. The JavaScript version is a bit complicated.