I'm giving a short talk to the experimental economics class I'm taking on how to use R to do the most common tests. This is a short tutorial in order to get R up and running with R studio. I assume that you've followed this tutorial already for my talk. If you're looking for my slides from the talk or the dataset I worked with, they're at the bottom of this page along with some other useful resources.
Let's get started. First, go to R's website, which should look something like this:
On the left hand side, click on the CRAN link under "Download, Packages" next to the arrow in the image above. This will bring up a page of links to download R from, like the image below. Choose a location close to you so that the download is faster.
This should bring you to a page that looks like the image below. Now click on the link to download R for whatever operating system you're running. I.e. if your computer has Windows, click on "Download R for Windows."
This should bring up another page with a few more links, like the picture below. Click on "base."
Finally we get to the page with the actual download link, like below. Click on "Download R Z for X" where X is whatever operating system you have and Z is the latest version of R, i.e. some number like "2.15.1". At the time I created this post, it was "Download R 2.15.1 for Windows" since I was installing R on a windows machine.
This will download the executable to install R - a file named "R-2.15.1-win.exe" or something similar. Make sure you save it somewhere you can find it easily. When it finishes, run the file. Just follow the on-screen instructions that pop up. You shouldn't have to change anything in order for R to properly install.
Now you're all set to start using R... except the GUI that R comes with out of the box isn't very good. Rstudio is a free IDE that improves on the base R GUI substantially. Go here to download it. Download the version of Rstudio that their website recommends for your machine somewhere that you can easily find. Once this completes, open the file - it should be called "RStudio-0.96.316.exe" or something similar. From this point, just follow the on-screen instructions to complete the installation.
Now we'll install a couple of useful packages that exist for R. First we'll install the R package "ggplot2". ggplot2 is a package for creating statistical graphics that drastically improves upon R's base graphics.
In order to install this, open up R Studio and make sure you're connected to the internet. Then type install.packages("ggplot2") into the R console and hit enter, as below.
R will output the following message, or something similar:
Installing package(s) into ‘C:/Users/Matt/R/win-library/2.15’
(as ‘lib’ is unspecified)
--- Please select a CRAN mirror for use in this session ---
Wait for a few seconds, then R will give you some options for which mirror to download from. Type the number for the mirror that is closest to you to download everything faster, then press enter. See the picture below.
R should be pretty busy for a few minutes while it downloads and installs several packages that ggplot2 depends on as well as ggplot2 itself. Once it finishes and you see the blue ">" in the bottom of the R console, the packages are installed.
There's another package you need to install using basically the same process. agricolae is a package that has a bunch of methods for agricultural statistics, but more importantly it has some useful nonparametric tests. To install it, type install.packages("agricolae") into the R console and follow the same process as before.
In addition, I've uploaded a dataset that we'll be using during my presentation. Download it here. Save it as diam.csv somewhere you can easily find it.
That's all the software you'll need to be up and running! Here are a bunch of useful resources, including the slides from my talk:
Slides from the presentation. (pdf) Probably more useful while you're sitting at your computer than while I was talking.
diam.csv. The dataset I used for most examples in my presentation.
An R script containing every R command from my presentation. Open with any text editor, though opening it in R Studio is best.
R Studio tutorial. (pdf) Useful for getting your bearings in R while using R Studio. It covers the basics of computation R, including some stuff I didn't cover such as dealing with vectors and matrices.
An R script containing an old presentation. This one has many more details about the basics in R as well as using ggplot2, plus some stuff about quickly using Bayesian methods to fit models. Note: enter install.packages("arm") into the R console to use the Bayesian stuff.
An R reference card. (pdf) Print it out and tape it on the wall next to your desk. Seriously. Do it now.
The ggplot2 website. This contains useful information for making complicated, informative and pretty graphics using the ggplot2 package.
Course website for the class I took to learn R. Some overlap, but there are many new things.
Knitr. This is a fantastic way to integrate your computations from R into a nice compiled Latex file, and it's relatively painless with R Studio.