% Setting Up R % DPI R Bootcamp Dec. 3rd to 5th, 2012 % Jared Knowles

Getting Ready for the DPI R Bootcamp

• How to get up and running with R in no time!

Materials

• Download R
• RStudio
• We'll also want to install a few basic packages. This can be done on the fly later, but a few include ggplot2;knitr;plyr

Step 1: Install R

• Simply run the installer. If you have a 64-bit machine, choose 64 bit R, if you don't know, just choose the default
• The install should be less than 100mb
• There is a short video on the next slide for help installing on Windows machines

Installing R on Windows

Installing R on a Mac

• Get the install file from CRAN
• Get RStudio
• Run the R installer, and then run the RStudio installer

Installing R on Ubuntu or Debian Linux

• Read the install notes online
• Get RStudio
• Run the R installer, and then run the RStudio installer

Step 2: RStudio

• Now run the RStudio installer
• RStudio will automatically find your R installation

Step 3: Install packages

First, watch this tutorial

Install Packages

• Are you running RStudio in administrator mode on Windows Vista or Windows 7?
• If not, you need to be to install packages (in most cases)
• When you launch RStudio, right click on the icon for RStudio and then click “Run as Administrator”
• If you are on a Mac or Windows XP you can disregard this

Install Packages (2)

• Now, copy and paste the code below into the bottom left window in RStudio (the R terminal):

install_new<-function(mypkg){
  if (mypkg %in% installed.packages()) cat("Package already installed")
  else{cat("Package not found, so installing with dependencies...\n 
           Press CTRL C to abort.")
    Sys.sleep(5)
    install.packages(mypkg,repos="http://cran.wustl.edu/")
}
}

install_new('plyr')
install_new('lmtest')
install_new('ggplot2')
install_new('gridExtra')
install_new('eeptools')
install_new('stringr')
install_new('knitr')
install_new('quantreg')
install_new('xtable')
install_new('lme4')
install_new('caret')

Alternate Install

• If you want or if you have troubles with the above, you can overwrite any existing versions of those packages and do the install in one line, shown below:

install.packages(c('plyr','lmtest','ggplot2','gridExtra','stringr',
                   'knitr','quantreg','xtable','lme4','eeptools','caret'),
                 repos="http://cran.wustl.edu/")

Run RStudio

• You should be able to run RStudio now and load any of the packages above.
• Run the code below in the RStudio terminal panel (bottom left) to test this:

library(lmtest)
example(gqtest)

And you should see…

## 
## gqtest> ## generate a regressor
## gqtest> x <- rep(c(-1,1), 50)
## 
## gqtest> ## generate heteroskedastic and homoskedastic disturbances
## gqtest> err1 <- c(rnorm(50, sd=1), rnorm(50, sd=2))
## 
## gqtest> err2 <- rnorm(100)
## 
## gqtest> ## generate a linear relationship
## gqtest> y1 <- 1 + x + err1
## 
## gqtest> y2 <- 1 + x + err2
## 
## gqtest> ## perform Goldfeld-Quandt test
## gqtest> gqtest(y1 ~ x)
## 
##  Goldfeld-Quandt test
## 
## data:  y1 ~ x 
## GQ = 2.542, df1 = 48, df2 = 48, p-value = 0.0007882
## 
## 
## gqtest> gqtest(y2 ~ x)
## 
##  Goldfeld-Quandt test
## 
## data:  y2 ~ x 
## GQ = 0.7892, df1 = 48, df2 = 48, p-value = 0.7924

And test the graphics…

Type in in the terminal (or copy and paste)

library(ggplot2)
y <- rt(200, df = 5)
qplot(sample = y, stat = "qq")

Results

library(ggplot2)
y <- rt(200, df = 5)
qplot(sample = y, stat = "qq")

That's all

• Did you see the output like the slides above?
• Did you see the plot in the lower right?

Optional

• If you want to see the cutting edge of R development–interactive web-based data applications, you can install the following beta software

options(repos = c(RStudio = "http://rstudio.org/_packages", getOption("repos")))
install.packages("shiny")

Optional Cont.

• To check if it worked (and it might not depending on your machine's configuration)

library(shiny)
runExample("06_tabsets")

• Does a new browser window or tab open up?
• In RStudio hit the “stop” button in the terminal panel to exit the demo

You are ready to go!

See you on December 3rd to the 5th to find out how to go from these basic steps to using R to learn from your data.

Can't wait to see you in Madison!

Attribution and License

This work (R Tutorial for Education, by Jared E. Knowles), in service of the Wisconsin Department of Public Instruction, is free of known copyright restrictions.