Installation

Having R and R Studio on your laptop will allow you to work on problem sets and explore the magnificent functionality of R outside the lab. R is the language and R Studio helps us interact with R. It is important that you install R before you install R Studio.

Install R

In your web browser, go to r-project.org and then click download R. At this point, you’ll be directed to a page with a list of institutions that host the Comprehensive R Archive Network (CRAN). The idea is to pick an institution near you. Scroll down to USA and click the link for either UC Berkeley or OSU. The choice is arbitrary, but you can maintain a Beaver-free machine by picking Berkeley. Also, the Ducks play Cal this week. Know your enemy.

Windows Instructions: If you have a Windows machine, click Download R for Windows then install R for the first time then Download R 3.6.1 for Windows. To complete installation, run the .exe file you downloaded.

Mac Instructions: If you have a Mac, click Download R for (Mac) OS X then R-3.6.1.pkg under “latest release.” To complete installation, run the .pkg file you downloaded.

Linux Instructions: If you run a Linux distro, note that installation instructions vary by distro. That said, you probably know what you’re doing.

Install R Studio

In your web browser, go to rstudio.com/products/rstudio/, scroll down to R Studio Desktop, and then click Download RStudio Desktop under “Open Source Edition.” Scroll down to “Installers for Supported Platforms” and click the link that corresponds with your operating system. To complete installation, run the installer you downloaded.

R Basics

Using the console

When you open R Studio for the first time, you should notice three panels. The large panel to the left is the console. This is where you run code that tells R what to do. You can also use the console as a calculator. For example, if you type 5+5*2-1 in the console and hit Enter, then R will return

## [1] 14

in the console.

Everything is an object

The upper-right panel is the global environment. This is where R Studio stores datasets, user-defined functions, and other objects.

To define an object, you use the assignment operator <-.1 For example, suppose that you want to assign the number 5 to an object called a. In the console, you would type

a <- 5

which reads “a gets 5.” When you execute this code (by hitting Enter), a will show up in the global environment. Hovering your cursor over a in the global environment tells you that a is a numeric object.

There are other kinds of objects, too. For example,

b <- "I Love Metrics"

is a character object, and

mat <- matrix(c(1, 2, 3, 4),
              nrow = 2)

is a matrix.

One of the nice features of R is that it can store multiple objects at a time. This is especially useful for analyzing data in a data.frame object and then storing the results in another data.frame object. It is also useful for cleaning and merging data. You might think that the ability to store multiple datasets is a trivial feature, but many other statistics packages can’t do this.2

Packages

R functions come in packages. When you open a fresh R session in R Studio, a number of packages come pre-loaded. These include packages with common math and statistics functions and are known collectively as “base R.” Base R is wonderful, but non-default packages offer a great deal of flexibility and functionality.

Install a package: install.packages("package.name.here")

  • Replace package.name.here with the name of the package you want to install.
  • Make sure the name of the package is in quotes.
  • You only have to install a package once.

Alternatively, you can click on the Packages tab of the bottom-right panel:

Load a package: library(package.name.here)

  • To use functions from a non-base package, you need to load the package at the beginning of each new R session.
  • Quotes around the package name are no longer needed.

pacman

We will often need to load several packages in a single session. One way to do this is to execute library(package.1), then library(package.2), then library(package.3), and so forth. A less cumbersome way to load multiple packages is to use the p_load function from the pacman package.

  • Use the instructions above to install pacman.
  • Load the pacman package with library(pacman).
  • When you want to load additional packages, you can then execute p_load(package.1, package.2, package.3).
  • Added bonus: p_load first checks to see if the packages are installed. If they aren’t, then it will install them for you.

R scripts

To produce reproducible3 R code, it is best to use scripts. Open a new R script file with the .R extension by clicking File then New File then R Script. We will write our first script to generate a histogram and scatter plot using ggplot2.

ggplot2

Start by writing code to install and load ggplot2.

library(pacman)
p_load(ggplot2)
  • To execute one line of your script, click the line you want to run and then click Run at the upper-right corner of your R script. A quicker alternative is to click the line you want to run and then use the keyboard shortcut Ctrl Enter.
  • To execute the entire script, click Source at the upper-right corner of your R script or use the keyboard shortcut Ctrl Alt R.

Aside: It is useful to leave comments in your code to explain to your future self what your code is doing and why. You can leave a comment by typing a hash #:

# This is a comment. R will ignore it.

Check out the example dataset midwest from ggplot2. You can view the first few rows of the dataset with variable names by using the head function.

head(midwest)
## # A tibble: 6 x 28
##     PID county state  area poptotal popdensity popwhite popblack
##   <int> <chr>  <chr> <dbl>    <int>      <dbl>    <int>    <int>
## 1   561 ADAMS  IL    0.052    66090      1271.    63917     1702
## 2   562 ALEXA~ IL    0.014    10626       759      7054     3496
## 3   563 BOND   IL    0.022    14991       681.    14477      429
## 4   564 BOONE  IL    0.017    30806      1812.    29344      127
## 5   565 BROWN  IL    0.018     5836       324.     5264      547
## 6   566 BUREAU IL    0.05     35688       714.    35157       50
## # ... with 20 more variables: popamerindian <int>, popasian <int>,
## #   popother <int>, percwhite <dbl>, percblack <dbl>, percamerindan <dbl>,
## #   percasian <dbl>, percother <dbl>, popadults <int>, perchsd <dbl>,
## #   percollege <dbl>, percprof <dbl>, poppovertyknown <int>,
## #   percpovertyknown <dbl>, percbelowpoverty <dbl>,
## #   percchildbelowpovert <dbl>, percadultpoverty <dbl>,
## #   percelderlypoverty <dbl>, inmetro <int>, category <chr>

Next, make a histogram of county poverty rates (measured by the variable percbelowpoverty) using the ggplot function. You will need to tell ggplot

  • The name of the dataset.
  • The name of the variable.
  • The type of graph (in this case, geom_histogram()).
ggplot(data = midwest, aes(x = percbelowpoverty)) + 
  geom_histogram()

How many counties have poverty rates over 40 percent? What is the modal poverty rate?

To visualize relationships between variables, you can make a scatter plot. Do poverty rates appear positively or negatively correlated with race, as measured by the variable percblack?

ggplot(data = midwest, aes(x = percblack, y = percbelowpoverty)) + 
  geom_point()

  1. Pro tip: You can type the assignment operator with the keyboard shortcut Alt -.↩

  2. Reason #142 the economics department switched from Stata to R.↩

  3. Reproducible = fewer headaches later.↩

 

Kyle Raze