library(leaflet)
leaflet() %>% addTiles() %>% setView(-1.264, 51.752, zoom = 17)

+−

Leaflet | © OpenStreetMap contributors, CC-BY-SA

What is R?

• R is a free statistical programming software. You can download R here.
• In this course, we will be running R in the RStudio IDE.

Why R?

• It's free!

• It's open-source!

• It's versatile! (In fact, all materials in this class have been created in R! )

This course

• This course is designed to introduce you to the basics of R programming.

• By the end, you will know how to:

  1. Generate and transform numeric, logical, and character vectors

  2. Deal with missing values

  3. Load and inspect data

  4. Generate descriptives

library(coronavirus)
corona_de <- coronavirus[coronavirus$country=="Germany" & 
                           coronavirus$type=="death", ]
corona_de$death_7 <-  zoo::rollmean(corona_de$cases, k = 7, fill = NA)
plot(x = corona_de$date, 
     y = corona_de$cases, 
     cex = 0.3)
lines(x = corona_de$date, 
     y = corona_de$death_7, 
     type = "l",
     cex = 1.5, 
     col = "red")

If you want to dive deeper

This course only scratches the surface of what you can do in R.

For more elaborate introductions and more advanced guides, see the following (free!) books:

• YaRrr! The Pirate’s Guide to R by _Nathaniel D. Phillips

  • This book is a great introduction to base R, and lots of material I draw from in this presentation is based on the book!

• R for Data Science by Gareth Golemund and Hadley Wickham

  • This book is fantastic if you want to learn about the tidyverse and tidy R, which is a different way of writting R code than base R.

• Advanced R by Hadley Wickham

  • Ths book is good for those who want to get a really in-depth understanding of how R operates.

If you want to dive deeper

The R community is very welcoming and inclusive. If you are feeling stuck, chances are someone has had the same issue before.

Here are some helpful resources and great groups to join:

• http://www.google.com
• http://www.r-bloggers.com
• http://www.stackoverflow.com
• https://rladies.org
• https://www.rstudio.com/resources/cheatsheets/ (list of cheatsheets)

Installing R and RStudio

• Download and install R from here https://cloud.r-project.org

• For Macs, you may have to download different versions:

  • Intel chip (R-4.1.1.pkg)

  • Apple Silicon M1 chip (R-4.1.1-arm64.pkg)

• Download RStudio here: https://www.rstudio.com/products/rstudio/download/

In this course we will be using R exclusively through RStudio.

Using RStudio

When you open RStudio it should look something like this:

Using RStudio

RStudio is an integrated development environment specifically developed for R, that lets you write code, run scripts, and view the results all in one.

• Source: This is the code editor, where you write and save your code.

• Console: This is where the output of your code will be printed.

• Environment/History: This is where any objects, such as vectors, matrices, or dataframes, will be stored.

• Viewer: This viewer previews any plots you create. You can also check your folder files and call for help here.

Tips for using RStudio

Always write code into the source code file, except for small checks and tests.

To execute the line of source code where the cursor currently resides you can press the Ctrl + Enter/Cmd + Enter key, rather than manually pressing the Run toolbar button.

Annotate your code using #

# Introduction to R
# 07.09.2021
# Use head(x, n = 2) to see the first two rows of a dataframe
head(mtcars, n = 2)
##               mpg cyl disp  hp drat    wt  qsec vs am gear carb
## Mazda RX4      21   6  160 110  3.9 2.620 16.46  0  1    4    4
## Mazda RX4 Wag  21   6  160 110  3.9 2.875 17.02  0  1    4    4

Send code from the source to the console

Using packages

R comes with a list of built in functions, but often you will want to use other functions written not by the original creators of R, but by other people.

If you want to use functions written by other people, you have to install it as a package.

To do this, we have to first install the package once, and then load it whenever we would like to use it.

# install the package (you only need to do this once)
install.packages('praise')

# load the package to use its functions
library(praise)
praise()
## [1] "You are awesome!"

Using packages

Sometimes you only need to use a very specific function of a package one time, and loading the entire package may seem unnecessary.

You can use package::function to call the function you are after. This tells R to only load the package for this specific chunk of code.

cowsay::say("Welcome to the course!", by = "cow")
## 
##  ----- 
## Welcome to the course! 
##  ------ 
##     \   ^__^ 
##      \  (oo)\ ________ 
##         (__)\         )\ /\ 
##              ||------w|
##              ||      ||

Feeling stuck?

You can use ? whenever you want to read the documentation of a particular command.

# how should I specify the mean and standard deviation of a normal distribution
?rnorm
# how does a histogram work
?hist
# how does the mean() function work
?mean

What we'll cover today

Introducing the Basics

1. Object types

2. Vectors

3. Missingness

4. Vector functions

5. Dataframes

6. Loading data

Objects and functions

Almost everything in R is either an object or a function.

• Object: number, vector, dataset, summary statistic, regression model, etc.

• Function: takes objects as arguments, does something, and returns an object.

# Create a vector object called height
height <- c(189, 178, 166, 178, 190)
# apply the mean() function to the object height
mean(height)
## [1] 180.2

→ The function mean() takes the object height, calculates the average, and returns a single number.

Objects and functions

When you use R, you will mostly:

1. Define objects

2. Apply functions to those objects

3. Repeat!

R as a calculator

3+5
## [1] 8
10/2
## [1] 5
sqrt(4)
## [1] 2

"Hello world!"
## [1] "Hello world!"
"1" + "3"
## Error in "1" + "3": non-numeric argument to binary operator

The <- operator

You can assign values to variables using the <- operator. You can then use the variable in subsequent operations.

x <- 9 + 11
x
## [1] 20
y <- x / 2
y
## [1] 10

greetings <- "Hello world!"
greetings
## [1] "Hello world!"

Vectors

We can create longer vectors by using c() (read: concatenate).

Numeric vectors

w <- 2 
y <- c(1, 2, 3)
z <- c(4, 5, 6)
z
## [1] 4 5 6

Character vectors

welcome <- c("Welcome", "to", "this", "course!")
welcome 
## [1] "Welcome" "to"      "this"    "course!"

Vectors

For longer vectors, writing out each element can be tedious. In addition to c(), there are other options.

Logical vectors

While numeric vectors can include any number and character values any character string, logical vectors can only take the values of either TRUE or FALSE.

Logical vectors are therefore often used to distinguish between two groups, or select a certain subset of variables.

In the example below, we create a logical vector that distinguishes between ages below and above the age of 18.

age <- c(14, 19, 23, 13, 16, 19, 18)
is_18 <- age >= 18
is_18
## [1] FALSE  TRUE  TRUE FALSE FALSE  TRUE  TRUE

The vector is_18 is TRUE when age is 18 or higher, and FALSE otherwise.

Logical operators

In the previous example we use >= to distinguish between ages below and above Some logical operators include:

Missing values

When we deal with data in the real world, there is often lots of missingness.

Missing values are denoted with NA.

NAs behave differently to other values.

num_vec <- c(5, NA, 15, 20, 25, NA)
num_vec / 5
## [1]  1 NA  3  4  5 NA
NA + 3
## [1] NA
c("hello", "my", "name", "is", NA)
## [1] "hello" "my"    "name"  "is"    NA

Vector functions

Length: Checks the length of a vector

x <- 2
y <- c(0.5, 45, 7, 45, 0.5)
z1 <- c("1 2 3 4 5 ", "6 7 8")
length(x)
## [1] 1
length(y)
## [1] 5
length(z)
## [1] 3

Sorting/unique: sorts or displays the unique values of a vector

sort(y)
## [1]  0.5  0.5  7.0 45.0 45.0
unique(y)
## [1]  0.5 45.0  7.0

Numeric vector functions

Character vector functions

If you look at the output of welcome, each word is included in separate quotation marks.

welcome
## [1] "Welcome" "to"      "this"    "course!"

This is because R treats each of these words as a separate element in a vector.

There are times when we might want to tell R to collapse the string into a single element. We can do this using the paste() function, specifying the option collapse = TRUE.

welcome2 <- paste(welcome, collapse = " ")
welcome2
## [1] "Welcome to this course!"
welcome3 <- paste(welcome, collapse = "")
welcome3
## [1] "Welcometothiscourse!"

Missing values

A lot of descriptive functions will throw up an error when there are missing values.

num_vec <- c(5, NA, 15, 20, 25, NA)
sum(num_vec)
## [1] NA
mean(num_vec)
## [1] NA

Descriptive functions include the argument na.rm = TRUE, which explicitly tells R to ignore missing values.

sum(num_vec, na.rm = TRUE)
## [1] 65
mean(num_vec, na.rm = TRUE)
## [1] 16.25

is.na is a logical operation that allows us to identify missing values.

is.na(num_vec)
## [1] FALSE  TRUE FALSE FALSE FALSE  TRUE
sum(num_vec, na.rm = TRUE)
## [1] 65

A note about vectors

Vectors can include either character values or numeric values, not both!

x <- rep(c(5, "a"), times = 2)
x
## [1] "5" "a" "5" "a"
x / 2
## Error in x/2: non-numeric argument to binary operator

With x_num, R automatically treats the vector as a character vector, because it includes some characters.

If we force R to treat x as a numeric vector, it will replace all non-numeric elements with NA.

as.numeric(x)
## Warning: NAs introduced by coercion
## [1]  5 NA  5 NA

Indexing vectors with []

Often we don't want to retrieve the whole vector, but only a specific element.

We can do this using [].

Numeric indexing

a[index], where a is the vector, and index is a vector of index values.

colors <- colors()
# What is the first color? 
colors[1]
## [1] "white"
# What are the first 5 colors
colors[1:3]
## [1] "white"        "aliceblue"    "antiquewhite"

Logical indexing

When indexing a vector with a logical index, R will only return values for which the index is TRUE.

years <- c(2010, 2005, 2012, 2013, 2001)
# select all years above 2010
years[years>2010]
## [1] 2012 2013
# select all years larger than 2002 and smaller than 2013
years[years > 2002 & years < 2013]
## [1] 2010 2005 2012

R actually interprets TRUE values as 1 and FALSE values as 0.

This allows us to quickly answer questions like:

#How many observations in years are greater than 2005?
sum(years > 2005)
## [1] 3
# What's the proportion of observations in years greater than 2005
mean(years > 2005)
## [1] 0.6

This is a really useful feature for quick calculations!

Changing values of a vector

In the example below, you know that the 4th value should have been 23, but was wrongly coded as NA.

age <- c(17, 21, 22, 25, NA)
age[5] <- 23
age 
## [1] 17 21 22 25 23
age[age >= 18] <- "18+"
age
## [1] "17"  "18+" "18+" "18+" "18+"

Dataframes

Most of the work we do as sociologists will involve playing around with rectangular data or dataframes.

While vectors are one dimensional, dataframes have two dimensions; rows and columns.

• columns: variables

• rows: observations

Dataframes

Load data

Rather than constructing your own dataframe in R, you will often work with data that already comes in a specific format, such as a csv, excel, .txt, Stata (.dta), or other file type.

R can load all these different file types.

# csv file 
data <- read.csv(data)
# To load stata files, you have to use the foreing package
# Note that you might need to install the package 
data <- foreign::read.dta(data)
# load an R data file 
load("data.Rda")

Inspect the data

Here are some of the most important ways to inspect a dataframe.

Inspect the data

# get the dimensions (rows and columns) of your data
dim(pop_df)
## [1] 4 3
# view the first 5 rows of data
head(pop_df)
##   country year pop_1m
## 1 Nigeria 2013  173.6
## 2  Gambia 2013    1.8
## 3 Finland 2013    5.4
## 4  Brazil 2013  200.4
# view the last 2 rows of data
tail(pop_df, 2)
##   country year pop_1m
## 3 Finland 2013    5.4
## 4  Brazil 2013  200.4

Inspect the data

# inspect the variable names in your dataframe
names(pop_df)
## [1] "country" "year"    "pop_1m"
# generate summary statistics for each variable in your dataframe. 
summary(pop_df)
##    country               year          pop_1m     
##  Length:4           Min.   :2013   Min.   :  1.8  
##  Class :character   1st Qu.:2013   1st Qu.:  4.5  
##  Mode  :character   Median :2013   Median : 89.5  
##                     Mean   :2013   Mean   : 95.3  
##                     3rd Qu.:2013   3rd Qu.:180.3  
##                     Max.   :2013   Max.   :200.4