A short introduction to R

Juan C. Rocha
Stockholm Resilience Centre, Stockholm University

About Juan

Slides: juanrocha.se/presentations/R_Intro
Recommended browser: Chrome (for nicer fonts)

Learning objectives

What you will learn

What you won’t learn

• Statistics*
• Machine learning
• Advanced programming

You will get a soft intro to each topic, but each topic is worth a course on its own

Why R?

• Wrong question!…However:
• Open source, dynamic language & welcoming community of developers
• Many at SRC / environmental science community speak R
• R is a programming language with strong rooting in Statistics and widely used in the social and natural sciences

Why programming?

• Empower you to do different kinds of science
  • Accessing data
  • Applying state-of-the-art statistical techniques
  • Updating as new data becomes available
  • Communicating to others
• Helps structuring your thinking
• Reproducibility
• Laziness is a core value: don’t waste time in repetition

There are >350 programming languages (Valverde 2016)

How many languages do you speak?

R language

R is an object oriented programming languange. Learning R takes time and practice as with any other language. The more you use it, the better you become at speaking it. Do you have R already installed? Do you have RStudio? If not please follow:

• R: Comprehensive R Archive Network
• RStudio

Outline

• How does R look?
• R as calculator
• Basic objects and classes
• Assignments and comparisons
• Functions & Packages
• Graphics
• Getting help

Good looking R

• In basic R you will have the console and some scripts
• In RStudio you have:
  1. Console
  2. Scripts
  3. Environment
  4. Plots

RStudio

R as calculator

You can use R to calculate simple and complex operations:

4 + 5

[1] 9

You can store the results of any operation by assigning it to an object. The assignment operator <- does it for you

a <- (35 * 43)/3
a

[1] 501.6667

Some times people uses = instead of <- for doing assignments

b = 45/9  # But <- is preferred
b

[1] 5

Note you can make comments on you code with #

Objects & Classes

One can think of a vector as a line of objects of the same kind

a <- c(1, 3, 5, 8)  # numeric vector

# Let's check out
class(a)

[1] "numeric"

vector: 1d object

b <- c("Celinda", "Juan", "Miriam")  # character vector
c <- c(TRUE, TRUE, FALSE)  # logical vector

# Check
class(b)

[1] "character"

class(c)

[1] "logical"

Vectors indexing

You can access the value of any element of your vectors by referring to their position using []

a[2]  # second element of vector a

[1] 3

a[2:4]  # you can use : to make sequence of numbers, here 2 to 4

[1] 3 5 8

b[c]  # you can pass an indexing vector to another vector!

[1] "Celinda" "Juan"   

Unlike other languages, R indexing starts at 1

matrix: 2d objects

m <- matrix(1:12, nrow = 3, ncol = 4, byrow = TRUE)
m

     [,1] [,2] [,3] [,4]
[1,]    1    2    3    4
[2,]    5    6    7    8
[3,]    9   10   11   12

dim(m)  # How big is it?

[1] 3 4

m[3, 4]  # Indexing

[1] 12

array: multi-dim objects

z <- array(1:12, dim = c(2, 2, 3))
z

, , 1

     [,1] [,2]
[1,]    1    3
[2,]    2    4

, , 2

     [,1] [,2]
[1,]    5    7
[2,]    6    8

, , 3

     [,1] [,2]
[1,]    9   11
[2,]   10   12

list: mixed objects

Mix of objects, don’t need to be the same class nor length

l <- list(a, b, c, m)  # unnamed list
l[[1]]

[1] 1 3 5 8

l <- list(a = a, b = b, c = c, m = m)  # named list
l$m  # you can use $ to access elements by name

     [,1] [,2] [,3] [,4]
[1,]    1    2    3    4
[2,]    5    6    7    8
[3,]    9   10   11   12

To access the elements of a list you should use [[]] by position, or $ by name

list: mixed objects

class(l)

[1] "list"

class(l[[4]])

[1] "matrix" "array" 

length(l)

[1] 4

If lists are mixed objects, can you put a list within a list? – Yes!

Data frames

Data frames are one of the most used objects in R. It can have different types of data

dat <- data.frame(books = a[2:4], student = b, pass = c)
dat

  books student  pass
1     3 Celinda  TRUE
2     5    Juan  TRUE
3     8  Miriam FALSE

str(dat)  # revels the structure of your data

'data.frame':   3 obs. of  3 variables:
 $ books  : num  3 5 8
 $ student: chr  "Celinda" "Juan" "Miriam"
 $ pass   : logi  TRUE TRUE FALSE

summary()

• str() and summary() are very useful functions to understand how your data looks like
  
• head() gives you the first 6 rows of your dataset and
  
• tail(n = 3) the last 3 ones.

summary(dat)

     books         student             pass        
 Min.   :3.000   Length:3           Mode :logical  
 1st Qu.:4.000   Class :character   FALSE:1        
 Median :5.000   Mode  :character   TRUE :2        
 Mean   :5.333                                     
 3rd Qu.:6.500                                     
 Max.   :8.000                                     

Tricks for later…

tibble::as_tibble(gapminder::gapminder)

# A tibble: 1,704 × 6
   country     continent  year lifeExp      pop gdpPercap
   <fct>       <fct>     <int>   <dbl>    <int>     <dbl>
 1 Afghanistan Asia       1952    28.8  8425333      779.
 2 Afghanistan Asia       1957    30.3  9240934      821.
 3 Afghanistan Asia       1962    32.0 10267083      853.
 4 Afghanistan Asia       1967    34.0 11537966      836.
 5 Afghanistan Asia       1972    36.1 13079460      740.
 6 Afghanistan Asia       1977    38.4 14880372      786.
 7 Afghanistan Asia       1982    39.9 12881816      978.
 8 Afghanistan Asia       1987    40.8 13867957      852.
 9 Afghanistan Asia       1992    41.7 16317921      649.
10 Afghanistan Asia       1997    41.8 22227415      635.
# … with 1,694 more rows

skimr::skim(dat)  # Results not shown

names(dat)

[1] "books"   "student" "pass"   

Indexing data frames

dat[2, 3]

[1] TRUE

dat

  books student  pass
1     3 Celinda  TRUE
2     5    Juan  TRUE
3     8  Miriam FALSE

dat$books

[1] 3 5 8

dat[, c("books", "pass")]

  books  pass
1     3  TRUE
2     5  TRUE
3     8 FALSE

Comparison

m  # our matrix with 1:12

     [,1] [,2] [,3] [,4]
[1,]    1    2    3    4
[2,]    5    6    7    8
[3,]    9   10   11   12

a  # our vector

[1] 1 3 5 8

m[1, ] > a  # is it greater than

[1] FALSE FALSE FALSE FALSE

m[2, ] == a  # is it equal to

[1] FALSE FALSE FALSE  TRUE

m[2, ] <= a  # is it less or equal than

[1] FALSE FALSE FALSE  TRUE

m[3, ] != a  # is not equal to

[1] TRUE TRUE TRUE TRUE

Comparison

x <- stats::rnorm(5)
x < 1

[1]  TRUE  TRUE FALSE  TRUE  TRUE

x[x > 0]

[1] 0.9917544 1.3410942

x1 <- 0.5 - 0.3
x2 <- 0.3 - 0.1
x1

[1] 0.2

x2

[1] 0.2

x1 == x2  # FALSE on most machines

[1] FALSE

isTRUE(all.equal(x1, x2))  # TRUE everywhere

[1] TRUE

Assignment & the pipe

l[[5]] <- l  # A list inside a list
length(l)

[1] 5

class(l)

[1] "list"

x <- mean(log(sqrt(a)))

y <- a |>
    sqrt() |>
    log() |>
    mean()
x == y

[1] TRUE

Functions

Where the magic starts…

What is magic?

Discuss in pairs for

`

`

Functions

Everything in R is a function call - John Chambers

# You can see how a function looks by calling it without ()
matrix

function (data = NA, nrow = 1, ncol = 1, byrow = FALSE, dimnames = NULL) 
{
    if (is.object(data) || !is.atomic(data)) 
        data <- as.vector(data)
    .Internal(matrix(data, nrow, ncol, byrow, dimnames, missing(nrow), 
        missing(ncol)))
}
<bytecode: 0x7ff0030aa320>
<environment: namespace:base>

• Information in () are arguments, the values above are the default values, e.g. byrow = FALSE
• <environment: namespace:base> means that matrix is a function of the package base

Functions & Packages 📦

We will see more about functions and how to write them in the future. For now, it’s good to know that functions are procedures or routines that were written by other R users and programmers to make your life simple (most of the times!). R comes with a number of basic funcitons by default, such as mean() or matrix(). Functions that work together for certain purposes (e.g. network analysis, linear regression) come in packages that you need to download, install and call every time you want to use them.

# install.packages ('network') # Intall a new package
library(network)  # Call it into your session

'network' 1.17.1 (2021-06-12), part of the Statnet Project
* 'news(package="network")' for changes since last version
* 'citation("network")' for citation information
* 'https://statnet.org' for help, support, and other information

Functions & Packages 📦

# Now we can build a network
m <- matrix(rbinom(25^2, 1, 0.1), 25, 25)
net <- network(m)
net

 Network attributes:
  vertices = 25 
  directed = TRUE 
  hyper = FALSE 
  loops = FALSE 
  multiple = FALSE 
  bipartite = FALSE 
  total edges= 60 
    missing edges= 0 
    non-missing edges= 60 

 Vertex attribute names: 
    vertex.names 

No edge attributes

plot(net)

Packages 📦

Currently CRAN package repository features 19021 available packages.

Graphics

plot(pressure)

Plots examples: base

Plots examples: ggplot

Plots examples: circlos

Plots examples: networks

Plots examples: ggplot or heatmaps

Getting help

Lerning R is an iterative process between doing something, getting stuck, asking for help, solving your problem, and keep going.

help("mean")

You can also type ?mean or ??mean

Getting help

All languages have dialects

stairs | escalators, lift | elevators, trucks | lorries, pants | trousers

dat[dat$books < 5 & dat$pass == TRUE, ]

  books student pass
1     3 Celinda TRUE

dat |>
    dplyr::filter(pass == TRUE & books < 5)

  books student pass
1     3 Celinda TRUE

Be patient, diversity requires open mind but is good: redundancy.

Resources

• Advanced R by Hadley Wickham
• Hands-On Programming with R by Garrett Grolemund
• R for Data Science by Wickham and Grolemund
• Big book of R by Oscar Baruffa

Exercise