R - Basics

Marcio Diniz | Michael Luu

Cedars Sinai Medical Center

06 September, 2022

Introduction

First steps in R

R is an object-oriented language;
Objects are boxes that we can use to store different classes of content;
Objects need names. Names should always start with letters and it also can contain numbers, but only a few special characters: underscore or period;

Types of objects

Objects can be classified in three classes of content: numeric, character, logical;
- Numeric objects contain numbers such that they can be double (dbl) or integers (int);
- Character (chr) objects contain strings, i.e., text between quotes;
  - If a character object contains only a pre-defined set of values, then it is considered a factor (fct).
- Logical objects (lgl) contain logical statements, i.e., TRUE or FALSE.

Let’s code our first objects:

# Who is it?
dwo_icecream <- 11 # days without icecream
favorite_food <- "risotto"
dog.lover = TRUE
enjoy_cold <- FALSE

dwo_icecream

[1] 11

favorite_food

[1] "risotto"

enjoy_cold

[1] FALSE

In order to store, we use the assignment operator <- or =.
Comments are preceded with #;
Objects can be accessed based on their names;
Objects are listed in the Environment tab.

Good practices for coding in R

It suggests to avoid . on names of objects and = as assignment operator:

dog_lover <- dog.lover

rm(dog.lover)
dog_lover

[1] TRUE

The function rm removes objects from the Environment tab.

Good practices for coding in R

Describe and comment your code for your future self!

Good practices for coding in R

Describe and comment your code for your colleagues!

Vectors

Vectors are stacked objects with different lengths, but the same type of objects
We can combine objects using the function c;
We calculate the length of a vector using the function length.

Let’s do it in R:

dwo_icecream <- c(5, 10, 8, 9, 3) # days without ice-cream
dwo_icecream

[1]  5 10  8  9  3

length(dwo_icecream)

[1] 5

All previous objects are also vectors, but with length 1:

length(favorite_food)

[1] 1

Combining vectors

If we combine different types of vectors, they will be converted following the hierarchy: character > numeric > logical:

# coerced all objects as characters
c(favorite_food, dog_lover, enjoy_cold, dwo_icecream)

[1] "risotto" "TRUE"    "FALSE"   "5"       "10"      "8"       "9"      
[8] "3"

c(dog_lover, dwo_icecream)  # TRUE is converted into 1

[1]  1  5 10  8  9  3

c(enjoy_cold, dwo_icecream)  # FALSE is converted into 0

[1]  0  5 10  8  9  3

Class of objects can be checked using the function class.

class(dwo_icecream)

[1] "numeric"

class(dog_lover)

[1] "logical"

class(favorite_food)

[1] "character"

Other functions to create vectors

Some numeric vectors can also be created with easier approaches than use the function c:
Using the function rep:

x <- c(1, 1, 1, 1, 1)
x

[1] 1 1 1 1 1

y <- rep(1, 5)
y

[1] 1 1 1 1 1

Using the function seq:

x <- c(1, 2, 3, 4, 5)
x

[1] 1 2 3 4 5

y <- 1:5
y

[1] 1 2 3 4 5

z <- seq(1, 5, by = 1)
z

[1] 1 2 3 4 5

Factors

Factors are special vectors that assume only pre-defined values defined using function factor:
It can be created based on a character vector:

dog_lover <- c("yes", "no", "yes", "yes", "yes")
dog_lover <- factor(dog_lover, 
                    levels = c("no", "yes"))
dog_lover

[1] yes no  yes yes yes
Levels: no yes

Or a numeric vector:

dog_lover <- c(1, 0, 1, 1, 1)
dog_lover <- factor(dog_lover,
                    levels = c(0, 1),
                    labels = c("no", "yes"))
dog_lover

[1] yes no  yes yes yes
Levels: no yes

Factors

Shortcut to transform a character vector into a factor.

dog_lover <- c("yes", "no", "yes", "yes", "yes")
dog_lover <- as.factor(dog_lover)
dog_lover

[1] yes no  yes yes yes
Levels: no yes

Factors have ordered levels:

dog_lover

[1] yes no  yes yes yes
Levels: no yes

levels(dog_lover)

[1] "no"  "yes"

nlevels(dog_lover)

[1] 2

Matrices

Matrix is a collection of vectors of the same type and length organized in a two-dimensional array;
A matrix has n rows and p columns.

Vectors can be bounded into two-dimensional arrays using the function cbind:

pushups <- c(1, 10, 0, 30, 25)

m <- cbind(dwo_icecream, pushups)
m

     dwo_icecream pushups
[1,]            5       1
[2,]           10      10
[3,]            8       0
[4,]            9      30
[5,]            3      25

The class of two-dimensional arrays generated by cbind is matrix:

class(m)

[1] "matrix" "array"

Matrices

When we bind vectors of different classes:

names <- c("m", "n", "p", "b", "t")
m <- cbind(names, dog_lover, 
           dwo_icecream, pushups)

     names dog_lover dwo_icecream pushups
[1,] "m"   "2"       "5"          "1"    
[2,] "n"   "1"       "10"         "10"   
[3,] "p"   "2"       "8"          "0"    
[4,] "b"   "2"       "9"          "30"   
[5,] "t"   "2"       "3"          "25"

Matrices are not appropriate to store a dataset.

Data frames

A data frame is a collection of vectors of the same/different types but of the same length organized in a two-dimensional array;
A data frame has n rows and p columns.

Data frames are created with the function data.frame:

df <- data.frame(names, dog_lover,
                 dwo_icecream, pushups)
df

  names dog_lover dwo_icecream pushups
1     m       yes            5       1
2     n        no           10      10
3     p       yes            8       0
4     b       yes            9      30
5     t       yes            3      25

class(df)

[1] "data.frame"

Subsetting

Vectors

Each vector component is located at a given position;
Components of a vector can be accessed using the operator position.

dwo_icecream

[1]  5 10  8  9  3

dwo_icecream[1]

[1] 5

dwo_icecream[c(1, 5)]

[1] 5 3

Matrices

Each matrix component is located at a given position;
Components of a matrix can be accessed using the operator [position row, position column].

We can access a specific component, a row and a column in a matrix:

# a specific component
m[1, 1]

names 
  "m"

# a row
m[5, ]

       names    dog_lover dwo_icecream      pushups 
         "t"          "2"          "3"         "25"

# a column
m[, 4]

[1] "1"  "10" "0"  "30" "25"

Data frames

Each data frame component is located at a given row and column;
Components of a data frame can be accessed similar to a matrix;
In addition, columns can be accessed be their names using the operator $.

We can access a specific component, a row and a column in a data frame:

df$names[5]

[1] "t"

df[5, 1]

[1] "t"

df$names

[1] "m" "n" "p" "b" "t"

df[, 1]

[1] "m" "n" "p" "b" "t"

df[5, ]

  names dog_lover dwo_icecream pushups
5     t       yes            3      25

Logical statements

Object classes

Logical statements about object’s class can be done as follow:

is.numeric(dwo_icecream)

[1] TRUE

is.logical(dwo_icecream)

[1] FALSE

is.character(dwo_icecream)

[1] FALSE

is.factor(dog_lover)

[1] TRUE

Conditions

Checking conditions can be done with logical statements:
Are these objects equal?

dog_lover == "yes"

[1]  TRUE FALSE  TRUE  TRUE  TRUE

Are these objects different?

dog_lover != "yes"

[1] FALSE  TRUE FALSE FALSE FALSE

Is this object greater (less) than 35?

dwo_icecream > 7 # for less, use <

[1] FALSE  TRUE  TRUE  TRUE FALSE

Is this object greater (less) or equal than 35?

dwo_icecream >= 7 # for less or equal, use <=

[1] FALSE  TRUE  TRUE  TRUE FALSE

Combining Conditions

Conditions can combined:
- AND

dwo_icecream > 7 & dog_lover == "yes"

[1] FALSE FALSE  TRUE  TRUE FALSE

dwo_icecream > 7 | dog_lover == "no"

[1] FALSE  TRUE  TRUE  TRUE FALSE

Functions

What is a function?

The function f takes objects from in X and transform them into objects in Y

A function takes inputs, make operations and gives back outputs;
Functions in R are organized in libraries;
Every function has a documentation that can be accessed with the operator ?.

First, let’s access the documentation of some of these functions:

?length
?c

Look at the tab Help to mode details of the function, including examples.

Functions

Let’s create our own function to convert Fahrenheit to Celsius.
How would we do this conversion manually?

# minimum temperature
temp_f <- 38
temp_c <- (temp_f - 32)*(5/9)
temp_c

[1] 3.333333

# range of temperatures
temp_f <- c(38, 105)
temp_c <- (temp_f - 32)*(5/9)
temp_c

[1]  3.333333 40.555556

How about with a function?

temp_f_c <- function(temp_f){

  temp_c <- (temp_f - 32)*(5/9)

  return(temp_c)
}

temp_f_c(38)

[1] 3.333333

temp_f_c(105)

[1] 40.55556

temp_f_c(c(38, 105))

[1]  3.333333 40.555556

For loops

How can we repeat tasks or perform them sequentially?

i <- 1
i

[1] 1

i <- i + 1
i

[1] 2

i <- i + 1
i

[1] 3

For loops

Control flow for iteration

for (i in 1:3){
  print(i)
}

[1] 1
[1] 2
[1] 3

Control flow for iteration

x <- rep(NA, 3)

for (i in 1:3){
  x[i] <- i
}
x

[1] 1 2 3