Outline

Introduction

• Who has used R before ?

Introduction

• Who has used R before ?

• What other programming language have you used before ?

Introduction

• Who has used R before ?

• What other programming language have you used before ?

• For those who are experts in R

Introduction

• Who has used R before ?

• What other programming language have you used before ?

• For those who are experts in R

  • please refrain to answer during this session...
  • help your neighbor...

Introduction

• Who has used R before ?

• What other programming language have you used before ?

• For those who are experts in R

  • please refrain to answer during this session...
  • help your neighbor...

• Two special slide formatting

Your turn...

Introduction

• Who has used R before ?

• What other programming language have you used before ?

• For those who are experts in R

  • please refrain to answer during this session...
  • help your neighbor...

• Two special slide formatting

Your turn...

Warning

Introduction

Computer languages

Introduction

History of R

• Mid 1970s - S Language for Statistical Computing conceived by John Chambers, Rick Becker, Trevor Hastie, Allan Wilks and others at Bell Labs

• Early 1990's - R was first implemented in the early 1990’s by Robert Gentleman and Ross Ihaka, both faculty members at the University of Auckland.

• 1995 - Open Source Project

• 1997 - Managed by the R Core Group

• 2000 - First release of R

• 2011 - First release of R studio

• Historical notes - Paper from 1998

Introduction

Why use R ?

• Script vs. Menu driven software (e.g. Excel)
  • Can be re-rerun with new data
  • Reproducible workflow

Introduction

Why use R ?

• Script vs. Menu driven software (e.g. Excel)
  • Can be re-rerun with new data
  • Reproducible workflow
• Open source
  • Huge number of libraries
  • Tidy "universe" : tidyverse and ggplot2
    • Very easy to manipulate tables (select columns, create new variables)
    • High quality graphics

Introduction

Why use R ?

• Script vs. Menu driven software (e.g. Excel)
  • Can be re-rerun with new data
  • Reproducible workflow
• Open source
  • Huge number of libraries
  • Tidy "universe" : tidyverse and ggplot2
    • Very easy to manipulate tables (select columns, create new variables)
    • High quality graphics
• Work environment
  • R studio

Introduction

Why use R ?

• Script vs. Menu driven software (e.g. Excel)
  • Can be re-rerun with new data
  • Reproducible workflow
• Open source
  • Huge number of libraries
  • Tidy "universe" : tidyverse and ggplot2
    • Very easy to manipulate tables (select columns, create new variables)
    • High quality graphics
• Work environment
  • R studio
• Document your data processing
  • R markdown
  • Create HTML, pdf, presentations

Introduction

Why use R ?

• Script vs. Menu driven software (e.g. Excel)
  • Can be re-rerun with new data
  • Reproducible workflow
• Open source
  • Huge number of libraries
  • Tidy "universe" : tidyverse and ggplot2
    • Very easy to manipulate tables (select columns, create new variables)
    • High quality graphics
• Work environment
  • R studio
• Document your data processing
  • R markdown
  • Create HTML, pdf, presentations
• Share your data and workflow
  • GitHub

Introduction

What can you do with R ?

Introduction

What can you do with R ?

Introduction

What can you do with R ?

Introduction

What can you do with R ?

Resources

Books and Manuals

• R intro : Very good introduction to R, short and clear
• R in a nutshell : Many many receipes to solve all your questions
• R graphics cook book : very good for graphics

Resources

On line courses and web sites

• Coursera
• Pluralsight
• Quick-R, very simple

Resources

Cheat sheets

• R basics
• ggplot2
• dplyr

Resources

Forum

• https://stackoverflow.com/
• http://r-statistics.co/
• https://stats.stackexchange.com/
• https://www.r-bloggers.com/

Let's get started

Setup

• Install R

• Install R studio

Let's get started

The R studio interface

Let's get started

Create a new project

• Open R studio
• Create new project for the course in a new directory
  • e.g. Data Science Class

Let's get started

Your first script

print("Hello world")

[1] "Hello world"

Two ways to proceed

1. Type directly in command window

Let's get started

Your first script

print("Hello world")

[1] "Hello world"

Two ways to proceed

1. Type directly in command window

2. Create a new script

Type in script window, select and execute (CTRL-R)

The R language

Everything in R is an object

• Assignement done with <-

> x <- 1
> y <- 2
> x + y

[1] 3

The R language

Everything in R is an object

• Assignement done with <-

> x <- 1
> y <- 2
> x + y

[1] 3

> z <- x + y
> z

[1] 3

The R language

= can be used instead of <- but refrain from it (not good style)

> z = x + y

The R language

= can be used instead of <- but refrain from it (not good style)

> z = x + y

You can view the values of the objects in R-studio environment window (top-right)

The R language

R is case sensitive

> Z

The R language

R is case sensitive

> Z

> Z

Error in eval(expr, envir, enclos): object 'Z' not found

The R language

Rules for naming objects

• Use
  • letters
  • numbers
  • the dot
  • the underscore (not the minus sign !)

• Start always with a letter
  • Myvariable, Myvariable1, Myvariable.1,Myvariable-01 are OK
  • 1Myvariable, My-variable, Myvariable@ are not OK

The R language

Use consistent naming

Five conventions

• alllowercase: e.g. adjustcolor
• period.separated: e.g. plot.new
• underscore_separated: e.g. numeric_version
• lowerCamelCase: e.g. addTaskCallback
• UpperCamelCase: e.g. SignatureMethod

Prefer third one, much more easy to read

• Use names for objects : last_name
• Use verbs for function : build_name
• Think about best order
  • e.g. prefer maybe name_last because then you can have name_first, name_full...
  • and you identify that all these objects are related to a name...

R objects

Data types

• character: "Daniel", "This is a course in R", 'Donald'

• numeric: 2, 15.5, 10e-3

• integer: 2L (the L tells R to store this as an integer)

• date: 2018-02-25

• logical: TRUE, FALSE

• complex: 1+4i (complex numbers with real and imaginary parts)

R objects

Data types

• character: "Daniel", "This is a course in R", 'Donald'

• numeric: 2, 15.5, 10e-3

• integer: 2L (the L tells R to store this as an integer)

• date: 2018-02-25

• logical: TRUE, FALSE

• complex: 1+4i (complex numbers with real and imaginary parts)

• No data "NA"

• Not a number "NaN" (e.g. division by zero)

R objects

Data structures

• Vector

• List

• Matrix

• Data frames

• Function

Vectors

The basic R structure is a vector: $$\begin{bmatrix}10 \\ 20 \\ 30 \end{bmatrix}$$

Vectors

The basic R structure is a vector: $$\begin{bmatrix}10 \\ 20 \\ 30 \end{bmatrix}$$

A vector can with a single element only $$\begin{bmatrix}10 \end{bmatrix}$$

Vectors

The basic R structure is a vector: $$\begin{bmatrix}10 \\ 20 \\ 30 \end{bmatrix}$$

A vector can with a single element only $$\begin{bmatrix}10 \end{bmatrix}$$

Assign a value to a vector

x <- 10
x

[1] 10

Vectors

Assign several elements

x <- c(10, 20, 30)
x

[1] 10 20 30

Vectors

Assign several elements

x <- c(10, 20, 30)
x

[1] 10 20 30

Assign range

x <- 10:30
x

 [1] 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Vectors

Assign characters

PoTU <- c("Donald", "Trump")
PoTU

[1] "Donald" "Trump"

Assign logical

flags <- c(TRUE, FALSE, TRUE)
flags

[1]  TRUE FALSE  TRUE

Vectors

Access specific elements of a vector

First

x[1]

[1] 10

Vectors

Access specific elements of a vector

First

x[1]

[1] 10

Range

x[1:5]

[1] 10 11 12 13 14

Vectors

Access specific elements of a vector

First

x[1]

[1] 10

Range

x[1:5]

[1] 10 11 12 13 14

Remove one element

x[-1]

 [1] 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Vectors

Determine object properties

Apply functions (we will come back to functions latter)

• typeof() - what is the object’s data type (low-level)?
• length() - how long is it? What about two dimensional objects?

typeof(x)
length(x)

Vectors

Determine object properties

Apply functions (we will come back to functions latter)

• typeof() - what is the object’s data type (low-level)?
• length() - how long is it? What about two dimensional objects?

typeof(x)
length(x)

[1] "integer"

[1] 21

Vectors

Determine object properties

Apply functions (we will come back to functions latter)

• typeof() - what is the object’s data type (low-level)?
• length() - how long is it? What about two dimensional objects?

typeof(x)
length(x)

[1] "integer"

[1] 21

Operators

Arithmetic Operators

Operators

Arithmetic Operators

We are performing vector operations !

$$\begin{bmatrix} 1\\2\\3\\..\end{bmatrix}+\begin{bmatrix}1\\2\\3\\..\end{bmatrix}=\begin{bmatrix}2\\4\\6\\..\end{bmatrix}$$

Operators

Arithmetic Operators

Vector one element

 x <- 1
 y <- 2
 z <- x + y
 z

[1] 3

Operators

Arithmetic Operators

Vector several elements

# Two instructions on the same line
 x <- 1:9;  y <- 1:9
 z <- x + y
 z

[1]  2  4  6  8 10 12 14 16 18

Operators

Arithmetic Operators

Vector several elements

# Two instructions on the same line
 x <- 1:9;  y <- 1:9
 z <- x + y
 z

[1]  2  4  6  8 10 12 14 16 18

Operators

Arithmetic Operators

Vector several elements

# Two instructions on the same line
 x <- 1:9;  y <- 1:9
 z <- x + y
 z

[1]  2  4  6  8 10 12 14 16 18

Operators

Arithmetic Operators

What happens when the vectors have different number of elements ?

x <- 1:9
y <- 1
z <- x + y
z

Operators

Arithmetic Operators

What happens when the vectors have different number of elements ?

x <- 1:9
y <- 1
z <- x + y
z

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

Operators

Arithmetic Operators

What happens when the vectors have different number of elements ?

x <- 1:9
y <- 1
z <- x + y
z

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

Equivalent to

y <- c(1, 1, 1, 1, 1, 1, 1, 1, 1)

The recycling rule...

Operators

Can we add logical ?

x <- TRUE
y <- FALSE
z <- x + y
z

Operators

Can we add logical ?

x <- TRUE
y <- FALSE
z <- x + y
z

[1] 1

Operators

Can we add logical ?

It does not give an error but...

The resulting variable is transformed to a numeric

Operators

Can we add logical ?

It does not give an error but...

The resulting variable is transformed to a numeric

typeof(x)

[1] "logical"

typeof(z)

[1] "integer"

Operators

Logical Operators

Operators

Logical Operators

x <- TRUE
y <- FALSE
z1 <- x | y
z2 <- x == y

Operators

Logical Operators

x <- TRUE
y <- FALSE
z1 <- x | y
z2 <- x == y

[1] TRUE

[1] FALSE

Operators

Can we add characters ?

first <- "Donald"
last <- "Trump"
full <- first + last

Operators

Can we add characters ?

first <- "Donald"
last <- "Trump"
full <- first + last

Generates an error

Error in first + last: non-numeric argument to binary operator

Operators

Can we add characters ?

first <- "Donald"
last <- "Trump"
full <- first + last

Generates an error

Error in first + last: non-numeric argument to binary operator

Functions

Function perform specific task on objects

• e.g. to concatanate strings we use paste0()

Functions

Function perform specific task on objects

• e.g. to concatanate strings we use paste0()

paste0(first, last)

[1] "DonaldTrump"

Functions

Function perform specific task on objects

• e.g. to concatanate strings we use paste0()

paste0(first, last)

[1] "DonaldTrump"

• Functions take arguments and return an object called result

• To know the arguments use ?

  ? paste0()  # Do not forget the parenthesis

Functions

Function perform specific task on objects

• e.g. to concatanate strings we use paste0()

paste0(first, last)

[1] "DonaldTrump"

• Functions take arguments and return an object called result

• To know the arguments use ?

  ? paste0()  # Do not forget the parenthesis

Functions

Function perform specific task on objects

• e.g. to concatanate strings we use paste0()

paste0(first, last)

[1] "DonaldTrump"

• Functions take arguments and return an object called result

• To know the arguments use ?

  ? paste0()  # Do not forget the parenthesis

• Can go directly to Help panel and type function name

Functions

Help

Functions

Help

Functions

Getting what you want

We would like to write "Donald Trump" but we have :

paste0(first, last)

[1] "DonaldTrump"

Functions

Getting what you want

We would like to write "Donald Trump" but we have :

paste0(first, last)

[1] "DonaldTrump"

paste(first, last)

[1] "Donald Trump"

Functions

Write your own function

my_sum <- function(a, b) {
    c <- a + b
    return(c)
}

Functions

Write your own function

my_sum <- function(a, b) {
    c <- a + b
    return(c)
}

my_sum(10, 20)

[1] 30

If you write 3 times the same piece of code write a function...

End of lecture one

Functions

Examples of functions

Most of the time you do not have to write functions because someone has already written one for what you want to do...

• Sum

x <- 1:100
sum(x)

[1] 5050

Functions

Examples of functions

Most of the time you do not have to write functions because someone has already written one for what you want to do...

• Sum

x <- 1:100
sum(x)

[1] 5050

• Normal distribution

y <- rnorm(100, mean = 0, sd = 1)
y[1:10]

 [1]  0.4885882 -0.6260146 -0.8855401 -1.2341267  0.3726551  0.8956950
 [7]  0.9124247  0.1755346  0.4628793 -1.5012981

Functions

Statistics

mean(y)

[1] 0.01007783

sd(y)

[1] 0.8875528

Functions

Statistics

mean(y)

[1] 0.01007783

sd(y)

[1] 0.8875528

Functions

Sample more points... 10,000 instead of 100

y <- rnorm(10000, mean = 0, sd = 1)
mean(y)

[1] 0.01251073

sd(y)

[1] 1.009886

Functions

Plot

library(graphics)
hist(y)

Packages

Packages are set of functions that have a common goal

They are really the strength of R

And these are only the "official"" packages. You can find more on GitHub

Packages

Installing a package

Download on your computer the package you need

Packages

Using a package

To use functions from the package

• use the syntax package::function

stringr::str_c(first, last, sep = " ")

[1] "Donald Trump"

Packages

Using a package

To use functions from the package

• use the syntax package::function

stringr::str_c(first, last, sep = " ")

[1] "Donald Trump"

• load the package with the library function

library(stringr)
str_c(first, last, sep = " ")

[1] "Donald Trump"

Packages

Using a package

To use functions from the package

• use the syntax package::function

stringr::str_c(first, last, sep = " ")

[1] "Donald Trump"

• load the package with the library function

library(stringr)
str_c(first, last, sep = " ")

[1] "Donald Trump"

Sometimes functions from different libraries have similar names

Packages

List installed packages

Other objects

• List

• Matrix

• Factors

• Data frames

Data frames

What is it ?

• Table mixing different types of columns (an Excel table...)
• However within a column all values are similar, e.g. numeric, logical, character

Data frames

What is it ?

• Table mixing different types of columns (an Excel table...)
• However within a column all values are similar, e.g. numeric, logical, character

df <- data.frame(label = letters[1:6], id = 1:6, value = rnorm(6, mean = 0, 
    sd = 1), flag = c(TRUE, FALSE), stringsAsFactors = FALSE)
df

  label id      value  flag
1     a  1  0.8002749  TRUE
2     b  2 -0.1723698 FALSE
3     c  3  1.0188527  TRUE
4     d  4 -1.4748408 FALSE
5     e  5  0.5381787  TRUE
6     f  6  0.8350807 FALSE

Data frames

Useful functions

dim(df)  # returns the dimensions of data frame

[1] 6 4

nrow(df)  # number of rows

[1] 6

ncol(df)  # number of columns

[1] 4

Data frames

Useful functions

str(df)  # structure of data frame - name, type and preview of data in each column

'data.frame':    6 obs. of  4 variables:
 $ label: chr  "a" "b" "c" "d" ...
 $ id   : int  1 2 3 4 5 6
 $ value: num  0.8 -0.172 1.019 -1.475 0.538 ...
 $ flag : logi  TRUE FALSE TRUE FALSE TRUE FALSE

colnames(df)  # columns names

[1] "label" "id"    "value" "flag"

Data frames

Access specific column

• Use $notation

df$value

[1]  0.8002749 -0.1723698  1.0188527 -1.4748408  0.5381787  0.8350807

Data frames

Access specific column

• Use $notation

df$value

[1]  0.8002749 -0.1723698  1.0188527 -1.4748408  0.5381787  0.8350807

• Use the df[i,j] notation

df[, 3]

[1]  0.8002749 -0.1723698  1.0188527 -1.4748408  0.5381787  0.8350807

df[, "value"]

[1]  0.8002749 -0.1723698  1.0188527 -1.4748408  0.5381787  0.8350807

Data frames

Access specific value

• Use $notation

df$label[5]

[1] "e"

df$label[1:5]

[1] "a" "b" "c" "d" "e"

Data frames

Access specific value

• Use $notation

df$label[5]

[1] "e"

df$label[1:5]

[1] "a" "b" "c" "d" "e"

• Use the df[i,j] notation

df[5, 1]

[1] "e"

df[1:5, "value"]

[1]  0.8002749 -0.1723698  1.0188527 -1.4748408  0.5381787

Data frames

Filter the data

df[df$id <= 3, ]

  label id      value  flag
1     a  1  0.8002749  TRUE
2     b  2 -0.1723698 FALSE
3     c  3  1.0188527  TRUE

Select lines for which the label is c

Data frames

Filter the data

df[df$id <= 3, ]

  label id      value  flag
1     a  1  0.8002749  TRUE
2     b  2 -0.1723698 FALSE
3     c  3  1.0188527  TRUE

Select lines for which the label is c

df[df$label == 3, ]

[1] label id    value flag 
<0 lignes> (ou 'row.names' de longueur nulle)

Next time: Markdown