Introduction

to R
CHAPTER I. Course Overview and
Preliminary Steps
Learning objectives

• Know what is R and how it works

• Learn basics of working with data in R
• Get familiar with basic commands/functions
• Learn how to do basic analysis on any dataset and be able to
create basic charts
What is R & Why we use it

• It’s a tool : Open-Source, cross platform, free programming

language designed to build statistical solutions
• Powerful : Gives access to CRAN repository containing over
10,000 packages with pre-defined functions for almost every
purpose
• Stays Relevant : Constantly being updated by users ( Scientists,
Statisticians, Researchers, Students!)
• More: Makes beautiful graphs, can create custom functions or
modify existing ones, can be integrated into many environments
and platforms such as Hadoop etc
Installing R
• Can be downloaded for free from
http://www.r-project.org/
• Download the version compatible with your OS
• Simple/Standard installation process
Installing R -Studio
• Can be downloaded for free from:
• https://www.rstudio.com/products/rstudio/download/
• Download the free version compatible with your OS
• R needs to be installed before installing R- Studio
R-Studio UI
Write your Global Environment-
code here See your datasets here

See your files,

graphs, help
documentation
and installed
packages here

Console - see
your code run
here
R Commands
• Assignments E.g.: x = 1, or x <- 1
• Functions E.g.: print(“Hello World”)
• Computations E.g.: 17 + 3 ; x + 5
• Mix E.g.: y = sqrt(16); y = 15 + 5

• Assignment queries will update objects in your

R environment
• Queries without assignment, as well as ‘call’ of
R objects will either generate an output in the
console, or in the plot tab
CHAPTER II. R Basics: Data Types
Variable Assignment in R
• A basic construct in programming is "variable"
• A variable allows you to store a piece of data (‘datum’, e.g.
6, ‘Hello’, etc.. ) or several pieces of data of a common
type, and assign them a unique name
• You can then later ‘call’ this variable's name to easily access
the value(s) that is/are stored within this variable.

Careful, R is case sensitive: The variables ‘x’ and ‘X’ can coexist
in R environment and have different values.
Basic data types in R

• R works with numerous data types. The most common

types are:
• Decimals values like 3.5, called 'numeric'
• Natural numbers like 3 are called 'integers'. Integers are
also numeric
• Boolean variables (TRUE or FALSE) are
classified as ‘logical’
• Text (or string) values are classified as
'character’
Basic data types in R

• Categorical variables are called ‘factors’. They

have a finite and defined set of values they can
take (e.g. eye_color can take have a value
contained in {‘blue’, ‘green’, ’brown’, ‘black’})
• Other variables can contain time data such as
dates, day of the week, hours, minutes, etc..
CHAPTER III. R Basics: Data
Structures
R Objects: Vectors

• To assign multiple values to a variable, we can use an R

object called a ‘vector’
• A vector is a sequence/collection of data elements of the
same basic type. Members in a vector are officially called
components. For Example: my_vector = c(14,26,38,30)
• To access a specific element in the vector, we simply need
to call variable_name[i], ‘i’ being the element’s position in
the vector. For example: vect[3] would return 38
R Objects: Matrices

• A matrix is a sequence/collection of data elements of the

same basic type arranged in a two-dimensional
rectangular layout.

• Being a 2-dimensional object, in order to obtain a specific

value within the matrix, 2 coordinates needs to be
entered. For example: my_matrix[i,j] would return the
element on the ith row, in the jth column

• my_matrix[i,] would return the entire ith row

• my_matrix[,j] would return the entire jth column

R Objects: Data Frames
• A data frame is used for storing data tables. It is a list of
vectors of equal length. Unlike matrices, it can gather
vectors containing different basic types

• Selection of specific elements in data frames works the

same way as for matrices. For example: my_dataframe[i,j]
would return the element on the ith row, in the jth
column
R Objects: Lists

• A list in R allows you to gather a variety of objects under

one name (that is, the name of the list) in an ordered way.
These objects can be matrices, vectors, data frames, even
other lists, etc. It is not even required that these objects
are related to each other.
• To access the ith object in the list, write my_list[[i]]
• If you want to access a variable in the ith object in the list,
write my_list[[i]] [variable coordinates]. See examples in R
CHAPTER IV. Importing Packages
and Datasets, Viewing Data
R: Packages
• R Packages are collections of R functions and data sets
• Some standard ones come with R installation
• Others can be installed in a few clicks in Rstudio, or
using install.packages(“package name”) function.
You can choose the CRAN Mirror closest to your
location, but the default Rstudio is consistently good
all over the world.
• Some have to be downloaded ( from
http://cran.r-project.org/, or through Google and
manually installed
• Once installed we need to call the package in when
needed using “library(“package name”)”
R: Importing Data

• More often than not, data is already available in

different formats ready to be imported to R.

• R accepts files of many formats, we will learn

importing files of the following formats:
• Text (.txt)
• CSV (.csv)
• Excel (.xls)
R: Importing Data

• Text files: use read.table() for space separated files,

comma separated files etc..
• CSV files: use read_csv() from readr package (used by
Rstudio interface)
• Excel files: use read_excel() from readxl package (used
by Rstudio interface)

See Rstudio examples to set Working Directory and import

different datasets
R: Importing Data

• For more formats (such as SPSS, SAS, STATA files etc…)

you can visit
http://cran.rproject.org/doc/manuals/R-data.pdf , here
you get information on how to import image files as
well !
Data Views
There are several ways to look at a data set:
• First, you can simply look at it entirely by double clicking
on it in the Global Environment, or by using
View(data_name) function
• You can look a specific column by calling it. E.g. data-
name$column_name
• Else, you can look at the first k rows, or the last k rows by
using head(data_name, k) or tail(data_name, k)
respectively
Data Overviews

You can also use functions to have a quick overview of the

data set you are working with:

• Try to use summary(data_name)

• You can also use str(data_name)

CHAPTER V. Data Manipulations
Filtering/Subsetting

• Use a Logical Operator

• ==, >, <, <=, >=, != are all logical operators.
• Note that the “equals” logical operator is two "==" signs, as one
"= " only is reserved for assignment.
• Result is a Logical variable
• To filter out rows in a dataset, place logic condition(s) in the
dataset’s squared brackets, before the coma
• You can filter using several conditions and separate them with
logical operators “|” (OR) and/or “&” (AND)

• See examples in Rstudio

Binding

• Binding columns: If 2 datasets, a dataset and a vector, or

2 vectors have the same number of values (rows in the
case of datasets), they can be placed together into one
same dataset using cbind()
• This is different from « merging » (see later chapter),
hence there is no row matching system: rows need to be
in the exact same order for the data to make sense.
• See example in Rstudio
• Binding rows: If 2 datasets have the same columns
(order, data types, names), one can be appended under
the other using rbind()
• See example in Rstudio
Transforming

• You can create new columns or modify existing ones by

applying transformations to them

• Transformations can be adding, subtracting,

multiplying, dividing, powering etc..
• But it can also be using functions such as log(), exp()
etc..

• See examples in R studio

Sorting

• In R, you can sort your dataset’s rows based on a

column’s alphabetical order (character variables), or
numerical order (numeric variables)
• You can apply an ascending or descending direction to
this order

• See examples in R studio

CHAPTER VI. Joins, Summary
Tables and Data Export
 Joins
• Joining consists in combining 2 or more datasets’ rows based
on a common column/field between them
• For a join to happen, 2 datasets need at least one same
column. It matches rows that have identical values in this
column.
Table 1 Table 2
Column A Column B Column B Column C
• Eg. A1
A2
B1
B1
B2
B1
C1
C2
A3 B2 B2 C3

Joined Tables
Column A Column B Column C
A1 B1 C2
A2 B1 C2
A3 B2 C1
A3 B2 C3

• Note: It is not like what the cbind() function does: cbind() fuses datasets by pasting
them one next to the other, regardless of what is in the data
Joins
• There are different types of joins :
Summary Tables
• Contingency tables: Use table(cat_var1,cat_var2)
(where cat_var1 and cat_var2 are categorical
variables) to obtain the observations count for
each combination of these variables’ levels.

• Diverse summary tables: Use data %>%

group_by(cat_var1) %>% summarise() from the
“Dplyr” package to aggregate datasets and obtain
the summary numbers you want.

• See examples in Rstudio

Export Data
• Export data to use outside of R: You can export
your datasets as .csv files using the write.csv()
function.

• Export data for later use in R: You can export your

datasets as R objects called .RDS files using
saveRDS(). You can import them into R using
readRDS(). These execute a lot faster.

• See examples in Rstudio

CHAPTER VII. Plots
Plots
• Plots (Graphs, Visualisations,..) are very powerful
tools. They allow you to quickly grasp trends and
patterns in data sets, some of which could not be
spotted by analysing summary tables only

• In R, ‘ggplot2’ package gives you endless

possibilities to create visualisations.

• In this video, we focus on qplot() function (from

‘ggplot2’), which can provide high quality graphs
with very little effort.
Plots
With qplot(), we can create:

• Histograms and Density plots to visualise Numerical

variables
• Bar plots to visualise categorical variables
• Box plots to visualise correlations between numerical and
categorical variables
• Dot Plots to visualise correlations between numerical
variables

We can also use color coding to add information to

graphs while keeping them easily interpretable. See
examples in R studio.
Plots
Finally, you can save your graphs as images.

Simply use the ggsave() function from the ggplot2

package

See examples in Rstudio

Thank you