Fundamental of Data Science for EESS

layout: true
background-image: url(img/course-logo.png), url(img/logo_SBR.png), url(img//NTU-Logo-full-colour.png)
background-position: right top 30px,  right 50px bottom 50px,left 50px bottom 50px, top 350px left 500px 
background-size: 45%, 25%, 20%

# Fundamentals of Data Science for EESS

---

## R session 01 - Introduction to R

2020-01-24

---

## R sessions

1 - Introduction to R  
2 - R markdown  
3 - Git  
4 - Data wrangling  
5 - Data visualisation  
6 - Data mapping  
]

---
layout: false
class: middle, inverse

# Outline

.font150[
* What is R and why use R ?
* Resources
* Get started
* Fundamentals of R
    - Data objects
    - Vectors
    - Operators
    - Functions
    - Packages
    - Data frames
]
---
layout: true

# Introduction

---

background-image: url(img/R-logo.png)
background-position: middle center 
background-size: 25%

- .font150[Who has used R before ?]

- .font150[What other programming language have you used before ?]

- .font150[For those who are experts in R]

--
    * please refrain to answer during this session...
    * help your neighbor...

---
exclude: false
background-image: url(img/computer-languages.png)
background-position: right 20px bottom 20px 
background-size: 70%

## Computer languages

---
exclude: false
background-image: url(img/R-logo.png)

## 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](https://www.stat.auckland.ac.nz/~ihaka/downloads/Interface98.pdf)

---
## 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

---
## What can you do with R ?

- **Science**
    * Statistics of course...
    * Data processing
    * Graphics
    * Time series analyses
    * Maps
    * Bioinformatics
]

- **But also**
    * Teach
    * Do a presentation
    * Write your CV
    * Build a web site
    * Write a book
    * Much more...
]

--
.center[
<img src="img/web-site-dv.png" width="30%" style="display: block; margin: auto;" />
]

---
layout: true

# Resources

---
background-image: url(img/R_nutshell.png), url(img/R_graphics_cookbook.png)
background-position: right 20px top 50px, right 300px top 250px 
background-size: 20%, 18%

## Books and Manuals

* [Applied Statistics with R](https://daviddalpiaz.github.io/appliedstats/index.html) : Quite simple introduction with emphasis on Stats
* [R intro](https://cran.r-project.org/doc/manuals/r-release/R-intro.pdf) : Very good introduction to R, short and clear
* [R in a nutshell](http://rbasicsworkshop.weebly.com/uploads/1/8/6/0/18603232/adler_2009_r-inanutshell.pdf) : Many many receipes to solve all your questions
* R graphics cook book : very good for graphics

---
background-image: url(img/web_quickR.png)
background-position: right 20px top 80px 
background-size: 60%

## On line courses and web sites

* [Coursera](https://www.coursera.org/learn/r-programming)
*	[Pluralsight](https://www.pluralsight.com/courses/r-programming-fundamentals) - Not free
* [Quick-R, very simple](http://www.statmethods.net/)

---

background-image: url(img/R_Studio-cheatsheets-01.png), url(img/R_Studio-cheatsheets-02.png)
background-position: right 600px top 100px, right 20px top 20px 
background-size: 30%, 50%

## Cheat sheets

* [R basics](http://github.com/rstudio/cheatsheets/raw/master/base-r.pdf)
* [ggplot2](https://github.com/rstudio/cheatsheets/raw/master/data-visualization-2.1.pdf)
* [dplyr](https://www.rstudio.com/wp-content/uploads/2015/02/data-wrangling-cheatsheet.pdf)

---
background-image: url(img/stackoverflow.png)
background-position: right 20px top 20px 
background-size: 40%

## Forum

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

---

---
background-image: url(img/R_Studio_interface.png)
background-position: right 20px top 20px 
background-size: 65%

## Setup

* Install [R](https://cran.r-project.org/index.html)

* Install [R studio](https://www.rstudio.com/products/rstudio/download/#download)

---
background-image: url(img/R_Studio_interface_numbered.png)
background-position: right 20px top 20px 
background-size: 65%

## The R studio interface

- **Bottom left**
    - Console
- **Top left**
    - File editor for .R and .Rmd files
    - Data frame visualization
- **Top right**
    - Environment (i.e. R objects)
    - History
- **Bottom right**
    - Files
    - Plots
    - Packages
    - Help
]

---
background-image: url(img/R-new-project.png)
background-position: right 20px top 20px 
background-size: 40%

## Create a new project

* Open R studio
* Create new project for the course in a new directory
    - e.g. `Experimental design course`

---

## Your first script

```r
print("Hello world")
```

```
[1] "Hello world"
```

### Two ways to proceed

1. Type directly in command window

2. Create a new script

---

---
## Everything in R is an **object**

* Assignement done with **<-**

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

```
[1] 3
```
--

```r
> z <- x + y
> z
```

```
[1] 3
```

---

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

```r
> z = x + y
```

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

---

## R is **case sensitive**

```r
> Z
```

```
Error in eval(expr, envir, enclos): objet 'Z' introuvable
```

---

## 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

---

## 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...

---
layout: true
# 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)

---

## Data structures

* **Vector**

* **List**

* **Matrix**

* **Data frames**

* **Function**

---
layout: true
# Vectors

---

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

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

--
## Assign a value to a vector

```r
x <- 10
x
```

```
[1] 10
```

---

## Assign several elements

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

```
[1] 10 20 30
```
--

## Assign range

```r
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
```

---

## Assign characters

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

```
[1] "Donald" "Trump" 
```

## Assign logical

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

```
[1]  TRUE FALSE  TRUE
```

---

## Access specific elements of a vector

### First

```r
x[1]
```

```
[1] 10
```

### Range

```r
x[1:5]
```

```
[1] 10 11 12 13 14
```

### Remove one element

```r
x[-1]
```

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

---

## 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?

```r
typeof(x)
length(x)
```
--

```
[1] "integer"
```

```
[1] 21
```

---
layout: true
# Operators

---

## Arithmetic Operators

| Operator |	Description | 
|---|---|---|
| +	| addition|   
| -	| subtraction |  
| *	| multiplication| 
| /	| division | 
| ^ or ** | 	exponentiation | 
| x %% y | 	modulus (x mod y) 5%%2 is 1 | 
| x %/% y	| integer division 5%/%2 is 2 |

---

## Arithmetic Operators

We are performing vector operations !

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

---
## Arithmetic Operators

Vector one element

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

```
[1] 3
```

---
## Arithmetic Operators

Vector several elements

```r
# 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
```
--

.warning[
* Several instructions on same line separate by **;**
* The hastag **#** indicate a comment -> Use heavily to document your code
* However, it is even better to use R markdown (see next class)
]
--
.student[
Use the other operators
]

---
## Arithmetic Operators

What happens when the vectors have different number of elements ?

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

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

Equivalent to

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

The recycling rule...

---

## Can we add logical ?

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

```
[1] 1
```

---

## Can we add logical ?

No error but...

The resulting variable is transformed to a **numeric**

```r
typeof(x)
```

```
[1] "logical"
```

```r
typeof(z)
```

```
[1] "integer"
```

---

## Logical Operators
| Operator |	Description | 
|---|---|---|
| <	| less than	|
| <=	| less than or equal to	|
| >	| greater than	|
| >=	| greater than or equal to	|
| ==	| exactly equal to	|
| !=	| not equal to	|
| !x	| Not x	|
| x &#124; y	| x OR y	|
| x & y	| x AND y	|
| isTRUE(x)	| test if X is TRUE	|

---

## Logical Operators

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

```
[1] TRUE
```

```
[1] FALSE
```

---

## Can we add characters ?

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

Generates an error

```
Error in first + last: argument non numérique pour un opérateur binaire
```

--
.student[
What can we do ?
]

---
layout: true
# Functions

---

Functions perform specific task on objects

* e.g. to concatanate strings we use **paste()**

```r
paste(first, last)
```

```
[1] "Donald Trump"
```
--

* Functions take **arguments** and return an object called **result**

* To know the arguments use ?

```r
  ? paste()  # Do not forget the parenthesis
```
--

* Can go directly to Help panel and type function name

---
background-image: url(img/R-help-paste-01.png)
background-position: right 20px top 20px 
background-size: 50%

## Help

---
background-image: url(img/R-help-paste-02.png)
background-position: right 20px top 20px 
background-size: 50%

## Help

---
## Getting what you want

Let's apply paste :

```r
paste(first, last)
```

```
[1] "Donald Trump"
```

.student[
* We would like to get "Donald_Trump"
* Can you read the help and suggest a change in the way we call the function ?
]

```r
paste(first, last, sep = "_")
```

```
[1] "Donald_Trump"
```

---
## Write your own function

```r
my_sum <- function(first_number, second_number) {
 c <- first_number + second_number
 return(c)
}
```
* __my_sum__ : function name
* __first_number, second_number__ : arguments
* instructions are enclosed by braces ({})
* return() : the value(s) returned

#### More compact way

```r
 my_sum <- function(first_number, second_number) {first_number + second_number}
```

---
## Call your function

```r
my_sum(10, 20)
```

```
[1] 30
```

--
* better

```r
my_sum(first_number = 10, second_number = 20)
```

```
[1] 30
```

---
## 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

```r
x <- 1:100
sum(x)
```

```
[1] 5050
```

* Normal distribution

```r
y <- rnorm(10, mean = 0, sd = 1)
y
```

```
 [1]  1.6731915 -0.2498182  1.0774267  0.7086024 -0.3065056  2.2168636
 [7]  0.7236404  0.2397608 -0.1206248 -1.2904100
```

---

## Statistics

```r
mean(y)
```

```
[1] 0.4672127
```

```r
sd(y)
```

```
[1] 1.033406
```

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

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

```
[1] -0.005369128
```

```r
sd(y)
```

```
[1] 0.9956479
```

---

## Plot

```r
library(graphics)
hist(y)
```
]

<img src="R-session-01-intro_files/figure-html/unnamed-chunk-47-1.png" style="display: block; margin: auto;" />
]

---

---
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

---

## Installing a package

Download on your computer the package you need

.center[
<img src="img/R_studio_package_01.png" width="45%" /><img src="img/R_studio_package_02.png" width="35%" />
]

---

## Using a package

To use functions from the package

- use the syntax `package::function`

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

```
[1] "Donald Trump"
```

- load the package with the library function

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

```
[1] "Donald Trump"
```

---
background-image: url(img/R_studio_package_03.png)
background-position: right 20px top 20px 
background-size: 50%

## 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

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

```
  label id      value  flag
1     a  1 -0.5324537  TRUE
2     b  2 -0.2858342 FALSE
3     c  3 -0.7311013  TRUE
4     d  4 -1.2440367 FALSE
5     e  5 -0.8671309  TRUE
6     f  6 -2.4274570 FALSE
```
.warning[
* We will NOT use factors: `stringsAsFactors = FALSE` 
]

---

## Useful functions

```r
dim(df)  # returns the dimensions of data frame
```

```
[1] 6 4
```

```r
nrow(df)  # number of rows
```

```
[1] 6
```

```r
ncol(df)  # number of columns
```

```
[1] 4
```

---

## Useful functions

```r
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  -0.532 -0.286 -0.731 -1.244 -0.867 ...
 $ flag : logi  TRUE FALSE TRUE FALSE TRUE FALSE
```

```r
colnames(df)  # columns names
```

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

---
## Access specific value

* Use the `df[i,j]` notation, first index corresponds to row, second index to column

```r
df[5, 3]
```

```
[1] -0.8671309
```
--

* Specifiy the name of the column

```r
df[5, "value"]
```

```
[1] -0.8671309
```
.warning[
* The result is a **vector**
]
---

## Access specific column

* Use the `df[i,j]` notation

```r
df[, 3]
```

```
[1] -0.5324537 -0.2858342 -0.7311013 -1.2440367 -0.8671309 -2.4274570
```

```r
df[, "value"]
```

```
[1] -0.5324537 -0.2858342 -0.7311013 -1.2440367 -0.8671309 -2.4274570
```
.warning[
* The result is a **vector**
]

---

## Access specific column

* Use `$`notation

```r
df$value
```

```
[1] -0.5324537 -0.2858342 -0.7311013 -1.2440367 -0.8671309 -2.4274570
```

## This can be used to access a specific value

*  `$` for the column, `[i]` for the row

```r
df$value[5]
```

```
[1] -0.8671309
```

---
## Access row

* Use the `df[i,j]` notation

```r
df[1, ]
```

```
  label id      value flag
1     a  1 -0.5324537 TRUE
```
.warning[
* The result is a **data frame**
]

---

## Access specific rows
* Rows for which the value of id <= 3

```r
 df[df$id <= 3,]
```

```
  label id      value  flag
1     a  1 -0.5324537  TRUE
2     b  2 -0.2858342 FALSE
3     c  3 -0.7311013  TRUE
```

```r
df[df$label == "c", ]
```

```
  label id      value flag
3     c  3 -0.7311013 TRUE
```

.warning[This syntax is complicated - tidyverse packages make it much more easy to manipulate and remember]

---
layout: false

# Recap

.font150[
- R is case sensitive: Z !=  z
- Objects: data types vs data structures
- Vectors: think in vector operations
- Operators: arithmetic vs. logical
- Functions: try to practice
- Data frames: most useful objects (Excel type)
]
---
exclude: false
layout: false
background-image: url(img/R-markdown-book.png)
background-position: right 20px top 20px
background-size: 25%

# Next class: 02 - Markdown

What you will learn :

* Mix text, R code and R output in a single document
* Produce documents as HTML, pdf or even Word from the same template

.student[
* Please install the following packages and their dependencies
    * rmarkdown (will install also knitr)
    * tinytex (Latex)

* Installation : https://bookdown.org/yihui/rmarkdown/installation.html
* Have a look at the book https://bookdown.org/yihui/rmarkdown
]

---
exclude: false
layout: false
background-image: url(img/Git-progit2.png)
background-position: right 20px top 20px
background-size: 30%

# Next class: 03 - Git

## Resources

* Pro GIT: https://git-scm.com/book/en/v2
* GitHub guide: https://guides.github.com
* Happy Git with R: https://happygitwithr.com/

## Software

* Git: https://git-scm.com
* Git desktop: https://github.com
* GitHub account: https://desktop.github.com