Dynamic data visualization with R

Laurent Rouvière
2022-12-05

1
Overview

• Materials: available at
https://lrouviere.github.io/page_perso/visualisationR.html

2
Overview

• Materials: available at
https://lrouviere.github.io/page_perso/visualisationR.html
• Prerequisites: basics on R, probability, statistics and computer
programming

2
Overview

• Materials: available at
https://lrouviere.github.io/page_perso/visualisationR.html
• Prerequisites: basics on R, probability, statistics and computer
programming
• Objectives:
• understand the importance of visualization in datascience
• visualize data, models and results of a datascience project
• discover (and master) some R visualization packages

2
Overview

• Materials: available at
https://lrouviere.github.io/page_perso/visualisationR.html
• Prerequisites: basics on R, probability, statistics and computer
programming
• Objectives:
• understand the importance of visualization in datascience
• visualize data, models and results of a datascience project
• discover (and master) some R visualization packages
• Teacher: Laurent Rouvière, [email protected]
• Research interests: nonparametric statistics, statistical learning
• Teaching: statistics and probability (University and engineer school)
• Responsabilities: head of the Master Mathématiques Appliquées,
Statistique of Rennes
• Consulting: energy (ERDF), banks, marketing, sport
2
Resources

• Slides and tutorials (supplement materials + exercises) available at

https://lrouviere.github.io/page_perso/visualisationR.html
• The web

3
Resources

• Slides and tutorials (supplement materials + exercises) available at

https://lrouviere.github.io/page_perso/visualisationR.html
• The web
• Book: R for statistics, Chapman & Hall

3
Why data visualization in your Master?

• Data are more and more complex

• Models are more and more complex
• Conclusions are more and more complex.

4
Why data visualization in your Master?

• Data are more and more complex

• Models are more and more complex
• Conclusions are more and more complex.
• We need visualization tools to:
• describe data
• calibrate models
• present results and conclusions of the study.

4
Why data visualization in your Master?

• Data are more and more complex

• Models are more and more complex
• Conclusions are more and more complex.
• We need visualization tools to:
• describe data
• calibrate models
• present results and conclusions of the study.

Consequence
Visualization reveals crucial throughout a statistical study.

4
How to make visualization?

• (at least) 2 ways to understand visualization:

1. Statistical methods or algorithms: PCA, LDA, trees…
2. Computing tools: R packages.

5
How to make visualization?

• (at least) 2 ways to understand visualization:

1. Statistical methods or algorithms: PCA, LDA, trees…
2. Computing tools: R packages.

• In this workshop, we will focus on some R tools:

1. ggplot2: system for declaratively creating graphics ⟹ 3-4h.

2. Mapping with ggmap, sf (static) leaflet (dynamic) ⟹ 3-4h.

3. Dynamic or interactive tools

• data with rAmCharts and plotly ⟹ 1h

• dashboard with flexdashboard ⟹ 1h
• web applications with shiny ⟹ 5h

5
 Boxplot for the iris dataset

> data(iris)
> summary(iris)
Sepal.Length Sepal.Width Petal.Length Petal.Width
Min. :4.300 Min. :2.000 Min. :1.000 Min. :0.100
1st Qu.:5.100 1st Qu.:2.800 1st Qu.:1.600 1st Qu.:0.300
Median :5.800 Median :3.000 Median :4.350 Median :1.300
Mean :5.843 Mean :3.057 Mean :3.758 Mean :1.199
3rd Qu.:6.400 3rd Qu.:3.300 3rd Qu.:5.100 3rd Qu.:1.800
Max. :7.900 Max. :4.400 Max. :6.900 Max. :2.500
Species
setosa :50
versicolor:50
virginica :50

6
 Classical tool
> boxplot(Sepal.Length~Species,data=iris)
7.5
Sepal.Length

6.5
5.5
4.5

setosa versicolor virginica

7
Species
 Ggplot tools

> library(tidyverse) #ggplot2 in tidyverse

> ggplot(iris)+aes(x=Species,y=Sepal.Length)+geom_boxplot()

7
Sepal.Length

setosa versicolor virginica

Species

8
A temperature map

temp

17.5

15.0

12.5

Many informations
• Background map with boundaries of departments;
• Temperatures in each departments (meteofrance website).

9
 Mapping with sf
> library(sf)
> dpt <- read_sf("./DATA/dpt")
> dpt |> select(NOM_DEPT,geometry) |> head()
Simple feature collection with 6 features and 1 field
Geometry type: MULTIPOLYGON
Dimension: XY
Bounding box: xmin: 644570 ymin: 6272482 xmax: 1077507 ymax: 6997000
Projected CRS: RGF93 v1 / Lambert-93
# A tibble: 6 x 2
NOM_DEPT geometry
<chr> <MULTIPOLYGON [m]>
1 AIN (((919195 6541470, 918932 6541203, 918628 6~
2 AISNE (((735603 6861428, 735234 6861392, 734504 6~
3 ALLIER (((753769 6537043, 753554 6537318, 752879 6~
4 ALPES-DE-HAUTE-PROVENCE (((992638 6305621, 992263 6305688, 991610 6~
5 HAUTES-ALPES (((1012913 6402904, 1012577 6402759, 101085~
6 ALPES-MARITIMES (((1018256 6272482, 1017888 6272559, 101677~

10
 Background map

> ggplot(dpt)+geom_sf()

50°N

48°N

46°N

44°N

42°N

0° 5°E 10°E

11
 Temperature map

> ggplot(dpt) + geom_sf(aes(fill=temp)) +

+ scale_fill_continuous(low="yellow",high="red")+
+ theme_void()

temp

17.5

15.0

12.5

12
 Interactive charts with rAmCharts
> library(rAmCharts)
> amBoxplot(Sepal.Length~Species,data=iris)

13
Dashboard

• Useful to publish groups of related data visualizations (dataset,

classical charts, simple models…)
• Package flexdahboard:
https://rmarkdown.rstudio.com/flexdashboard/index.html

14
Dashboard

• Useful to publish groups of related data visualizations (dataset,

classical charts, simple models…)
• Package flexdahboard:
https://rmarkdown.rstudio.com/flexdashboard/index.html
• Based on Rmarkdown syntax
• Example: https://lrouviere.shinyapps.io/dashboard/

14
Interactive web apps with shiny

• Shiny is a R package that makes it easy to build interactive web apps

straight from R.

• Examples:
• understand overfitting in machine learning:
https://lrouviere.shinyapps.io/overfitting_app/
• bike stations in Rennes: https://lrouviere.shinyapps.io/velib/

15
To summarize

• 15 hours for 3 (or 4) topics.

• 1 topic = slides + tutorial (supplement material + exercises).
• Require personal efforts.
• To Practice, to make mistakes and to correct these mistakes: only way
to learn computer tools.

16
To summarize

• 15 hours for 3 (or 4) topics.

• 1 topic = slides + tutorial (supplement material + exercises).
• Require personal efforts.
• To Practice, to make mistakes and to correct these mistakes: only way
to learn computer tools.
• You need to work alone between the sessions.
• Everyone can develop at its own pace (the goal is to progress), and
ask questions during the sessions.
• I’m here to (try) to answer.

16
 Outline
1. Data visualization with ggplot2

Conventional graphical functions (a reminder)

ggplot2 grammar
2. Mapping

ggmap

Shapefile contours with sf

Interactive maps with leaflet

3. Some Dynamic visualization tools

rAmCharts and plotly

Graphs with visNetwork

Dashboard with flexdasboard 17

Data visualization with ggplot2

18
• Graphs are often the starting point for statistical analysis.
• One of the main advantages of R is how easy it is for the user to
create many different kinds of graphs.
• We begin by a (short) review on conventional graphs,
• followed by an examination of some more complex representations,
especially with ggplot2 package.

19
Data visualization with ggplot2

Conventional graphical functions (a

reminder)

20
The plot function

• It is a generic function to represent all kind of data.

• For a scatter plot, we have to specify a vector for the 𝑥-axis and a
vector for the 𝑦-axis.
> x <- seq(-2*pi,2*pi,by=0.1)
> plot(x,sin(x),type="l",xlab="x",ylab="sin(x)")
> abline(h=c(-1,1))
1.0
sin(x)

−1.0 0.0

−6 −4 −2 0 2 4 6
21
x
Graphs for datasets

• Many kind of representations are needed according to the variables we

want to visualize.

• Histogram for continuous variables, barplot for categorical variables.

• Scatterplot for 2 continuous variables.

• Boxplot to visualize distributions.

22
Graphs for datasets

• Many kind of representations are needed according to the variables we

want to visualize.

• Histogram for continuous variables, barplot for categorical variables.

• Scatterplot for 2 continuous variables.

• Boxplot to visualize distributions.

Fortunately
There is a R function for all representations.

22
 Scatterplot with dataset

> plot(Sepal.Length~Sepal.Width,data=iris)

7.5
Sepal.Length

6.0
4.5

2.0 2.5 3.0 3.5 4.0

Sepal.Width
> plot(iris$Sepal.Width,iris$Sepal.Length) #similar

23
 Histogram for continous variable
> hist(iris$Sepal.Length,probability=TRUE,
+ col="red",xlab="Sepal.Length",main="Histogram")

Histogram
0.4
Density

0.2
0.0

4 5 6 7 8

Sepal.Length
24
 Barplot for categorical variables

> barplot(table(iris$Species))
40
20
0

setosa versicolor virginica

25
 Boxplot

> boxplot(Sepal.Length~Species,data=iris)

7.5
Sepal.Length

6.0
4.5

setosa versicolor virginica

Species

26
Data visualization with ggplot2

ggplot2 grammar

27
• ggplot2 is a plotting system for R based on the grammar of graphics
(as dplyr to manipulate data).

• The goal is to provide a clear syntax for an efficient visualization.

• Ggplot provides
• ”nice” graphs (nor always the case for conventional R graphs).
• ”complex” graphs with few command lines.

• Documentation: tutorial, book

28
• ggplot2 is a plotting system for R based on the grammar of graphics
(as dplyr to manipulate data).

• The goal is to provide a clear syntax for an efficient visualization.

• Ggplot provides
• ”nice” graphs (nor always the case for conventional R graphs).
• ”complex” graphs with few command lines.

• Documentation: tutorial, book

Remark
Today, most of the charts are made with ggplot ⟹ important to master
the syntax.

28
For a given dataset, a graph is defined from many layers. We have to
specify:

• the data
• the variables we want to plot
• the type of representation (scatterplot, boxplot…).

29
For a given dataset, a graph is defined from many layers. We have to
specify:

• the data
• the variables we want to plot
• the type of representation (scatterplot, boxplot…).

Ggplot graphs are defined from these layers. We indicate

• the data with ggplot

• the variables with aes (aesthetics)
• the kind of representation with geom_...

29
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

30
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

• Aesthetics (aes): to describe the way that variables in the data are
mapped. All the variables used in the graph should be specified in aes.

30
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

• Aesthetics (aes): to describe the way that variables in the data are
mapped. All the variables used in the graph should be specified in aes.
• Geometrics (geom_...): to control the type of plot.

30
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

• Aesthetics (aes): to describe the way that variables in the data are
mapped. All the variables used in the graph should be specified in aes.
• Geometrics (geom_...): to control the type of plot.
• Statistics (stat_...): to describe transformation of the data.

30
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

• Aesthetics (aes): to describe the way that variables in the data are
mapped. All the variables used in the graph should be specified in aes.
• Geometrics (geom_...): to control the type of plot.
• Statistics (stat_...): to describe transformation of the data.
• Scales (scale_...): to control the mapping from data to aesthetic
attributes (change colors, size…).

30
 The grammar

Main elements of the grammar are:

• Data (ggplot): the dataset, it should be a dataframe or a tibble.

• Aesthetics (aes): to describe the way that variables in the data are
mapped. All the variables used in the graph should be specified in aes.
• Geometrics (geom_...): to control the type of plot.
• Statistics (stat_...): to describe transformation of the data.
• Scales (scale_...): to control the mapping from data to aesthetic
attributes (change colors, size…).

All these elements are gathered with the operator +.

30
 An example

> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width)+geom_point()

4.5

4.0
Sepal.Width

3.5

3.0

2.5

2.0
5 6 7 8
Sepal.Length

31
 Color and size
> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width)+
+ geom_point(color="blue",size=2)

4.5

4.0
Sepal.Width

3.5

3.0

2.5

2.0
5 6 7 8
Sepal.Length

32
 Color by (categorical) variable
> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width,
+ color=Species)+geom_point()

4.5

4.0

Species
Sepal.Width

3.5
setosa

versicolor
3.0
virginica

2.5

2.0
5 6 7 8
Sepal.Length

33
 Changing the color
> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width,
+ color=Species)+geom_point()+
+ scale_color_manual(values=c("setosa"="blue","virginica"="green",
+ "versicolor"="red"))
4.5

4.0

Species
Sepal.Width

3.5
setosa

versicolor
3.0
virginica

2.5

2.0
5 6 7 8
Sepal.Length 34
 Color by (continous) variable
> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width,
+ color=Petal.Width)+geom_point()

4.5

4.0 Petal.Width
2.5
Sepal.Width

3.5 2.0

1.5
3.0
1.0

0.5
2.5

2.0
5 6 7 8
Sepal.Length

35
 Color by (continous) variable
> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width,
+ color=Petal.Width)+geom_point()+
+ scale_color_continuous(low="yellow",high="red")

4.5

4.0 Petal.Width
2.5
Sepal.Width

3.5 2.0

1.5
3.0
1.0

0.5
2.5

2.0
5 6 7 8
Sepal.Length
36
 Histogram

> ggplot(iris)+aes(x=Sepal.Length)+geom_histogram(fill="red")

12.5

10.0

7.5
count

5.0

2.5

0.0
5 6 7 8
Sepal.Length

37
 Barplot

> ggplot(iris)+aes(x=Species)+geom_bar(fill="blue")

50

40

30
count

20

10

0
setosa versicolor virginica
Species

38
 Facetting (more “complex”)

> ggplot(iris)+aes(x=Sepal.Length,y=Sepal.Width)+geom_point()+
+ geom_smooth(method="lm")+facet_wrap(~Species)

setosa versicolor virginica

4.5

4.0
Sepal.Width

3.5

3.0

2.5

2.0
5 6 7 8 5 6 7 8 5 6 7 8
Sepal.Length

39
Combining ggplot with dplyr

• We often have to work on the dataframe to obtain an efficient ggplot

syntax.

• For instance
> head(df)
# A tibble: 6 x 3
size weight.20 weight.50
<dbl> <dbl> <dbl>
1 153 61.2 81.4
2 169 67.5 81.4
3 168 69.4 80.3
4 169 66.1 81.9
5 176 70.4 79.2
6 169 67.6 88.9

40
Goal

90

80 age
weight

20
70 50

60

150 160 170

size

41
 dplyr step

Gather column weight.M and weight.W into one column weight with
pivot_longer:
> df1 <- df |> pivot_longer(-size,names_to="age",values_to="weight")
> df1 |> head()
# A tibble: 6 x 3
size age weight
<dbl> <chr> <dbl>
1 153 weight.20 61.2
2 153 weight.50 81.4
3 169 weight.20 67.5
4 169 weight.50 81.4
5 168 weight.20 69.4
6 168 weight.50 80.3
> df1 <- df1 |>
+ mutate(age=recode(age,"weight.20"="20","weight.50"="50"))

42
 ggplot step

> ggplot(df1)+aes(x=size,y=weight,color=age)+
+ geom_point()+geom_smooth(method="lm")+theme_classic()

90

80 age
weight

20
70 50

60

150 160 170

size

43
Statistics

• Some charts requires to compute statistics

44
Statistics

• Some charts requires to compute statistics

• Histogram example: count the number of observations in each bins

(or the density)

44
 Statistics

• Some charts requires to compute statistics

• Histogram example: count the number of observations in each bins

(or the density)
Consequence
geom_histogram calls stat_bin function to compute these statistics.

> geom_histogram(...,stat = "bin",...)

help(stat_bin)
Computed variables
count
number of points in bin

density
density of points in bin, scaled to integrate to 1
...
44
Visualize another statistics

> ggplot(diamonds)+aes(x=carat)+ > ggplot(diamonds)+

+ geom_histogram() + aes(x=carat,y=after_stat(density)
15000 + #or aes(x=carat,y=..density..)
+ geom_histogram()

10000 1.5
count

density
5000 1.0

0.5
0
0 1 2 3 4 5
carat 0.0
0 1 2 3 4 5
carat

45
stat_summary

• More generally, stat_summary allows to computes any statistics for a

graph.
> ggplot(diamonds)+aes(x=carat)+
+ geom_histogram(aes(y=after_stat(density)))+
+ stat_summary(aes(y=0,xintercept=after_stat(x)),
+ fun="quantile",geom="vline",
+ orientation = "y",color="red")

1.5
density

1.0

0.5

0.0
46
0 1 2 3 4 5
 Complement: some demos

> demo(image)
> example(contour)
> demo(persp)
> library("lattice");demo(lattice)
> example(wireframe)
> library("rgl");demo(rgl)
> example(persp3d)
> demo(plotmath);demo(Hershey)

47
 Complement: some demos

> demo(image)
> example(contour)
> demo(persp)
> library("lattice");demo(lattice)
> example(wireframe)
> library("rgl");demo(rgl)
> example(persp3d)
> demo(plotmath);demo(Hershey)

⟹ Work on this part of the tutorial.

47
Mapping

48
Introduction

• Many applications require maps to visualize data or results of a model;

• Many R packages: ggmap, RgoogleMaps, maps…

• In this part: ggmap, sf (static mapping) and leaflet (interactive

mapping).

49
Mapping

ggmap

50
Syntax

• Similar to ggplot…

51
Syntax

• Similar to ggplot…

• Instead of
> ggplot(data)+...

• use
> ggmap(backgroundmap)+...

51
 Background map

> library(ggmap)
> us <- c(left = -125, bottom = 25.75, right = -67, top = 49)
> map <- get_stamenmap(us, zoom = 5, maptype = "toner-lite")
> ggmap(map)

45

40
lat

35

30

−120 −110 −100 −90 −80 −70

lon

52
 Adding informations with ggplot
> fr <- c(left = -6, bottom = 41, right = 10, top = 52)
> fond <- get_stamenmap(fr, zoom = 5,"toner-lite")
> Paris <- data.frame(lon=2.351499,lat=48.85661)
> ggmap(fond)+geom_point(data=Paris,aes(x=lon,y=lat),color="red")

52

50

48
lat

46

44

42

−5 0 5 10
lon 53
Mapping

Shapefile contours with sf

54
sf package

• Ggmap: ok for easy maps (background with some points).

• Not sufficient for more complex representations (color countries

according to variables).

55
sf package

• Ggmap: ok for easy maps (background with some points).

• Not sufficient for more complex representations (color countries

according to variables).

• sf allows to manage specific tools for mapping: boundaries for

countries or department, coordinate systems (latitudes-longitudes,
World Geodesic System 84…)

• Background map with format shapefile (contours = polygons)

• Compatible with ggplot (geom_sf verb).

55
sf package

• Ggmap: ok for easy maps (background with some points).

• Not sufficient for more complex representations (color countries

according to variables).

• sf allows to manage specific tools for mapping: boundaries for

countries or department, coordinate systems (latitudes-longitudes,
World Geodesic System 84…)

• Background map with format shapefile (contours = polygons)

• Compatible with ggplot (geom_sf verb).

References
• https://statnmap.com/fr/2018-07-14-initiation-a-la-cartographie-
avec-sf-et-compagnie/
• Vignettes on the cran: 55
 Example

> library(sf)
> dpt <- read_sf("./DATA/dpt")
> dpt[1:5,3]
Simple feature collection with 5 features and 1 field
Geometry type: MULTIPOLYGON
Dimension: XY
Bounding box: xmin: 644570 ymin: 6290136 xmax: 1022851 ymax: 6997000
Projected CRS: RGF93 v1 / Lambert-93
# A tibble: 5 x 2
NOM_DEPT geometry
<chr> <MULTIPOLYGON [m]>
1 AIN (((919195 6541470, 918932 6541203, 918628 6~
2 AISNE (((735603 6861428, 735234 6861392, 734504 6~
3 ALLIER (((753769 6537043, 753554 6537318, 752879 6~
4 ALPES-DE-HAUTE-PROVENCE (((992638 6305621, 992263 6305688, 991610 6~
5 HAUTES-ALPES (((1012913 6402904, 1012577 6402759, 101085~

56
 Visualize with plot
> plot(st_geometry(dpt))

57
 Visualize with ggplot

> ggplot(dpt)+geom_sf()

50°N

48°N

46°N

44°N

42°N

0° 5°E 10°E

58
Adding points on the map

• Define coordinates with st_point

> point <- st_sfc(st_point(c(2.351462,48.85670)),
+ st_point(c(4.832011,45.75781)),
+ st_point(c(5.369953,43.29617)))

59
Adding points on the map

• Define coordinates with st_point

> point <- st_sfc(st_point(c(2.351462,48.85670)),
+ st_point(c(4.832011,45.75781)),
+ st_point(c(5.369953,43.29617)))

• Specify the coordinate system (4326 for lat-lon)

> st_crs(point) <- 4326 #lat-lon
> point
Geometry set for 3 features
Geometry type: POINT
Dimension: XY
Bounding box: xmin: 2.351462 ymin: 43.29617 xmax: 5.369953 ymax: 48.856
Geodetic CRS: WGS 84
POINT (2.351462 48.8567)
POINT (4.832011 45.75781)
POINT (5.369953 43.29617)
59
 ggplot step

> ggplot(dpt) + geom_sf(fill="white")+

+ geom_sf(data=point,color="red",size=4)+theme_void()

60
 Coloring polygons

> set.seed(1234)
> dpt1 <- dpt |> mutate(temp=runif(96,10,20))
> ggplot(dpt1) + geom_sf(aes(fill=temp)) +
+ scale_fill_continuous(low="yellow",high="red")+
+ theme_void()

temp

17.5

15.0

12.5

61
Supplement: geometry class

• One of the main advantage of sf is the geometry class.

• It allows to conduct the representation with plot or geom_sf:
• point or multipoint ⟹ points to locate a place;
• polygon or multipolygon ⟹ contours to visualize boundaries.
• Some useful functions:
• st_point and st_multipoint: create points or sequence of points
• st_sfc: create a list of sf objects
• st_crs: specify coordinate reference system
• st_cast: cast geometry to another type (convert a multipoint object to
many point objects)
• …

62
• Creation of a sf object
> b1 <- st_point(c(3,4))
> b1
POINT (3 4)
> class(b1)
[1] "XY" "POINT" "sfg"

• Creation of a sfc object (list of sf objects)

> b2 <- st_sfc(st_point(c(1,2)),st_point(c(3,4)))
> b2
Geometry set for 2 features
Geometry type: POINT
Dimension: XY
Bounding box: xmin: 1 ymin: 2 xmax: 3 ymax: 4
CRS: NA
POINT (1 2)
POINT (3 4)
> class(b2)
[1] "sfc_POINT" "sfc"
63
 • Extract geometry in a sf object
> class(dpt)
[1] "sf" "tbl_df" "tbl" "data.frame"
> b3 <- st_geometry(dpt)
> b3
Geometry set for 96 features
Geometry type: MULTIPOLYGON
Dimension: XY
Bounding box: xmin: 99226 ymin: 6049647 xmax: 1242375 ymax: 7110524
Projected CRS: RGF93 v1 / Lambert-93
First 5 geometries:
MULTIPOLYGON (((919195 6541470, 918932 6541203,...
MULTIPOLYGON (((735603 6861428, 735234 6861392,...
MULTIPOLYGON (((753769 6537043, 753554 6537318,...
MULTIPOLYGON (((992638 6305621, 992263 6305688,...
MULTIPOLYGON (((1012913 6402904, 1012577 640275...
> class(b3)
[1] "sfc_MULTIPOLYGON" "sfc"

⟹ Work on this part of the tutorial. 64

Mapping

Interactive maps with leaflet

65
Background map

• Leaflet is one of the most popular open-source JavaScript libraries for

interactive maps.

• Documentation: here
> library(leaflet)
> leaflet() |> addTiles()

+
−

66
Leaflet | © OpenStreetMap contributors, CC-BY-SA
 Many background styles
> Paris <- c(2.35222,48.856614)
> leaflet() |> addTiles() |>
+ setView(lng = Paris[1], lat = Paris[2],zoom=12)

+
−

Leaflet | © OpenStreetMap contributors, CC-BY-SA

67
> leaflet() |> addProviderTiles("Stamen.Toner") |>
+ setView(lng = Paris[1], lat = Paris[2], zoom = 12)

+
−

Leaflet | Map tiles by Stamen Design, CC BY 3.0 — Map data ©

OpenStreetMap contributors

68
Leaflet with data

• Location of 1000 seismic events near Fiji

> data(quakes)
> head(quakes)
lat long depth mag stations
1 -20.42 181.62 562 4.8 41
2 -20.62 181.03 650 4.2 15
3 -26.00 184.10 42 5.4 43
4 -17.97 181.66 626 4.1 19
5 -20.42 181.96 649 4.0 11
6 -19.68 184.31 195 4.0 12

69
 Visualize seismics with magnitude more than 5.5
> quakes1 <- quakes |> filter(mag>5.5)
> leaflet(data = quakes1) |> addTiles() |>
+ addMarkers(~long, ~lat, popup = ~as.character(mag))

+
−

Leaflet | © OpenStreetMap contributors, CC-BY-SA

Remark
70
When you click on a marker, the magnitude appears.
 addCircleMarkers
> leaflet(data = quakes1) |> addTiles() |>
+ addCircleMarkers(~long, ~lat, popup=~as.character(mag),
+ radius=3,fillOpacity = 0.8,color="red")

+
−

Leaflet | © OpenStreetMap contributors, CC-BY-SA

71
 Color polygon (combining leaflet and sf)

> leaflet() |> addTiles() |>

+ addPolygons(data = dpt2,color=~pal1(t_prev),fillOpacity = 0.6,
+ stroke = TRUE,weight=1,
+ popup=~paste(as.character(NOM_DEPT),
+ as.character(t_prev),sep=" : "))

⟹ Work on this part of the tutorial.

72
Some Dynamic visualization tools

73
Some R tools for dynamic visualization

• Classical charts with rAmCharts and plotly.

• Graphs with visNetwork.

• Dashboard with flexdashboard.

74
Some Dynamic visualization tools

rAmCharts and plotly

75
rAmCharts

• user-friendly for standard graphs (scatterplot, times series,

histogram…).

• We just have to use classical R functions with the prefix am.

• Examples: amPlot, amHist, amBoxplot.

• References:
https://datastorm-open.github.io/introduction_ramcharts/

76
 rAmCharts Histogram

> library(rAmCharts)
> amHist(iris$Petal.Length)

40

35

30

25
Frequency

20

15

10

0
1.25 2.75 4.25 5.75

iris$Petal.Length

77
 rAmcharts Boxplot

> amBoxplot(iris)

0
Sepal.LengthSepal.WidthPetal.LengthPetal.Width

78
Plotly

• R package for creating interactive web-based graphs via the open

source JavaScript graphing library plotly.js.

• Commands are expanded into 3 parts:

• dataset and variables (plot_ly) ;
• additional representations (add_trace, add_markers…) ;
• options (axis, titles…) (layout).

• References: https://plot.ly/r/reference/

79
 Scatter plot
> library(plotly)
> iris |> plot_ly(x=~Sepal.Width,y=~Sepal.Length,color=~Species) |>
+ add_markers(type="scatter")

80
 Plotly boxplot

> iris |> plot_ly(x=~Species,y=~Petal.Length) |> add_boxplot()

81
Some Dynamic visualization tools

Graphs with visNetwork

82
Connections between individuals

• Many datasets can be visualized with graphs, especially when one has
to study connections between individuals (genomic, social network…).

83
Connections between individuals

• Many datasets can be visualized with graphs, especially when one has
to study connections between individuals (genomic, social network…).

• One individual = one node and one connection = one edge.

> set.seed(123)
> nodes <- data.frame(id = 1:15, label = paste("Id", 1:15))
> edges <- data.frame(from = trunc(runif(15)*(15-1))+1,
+ to = trunc(runif(15)*(15-1))+1)
> head(edges)
from to
1 5 13
2 12 4
3 6 1
4 13 5
5 14 14
6 1 13
83
Static graph: igraph package

• References: http://igraph.org/r/,
http://kateto.net/networks-r-igraph
> library(igraph)
> net <- graph_from_data_frame(d=edges, vertices=nodes, directed=F)
> plot(net,vertex.color="green",vertex.size=25)

84
Dynamic graph: visNetwork Package

• Reference:
https://datastorm-open.github.io/visNetwork/interaction.html
> library(visNetwork)
> visNetwork(nodes,edges)

85
 Nodes color
> nodes$group <- c(rep("A",8),rep("B",7))
> visNetwork(nodes,edges) |>
+ visGroups(groupname = "A", color = "darkblue",
+ shape = "square") |>
+ visGroups(groupname = "B", color = "red",
+ shape = "triangle")

86
 Edeges width
> edges$width <- round(runif(nrow(edges),1,10))
> visNetwork(nodes,edges) |>
+ visEdges(shadow = TRUE,
+ arrows =list(to = list(enabled = TRUE)),
+ color = list(color = "black", highlight = "red"))

87
Some Dynamic visualization tools

Dashboard with flexdasboard

88
• Just a tool… but an important visualization tool in datascience

• Allow to gather important messages on datasets and/or models

89
• Just a tool… but an important visualization tool in datascience

• Allow to gather important messages on datasets and/or models

• Package: flexdasboard

• Syntax: simple… only Rmarkdown

• Reference: https://pkgs.rstudio.com/flexdashboard/index.html

89
 Header

---
title: "My title"
output:
flexdashboard::flex_dashboard:
orientation: columns
vertical_layout: fill
theme: default
---

• Default theme could be replaced by other themes (cosmo, bootstrap,

cerulean…) (see here). You just have to add
theme: yeti

90
 Flexdashboard | code
Descriptive statistics
=====================================
Column {data-width=650}
-----------------------------------------------------------------------
### Dataset
```{r}
DT::datatable(df, options = list(pageLength = 25))
```
Column {data-width=350}
-----------------------------------------------------------------------
### Correlation matrix
```{r}
cc <- cor(df[,1:11])
mat.cor <- corrplot::corrplot(cc)
```
### Histogram
```{r}
amHist(df$maxO3)
91
```
 Flexdashboard | dashboard

⟹ Work on this part of the tutorial.

92
Assessment

93
Visualization project

• Group (2 members)

• Find a visualization problem (for instance a dataset with a statistical

problem: classification, regression…)

• Build a shiny web application to visualize important informations on

the problem

• Don’t hesitate to use tools presented in the lecture (you can also use
other tools)

• Deploy the application on the web with shinyapps, see

https://docs.rstudio.com/shinyapps.io/index.html

• Make a markdwon document (4 or 5 pages) which presents your work,

use pdf or html output.
94
 Process

You have to submit on moodle:

• Rmarkdown document (html or pdf output) which presents your work

and the url of the shiny application

• shiny files (app.R or ui.R and server.R)

• the dataset (or an url where we can download the data)

• all files needed to test your application

before ???.

95
Some examples (Smart Data, 2021)

• https://jmlascar.shinyapps.io/Fraisse_Lascar_App/

• https://mssdprojectriemerleroy.shinyapps.io/MSSD-Project-Riemer-
Le-Roy/

• https://abdessamadmarc.shinyapps.io/R_viz_project/

• https://razvanvisoiu.shinyapps.io/USA_Election_Analysis/

96