Getting started with ggplot2 - Gaston Sanchez · Getting started with ggplot2 STAT 133 Gaston Sanchez Department of Statistics, UC ... I Variables: height, weight, jedi I Geoms: points

Getting started with ggplot2STAT 133

Gaston Sanchez

Department of Statistics, UC–Berkeley

gastonsanchez.com

github.com/gastonstat/stat133

Course web: gastonsanchez.com/stat133

http://www.gastonsanchez.com

http://www.gastonsanchez.com

http://github.com/gastonstat/stat133

http://www.gastonsanchez.com/stat133

ggplot2

2

Resources for "ggplot2"

I Documentation: http://docs.ggplot2.org/

I Book: ggplot2: Elegant Graphics for Data Analysis(by Hadley Wickham)

I Book: R Graphics Cookbook (by Winston Chang)

I RStudio ggplot2 cheat sheethttps://www.rstudio.com/wp-content/uploads/2015/03/ggplot2-cheatsheet.pdf

3

http://docs.ggplot2.org/

https://www.rstudio.com/wp-content/uploads/2015/03/ggplot2-cheatsheet.pdf

package "ggplot2"

# remember to install ggplot2

# (just once)

install.packages("ggplot2")

# load ggplot2

library(ggplot2)

# see basic documentation

?ggplot

4

ggplot2 book

5

R Graphics Cookbook

6

●●

●

●

●

●

●

●

●

●●

● ●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

100

200

300

10 15 20 25 30 35mpg

hp

cyl

●

●

●

4

6

8

Miles per gallon −vs− Horsepower

7

●●●

●

●

●

●

●

●

●●

● ●●

●●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

10 15 20 25 30

5015

025

0


mpg

hp

●

●

●

468

8

About "ggplot2"

I "ggplot2" (by Hadley Wickham) is an R package forproducing statistical graphics

I It provides a framework based on Leland Wilkinson’sGrammar of Graphics

I "ggplot2" provides beautiful plots while taking care offiddly details like legends, axes, colors, etc.

I "ggplot2" is built on the R graphics package "grid"

I Underlying philosophy is to describe a wide range ofgraphics with a compact syntax and independentcomponents

9

The Grammar of Graphics

10

About the Grammar of Graphics

I The Grammar of Graphics is Wilkinson’s attempt to definea theoretical framework for graphics

I Grammar: Formal system of rules for generating graphics– Some rules are mathematic– Some rules are aesthetic

11


3 Stages of Graphic Creation

I Specification: link data to graphic objects

I Assembly: put everything together

I Display: render of a graphic

12


SpecificationLink data to graphic objects

I Data

I Transformation of variables (e.g. aggregation)

I Scale transformations (e.g. log)

I Coordinate system (e.g. cartesian)

I Graphic Elements (e.g. points, lines)

I Guides (e.g. labels, legends)

13

R package "ggplot2"

About "ggplot2"

I Default appearance of plots carefully chosen

I Designed with visual perception in mind

I Inclusion of some components, like legends, are automated

I Great flexibility for annotating, editing, and embeddingoutput

14

Base graphics -vs- "ggplot2"

base graphics

●●

●

●

●

●

●

●

●

●●

● ●●

●●

●

●

●●

●

●●

●

●

●

●

●

●

●

●

●

10 15 20 25 30

5010

015

020

025

030

0

mpg

hp

ggplot2

●●

●

●

●

●

●

●

●

●●

● ●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

100

200

300

10 15 20 25 30 35mpg

hp

15

About "ggplot2"

I "ggplot2" is the name of the package

I The gg in "ggplot2" stands for Grammar of Graphics

I Inspired in the Grammar of Graphics by Lee Wilkinson

I "ggplot" is the class of objects (plots)

I ggplot() is the main function in "ggplot2"

16

What is a Statistical Graphic?

17

Some Data set

mtcars

## mpg hp cyl

## Mazda RX4 21.0 110 6

## Mazda RX4 Wag 21.0 110 6

## Datsun 710 22.8 93 4

## Hornet 4 Drive 21.4 110 6

## Hornet Sportabout 18.7 175 8

## Valiant 18.1 105 6

## Duster 360 14.3 245 8

## Merc 240D 24.4 62 4

## Merc 230 22.8 95 4

## Merc 280 19.2 123 6

18

What is a statistical graphic?

●●

●

●

●

●

●

●

●

●●

● ●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

0

100

200

300

10 15 20 25 30 35mpg

hp

cyl

●

●

●

4

6

8


19


Elements to draw the chart “manually”

I coordinate system

I x and y axis (intervals)

I axis tick marks

I axis labels, and title

I points (with colors)

I regression line (and ribbon)

I legend

20


Elements to draw the chart “manually”

I coordinate system

I x and y axis (intervals)

I axis tick marks

I axis labels, and title

I points (with colors)

I regression line (and ribbon)

I legend

20


Simply put, a statistical graphic is:I A mapping from data to aesthetic attributes (color, shape,

size) of geometric objects (points, lines, bars)

I A plot may also contain statistical transformations of thedata

I A plot is drawn on a specific coordinate system

I Sometimes faceting can be used to get the same plot fordifferent subsets of the dataset

21

Starting with "ggplot2"

22

starwarstoy.csv

## Warning in file(file, "rt"): cannot open file

’/Users/gaston/Documents/stat133/stat133/datasets/starwarstoy.csv’: No

such file or directory

## Error in file(file, "rt"): cannot open the connection

## Error in eval(expr, envir, enclos): object ’starwars’ not found

23

Scatterplot

## Error in ggplot(data = starwars): object ’starwars’ not found

24

Main steps in creating ggplot graphics

A B C D E F

Dataset Which variables

A B C D E F

Geometric objects

abcd

points

text

lines

bars

Aesthetics

x = Ay = Bcolor = Csize = defaultshape = default

1 2

3 4

25

Building a scatterplot

User specifications

I Dataset: starwars

I Variables: height, weight, jedi

I Geoms: pointsI Aesthetics (attributes):

– x: height

– y: weight

– color: jedi

26

Scatterplot with "ggplot2"

ggplot(data = starwars) +

geom_point(aes(x = height, y = weight, color = jedi))

I ggplot() initializes a "ggplot" object

I specify the dataset with data

I type of geometric object: geom point()I mapping aesthetic attributes to variables with aes()

– x-position: height

– y-position: weight

– color: jedi

27




I ggplot() initializes a "ggplot" object

I specify the dataset with data

I type of geometric object: geom point()I mapping aesthetic attributes to variables with aes()

– x-position: height

– y-position: weight

– color: jedi

27





28


Automated things in "ggplot2"

I Axis labels

I Legends (position, labels, symbols)

I Choose of colors for points

I Background color (e.g. gray)

I Grid lines (major and minor)

I Axis tick marks

you can always change the automated elements

29

"ggplot2" graphics

Philosophy of "ggplot2"A graphic is a mapping from data to aesthetic attributes(color, shape, size) of geometric objects (points, lines, bars)

30





31

Mapping

height weight jedi

1.721.501.821.800.961.670.662.28

77497780327517

112

jedino_jedi

jedino_jedino_jedino_jedi

jedino_jedi

x y color

x1x2x3x4x5x6x7x8

y1y2y3y4y5y6y7y8

#F8766D#00BFC4#F8766D#00BFC4#00BFC4#00BFC4#F8766D#00BFC4

data values aesthetic attributes

mapping

32

"ggplot2" graphics

Philosophy of "ggplot2"A graphic is a mapping from data to aesthetic attributes(color, shape, size) of geometric objects (points, lines, bars)

I ggplot(data, ...)

I aes()

I geom objects()

33


How does "ggplot2" work?

I plots are created piece-by-piece

I plot components added with + operator

I aesthetic attributes mapped to data values

I computation of scales for aesthetic attributes

34

How does it work?

Usually, we specify the data and variables inside the functionggplot()

ggplot(data = mtcars, aes(x = mpg, y = hp))

Note the use of the internal function aes() to map x to mpg,and y to hp.

Then we add a layer of geometric objects: points in this case

+ geom_point()

35

Some alternative options

# option A

ggplot(data = starwars,

aes(x = height, y = weight, color = jedi)) +

geom_point()

# option B



# option C

ggplot() +

geom_point(data = starwars,

aes(x = height, y = weight, color = jedi))

36


# option A



geom_point()

# option B



# option C

ggplot() +



36


# option A



geom_point()

# option B



# option C

ggplot() +



36

Main inquiries

Always ask yourself ...

I What is the data set of interest?

I What variables will be used to make the plot?

I What graphics shapes will be used to display?

I What features of the shapes will be used to represent thedata values?

37

"ggplot2" basics

I The data must be in a data.frame

I Variables are mapped to aesthetic attributes

I Aesthetic attributes belong to geometric objects geoms(points, lines, polygons)

38

Basic Terminology

I ggplot() - The main function where you specify thedataset and variables to plot

I geoms - geometric objetcs– geom point(), geom bar(), geom line(), geom density()

I aes - aesthetics (i.e. attributes)– shape, color, fill, linetype

39

Warning

"ggplot2" comes with the function qplot() (i.e. quick plot).Avoid using it!

As Karthik Ram says: “you’ll end up unlearning and relearning agood bit”

40