Recitation 07 - Introduction to ggplot2

Introduction to GGPLOT2

ggplot2 is Hadley Wickham’s R package for producing “elegant graphics for data analysis”

The gg in ggplot2 stands for “Grammar of Graphics”

The concept behind ggplot2 divides plot into three different fundamental parts: Plot = Data + Aesthetics + Geometry.

• Data is a data frame
• Aesthetics is used to indicate x and y variables and fine tune other variables of the graphs
• Geometry defines the type of graphics (histogram, box plot, line plot, density plot)

An aesthetic is a visual property of the objects in your plot. Aesthetics include things like the size, the shape, or the color of your points. You can display a point (like the one below) in different ways by changing the values of its aesthetic properties.

There are two major functions in ggplot2: qplot() and ggplot() - qplot() stands for quick plot, which can be used to produce easily simple plots. - ggplot() function is more flexible and robust than qplot for building a plot piece by piece and will be our focus

The seven ingredients of a ggplot

#ggplot(data = <DATA>) + 
#  <GEOM_FUNCTION>(
#     mapping = aes(<MAPPINGS>),
#     stat = <STAT>, 
#     position = <POSITION>
#  ) +
#  <COORDINATE_FUNCTION> +
#  <FACET_FUNCTION>

Load the required Packages

if (!require("pacman")) install.packages("pacman")
pacman::p_load(ggplot2, gridExtra)

Let’s clean the global environment before moving further

rm(list=ls())
cat("\014")

A major disadvantage of ggplot2 is that the data can only be a data.frame

Let’s load the dataset we’ve already worked with in the past

#Load the dataset
wine<-read.csv("~/Downloads/WINE.csv")
wine$QUALITY<-as.factor(wine$QUALITY)
head(wine)

##     PH ALCOHOL QUALITY   COUNTRY  PRICE RATE YEAR
## 1 3.57    10.2       5     Italy 145.65   93 1982
## 2 3.20     9.8       5 Argentina  17.12   53 2000
## 3 3.42    11.0       6 Kazahstan  63.31    8 2003
## 4 3.52    11.2       6 Argentina  89.87   82 2002
## 5 3.45    10.5       5 Argentina 128.45   93 2007
## 6 3.51     9.4       5 Kazahstan  20.38   66 1973

Barplot

# Basic barplot
p<-ggplot(data=wine, aes(x=QUALITY, y=PRICE)) +theme_minimal()+
   geom_bar(stat="identity", fill = wine$QUALITY)

q<-p+coord_flip() #horizontal barplot

#Let's change the width of the bar
r<-ggplot(data=wine, aes(x=QUALITY, y=PRICE)) +theme_minimal()+
   geom_bar(stat="identity", fill = wine$QUALITY, width = 0.4)

s<- ggplot(data=wine, aes(x=QUALITY, y=PRICE)) +theme_minimal()+
    geom_bar(stat="identity")+ 
    scale_x_discrete(limits=c(5,8))#Let's look at only the Quality 5,8

grid.arrange(p,q,r,s, nrow=2)

It is also possible to change manually barplot fill colors using the functions :

scale_fill_manual() : to use custom colors scale_fill_brewer() : to use color palettes from RColorBrewer package scale_fill_grey() : to use grey color palettes

# Use custom color palettes
a<-ggplot(data=wine, aes(x=QUALITY, y=PRICE, fill = QUALITY))+
   geom_bar(stat="identity")+
   theme_minimal()+
   theme(legend.position = "top")+ #Change legend position
  scale_fill_manual(values=c("red","orange","blue","thistle","cyan","green"))

# use brewer color palettes
b<-ggplot(data=wine, aes(x=QUALITY, y=PRICE, fill = QUALITY))+
   geom_bar(stat="identity")+
   theme_minimal()+
   theme(legend.position = "bottom")+ #Change legend position
   scale_fill_brewer(palette="Dark2")

# Use grey scale
c<-ggplot(data=wine, aes(x=QUALITY, y=PRICE, fill = QUALITY))+
   geom_bar(stat="identity")+
   theme_minimal()+
   scale_fill_grey()

grid.arrange(a,b,c, nrow = 2)

Stacked Barplot

It is created by default. You can use the function position_dodge() to change this.

wine$COUNTRY<-as.factor(wine$COUNTRY)
a<-ggplot(data=wine, aes(x=QUALITY, y=PRICE, fill = COUNTRY))+
   geom_bar(stat="identity")+
   theme_light()

# use brewer color palettes
b<-ggplot(data=wine, aes(x=QUALITY, y=PRICE, fill = COUNTRY))+
   geom_bar(stat="identity", position = position_dodge())+
   theme_minimal()+
   scale_fill_brewer(palette="Dark2")

grid.arrange(a,b, nrow=2)

Box Plot

The function geom_boxplot() is used

geom_boxplot(outlier.colour="black", outlier.shape=16,outlier.size=2, notch=FALSE)

## geom_boxplot: outlier.colour = black, outlier.fill = NULL, outlier.shape = 16, outlier.size = 2, outlier.stroke = 0.5, outlier.alpha = NULL, notch = FALSE, notchwidth = 0.5, varwidth = FALSE, na.rm = FALSE, orientation = NA
## stat_boxplot: na.rm = FALSE, orientation = NA
## position_dodge2

Details

outlier.colour, outlier.shape, outlier.size : The color, the shape and the size for outlying points notch : logical value. If TRUE, make a notched box plot. The notch displays a confidence interval around the median which is normally based on the median +/- 1.58*IQR/sqrt(n). Notches are used to compare groups; if the notches of two boxes do not overlap, this is a strong evidence that the medians differ.

Basic box plot

p <- ggplot(wine, aes(x=QUALITY, y=PRICE)) + 
     geom_boxplot()

# Rotate the box plot
b<-p + coord_flip()

# Notched box plot
c<-ggplot(wine, aes(x=QUALITY, y=PRICE)) + 
   geom_boxplot(notch=TRUE)

# Change outlier, color, shape and size
d<-ggplot(wine, aes(x=QUALITY, y=PRICE)) + 
   geom_boxplot(outlier.colour="red",outlier.shape=8,outlier.size=4)

grid.arrange(p,b,c,d)

The function stat_summary() can be used to add mean points to a box plot :

# Box plot with mean points
a<-ggplot(wine, aes(x=QUALITY, y=PRICE, fill = QUALITY)) + 
   geom_boxplot() + 
   stat_summary(fun=mean, geom="point", shape=23, size=4)
a

Histogram

For histogram we will be using the inbuilt mtcars dataset

data(mtcars)
cars<-mtcars
head(cars)

##                    mpg cyl disp  hp drat    wt  qsec vs am gear carb
## Mazda RX4         21.0   6  160 110 3.90 2.620 16.46  0  1    4    4
## Mazda RX4 Wag     21.0   6  160 110 3.90 2.875 17.02  0  1    4    4
## Datsun 710        22.8   4  108  93 3.85 2.320 18.61  1  1    4    1
## Hornet 4 Drive    21.4   6  258 110 3.08 3.215 19.44  1  0    3    1
## Hornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2
## Valiant           18.1   6  225 105 2.76 3.460 20.22  1  0    3    1

We will use geom_histogram function

geom_histogram(mapping = NULL, data = NULL, binwidth = 0.5)

## geom_bar: na.rm = FALSE, orientation = NA
## stat_bin: binwidth = 0.5, bins = NULL, na.rm = FALSE, orientation = NA, pad = FALSE
## position_stack

gp0 <- ggplot(data = cars, 
              aes(x = mpg))+ 
               geom_histogram(
               aes(y=..density..),
               fill = "thistle",
               binwidth = 0.5)+
               theme_minimal()
               
gp1 <- ggplot(data = cars, 
              aes(x = mpg))+ 
               geom_histogram(
               aes(y=..density..), 
               colour = "black", #color of the border 
               fill = "thistle",#color inside the bin
               binwidth = 1) + 
               geom_density(alpha=.2)+ #Add density curve
               theme_minimal()

#Overlay with density plot. The value of alpha controls the level of transparency
gp2 <- ggplot(data = cars, 
              aes(x = mpg, fill= as.factor(cyl)))+ 
               geom_histogram(
               aes(y=..density..), 
               binwidth = 1) + 
               geom_density(alpha=.2) #Add density curve
gp2<-gp2+ scale_fill_discrete(name = "No. of Cylinders")+ theme_minimal()
grid.arrange(gp0,gp1,gp2)

Let’s add mean line and density plot on the histogram

# Add mean line
a<-ggplot(data = cars,aes(x = mpg))+ 
               geom_histogram(aes(y=..density..), 
               colour = "black", 
               fill = "thistle",
               binwidth = 0.5) + 
               theme_minimal()+
               geom_density(alpha=.2)+ #Add density curve
               geom_vline(aes(xintercept=mean(mpg)),color="red", linetype="dashed", size=1)

b<-ggplot(data = cars,aes(x = mpg))+ 
               geom_histogram(aes(y=..density..), 
               colour = "black", 
               fill = "thistle",
               linetype = "dashed",
               binwidth = 0.5) + 
               theme_minimal()+
               geom_density(alpha=.2)+ #Add density curve
               geom_vline(aes(xintercept=mean(mpg)),color="red", linetype="dashed", size=1)
grid.arrange(a,b)

Let’s Use facets to Split the plots into multiple panels:

b<-b+facet_grid(cyl~.)
plot(b)

Scatter plot

The function geom_point() is used

gp1 <- ggplot(data=cars, aes(x=wt, y=mpg)) +
    geom_point(size=2, shape=21, fill= "thistle")+theme_minimal() +
    labs(title = "Weight of Cars V/S Miles Per Gallon") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")
gp1

As expected we can see that with increasing weight of the cars, the miles per gallon decreases

Let’s explore few of the shapes and point sizes in the scatter plot

gp2 <- ggplot(data=cars, aes(x=wt, y=mpg)) +
    geom_point(size=2, shape=21)+theme_minimal() +
    labs(title = "Shape 1, Size 2") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

gp3 <- ggplot(data=cars, aes(x=wt, y=mpg)) +
    geom_point(size=3, shape=18)+theme_minimal() +
    labs(title = "Shape 18, Size 3") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

gp4 <- ggplot(data=cars, aes(x=wt, y=mpg)) +
    geom_point(size=1, shape=13)+theme_minimal() +
    labs(title = "Shape 13, Size 1") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

gp5 <- ggplot(data=cars, aes(x=wt, y=mpg)) +
    geom_point(size=4, shape=1)+theme_minimal() +
    labs(title = "Shape 1, Size 4") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

grid.arrange(gp2,gp3,gp4,gp5, nrow=2, ncol=2)

Let’s color the points based on the number of cylinders in the car

cars$cyl<- as.factor(cars$cyl)
head(cars)

##                    mpg cyl disp  hp drat    wt  qsec vs am gear carb
## Mazda RX4         21.0   6  160 110 3.90 2.620 16.46  0  1    4    4
## Mazda RX4 Wag     21.0   6  160 110 3.90 2.875 17.02  0  1    4    4
## Datsun 710        22.8   4  108  93 3.85 2.320 18.61  1  1    4    1
## Hornet 4 Drive    21.4   6  258 110 3.08 3.215 19.44  1  0    3    1
## Hornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2
## Valiant           18.1   6  225 105 2.76 3.460 20.22  1  0    3    1

gp6 <- ggplot(data=cars, aes(x=wt, y=mpg, fill = cyl)) +
    geom_point(size=2, shape=21)+theme_minimal() +
    geom_label(label=rownames(cars), position=position_dodge(width=0.5),hjust=1,)+ #Add text next to each point
    labs(title = "Weight of Cars V/S Miles Per Gallon") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")
gp6

Add regression lines to the scatter plot

geom_smooth(method="auto", se=TRUE, fullrange=FALSE, level=0.95)

## geom_smooth: na.rm = FALSE, orientation = NA, se = TRUE
## stat_smooth: na.rm = FALSE, orientation = NA, se = TRUE, fullrange = FALSE, level = 0.95, method = auto
## position_identity

• method : smoothing method to be used. Possible values are lm, glm, gam, loess, rlm.
• method = “loess”: This is the default value for small number of observations. It computes a smooth local regression. You can read more about loess using the R code ?loess.
• method =“lm”: It fits a linear model. Note that, it’s also possible to indicate the formula as formula = y ~ poly(x, 3) to specify a degree 3 polynomial.
• se : logical value. If TRUE, confidence interval is displayed around smooth. A simplified format is :
• fullrange : logical value. If TRUE, the fit spans the full range of the plot
• level : level of confidence interval to use. Default value is 0.95

gp7<-ggplot(data=cars, aes(x=wt, y=mpg)) +
     geom_point(size=1, shape=22)+theme_minimal() +  
     geom_smooth(method=lm)+
     labs(title = "Weight of Cars V/S Miles Per Gallon") +
     labs(x="Weight of Car")+labs(y="Miles Per Gallon")

#Let's get rid of the confidence interval
gp8<-ggplot(data=cars, aes(x=wt, y=mpg)) +
     geom_point(size=1, shape=22)+theme_minimal() +  
     geom_smooth(method=lm, se = FALSE, linetype = "dashed")+
     labs(title = "Weight of Cars V/S Miles Per Gallon") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

# Add regression lines
gp9<-ggplot(data=cars, aes(x=wt, y=mpg, col=cyl, shape =cyl)) +
     geom_point(size=1, shape=22)+theme_minimal() +  
     geom_smooth(method=lm)+
     labs(title = "Weight of Cars V/S Miles Per Gallon") +
    labs(x="Weight of Car")+labs(y="Miles Per Gallon")

# Remove confidence intervals
gp10<-ggplot(data=cars, aes(x=wt, y=mpg, col=cyl, shape =cyl)) +
      geom_point(size=1, shape=22)+theme_minimal() +  
      geom_smooth(method=lm, se = FALSE)+
      labs(title = "Weight of Cars V/S Miles Per Gallon") +
      labs(x="Weight of Car")+labs(y="Miles Per Gallon")

grid.arrange(gp7,gp8,gp9,gp10, nrow =2)

It is also possible to change manually point and line colors using the functions :

scale_color_brewer() : to use color palettes from RColorBrewer package scale_color_grey() : to use grey color palettes

p <- ggplot(cars, aes(x=wt, y=mpg, color=cyl, shape=cyl)) +
     geom_point() + 
     geom_smooth(method=lm, se=FALSE, fullrange=TRUE)+
     labs(x="Weight of Car")+labs(y="Miles Per Gallon")+
     theme_minimal()

# Use brewer color palettes
a<-p+scale_color_brewer(palette="Dark2")

# Use grey scale
b<-p + scale_color_grey()

grid.arrange(a,b, nrow =1)

Further Resources

• https://r4ds.had.co.nz/data-visualisation.html
• https://www.maths.usyd.edu.au/u/UG/SM/STAT3022/r/current/Misc/data-visualization-2.1.pdf
• http://r-statistics.co/Complete-Ggplot2-Tutorial-Part1-With-R-Code.html