Introduction

This is the last post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to combine multiple plots. In this post, we will learn to modify the appearance of all non data components of the plot such as:

• axis
• legend
• panel
• plot area
• background
• margin
• facets

Introduction

This is the 19th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to modify the title, label and bar of a legend. In this post, we will learn about faceting i.e. combining plots.

Introduction

This is the 18th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to modify the legend of plot when alpha is mapped to a categorical variable. In this post, we will learn to modify legend

• title
• label
• and bar

So far, we have learnt to modify the components of a legend using scale_* family of functions. Now, we will use the guide argument and supply it values using the guide_legend() function.

Introduction

This is the 17th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to modify the legend of plot when size is mapped to continuous variable. In this post, we will learn to modify the following using scale_alpha_continuous() when alpha or transparency is mapped to variables:

• title
• breaks
• limits
• range
• labels
• values

Introduction

This is the 16th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to modify the legend of plot when shape is mapped to categorical variables. In this post, we will learn to modify the following using scale_size_continuous when size aesthetic is mapped to variables:

• title
• breaks
• limits
• range
• labels
• values

Introduction

This is the 15th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to modify the legend of plots when aesthetics are mapped to variables..In this post, we will learn to modify the following using scale_shape_manual when shape is mapped to categorical variables:

• title
• breaks
• limits
• labels
• values

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

Introduction

This is the 14th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to modify the legend of plots when aesthetics are mapped to variables. In this post, we will continue to explore different ways to modify/customize the legend of plots.

Specifically, we will learn to modify the following using scale_fill_manual() when fill is mapped to categorical variables:

• title
• breaks
• limits
• labels
• values

Introduction

This is the 13th post in the series Elegant Data Visualization with ggplot2. In the previos post, we learnt how to modify the axis of plots. In this post, we will focus on modifying the appearance of legend of plots when the aesthetics are mapped to variables. Specifically, we will learn to modify the following when color is mapped to categorical variables:

• title
• breaks
• limits
• labels
• values

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

Basic Plot

Let us start with a scatter plot examining the relationship between displacement and miles per gallon from the mtcars data set. We will map the color of the points to the cyl variable.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl)))

As you can see, the legend acts as a guide for the color aesthetic. Now, let us learn to modify the different aspects of the legend.

Values

To change the default colors in the legend, use the values argument and supply a character vector of color names. The number of colors specified must be equal to the number of levels in the categorical variable mapped. In the below example, cyl has 3 levels (4, 6, 8) and hence we have specified 3 colors.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(values = c("red", "blue", "green"))

Title

In the previous example, the title of the legend (factor(cyl)) is not very intuitive. If the user does not know the underlying data, they will not be able to make any sense out of it. Let us change it to Cylinders using the name argument.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(name = "Cylinders", 
    values = c("red", "blue", "green"))

Now, the user will know that the different colors represent number of cylinders in the car.

Limits

Let us assume that we want to modify the data to be displayed i.e. instead of examining the relationship between mileage and displacement for all cars, we desire to look at only cars with at least 6 cylinders. One way to approach this would be to filter the data using filter from dplyr and then visualize it. Instead, we will use the limits argument and filter the data for visualization.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(values = c("red", "blue", "green"), limits = c(6, 8))

## Warning: Removed 11 rows containing missing values (geom_point).

As you can see above, ggplot2 returns a warning message indicating data related to 4 cylinders has been dropped. If you observe the legend, it now represents only 4 and 6 cylinders.

Labels

The labels in the legend can be modified using the labels argument. Let us change the labels to Four, Six and Eight in the next example. Ensure that the labels are intuitive and easy to interpret for the end user of the plot.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(values = c("red", "blue", "green"),
    labels = c('Four', 'Six', 'Eight'))

Breaks

When there are large number of levels in the mapped variable, you may not want the labels in the legend to represent all of them. In such cases, we can use the breaks argument and specify the labels to be used. In the below case, we use the breaks argument to ensure that the labels in legend represent two levels (4, 8) of the mapped variable.

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(values = c("red", "blue", "green"),
    breaks = c(4, 8))

Putting it all together…

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl))) +
  scale_color_manual(name = "Cylinders", values = c("red", "blue", "green"),
    labels = c('Four', 'Six', 'Eight'), limits = c(4, 6, 8), breaks = c(4, 6, 8))

Summary

In this post, we learnt to modify the following aspects of legends:

• title
• breaks
• limits
• labels
• values

Up Next..

In the next post, we will learn how to modify legend when fill is mapped to variables.

Introduction

This is the twelfth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to build histograms. Now that we have learnt to build different plots, let us look at different ways to modify the axis. Along the way, we will also explore the scale_*() family of functions.

Modify X and Y axis

• title
• labels
• limits
• breaks
• position

In this module, we will learn how to modify the X and Y axis using the following functions:

• Continuous Axis
  • scale_x_continuous()
  • scale_y_continuous()
• Discrete Axis
  • scale_x_discrete()
  • scale_y_discrete()

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

Continuous Axis

If the X and Y axis represent continuous data, we can use scale_x_continuous() and scale_y_continuous() to modify the axis. They take the following arguments:

• name
• limits
• breaks
• labels
• position

Let us continue with the scatter plot we have used in previous posts.

ggplot(mtcars) +
  geom_point(aes(disp, mpg))

The name argument is used to modify the X axis label. In the below example, we change the X axis label to ‘Displacement’. In previous posts, we have used xlab() to work with the X axis label.

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  scale_x_continuous(name = "Displacement")

To modify the range, use the limits argument. It takes a vector of length 2 i.e. 2 values, the lower and upper limit of the range. It is an alternative for xlim().

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  scale_x_continuous(limits = c(0, 600))

In the above plot, the ticks on the X axis appear at 0, 200, 400 and 600. Let us say we want the ticks to appear more closer i.e. the difference between the tick should be reduced by 50. The breaks argument will allow us to specify where the ticks appear. It takes a numeric vector equal to the length of the number of ticks.

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  scale_x_continuous(breaks = c(150, 300, 450))

We can change the tick labels using the labels argument. In the below example, we use words instead of numbers. When adding labels, we need to ensure that the length of the breaks and labels are same.

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  scale_x_continuous(breaks = c(150, 300, 450),
    labels = c('One Hundred Fifty', 'Three Hundred', 'Four Hundred Fifity'))

The position of the axes can be changed using the position argument. In the below example, we can move the axes to the top of the plot by supplying the value ‘top’.

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  scale_x_continuous(position = 'top')

Putting it all together..

ggplot(mtcars) + geom_point(aes(disp, mpg)) +
  scale_x_continuous(name = "Displacement", limits = c(0, 600),
                     breaks = c(0, 150, 300, 450, 600), position = 'top',
                     labels = c('0', '150', '300', '450', '600'))

Y Axis - Continuous

ggplot(mtcars) + geom_point(aes(disp, mpg)) +
  scale_y_continuous(name = "Miles Per Gallon", limits = c(0, 45),
                     breaks = c(0, 15, 30, 45), position = 'right',
                     labels = c('0', '15', '30', '45'))

Discrete Axis

If the X and Y axis represent discrete or categorical data, scale_x_discrete() and scale_y_discrete() can be used to modify them. They take the following arguments:

• name
• labels
• breaks
• position

The above options serve the same purpose as in the case of continuous scales.

ggplot(mtcars) +
  geom_bar(aes(factor(cyl))) +
  scale_x_discrete(name = "Number of Cylinders")

ggplot(mtcars) +
  geom_bar(aes(factor(cyl))) +
  scale_x_discrete(labels = c("4" = "Four", "6" = "Six", "8" = "Eight"))

ggplot(mtcars) +
  geom_bar(aes(factor(cyl))) +
  scale_x_discrete(breaks = c("4", "6", "8"))

ggplot(mtcars) +
  geom_bar(aes(factor(cyl))) +
  scale_x_discrete(position = 'bottom')

ggplot(mtcars) + geom_bar(aes(factor(cyl))) +
  scale_x_discrete(name = "Number of Cylinders",
    labels = c("4" = "Four", "6" = "Six", "8" = "Eight"),
    breaks = c("4", "6", "8"), position = "bottom")

Summary

In this post, we learnt to modify

• title
• labels
• limits
• breaks
• position

Up Next..

In the next post, we will learn to modify the legend when color is mapped to a variable.

Introduction

This is the eleventh post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to build box plots. In this post, we will learn to

• build histogram
• specify bins
• modify
  • color
  • fill
  • alpha
  • bin width
  • line type
  • line size
• map aesthetics to variables

A histogram is a plot that can be used to examine the shape and spread of continuous data. It looks very similar to a bar graph and can be used to detect outliers and skewness in data. The histogram graphically shows the following:

• center (location) of the data
• spread (dispersion) of the data
• skewness
• outliers
• presence of multiple modes

To construct a histogram, the data is split into intervals called bins. The intervals may or may not be equal sized. For each bin, the number of data points that fall into it are counted (frequency). The Y axis of the histogram represents the frequency and the X axis represents the variable.

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

Data

ecom <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/web.csv')
ecom

## # A tibble: 1,000 x 11
##       id referrer device bouncers n_visit n_pages duration country purchase
##    <dbl> <chr>    <chr>  <lgl>      <dbl>   <dbl>    <dbl> <chr>   <lgl>   
##  1     1 google   laptop TRUE          10       1      693 Czech ~ FALSE   
##  2     2 yahoo    tablet TRUE           9       1      459 Yemen   FALSE   
##  3     3 direct   laptop TRUE           0       1      996 Brazil  FALSE   
##  4     4 bing     tablet FALSE          3      18      468 China   TRUE    
##  5     5 yahoo    mobile TRUE           9       1      955 Poland  FALSE   
##  6     6 yahoo    laptop FALSE          5       5      135 South ~ FALSE   
##  7     7 yahoo    mobile TRUE          10       1       75 Bangla~ FALSE   
##  8     8 direct   mobile TRUE          10       1      908 Indone~ FALSE   
##  9     9 bing     mobile FALSE          3      19      209 Nether~ FALSE   
## 10    10 google   mobile TRUE           6       1      208 Czech ~ FALSE   
## # ... with 990 more rows, and 2 more variables: order_items <dbl>,
## #   order_value <dbl>

Data Dictionary

• id: row id
• referrer: referrer website/search engine
• os: operating system
• browser: browser
• device: device used to visit the website
• n_pages: number of pages visited
• duration: time spent on the website (in seconds)
• repeat: frequency of visits
• country: country of origin
• purchase: whether visitor purchased
• order_value: order value of visitor (in dollars)

ggplot(ecom) +
  geom_histogram(aes(n_visit))

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 7)

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 7, fill = 'blue')

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 7, fill = 'blue', alpha = 0.3)

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 7, fill = 'white', color = 'blue')

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 7, fill = 'blue', color = 'white')

ggplot(ecom) +
  geom_histogram(aes(n_visit), binwidth = 2, fill = 'blue', color = 'black')

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 5, fill = 'white', 
    color = 'blue', linetype = 3)

ggplot(ecom) +
  geom_histogram(aes(n_visit), bins = 5, fill = 'white', 
    color = 'blue', size = 1.25)

ggplot(ecom) +
  geom_histogram(aes(n_visit, fill = device), bins = 7)

Summary

In this post, we learnt to:

• build histogram
• specify bins
• modify
  • color
  • fill
  • alpha
  • bin width
  • line type
  • line size
• map aesthetics to variables

Up Next..

In the next post, we will learn to modify the axes of a plot.

Introduction

This is the 9th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to build bar charts. In this post, we will learn to:

• build box plots
• modify box
  • color
  • fill
  • alpha
  • line size
  • line type
• modify outlier
  • color
  • shape
  • size
  • alpha

The box plot is a standardized way of displaying the distribution of data. It is useful for detecting outliers and for comparing distributions and shows the shape, central tendancy and variability of the data.

Structure

• the body of the boxplot consists of a “box” (hence, the name), which goes from the first quartile (Q1) to the third quartile (Q3)
• within the box, a vertical line is drawn at the Q2, the median of the data set
• two horizontal lines, called whiskers, extend from the front and back of the box
• the front whisker goes from Q1 to the smallest non-outlier in the data set, and the back whisker goes from Q3 to the largest non-outlier
• if the data set includes one or more outliers, they are plotted separately as points on the chart

Data

We are going to use two different data sets in this post. Both the data sets have the same data but are in different formats.

daily_returns <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/tickers.csv')
daily_returns

## # A tibble: 250 x 5
##       AAPL   AMZN      FB    GOOG    MSFT
##      <dbl>  <dbl>   <dbl>   <dbl>   <dbl>
##  1  1.38    24.2   2.12    22.4    1.12  
##  2  2.83     3.25 -0.860    5.99   0.767 
##  3 -0.0394   9.91  1.45     6.75   0.973 
##  4  0.108    3.76 -0.770  -10.7   -0.285 
##  5  1.64    19.8   4.75     8.66   0.501 
##  6  0.0689   5.33 -0.300   -0.930  0.256 
##  7 -0.561   -5.21 -0.630   -7.28  -0.708 
##  8  0.551    0.25 -0.460    0.690  0.128 
##  9 -0.217  -13.6   0.0300   6.56   0.0786
## 10 -0.108   -4.25  0.460    2.60   0.472 
## # ... with 240 more rows

Univariate Box Plot

If you are not comparing the distribution of continuous data, you can create box plot for a single variable. Unlike plot(), where we could just use 1 input, in ggplot2, we must specify a value for the X axis and it must be categorical data. Since we are not comparing distributions, we will use 1 as the value for the X axis and wrap it inside factor() to treat it as a categorical variable. In the below example, we examine the distribution of stock returns of Apple.

ggplot(daily_returns) +
  geom_boxplot(aes(x = factor(1), y = AAPL))

Data

For the rest of the post, we will use the below data set. Instead of 5 columns, we have two columns. One for the stock names and another for returns.

tidy_returns <- 
  read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/tidy_tickers.csv',
  col_types = list(col_factor(levels = c('AAPL', 'AMZN', 'FB', 'GOOG', 'MSFT')), col_double()))
tidy_returns

## # A tibble: 1,254 x 2
##    stock returns
##    <fct>   <dbl>
##  1 AAPL   1.38  
##  2 AAPL   2.83  
##  3 AAPL  -0.0394
##  4 AAPL   0.108 
##  5 AAPL   1.64  
##  6 AAPL   0.0689
##  7 AAPL  -0.561 
##  8 AAPL   0.551 
##  9 AAPL  -0.217 
## 10 AAPL  -0.108 
## # ... with 1,244 more rows

Box Plot

With the above data, let us create a box plot where we compate the distribution of stock returns of different companies. We map X axis to the column with stock names and Y axis to the column with stock returns. Note that, the column names are wrapped inside aes().

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns))

To create a horizontal bar plot, we can use coord_flip() which will flip the coordinate axes.

Horizontal Box Plot

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns)) +
  coord_flip()

Notch

Notches are used to compare medians. You can use the notch argument and set it to TRUE.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns),
    notch = TRUE) 

Jitter

Just for comparison, let us plot the returns as points on top of the box plot using geom_jitter(). We modify the color of the points using the color argument and the spread using the width argument.

ggplot(tidy_returns, aes(x = stock, y = returns)) +
  geom_boxplot() +
  geom_jitter(width = 0.2, color = 'blue')

Outliers

To highlight extreme observations, we can modify the appearance of outliers using the following:

• color
• shape
• size
• alpha

To modify the color of the outliers, use the outlier.color argument. The color can be specified either using its name or the associated hex code.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), outlier.color = 'red')

The shape of the outlier can be modified using the outlier.shape argument. It can take values between 0 and 25.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), outlier.shape = 23) 

The size of the outlier can be modified using the outlier.size argument. It can take any value greater than 0.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), outlier.size = 3) 

You can play around with the transparency of the outlier using the outlier.alpha argument. It can take values between 0 and 1.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), outlier.color = 'blue', outlier.alpha = 0.3) 

Box Aesthetics

The appearance of the box can be controlled using the following:

• color
• fill
• alpha
• line type
• line width

Specify Values

The background color of the box can be modified using the fill argument. The color can be specified either using its name or the associated hex code.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), fill = c('blue', 'red', 'green', 'yellow', 'brown')) 

To modify the transparency of the background color, use the alpha argument. It can take any value between 0 and 1.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), fill = 'blue', alpha = 0.3) 

The color of the border can be modified using the color argument. The color can be specified either using its name or the associated hex code.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), color = c('blue', 'red', 'green', 'yellow', 'brown')) 

The width of the border can be changed using the size argument. It can take any value greater than 0.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), size = 1.5) 

To change the line type of the border, use the linetype argument. It can take any value between 0 and 6.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns), linetype = 2) 

Map Variables

Instead of specifying values, we can map fill and color to variables as well. In the below example, we map fill to the variable stock. It assigns different colors to the different stocks.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns, fill = stock)) 

Let us map color to the variable stock. It will assign different colors to the box borders.

ggplot(tidy_returns) +
  geom_boxplot(aes(x = stock, y = returns, color = stock)) 

Introduction

This is the ninth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to build line charts. In this post, we will learn to:

• build
  • simple bar plot
  • stacked bar plot
  • grouped bar plot
  • proportional bar plot
• map aesthetics to variables
• specify values for
  • bar color
  • bar line color
  • bar line type
  • bar line size

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

Data

ecom <- read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/ecom.csv',
  col_types = list(col_factor(levels = c('Desktop', 'Mobile', 'Tablet')), 
  col_factor(levels = c(TRUE, FALSE)), col_factor(levels = c(TRUE, FALSE)), 
  col_factor(levels = c('Affiliates', 'Direct', 'Display', 'Organic', 'Paid', 'Referral', 'Social'))))
ecom

## # A tibble: 5,000 x 4
##    device  bouncers purchase referrer  
##    <fct>   <fct>    <fct>    <fct>     
##  1 Desktop FALSE    FALSE    Affiliates
##  2 Mobile  FALSE    FALSE    Affiliates
##  3 Desktop TRUE     FALSE    Organic   
##  4 Desktop FALSE    FALSE    Organic   
##  5 Mobile  TRUE     FALSE    Direct    
##  6 Desktop TRUE     FALSE    Direct    
##  7 Desktop FALSE    FALSE    Referral  
##  8 Tablet  TRUE     FALSE    Organic   
##  9 Mobile  TRUE     FALSE    Social    
## 10 Desktop TRUE     FALSE    Organic   
## # ... with 4,990 more rows

Data Dictionary

• device: device used to visit the website
• bouncers: whether visit was a bouncer (exit website from landing page)
• purchase: whether visitor purchased
• referrer: referrer website/search engine

Aesthetics

• fill
• color
• linetype
• size
• position

ggplot(ecom) +
  geom_bar(aes(device))

ggplot(ecom) +
  geom_bar(aes(device), fill = c('red', 'blue', 'green'))

ggplot(ecom) +
  geom_bar(aes(device, fill = referrer))

ggplot(ecom) +
  geom_bar(aes(device, fill = referrer), position = 'dodge')

ggplot(ecom) +
  geom_bar(aes(device, fill = referrer), position = 'fill')

ggplot(ecom) +
  geom_bar(aes(device, fill = referrer)) +
  coord_flip()

ggplot(ecom) +
  geom_bar(aes(device), fill = 'white', color = c('red', 'blue', 'green'))

ggplot(ecom) +
  geom_bar(aes(device), fill = 'white',  color = 'black', linetype = 2)

ggplot(ecom) +
  geom_bar(aes(device), fill = 'white', color = 'black', size = 2)

Summary

In this post, we learnt to:

• build
  • simple bar plot
  • stacked bar plot
  • grouped bar plot
  • proportional bar plot
• map aesthetics to variables
• specify values for
  • bar color
  • bar line color
  • bar line type
  • bar line size

Up Next..

In the next post, we will learn to build box plots.

Introduction

This is the 8th post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to build scatter plots. In this post, we will learn to:

• build
  • simple line chart
  • grouped line chart
• map aesthetics to variables
• modify line
  • color
  • type
  • size

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

gdp <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/gdp.csv')

## Warning: Missing column names filled in: 'X1' [1]

gdp

## # A tibble: 6 x 6
##      X1     X year       growth india china
##   <dbl> <dbl> <date>      <dbl> <dbl> <dbl>
## 1     1     1 2000-01-01      6     5     8
## 2     2     2 2001-01-01      9     9     5
## 3     3     3 2002-01-01      8     8     6
## 4     4     4 2003-01-01      9     8     8
## 5     5     5 2004-01-01      9     5     9
## 6     6     6 2005-01-01      8     7     8

ggplot(gdp, aes(year, india)) +
  geom_line()

ggplot(gdp, aes(year, india)) +
  geom_line(color = 'blue')

ggplot(gdp, aes(year, india)) +
  geom_line(linetype = 2)

ggplot(gdp, aes(year, india)) +
  geom_line(linetype = 'dashed')

ggplot(gdp, aes(year, india)) +
  geom_line(size = 2)

gdp2 <- gdp %>% 
  select(year, growth, india, china) %>% 
  gather(key = country, value = gdp, -year)

gdp2

## # A tibble: 18 x 3
##    year       country   gdp
##    <date>     <chr>   <dbl>
##  1 2000-01-01 growth      6
##  2 2001-01-01 growth      9
##  3 2002-01-01 growth      8
##  4 2003-01-01 growth      9
##  5 2004-01-01 growth      9
##  6 2005-01-01 growth      8
##  7 2000-01-01 india       5
##  8 2001-01-01 india       9
##  9 2002-01-01 india       8
## 10 2003-01-01 india       8
## 11 2004-01-01 india       5
## 12 2005-01-01 india       7
## 13 2000-01-01 china       8
## 14 2001-01-01 china       5
## 15 2002-01-01 china       6
## 16 2003-01-01 china       8
## 17 2004-01-01 china       9
## 18 2005-01-01 china       8

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line()

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(color = country))

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(linetype = country))

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(size = country))

## Warning: Using size for a discrete variable is not advised.

Summary

In this post, we learnt to:

• build
  • simple line chart
  • grouped line chart
• map aesthetics to variables
• modify line
  • color
  • type
  • size

Up Next..

In the next post, we will learn to build bar plots.

Introduction

This is the fifth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt about text annotations. In this post, we will:

• build scatter plots
• modify point
  • color
  • fill
  • alpha
  • shape
  • size
• fit regression line

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

ggplot(mtcars) +
  geom_point(aes(disp, mpg))

ggplot(mtcars) +
  geom_point(aes(disp, mpg), position = 'jitter')

ggplot(mtcars) +
  geom_jitter(aes(disp, mpg))

ggplot(mtcars) +
  geom_point(aes(disp, mpg), color = 'blue', position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg), color = 'blue', alpha = 0.4, position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg), shape = 3, position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg), size = 3, position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = factor(cyl)), 
             position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg, color = hp), 
             position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg, shape = factor(cyl)), position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg, size = hp), color = 'blue', position = 'jitter')

ggplot(mtcars) +
  geom_point(aes(disp, mpg, size = factor(cyl)), color = 'blue', position = 'jitter')

## Warning: Using size for a discrete variable is not advised.

ggplot(mtcars, aes(disp, mpg)) +
  geom_point(position = 'jitter') +
  geom_smooth(method = 'lm', se = FALSE)

## `geom_smooth()` using formula 'y ~ x'

lm(mpg ~ disp, data = mtcars)

## 
## Call:
## lm(formula = mpg ~ disp, data = mtcars)
## 
## Coefficients:
## (Intercept)         disp  
##    29.59985     -0.04122

ggplot(mtcars, aes(disp, mpg)) +
  geom_point(position = 'jitter') +
  geom_abline(slope = -0.04122, intercept = 29.59985)

ggplot(mtcars, aes(disp, mpg)) +
  geom_point(position = 'jitter') +
  geom_smooth(method = 'lm', se = TRUE)

## `geom_smooth()` using formula 'y ~ x'

ggplot(mtcars, aes(disp, mpg)) +
  geom_point(position = 'jitter') +
  geom_smooth(method = 'loess', se = FALSE)

## `geom_smooth()` using formula 'y ~ x'

Summary

In this post, we learnt to:

• build scatter plots
• map aesthetics to variables
• fit regression line

Up Next..

In the next post, we will learn to build line charts.

Introduction

This is the sixth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt to modify the axis and plot labels. In this post, we will learn to add text to the plots.

• add custom text
• modify color
• modify size
• modify fontface
• modify angle

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text', color = 'red')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text', size = 6)

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text', fontface = 'bold')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text', angle = 25)

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  annotate('text', x = 200, y = 30, label = 'Sample Text',
           color = 'red', size = 6, fontface = 'bold', angle = 25)

Introduction

This is the fifth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt about aesthetics. In this post, we will learn to:

• add title and subtitle to the plot
• modify axis labels
• modify axis range
• remove axis
• format axis

ggplot(mtcars) +
  geom_point(aes(disp, mpg))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  ggtitle(label = 'Displacement vs Mileage', subtitle = 'disp vs mpg')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  xlab('Displacement') + ylab('Miles Per Gallon')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  labs(title = 'Displacement vs Mileage', subtitle = 'disp vs mpg', 
    x = 'Displacement', y = 'Miles Per Gallon')

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  xlim(c(0, 600))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  ylim(c(0, 40))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  expand_limits(x = c(0, 600), y = c(0, 40))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) +
  theme(axis.title.x = element_blank(), axis.title.y = element_blank())

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(color = 'blue'))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(family = 'Arial'))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(face = 'bold'))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(size = 8))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(hjust = 0.3))

Putting it all together…

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ggtitle('Diaplacement vs Mileage') +
  theme(plot.title = element_text(color = 'blue', family = 'Arial',
    face = 'bold', size = 12, hjust = 0.5))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + xlab('Diaplacement') +
  theme(axis.title.x = element_text(color = 'blue', family = 'Arial',
    face = 'bold', size = 8, hjust = 0.5, angle = 15))

ggplot(mtcars) +
  geom_point(aes(disp, mpg)) + ylab('Miles Per Gallon') +
  theme(axis.title.y = element_text(color = 'blue', family = 'Arial',
    face = 'italic', size = 8, vjust = 0.3, angle = 90))

Summary

In this post, we learnt to:

• add title and subtitle to the plot
• modify axis labels
• modify axis range
• remove axis
• format axis

Up Next..

In the next post, we will learn to add text annotations to plots.

Introduction

This is the fourth post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt about geoms and how we can use them to build different plots. In this post, we will focus on the aesthetics i.e.  color, shape, size, alpha, line type, line width etc. We can map these to variables or specify values for them. If we want to map the above to variables, we have to specify them within the aes() function. We will look at both methods in the following sections.

Explore aesthetics such as

• color
• shape
• size
• fill
• alpha
• width

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2
• dplyr
• tidyr

All the data sets used in this post can be found here and code can be downloaded from here.

Data

ecom <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/web.csv')
ecom

## # A tibble: 1,000 x 11
##       id referrer device bouncers n_visit n_pages duration country purchase
##    <dbl> <chr>    <chr>  <lgl>      <dbl>   <dbl>    <dbl> <chr>   <lgl>   
##  1     1 google   laptop TRUE          10       1      693 Czech ~ FALSE   
##  2     2 yahoo    tablet TRUE           9       1      459 Yemen   FALSE   
##  3     3 direct   laptop TRUE           0       1      996 Brazil  FALSE   
##  4     4 bing     tablet FALSE          3      18      468 China   TRUE    
##  5     5 yahoo    mobile TRUE           9       1      955 Poland  FALSE   
##  6     6 yahoo    laptop FALSE          5       5      135 South ~ FALSE   
##  7     7 yahoo    mobile TRUE          10       1       75 Bangla~ FALSE   
##  8     8 direct   mobile TRUE          10       1      908 Indone~ FALSE   
##  9     9 bing     mobile FALSE          3      19      209 Nether~ FALSE   
## 10    10 google   mobile TRUE           6       1      208 Czech ~ FALSE   
## # ... with 990 more rows, and 2 more variables: order_items <dbl>,
## #   order_value <dbl>

ggplot(mtcars, aes(x = disp, y = mpg, color = factor(cyl))) +
  geom_point()

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(aes(color = factor(cyl)))

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(color = 'blue')

ggplot(mtcars, aes(x = disp, y = mpg, shape = factor(cyl))) +
  geom_point()

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(aes(shape = factor(cyl)))

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(shape = 5)

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(shape = 5, color = 'blue')

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(shape = 5, fill = 'blue')

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(shape = 22, fill = 'blue')

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(shape = 22, color = 'blue')

ggplot(mtcars, aes(x = disp, y = mpg, size = disp)) +
  geom_point()

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(size = 4)

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(aes(alpha = factor(cyl)), color = 'blue')

## Warning: Using alpha for a discrete variable is not advised.

Data

gdp <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/gdp.csv')

## Warning: Missing column names filled in: 'X1' [1]

A line chart can be created using geom_line(). In the below example, we examine the GDP trend of India and modify the color of the line to ‘blue’.

ggplot(gdp, aes(year, india)) +
  geom_line(color = 'blue')

To modify the line type, use the linetype argument. It can take values between 1 and 5.

ggplot(gdp, aes(year, india)) +
  geom_line(linetype = 2)

The line type can also be mentioned in the following way:

ggplot(gdp, aes(year, india)) +
  geom_line(linetype = 'dashed')

The width of the line can be modified using the size argument.

ggplot(gdp, aes(year, india)) +
  geom_line(size = 2)

Now let us map the aesthetics to the variables. The data used in the above example cannot be used as we need a variable with country names. We will use gather() function from the tidyr package to reshape the data.

gdp2 <- 
  gdp %>% 
  select(year, growth, india, china) %>% 
  gather(key = country, value = gdp, -year)

gdp2

## # A tibble: 18 x 3
##    year       country   gdp
##    <date>     <chr>   <dbl>
##  1 2000-01-01 growth      6
##  2 2001-01-01 growth      9
##  3 2002-01-01 growth      8
##  4 2003-01-01 growth      9
##  5 2004-01-01 growth      9
##  6 2005-01-01 growth      8
##  7 2000-01-01 india       5
##  8 2001-01-01 india       9
##  9 2002-01-01 india       8
## 10 2003-01-01 india       8
## 11 2004-01-01 india       5
## 12 2005-01-01 india       7
## 13 2000-01-01 china       8
## 14 2001-01-01 china       5
## 15 2002-01-01 china       6
## 16 2003-01-01 china       8
## 17 2004-01-01 china       9
## 18 2005-01-01 china       8

To map the aesthetics to a variable, we must use the group argument. In the below example, we map the aesthetics to country. But we cannot distinguish between the lines as their color, width and line type are the same. We have easily plotted the GDP trend of all countries using the group argument. Now, let us ensure that we can distinguish and identidy them using different aesthetics.

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line()

Let us begin by ensuring that the lines have different color using the color argument within aes() and assigning it the variable country.

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(color = country))

Instead of color, now we modify the line type using the linetype argument.

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(linetype = country))

In the below instance, we assign different width to the lines using the size argument.

ggplot(gdp2, aes(year, gdp, group = country)) +
  geom_line(aes(size = country))

## Warning: Using size for a discrete variable is not advised.

Before we wrap up, let us quickly see how we can map aesthetics to variables for different plots.

ggplot(ecom, aes(device, fill = purchase)) +
  geom_bar()

ggplot(ecom) +
  geom_histogram(aes(duration, fill = purchase), bins = 10)

ggplot(ecom) +
  geom_boxplot(aes(device, duration, fill = purchase))

Summary

In this post, we learnt about aesthetics i.e. how to modify the properties of geoms such as:

• color
• shape
• size
• fill
• alpha
• width

Up Next..

In the next post, we will learn to modify the axis and labels of a plot.

Introduction

This is the third post in the series Elegant Data Visualization with ggplot2. In the previous post, we learnt how to create plots using the qplot() function. In this post, we will create some of the most routinely used plots to explore data using the geom_* functions.

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2
• tibble
• dplyr

All the data sets used in this post can be found here and code can be downloaded from here.

Data

ecom <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/web.csv')
ecom

## # A tibble: 1,000 x 11
##       id referrer device bouncers n_visit n_pages duration country purchase
##    <dbl> <chr>    <chr>  <lgl>      <dbl>   <dbl>    <dbl> <chr>   <lgl>   
##  1     1 google   laptop TRUE          10       1      693 Czech ~ FALSE   
##  2     2 yahoo    tablet TRUE           9       1      459 Yemen   FALSE   
##  3     3 direct   laptop TRUE           0       1      996 Brazil  FALSE   
##  4     4 bing     tablet FALSE          3      18      468 China   TRUE    
##  5     5 yahoo    mobile TRUE           9       1      955 Poland  FALSE   
##  6     6 yahoo    laptop FALSE          5       5      135 South ~ FALSE   
##  7     7 yahoo    mobile TRUE          10       1       75 Bangla~ FALSE   
##  8     8 direct   mobile TRUE          10       1      908 Indone~ FALSE   
##  9     9 bing     mobile FALSE          3      19      209 Nether~ FALSE   
## 10    10 google   mobile TRUE           6       1      208 Czech ~ FALSE   
## # ... with 990 more rows, and 2 more variables: order_items <dbl>,
## #   order_value <dbl>

Scatter Plot

A scatter plot displays the relationship between two continuous variables. In ggplot2, we can build a scatter plot using geom_point(). Scatterplots can show you visually

• the strength of the relationship between the variables
• the direction of the relationship between the variables
• and whether outliers exist

ggplot(ecom, aes(x = n_pages, y = duration)) + 
  geom_point()

ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point() + 
  geom_abline(intercept = 37.285, slope = -5.344)

ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_smooth(method = 'lm', se = TRUE)

## `geom_smooth()` using formula 'y ~ x'

ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_smooth(method = 'loess', se = FALSE)

## `geom_smooth()` using formula 'y ~ x'

ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point() +
  geom_hline(yintercept = 30) 

ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point() +
  geom_vline(xintercept = 5) 

ggplot(ecom, aes(x = factor(device))) +
  geom_bar()

ggplot(ecom, aes(x = factor(device))) +
  geom_bar(aes(weight = order_value))

ggplot(ecom, aes(x = factor(device))) +
  geom_bar(aes(fill = purchase))

ggplot(ecom, aes(x = factor(device))) +
  geom_bar(aes(fill = purchase)) +
  coord_flip()

device <- c('laptop', 'mobile', 'tablet')
visits <- c(30000, 12000, 5000)
traffic <- tibble::tibble(device, visits)
ggplot(traffic, aes(x = device, y = visits)) +
  geom_col(fill = 'blue') 

ggplot(ecom, aes(x = factor(device), y = n_pages)) +
  geom_boxplot()

ggplot(ecom, aes(x = duration)) +
  geom_histogram()

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

ggplot(ecom, aes(x = duration)) +
  geom_histogram(bins = 5)

gdp <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/gdp.csv')

## Warning: Missing column names filled in: 'X1' [1]

gdp

## # A tibble: 6 x 6
##      X1     X year       growth india china
##   <dbl> <dbl> <date>      <dbl> <dbl> <dbl>
## 1     1     1 2000-01-01      6     5     8
## 2     2     2 2001-01-01      9     9     5
## 3     3     3 2002-01-01      8     8     6
## 4     4     4 2003-01-01      9     8     8
## 5     5     5 2004-01-01      9     5     9
## 6     6     6 2005-01-01      8     7     8

ggplot(gdp, aes(year, india)) +
  geom_line()

ggplot(gdp, aes(year, india)) +
  geom_line(color = 'blue', linetype = 'dashed')

ggplot(mtcars, aes(disp, mpg, label = rownames(mtcars))) +
  geom_label()

ggplot(mtcars, aes(disp, mpg, label = rownames(mtcars))) +
  geom_text(check_overlap = TRUE, size = 2)

Summary

In this post, we learnt about different geoms such as

• geom_point()
• geom_line()
• geom_histogram()
• geom_bar()
• geom_boxplot()
• geom_abline()
• geom_text()

Up Next..

In the next post, we will learn about aesthetics.

Introduction

This is the second post in the series Elegant Data Visualization with ggplot2. In the previous post, we understood the concept of grammar of graphics and even built a bar plot step by step while exploring the different components of a plot/chart. In this post, we will learn to quickly build a set of plots that are routinely used to explore data using qplot(). It can be used to quickly create plots but also has certain limitations. Nevertheless, if you want to quickly explore data using a single function, qplot() is your friend.

Libraries, Code & Data

We will use the following libraries in this post:

• readr
• ggplot2

All the data sets used in this post can be found here and code can be downloaded from here.

qplot(disp, mpg, data = mtcars)

qplot(disp, mpg, data = mtcars, geom = c('point', 'line'))

qplot(disp, mpg, data = mtcars, color = factor(cyl))

qplot(disp, mpg, data = mtcars, shape = factor(cyl))

qplot(disp, mpg, data = mtcars, size = qsec)

qplot(factor(cyl), data = mtcars, geom = c('bar'))

qplot(factor(cyl), data = mtcars, geom = c('bar'), fill = factor(am))

qplot(factor(cyl), mpg, data = mtcars, geom = c('boxplot'))

qplot(factor(1), mpg, data = mtcars, geom = c('boxplot'))

qplot(x = date, y = unemploy, data = economics, geom = c('line'))

qplot(x = date, y = unemploy, data = economics, geom = c('line'),
      color = 'red')

qplot(mpg, data = mtcars, bins = 5)

Summary

In this post, we learnt to quickly create plots using the qplot() function. While useful, it has limitations and can be used only to quickly visualize data.

Up Next..

In the next post, we will create the same set of plots but using geoms.

Introduction

This is the tenth post in the series Data Visualization With R. In the previous post, we learnt how to add text annotations to plots. In this post, we will learn how to combine multiple plots. Often, it is useful to have multiple plots in the same frame as it allows us to get a comprehensive view of a particular variable or compare among different variables. The Graphics package offers two methods to combine multiple plots. par() can be used to set graphical parameters regarding plot layout using the mfcol and mfrow arguments. layout() serves the same purpose but offers more flexibility by allowing us to modify the height and width of rows and columns.

par() allows us to customize the graphical parameters(title, axis, font, color, size) for a particular session. For combining multiple plots, we can use the graphical parameters mfrow and mfcol. These two parameters create a matrix of plots filled by rows and columns respectively. Let us combine plots using both the above parameters.

mfrow combines plots filled by rows i.e it takes two arguments, the number of rows and number of columns and then starts filling the plots by row. Below is the syntax for mfrow.

# mfrow syntax
mfrow(number of rows, number of columns)

Let us begin by combining 4 plots in 2 rows and 2 columns:

Libraries, Code & Data

All the data sets used in this post can be found here and code can be downloaded from here.

Case Study 1

Let us begin by combining 4 plots in 2 rows and 2 columns. The plots will be filled by rows as we are using the mfrow function:

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 4 graphs to be combined and filled by rows
par(mfrow = c(2, 2))
 
# specify the graphs to be combined
plot(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

Case Study 2

Combine 2 plots in 1 row and 2 columns.

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 2 graphs to be combined and filled by rows
par(mfrow = c(1, 2))
 
# specify the graphs to be combined
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

Case Study 3

Combine 2 plots in 2 rows and 1 column.

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 2 graphs to be combined and filled by rows
par(mfrow = c(2, 1))
 
# specify the graphs to be combined
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

Case Study 4

Combine 3 plots in 1 row and 3 columns.

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 3 graphs to be combined and filled by rows
par(mfrow = c(1, 3))
 
# specify the graphs to be combined
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

Case Study 5

Combine 3 plots in 3 rows and 1 column.

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 3 graphs to be combined and filled by rows
par(mfrow = c(3, 1))
 
# specify the graphs to be combined
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

mfcol

mfcol combines plots filled by columns i.e it takes two arguments, the number of rows and number of columns and then starts filling the plots by columns. Below is the syntax for mfrow:

# mfcol syntax
mfcol(number of rows, number of columns)

Let us begin by combining 4 plots in 2 rows and 2 columns:

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 4 graphs to be combined and filled by columns 
par(mfcol = c(2, 2))
 
# specify the graphs to be combined
plot(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

 
# restore the setting stored in init
par(init)

Special Cases

What happens if we specify lesser or more number of graphs? In the next two examples, we will specify lesser or more number of graphs than we ask the par() function to combine. Let us see what happens in such instances:

Case 1: Lesser number of graphs specified We will specify that 4 plots need to be combined in 2 rows and 2 columns but provide only 3 graphs.

Case 2: Extra graph specified We will specify that 4 plots need to be combined in 2 rows and 2 columns but specify 6 graphs instead of 4.

Case Study 7

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 4 graphs to be combined and filled by rows
par(mfrow = c(2, 2))
 
# specify the graphs to be combined
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
 
# restore the setting stored in init
par(init)

Case Study 8

# store the current parameter settings in init
init <- par(no.readonly=TRUE)
 
# specify that 4 graphs to be combined and filled by rows
par(mfrow = c(2, 2))
 
# specify the graphs to be combined
plot(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

plot(mtcars$disp, mtcars$mpg)
boxplot(mtcars$mpg)

# restore the setting stored in init
par(init)

Layout

At the core of the layout() function is a matrix. We communicate the structure in which the plots must be combined using a matrix. As such, the layout function is more flexible compared to the par() function.

Let us begin by combining 4 plots in a 2 row/2 column structure. We do this by creating a layout using the matrix function.

Case Study 1

Combine 4 plots in 2 rows/2 columns filled by rows.

# specify the layout
# 4 plots to be combined in 2 row/ 2 columns and arranged by row
layout(matrix(c(1, 2, 3, 4), nrow = 2, byrow = TRUE))
 
# specify the 4 plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)

Case Study 2

Combine 4 plots in 2 rows/2 columns filled by columns

To fill the plots by column, we specify byrow = FALSE in the matrix.

# specify the layout
# 4 plots to be combined in 2 row/ 2 columns and filled by columns
layout(matrix(c(1, 2, 3, 4), nrow = 2, byrow = FALSE))
 
# specify the 4 plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)

Case Study 3

Combine 3 plots in 2 rows/2 columns filled by rows

The magic of the layout() function begins here. We want to combine 3 plots and the first plot should occupy both the columns in row 1 and the next 2 plots should be in row 2. If you look at the matrix below, 1 is specified twice and since the matrix is filled by row, it will occupy both the columns in the first row. Similarly the first plot will occupy the entire first row. It will be crystal clear when you see the plot.

# specify the matrix
matrix(c(1, 1, 2, 3), nrow = 2, byrow = TRUE)
##      [,1] [,2]
## [1,]    1    1
## [2,]    2    3

# 3 plots to be combined in 2 row/ 2 columns and arranged by row
layout(matrix(c(1, 1, 2, 3), nrow = 2, byrow = TRUE))
 
# specify the 3 plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

Case Study 4

Combine 3 plots in 2 rows/2 columns filled by rows

The plots must be filled by rows and the third plot must occupy both the columns of the second row while the other two plots will be placed in the first row. The matrix would look like this:

# specify the matrix
matrix(c(1, 2, 3, 3), nrow = 2, byrow = TRUE)
##      [,1] [,2]
## [1,]    1    2
## [2,]    3    3

# 3 plots to be combined in 2 row/ 2 columns and arranged by row
layout(matrix(c(1, 2, 3, 3), nrow = 2, byrow = TRUE))
 
# specify the 3 plots
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)

Case Study 5

Combine 3 plots in 2 rows/2 columns filled by columns

The plots must be filled by columns and the first plot must occupy both the rows of the first column while the other two plots will be placed in the second column in two rows. The matrix would look like this:

# specify the matrix
matrix(c(1, 1, 2, 3), nrow = 2, byrow = FALSE)
##      [,1] [,2]
## [1,]    1    2
## [2,]    1    3

# 3 plots to be combined in 2 row/ 2 columns and arranged by columns
layout(matrix(c(1, 1, 2, 3), nrow = 2, byrow = FALSE))
 
# specify the 3 plots
hist(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)
boxplot(mtcars$mpg)

Case Study 6

Combine 3 plots in 2 rows/2 columns filled by columns

The plots must be filled by columns and the first plot must occupy both the rows of the second column while the other two plots will be placed in the first column in two rows. The matrix would look like this:

# specify the matrix
matrix(c(1, 2, 3, 3), nrow = 2, byrow = FALSE)
##      [,1] [,2]
## [1,]    1    3
## [2,]    2    3

# 3 plots to be combined in 2 row/ 2 columns and arranged by columns
layout(matrix(c(1, 2, 3, 3), nrow = 2, byrow = FALSE))
 
# specify the 3 plots
boxplot(mtcars$mpg)
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)

Widths

In all the layouts created so far, we have kept the size of the rows and columns equal. What if you want to modify the width and height of the columns and rows? The widths and heights arguments in the layout() function address the above mentioned issue. Let us check them out one by one: The widths argument is used for specifying the width of the columns. Based on the number of columns in the layout, you can specify the width of each column. Let us look at some examples.

Case Study 7

Width of the 2nd column is twice the width of the 1st column

# specify the matrix
matrix(c(1, 2, 3, 4), nrow = 2, byrow = TRUE)
##      [,1] [,2]
## [1,]    1    2
## [2,]    3    4

# 4 plots to be combined in 2 row/ 2 columns and arranged by columns
layout(matrix(c(1, 2, 3, 4), nrow = 2, byrow = TRUE), widths = c(1, 3))
 
# specify the plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)

Case Study 8

Width of the 2nd column is twice that of the first and last column

# specify the matrix
matrix(c(1, 2, 3, 4, 5, 6), nrow = 2, byrow = TRUE)
##      [,1] [,2] [,3]
## [1,]    1    2    3
## [2,]    4    5    6

# 6 plots to be combined in 2 row/ 3 columns and filled by rows
layout(matrix(c(1, 2, 3, 4, 5, 6), nrow = 2, byrow = TRUE), widths = c(1, 2, 1))
 
# specify the plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)

Heights

The heights arguments is used to modify the height of the rows and based on the number of rows specified in the layout, we can specify the height of each row.

Case Study 9

Height of the 2nd row is twice that of the first row

# 4 plots to be combined in 2 row/ 2 columns and filled by rows
layout(matrix(c(1, 2, 3, 4), nrow = 2, byrow = TRUE), heights= c(1, 2))
 
# specify the 4 plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)

Putting it all together…

Before we end this section, let us combine plots using both the widths and heights option

# specify the matrix
matrix(c(1, 2, 3, 4, 5, 6), nrow = 3, byrow = TRUE)
##      [,1] [,2]
## [1,]    1    2
## [2,]    3    4
## [3,]    5    6

# 6 plots to be combined in 3 row/ 2 columns and arranged by rows
layout(matrix(c(1, 2, 3, 4, 5, 6), nrow = 3, byrow = TRUE), heights= c(1, 2, 1),
widths = c(2, 1))
 
# specify the 6 plots
plot(mtcars$disp, mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)
plot(mtcars$mpg)
hist(mtcars$mpg)
boxplot(mtcars$mpg)