Chapter 5. Visualizing Data with ggplot2

This chapter covers the following topics:

• Creating basic plots with ggplot2

• Changing aesthetics mapping

• Introducing geometric objects

• Performing transformations

• Adjusting plot scales

• Faceting

• Adjusting themes

• Combining plots

• Creating maps

Introduction

When analyzing data, our primary goal is to efficiently and precisely deliver the findings to our audience. An easy way to present data is to display it in a table format. However, for larger datasets, it becomes challenging to visualize data in this format.

For example, the following table contains regional sales data:

Region	Jul-12	Aug-12	Sep-12	Oct-12	Nov-12	Dec-12
Alberta	22484.08	65244.19	15946.36	38593.39	34123.56	34753.98
British Columbia	23785.05	51533.77	44508.33	57687.6	19308.37	43234.77

In table format, it is hard to see which region's sales performed best. Thus, to make the data easier to read, it may be preferable to present the data in a chart or other graphical format. The following figure is a graph of the data from the table, which makes it much easier to determine which region performed best each month in terms of sales:

Figure 1: Sales amount by region

One of the most attractive features of R is that it already has many visualization packages...

Creating basic plots with ggplot2

In this recipe, we demonstrate how to use The Grammar of Graphics to construct our very first ggplot2 chart with the superstore sales dataset.

Getting ready

First, download the superstore_sales.csv dataset from the https://github.com/ywchiu/rcookbook/raw/master/chapter7/superstore_sales.csv GitHub link.

Next, you can use the following code to download the CSV file to your working directory:

> download.file('https://github.com/ywchiu/rcookbook/raw/master/chapter7/superstore_sales.csv', 'superstore_sales.csv')

You will also need to load the dplyr package to manipulate the superstore_sales dataset.

How to do it…

Please perform the following steps to create a basic chart with ggplot2:

1. First, install and load the ggplot2 package:

   > install.packages("ggplot2")
   > library(ggplot2)
   

2. Import superstore_sales.csv into an R session:

   > superstore <-read.csv('superstore_sales.csv', header=TRUE)
   > superstore$Order.Date <- as.Date(superstore$Order.Date)
   >...

Changing aesthetics mapping

Aesthetics mapping describes how data variables are mapped to the visual property of a plot. In this recipe, we discuss how to modify aesthetics mapping on geometric objects.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Also, you need to complete the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to add aesthetics to the plot:

1. First, create a scatterplot by mapping Year_Month to the x axis, Total_Sales to the y axis, and Province to color:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour=Province)) + ggtitle('With geom_point')
   > g + geom_point()
   

2. Set the aesthetics mapping on the geometric object:

   > g2 <- ggplot(data=sample_sum) + geom_point(mapping=aes(x=Year_Month, y=Total_Sales, colour=Province)) + ggtitle('With Aesthetics Mapping')
   > g2
   

   

   Figure 4: Scatterplots using geom_point and aesthetics mapping

3. Adjust the point size...

Introducing geometric objects

Geometric objects are elements that we mark on the plot. One can use the geometric object in ggplot2 to create either a line, bar, or box chart. Moreover, one can integrate these simple geometric objects and aesthetic mapping to create a more professional plot. In this recipe, we introduce how to use geometric objects to create various charts.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Perform the following steps to create a geometric object in ggplot2:

1. First, create a scatterplot with the geom_point function:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, col=Province )) + ggtitle('Scatter Plot')
   > g + geom_point() 
   

2. Use the geom_line function to plot a line chart:

   > g+ geom_line(linetype="dashed")
   

   

   Figure 9: Scatterplot and dashed line chart

3. Use the geom_bar function to make a stack bar chart:

   > g+geom_bar(stat = "identity", aes(fill=Province) , position...

Performing transformations

Besides mapping particular variables to either the x or y axis, one can first perform statistical transformations on variables, and then remap the transformed variable to a specific position. In this recipe, we introduce how to perform variable transformations with ggplot2.

Getting ready

Ensure you have completed the previous steps by storing sum_price_by_province, and sample_sum in your R environment.

How to do it…

Perform the following steps to perform statistical transformation in ggplot2:

1. First, create a dataset named sample_sum2 by filtering sales data from Alberta and British Columbia:

   > sample_sum2 <- sum_price_by_province %>% filter(Province %in% c('Alberta', 'British Columbia' ) )
   

2. Create a line plot with a regression line, using the geom_point and geom_smooth functions:

   > g <- ggplot(data=sample_sum2, mapping=aes(x=Year_Month, y=Total_Sales, col=Province ))
   > g + geom_point(size=5) + geom_smooth() + ggtitle('Adding Smoother')
   

   

   Figure 14: Adding...

Adjusting scales

Besides setting aesthetic mapping for each plot or geometric object, one can use scale to control how variables are mapped to the visual property. In this recipe, we introduce how to adjust the scale of aesthetics in ggplot2.

Getting ready

Ensure you have completed the previous steps by storing sample_mean in your R environment.

How to do it…

Perform the following steps to adjust the scale of aesthetic magnitude in ggplot2:

1. First, make a scatterplot by setting size=Total_Sales, colour=Province, y=Province, and conditional on Year_Month. Resize the point with the scale_size_continuous function:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Province, size=Total_Sales, colour = Province ))
   > g + geom_point(aes(size=Total_Sales)) + scale_size_continuous(range=c(1,10)) + ggtitle('Resize The Point')
   

2. Repaint the point in gradient color with the scale_color_gradient function:

   > g + geom_point(aes(colour=Total_Sales)) + scale_color_gradient()+ ggtitle('Repaint...

Faceting

When performing data exploration, it is essential to compare data across different groups. Faceting is a technique that enables the user to create graphs for subsets of data. In this recipe, we demonstrate how to use the facet function to create a chart for multiple subsets of data.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to create a chart for multiple subsets of data:

1. First, use the facet_wrap function to create multiple subplots by using the Province variable as the condition:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   >  g+geom_point(size = 5) +  facet_wrap(~Province) + ggtitle('Create Multiple Subplots by Province')
   

   

   Figure 19: Create multiple subplots by province

2. On the other hand, we can change the layout of the plot in a vertical direction if we set the number of columns to 1:

   > g+geom_point() +  facet_wrap...

Adjusting themes

Besides deciding the visual property of a geometric object with aesthetic mapping, one can adjust the background color, grid lines, and other non-data properties with the theme. We introduce how to change the theme in this recipe.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to adjust the theme in ggplot2:

1. We can use different theme functions to adjust the theme of the plot:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   > g+geom_point(size=5) + theme_bw()+ ggtitle('theme_bw Example')
   > g+geom_point(size=5) + theme_dark()+ ggtitle('theme_dark Example')
   

   

   Figure 23: Scatterplot with different themes

2. We can set the theme freely with the theme function:

   > g +geom_point(size=5) + 
   + theme(
   +   axis.text = element_text(size = 12),
   +   legend.background = element_rect(fill = "white"),
   +   panel.grid.major = element_line...

Combining plots

To create an overview of a dataset, we may need to combine individual plots into one. In this recipe, we introduce how to combine individual subplots into one plot.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Also, you need to complete the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to combine plots in ggplot2:

1. First, we need to load the grid library into an R session:

   > library(grid)
   

2. We can now create a new page:

   > grid.newpage()
   

3. Moving on, we can create two ggplot2 plots:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   >   plot1 <- g + geom_point(size=5) + ggtitle('Scatter Plot')
   >   plot2 <- g + geom_line(size=3) + ggtitle('Line Chart')
   

4. Next, we can push the visible area with a layout of two columns in one row, using the pushViewport function:

   > pushViewport(viewport(layout = grid.layout(1, 2)))
   

5. Last, we...

Creating maps

One can use a map to visualize the geographical relationship of spatial data. Here, we introduce how to create a map from a shapefile with ggplot2. Moreover, we introduce how to use ggmap to download map data from an online mapping service.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Please download all files from the following GitHub link folder:

https://github.com/ywchiu/rcookbook/tree/master/chapter7

How to do it…

Perform the following steps to create a map with ggmap:

1. First, load the ggmap and maptools libraries into an R session:

   > install.packages("ggmap")
   > install.packages("maptools")
   > library(ggmap)
   > library(maptools)
   

2. We can now read the .shp file with the readShapeSpatial function:

   > nyc.shp <- readShapeSpatial("nycc.shp")
   > class(nyc.shp)
   [1] "SpatialPolygonsDataFrame"
   attr(,"package")
   [1] "sp"
   

3. At this point, we can plot the map with the geom_polygon function:

   > ggplot() +  geom_polygon(data = nyc.shp, aes(x...

The rest of the chapter is locked

Creating basic plots with ggplot2

In this recipe, we demonstrate how to use The Grammar of Graphics to construct our very first ggplot2 chart with the superstore sales dataset.

Getting ready

First, download the superstore_sales.csv dataset from the https://github.com/ywchiu/rcookbook/raw/master/chapter7/superstore_sales.csv GitHub link.

Next, you can use the following code to download the CSV file to your working directory:

> download.file('https://github.com/ywchiu/rcookbook/raw/master/chapter7/superstore_sales.csv', 'superstore_sales.csv')

You will also need to load the dplyr package to manipulate the superstore_sales dataset.

How to do it…

Please perform the following steps to create a basic chart with ggplot2:

1. First, install and load the ggplot2 package:

   > install.packages("ggplot2")
   > library(ggplot2)
   

2. Import superstore_sales.csv into an R session:

   > superstore <-read.csv('superstore_sales.csv', header=TRUE)
   > superstore$Order.Date <- as.Date(superstore$Order.Date)
   >...

Changing aesthetics mapping

Aesthetics mapping describes how data variables are mapped to the visual property of a plot. In this recipe, we discuss how to modify aesthetics mapping on geometric objects.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Also, you need to complete the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to add aesthetics to the plot:

1. First, create a scatterplot by mapping Year_Month to the x axis, Total_Sales to the y axis, and Province to color:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour=Province)) + ggtitle('With geom_point')
   > g + geom_point()
   

2. Set the aesthetics mapping on the geometric object:

   > g2 <- ggplot(data=sample_sum) + geom_point(mapping=aes(x=Year_Month, y=Total_Sales, colour=Province)) + ggtitle('With Aesthetics Mapping')
   > g2
   

   

   Figure 4: Scatterplots using geom_point and aesthetics mapping

3. Adjust the point size...

Introducing geometric objects

Geometric objects are elements that we mark on the plot. One can use the geometric object in ggplot2 to create either a line, bar, or box chart. Moreover, one can integrate these simple geometric objects and aesthetic mapping to create a more professional plot. In this recipe, we introduce how to use geometric objects to create various charts.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Perform the following steps to create a geometric object in ggplot2:

1. First, create a scatterplot with the geom_point function:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, col=Province )) + ggtitle('Scatter Plot')
   > g + geom_point() 
   

2. Use the geom_line function to plot a line chart:

   > g+ geom_line(linetype="dashed")
   

   

   Figure 9: Scatterplot and dashed line chart

3. Use the geom_bar function to make a stack bar chart:

   > g+geom_bar(stat = "identity", aes(fill=Province) , position...

Performing transformations

Besides mapping particular variables to either the x or y axis, one can first perform statistical transformations on variables, and then remap the transformed variable to a specific position. In this recipe, we introduce how to perform variable transformations with ggplot2.

Getting ready

Ensure you have completed the previous steps by storing sum_price_by_province, and sample_sum in your R environment.

How to do it…

Perform the following steps to perform statistical transformation in ggplot2:

1. First, create a dataset named sample_sum2 by filtering sales data from Alberta and British Columbia:

   > sample_sum2 <- sum_price_by_province %>% filter(Province %in% c('Alberta', 'British Columbia' ) )
   

2. Create a line plot with a regression line, using the geom_point and geom_smooth functions:

   > g <- ggplot(data=sample_sum2, mapping=aes(x=Year_Month, y=Total_Sales, col=Province ))
   > g + geom_point(size=5) + geom_smooth() + ggtitle('Adding Smoother')
   

   

   Figure 14: Adding...

Adjusting scales

Besides setting aesthetic mapping for each plot or geometric object, one can use scale to control how variables are mapped to the visual property. In this recipe, we introduce how to adjust the scale of aesthetics in ggplot2.

Getting ready

Ensure you have completed the previous steps by storing sample_mean in your R environment.

How to do it…

Perform the following steps to adjust the scale of aesthetic magnitude in ggplot2:

1. First, make a scatterplot by setting size=Total_Sales, colour=Province, y=Province, and conditional on Year_Month. Resize the point with the scale_size_continuous function:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Province, size=Total_Sales, colour = Province ))
   > g + geom_point(aes(size=Total_Sales)) + scale_size_continuous(range=c(1,10)) + ggtitle('Resize The Point')
   

2. Repaint the point in gradient color with the scale_color_gradient function:

   > g + geom_point(aes(colour=Total_Sales)) + scale_color_gradient()+ ggtitle('Repaint...

Faceting

When performing data exploration, it is essential to compare data across different groups. Faceting is a technique that enables the user to create graphs for subsets of data. In this recipe, we demonstrate how to use the facet function to create a chart for multiple subsets of data.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to create a chart for multiple subsets of data:

1. First, use the facet_wrap function to create multiple subplots by using the Province variable as the condition:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   >  g+geom_point(size = 5) +  facet_wrap(~Province) + ggtitle('Create Multiple Subplots by Province')
   

   

   Figure 19: Create multiple subplots by province

2. On the other hand, we can change the layout of the plot in a vertical direction if we set the number of columns to 1:

   > g+geom_point() +  facet_wrap...

Adjusting themes

Besides deciding the visual property of a geometric object with aesthetic mapping, one can adjust the background color, grid lines, and other non-data properties with the theme. We introduce how to change the theme in this recipe.

Getting ready

Ensure you have completed the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to adjust the theme in ggplot2:

1. We can use different theme functions to adjust the theme of the plot:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   > g+geom_point(size=5) + theme_bw()+ ggtitle('theme_bw Example')
   > g+geom_point(size=5) + theme_dark()+ ggtitle('theme_dark Example')
   

   

   Figure 23: Scatterplot with different themes

2. We can set the theme freely with the theme function:

   > g +geom_point(size=5) + 
   + theme(
   +   axis.text = element_text(size = 12),
   +   legend.background = element_rect(fill = "white"),
   +   panel.grid.major = element_line...

Combining plots

To create an overview of a dataset, we may need to combine individual plots into one. In this recipe, we introduce how to combine individual subplots into one plot.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Also, you need to complete the previous steps by storing sample_sum in your R environment.

How to do it…

Please perform the following steps to combine plots in ggplot2:

1. First, we need to load the grid library into an R session:

   > library(grid)
   

2. We can now create a new page:

   > grid.newpage()
   

3. Moving on, we can create two ggplot2 plots:

   > g <- ggplot(data=sample_sum, mapping=aes(x=Year_Month, y=Total_Sales, colour = Province ))
   >   plot1 <- g + geom_point(size=5) + ggtitle('Scatter Plot')
   >   plot2 <- g + geom_line(size=3) + ggtitle('Line Chart')
   

4. Next, we can push the visible area with a layout of two columns in one row, using the pushViewport function:

   > pushViewport(viewport(layout = grid.layout(1, 2)))
   

5. Last, we...

Creating maps

One can use a map to visualize the geographical relationship of spatial data. Here, we introduce how to create a map from a shapefile with ggplot2. Moreover, we introduce how to use ggmap to download map data from an online mapping service.

Getting ready

Ensure you have installed and loaded ggplot2 into your R session. Please download all files from the following GitHub link folder:

https://github.com/ywchiu/rcookbook/tree/master/chapter7

How to do it…

Perform the following steps to create a map with ggmap:

1. First, load the ggmap and maptools libraries into an R session:

   > install.packages("ggmap")
   > install.packages("maptools")
   > library(ggmap)
   > library(maptools)
   

2. We can now read the .shp file with the readShapeSpatial function:

   > nyc.shp <- readShapeSpatial("nycc.shp")
   > class(nyc.shp)
   [1] "SpatialPolygonsDataFrame"
   attr(,"package")
   [1] "sp"
   

3. At this point, we can plot the map with the geom_polygon function:

   > ggplot() +  geom_polygon(data = nyc.shp, aes(x...