Graphics with R: an introduction to ggplot2

Figure. An example plot, made with the R graphics package “ggplot2”.

By Christina Suzanne

ggplot2 is a flexible graphing package that makes it easier to produce good plots without having to program every single detail. However, using ggplot2 requires a mental shift in the way you think about making plots, including getting familiar with the terminology:

• Data – what we want to visualize; for ggplot() data must be in a data.frame in a long format
• Geoms – geometric objects that represent data such as points, bars, lines
• Aesthetics – visual properties of geoms, e.g. (x,y) position, symbol
• Mappings – connect data values to aesthetic

Beginning to plot in ggpplot2 can be tricky at first; however, because of its growing popularity, there is a lot of support for it on R forums. Another useful resource is the R Graphics Cookbook (Chang 2013).

You can run the script provided below by cutting and pasting it onto the R command line and pressing <enter>.

R Script

#Script written by Christina Suzanne, June 2013
#Note: after each change the the graph, entering “p” is required to view the new graph.
#In the following, the function “quote” can be replaced with quotes. E.g. quote(Weight)

#can be entered as “Weight”. R uses only straight quotes, rather than left and right quotes;
#Wordpress prefers to use left and right quotes. Hence the use of the “quote” function.

chickwts         #Look at data

#Make a Boxplot using chickwts data:
p <- ggplot(chickwts, aes(x=feed,y=weight)) + geom_boxplot( ) #defining a boxplot
p    #Viewing the plot. After each change the the graph, entering “p” is required to view the new graph.

#Make a legend filled with default colours:
p <- ggplot(chickwts, aes(x=feed, y=weight, fill = feed)) + geom_boxplot( );

#Capitalize the legend title to “Feed” (with a capital) and abbreviate the labels:
p <- p + scale_fill_discrete(name=quote(Feed), labels = c(quote(C), quote(HB), quote(LS), quote(MM), quote(SB), quote(SF)));

#Change the y axis title and set y axis limits; make breaks at every 100:
p <- p + scale_y_continuous(name=quote(Weight), limits = c(0, 700), breaks = 0:700*100);

#Make the y axis title black and size 17:
p = p + theme(axis.title.y = element_text(size=17, colour = rgb(0,0,0)))

#Make the x axis title black and size 17:
p = p + theme(axis.title.x = element_text(size=17, colour=rgb(0,0,0)));

#Change x axis labels to size 12 and colour to black:
p = p + theme(axis.text.x = element_text(size=12, colour=rgb(0,0,0)))

#Make the y axis labels black and change the size to 12:
p = p + theme(axis.text.y = element_text(size=12, colour = rgb(0,0,0)));

#Change legend colour to black and the size to 15:
p = p + theme(legend.title = element_text(colour=rgb(1,0,0), size=15 ));

#Change legend labels to black and size to 11:
p <- p + theme(legend.text = element_text(colour=rgb(0,0,0), size = 11))

#Add a box around the legend:
p <- p + theme(legend.background = element_rect( ))

#Modify legend box to be filled in “white”, with a size = “0.5”, and linetype = “dotted”:
p = p + theme(legend.background = element_rect(fill=rgb(1,1,1), size=0.5, linetype=2));

#Position legend on plot:
p <- p + theme(legend.position=c(0.27,0.8))

p <-  p + ggtitle(quote(Chickens))

# Finally, modify title, reduce line spacing and make bold
p <- p + theme(plot.title = element_text(lineheight=0.8, face=quote(bold)))

p #view the final plot

References

Cookbook for R. Winston Chang. http://www.cookbook-r.com/Graphs/

ggplot2: An implementation of the grammar of graphics in R. Hadley Wickham. (H. Wickham. ggplot2: elegant graphics for data analysis. Springer New York, 2009.)