Data Visualization: ggplot part 6

Learning Goals

At the end of this exercise, you will be able to:
1. Adjust the aesthetics of plots in ggplot using themes, RColorBrewer, and paletteer.
2. Use faceting to produce multi-panel plots.

Resources

• ggplot2 cheatsheet
• R for Data Science
• R Cookbook
• ggplot themes
• Rebecca Barter ggplot Tutorial
• R Color Brewer

Load the libraries

library(tidyverse)
library(RColorBrewer)
library(paletteer)
library(janitor)

options(scipen=999) #cancels the use of scientific notation for the session

Data

For this tutorial, we will use:
Homerange. The data are from: Tamburello N, Cote IM, Dulvy NK (2015) Energy and the scaling of animal space use. The American Naturalist 186(2):196-211. http://dx.doi.org/10.1086/682070.

homerange <- read_csv("data/Tamburelloetal_HomeRangeDatabase.csv", na = c("", "NA", "\\"))

ggplot themes

There are many options to change the theme of your plots within ggplot. Have a look [here]https://ggplot2.tidyverse.org/reference/ggtheme.html) for a list of the themes.

Let’s start by building a simple barplot.

p <- homerange %>% 
  ggplot(aes(x=taxon, fill=trophic.guild))+
  geom_bar(na.rm=T, position="dodge")

Have a look at the linedraw theme; I am adding it as another layer.

p + theme_linedraw()+
  theme(axis.text.x = element_text(angle = 60, hjust=1))+
  labs(title = "Observations by Taxon in Homerange Data",
       x = NULL,
       y= "n",
       fill= "Trophic Guild")

Practice

1. Build a scatterplot that compares log10.mass and log10.preymass. Color the points by taxon. Store this plot as object q.

q <- homerange %>% 
  ggplot(aes(x=log10.mass, y=log10.preymass, color=taxon))+
  geom_point(na.rm=T)

2. Use plot q but add the classic theme.

q + theme_classic()

Legends

There are lots of options to manipulate legends. Have a look here.

p+theme_linedraw()+
  theme(legend.position = "bottom",
        axis.text.x = element_text(angle = 60, hjust=1))+
  labs(title = "Observations by Taxon in Homerange Data",
       x = NULL,
       y= "n",
       fill= "Trophic Guild")

Not enough? Try using ggthemes

There are many packages that include additional themes, one of which is ggthemes. Some of these are nice because they are designed to mimic the look of popular publications.

#install.packages("ggthemes")
library(ggthemes)

Here is a list of the ggthemes

ls("package:ggthemes")[grepl("theme_", ls("package:ggthemes"))]

##  [1] "theme_base"            "theme_calc"            "theme_clean"          
##  [4] "theme_economist"       "theme_economist_white" "theme_excel"          
##  [7] "theme_excel_new"       "theme_few"             "theme_fivethirtyeight"
## [10] "theme_foundation"      "theme_gdocs"           "theme_hc"             
## [13] "theme_igray"           "theme_map"             "theme_pander"         
## [16] "theme_par"             "theme_solarized"       "theme_solarized_2"    
## [19] "theme_solid"           "theme_stata"           "theme_tufte"          
## [22] "theme_wsj"

p + 
  theme_stata()+
  theme(legend.position = "bottom",
        axis.text.x = element_text(angle = 60, hjust=1))+
  labs(title = "Observations by Taxon in Homerange Data",
       x = NULL,
       y= "n",
       fill= "Trophic Guild")

Practice

1. Use plot q and try a theme of your choice.

q+theme_clean()

RColorBrewer

The default colors used by ggplot are uninspiring. They don’t make plots pop out in presentations or publications, and you may want to use a customized palette to make things visually consistent.

Access the help for RColorBrewer.

?RColorBrewer

The thing to notice is that there are three different color palettes: 1) sequential, 2) diverging, and 3) qualitative. Within each of these there are several selections. You can bring up the colors by using display.brewer.pal(). Specify the number of colors that you want and the palette name.

display.brewer.pal(5,"Accent") #qualitative palette

The R Color Brewer website is very helpful for getting an idea of the color palettes. To make things easy, use these two guidelines:

+scale_colour_brewer() is for points
+scale_fill_brewer() is for fills

Here I chose the Set2 palette. Take a moment and experiment with other options.

p+scale_fill_brewer(palette = "Set2")+
  theme(legend.position = "bottom",
        axis.text.x = element_text(angle = 60, hjust=1))+
  labs(title = "Observations by Taxon in Homerange Data",
       x = NULL,
       y= "n",
       fill= "Trophic Guild")

Practice

1. Keep using the sameq` plot, but test out one of R Color Brewer palettes.

q+theme_classic()+
  scale_colour_brewer(palette = "Paired")

Manually Setting Colors

You can also use paleteer to build a custom palette for consistency. To access the paleteer collection, I add it to a new object.

colors <- paletteer::palettes_d_names

Now we can display the palettes. Assign the palette to my_palette and then build this base R bar plot. There are a lot of options; paleteer is a collection of popular palettes. I really like the [ggsci package] (https://cran.r-project.org/web/packages/ggsci/vignettes/ggsci.html)

my_palette <- paletteer_d("ggsci::nrc_npg")

barplot(rep(1,6), axes=FALSE, col=my_palette)

Now we just identify my_palette as part of scale_fill_manual()

p+scale_fill_manual(values=my_palette)+
  theme(legend.position = "bottom",
        axis.text.x = element_text(angle = 60, hjust=1))+
  labs(title = "Observations by Taxon in Homerange Data",
       x = NULL,
       y= "n",
       fill= "Trophic Guild")

Practice

1. Keep using the sameqplot, but test out one ofpaleteer` themes.

q+theme_classic()+
  scale_colour_manual(values=my_palette)

Faceting

Faceting is one of the amazing features of ggplot. It allows us to make multi-panel plots for easy comparison. Here is a boxplot that shows the range of log10.mass by taxon.

homerange %>% 
  ggplot(aes(x=taxon, y=log10.mass, fill=trophic.guild))+
  geom_boxplot()+
  theme(axis.text.x = element_text(angle = 60, hjust=1))

There are other categorical variables that might be interesting to overlay. facet_wrap() makes a ribbon of panels by a specified categorical variable and allows you to control how you want them arranged.

homerange %>% 
  ggplot(aes(x=taxon, y=log10.mass, fill=taxon))+
  geom_boxplot()+
  facet_wrap(~trophic.guild, ncol=2, labeller=label_both)+
  theme(axis.text.x = element_text(angle = 60, hjust=1),
        legend.position = "none",
        strip.text = element_text(size = 12, face = "bold"))

facet_grid() allows control over the faceted variable; it can be arranged in rows or columns. rows~columns.

homerange %>% 
  ggplot(aes(x=taxon, y=log10.mass))+
  geom_boxplot()+
  facet_grid(trophic.guild~.)+
  theme(axis.text.x = element_text(angle = 60, hjust=1))

homerange %>% 
  ggplot(aes(x=taxon, y=log10.mass))+
  geom_boxplot()+
  facet_grid(.~trophic.guild)+
  theme(axis.text.x = element_text(angle = 60, hjust=1))

facet_grid() will also allow the comparison of two categorical variables, just remember a~b where a is rows and b is columns.

homerange %>% 
  ggplot(aes(x=taxon, y=log10.mass))+
  geom_boxplot()+
  facet_grid(trophic.guild~thermoregulation)+
  theme(axis.text.x = element_text(angle = 60, hjust=1))

Practice

1. Use facet_wrap to compare the distribution of log10.mass by thermoregulation.

homerange %>% 
  ggplot(aes(x=log10.mass))+
  geom_density(fill="steelblue", alpha=0.3)+
  facet_wrap(~thermoregulation)

2. Use facet_grid to compare the distribution of log10.mass by locomotion.

homerange %>% 
  ggplot(aes(x=log10.hra))+
  geom_density(fill="steelblue", alpha=0.3)+
  facet_grid(.~locomotion)

Wrap-up

Please review the learning goals and be sure to use the code here as a reference when completing the homework.

–>Home