Lab 04b: Graphing using ggplot

About

This lab will introduce you to the ggplot2 package in R. This is a powerful and flexible package which is capable of producing publication quality graphics.

Preliminaries

Before we begin, make sure to download the data from D2L. The data we are using in this lab is a small simulated data set called ggplot_lab_invert_survey.csv. You will also need to download galapago-finches.csv for the homework assignment. Make sure to place both of these files in the data/ folder in your R Project, otherwise you will need to modify the code below to have it run correctly.

ggplot2 is a part of the tidyverse and was included when you previously ran install.packages("tidyverse")
Remember you only need to install packages on each machine once.

library(tidyverse)

Getting Started

So far we’ve been mostly working with data sets that are built into R, or making our own small data sets within our scripts.
Now we will need to read in the data sets into R.
But first, take a look at the ggplot_lab_invert_survey.csv data set by opening them.
- Right-Click on invert.csv in your file explorer program to Open With Notepad or text editor
This data is comma separated, which is the default type in the read_csv() function. read_csv() is a part of the tidyverse and is the the main function that we will throughout this course.
- There is also a read.csv() function that comes with the base installation of R. It behaves very similarly but is slightly less “smart” than read_csv().
We’ll call our data frame invert and assign it the output from read_csv()
We still to supply the file name in quotes as the first argument
to tell R that it’s located in the data/ folder, we add this to the beginning of the filename

invert <- read_csv("data/ggplot_lab_invert_survey.csv")

Did everyone successfully load the data into R?
- Are you in your R-project?
- Do you have a data/ folder?
- Is your ggplot_lab_invert_survey.csv in the data/ folder?

Whenever you load a dataset into R, it is always a good idea to take a look at it to familiarize yourself with the structure. Run the following lines of code in RStudio one by one (I have removed the output to save space here):

names(invert)

## [1] "month"          "taxa"           "body_length_mm" "body_mass_g"

head(invert)

## # A tibble: 6 × 4
##   month taxa   body_length_mm body_mass_g
##   <dbl> <chr>           <dbl>       <dbl>
## 1     3 beetle           26.9       17.4 
## 2     7 beetle           32.1       11.3 
## 3    11 beetle           21.4        4.40
## 4     3 beetle           25.0       26.2 
## 5     7 beetle           17.7       12.6 
## 6    11 beetle           31.9       15.2

tail(invert)

## # A tibble: 6 × 4
##   month taxa   body_length_mm body_mass_g
##   <dbl> <chr>           <dbl>       <dbl>
## 1     3 spider           23.2       15.2 
## 2     7 spider           20.2        2.02
## 3    11 spider           25.3       16.0 
## 4     3 spider           27.4        9.77
## 5     7 spider           25.4       12.6 
## 6    11 spider           26.5       11.9

str(invert)

## spc_tbl_ [30 × 4] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
##  $ month         : num [1:30] 3 7 11 3 7 11 3 7 11 3 ...
##  $ taxa          : chr [1:30] "beetle" "beetle" "beetle" "beetle" ...
##  $ body_length_mm: num [1:30] 26.9 32.1 21.4 25 17.7 ...
##  $ body_mass_g   : num [1:30] 17.4 11.3 4.4 26.2 12.6 ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   month = col_double(),
##   ..   taxa = col_character(),
##   ..   body_length_mm = col_double(),
##   ..   body_mass_g = col_double()
##   .. )
##  - attr(*, "problems")=<externalptr>

summary(invert)

##      month        taxa           body_length_mm   body_mass_g    
##  Min.   : 3   Length:30          Min.   :11.01   Min.   : 2.020  
##  1st Qu.: 3   Class :character   1st Qu.:20.89   1st Qu.: 6.397  
##  Median : 7   Mode  :character   Median :26.02   Median :11.931  
##  Mean   : 7                      Mean   :25.00   Mean   :11.671  
##  3rd Qu.:11                      3rd Qu.:28.35   3rd Qu.:15.630  
##  Max.   :11                      Max.   :33.83   Max.   :26.195

You can see the entire invert object by running the following command:
View(invert)
we can see that it includes information on:
- the month (as a number) of the sampling month
- the taxa (either beetle or spider) as a character vector
- the body length (in millimeters) of each individual
- and the body mass (in grams) of each individual

`ggplot2`

Very popular plotting package
Makes publication quality plots quickly
Declarative - describe what you want not how to build it
Contrasts w/Imperative - how to build it step by step

Basics

`ggplot` basic syntax

\[\underbrace{ggplot}_{initiate\; plot}(\underbrace{data = df}_{data\;frame},\; \underbrace{aes(x =\; , y = \;, color = \;)}_{plot\; attributes}) + \underbrace{geom\_line()}_{geometry}\] * To build a plot using ggplot we start with the ggplot() function

ggplot()

ggplot() creates a base ggplot object that we can then add things to
- Think of it as a blank canvas

Plot attributes, mapping and `aes()`

We can also add optional arguments for information to be shared across different components of the plot
The two main arguments we typically use here are
data - which is the name of the data frame we are working with, so invert
mapping - which describes which columns of the data are used for different aspects of the plot
We create a mapping by using the aes function, which stands for “aesthetic”, and then linking columns to pieces of the plot
We’ll start with telling ggplot what value should be on the x and y axes
Let’s plot the relationship between the body_length_mm of invertebrates to their body_mass_g
note that I start a new line after each , and that Rstudio automatically indents for us. This improves readability and makes it easier to look for typos.

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g))

Notice we now have “CIRC” on the x-axis and “HEIGHT” on the y-axis
This still doesn’t create a figure, it’s just a blank canvas and some information on the default values for the data and mapping columns to axes of the plots

Plot Layers

We can add data to the plot using layers
We do this by adding a + after the the ggplot() function and then adding something called a geom, which stands for geometry
To make a scatter plot we use geom_point()

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point()

It is standard to hit Enter after the plus so that each layer shows up on its own line
To change things about the layer we can pass additional arguments to the geom
We can do things like change
- the size of the points, we’ll set it to 3
- the color of the points, we’ll set it to "blue"
- the transparency of the points, which is called alpha, we’ll set it to 0.5

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point(size = 3, color = "blue", alpha = 0.5)

To add labels (like documentation for your graphs!) we use the labs function

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point(size = 3, color = "blue", alpha = 0.5) +
  labs(x = "Body length [mm]",
       y = "Body mass [g]",
       title = "Simulated ground-dwelling invertebrate survey")

One Continuous Variable

Throughout the semester, we will want to see the distribution of one continuous variable. The two main ways we will do that is through:

Histograms
Boxplots

Histograms

We can make a histogram of body_mass_g by modifying our earlier code.
Note only one variable, so we remove the y=... part.
We also change geom_point() to geom_histogram

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_histogram()

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

geom_histogram() groups observations into bins
- i.e., all the numbers between 10-12 g
automatically picks bin numbers based on data (graph above used 30 bins, see message output), but we can modify this using the bins argument

5 bins

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_histogram(bins = 5)

see that there are now 5 bins, but the “edges” of the bins are unclear

Binwidth

We can set the size of the bins with binwidth argument
for example, let’s say we want bins that are 2.5 g across

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_histogram(binwidth = 2.5)

Histogram attributes

We can change the attributes of the histogram output in similar ways as above
- Common arguments are: fill, color (outline), size (outline), and alpha (transparency).
- You can also explore different “themes”. Below is theme_bw(), but you can try theme_classic(), theme_dark(), theme_void(), etc.

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_histogram(binwidth = 2.5,
                 fill = "dodgerblue",
                 color = "black",
                 size = 2,
                 alpha = 0.5) +
  theme_bw()

## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.

Boxplots

Boxplots summarize the data distribution
AKA “Cat and Whisker” Plots

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_boxplot()

* whiskers: show minimum (left side) and maximum (right side) of data

Box: shows where 50% of the data lies
Line in middle of box is the median
Useful to loo for normal distributions:
- Symmetry of box, whiskers, etc.

Boxplot attributes

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_boxplot(fill = "red",
               color = "black")

Density Plots

Density plots show the full distribution
i.e., a histogram has discrete bins, but a density plot is continuous

ggplot(data = invert,
       mapping = aes(x = body_mass_g)) +
  geom_density(fill = "pink",
               alpha = 0.75) +
  theme_dark()

You should now be able to complete problem 01 in the homework assignment

One Continuous Variable, and Categories

Boxplots

We will often want to compare the distributions between two groups. We can do this by modifying our boxplot code above by adding a grouping variable as our y = ... in the aes() function

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa)) +
  geom_boxplot()

we can also vary the color/fill by taxa by including it in the aes() function

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot()

ggplot automatically picked two colors for us based on the values in the taxa column

Change fill colors

Manually

See a list of colors available in ggplot

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot() +
  scale_fill_manual(values = c("dodgerblue", "hotpink"))

note that this does not work if you put it in the geom_boxplot() argument:

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa)) +
  geom_boxplot(fill = taxa)

Likewise, if you put a fill/color argument in the geom_boxplot() function, it will overwrite what you mapped above

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot(fill = "black")

Categorical vs. continuous variables

Our data has the results of three surveys across months
we may want to see if there are different distributions across months

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     fill = month)) +
  geom_boxplot()

## Warning: The following aesthetics were dropped during statistical transformation: fill
## ℹ This can happen when ggplot fails to infer the correct grouping structure in
##   the data.
## ℹ Did you forget to specify a `group` aesthetic or to convert a numerical
##   variable into a factor?

this didn’t work because month is a number, which R interprets as a continuous variable.
notice the warning discusses “group” structure of the data
we can get around this a number of ways.
the easiest at this point is to force month to be a factor (what R calls a categorical variable).

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     fill = as.factor(month))) +
  geom_boxplot()

Histograms

We can also do a similar thing with histograms.
note that we need to remove the y=... argument, and we also need to add position = "identity" to the histogram call.
- otherwise the counts are “stacked” and the display is misleading

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     fill = taxa)) +
  geom_histogram(position = "identity",
                 alpha = 0.75) +
  scale_fill_manual(values = c("grey30", "goldenrod")) +
  theme_bw()

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

You should now be able to complete problem 02 in the homework assignment

Two Continuous Variables

Scatter Plots

Display two continuous variables by mapping them to x and y

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point(size = 3)

map attributes to other columns

Color

Using a categorical variable
Gives discrete colors
- i.e., “red”, or “blue”

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     color = taxa)) +
  geom_point(size = 3)

Using a continuous variable
gives “shades” of one color, based on value

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     color = month)) +
  geom_point(size = 3)

Shape

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     shape = taxa)) +
  geom_point(size = 4)

Shape + color

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     color = taxa, 
                     shape = as.factor(month))) +
  geom_point(size = 4)

`geom_smooth()`

We will often want to estimate the relationship between our predictor (x) and our response (y).
do this with geom_smooth()

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point(size = 4) +
  geom_smooth()

## `geom_smooth()` using method = 'loess' and formula = 'y ~ x'

the default is a “squiggly line”
we will limit ourselves to linear relationships, so add method = "lm"

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g)) +
  geom_point(size = 4) +
  geom_smooth(method = "lm")

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

geom_smooth() inherits other aspects from aes.
add a smooth for both taxa:

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     color = taxa)) +
  geom_point(size = 4) +
  geom_smooth(method = "lm")

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

In any of the plots, it is also possible to add facets, which gives each group their own panel:

ggplot(data = invert,
       mapping = aes(x = body_length_mm,
                     y = body_mass_g,
                     color = taxa)) +
  geom_point(size = 4) +
  geom_smooth(method = "lm") +
  facet_wrap(~taxa)

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

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot() +
  scale_fill_manual(values = c("dodgerblue", "hotpink")) +
  facet_wrap(~taxa)

can also add a different variable
this helps to visualize when we have 2+ grouping variables

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot() +
  scale_fill_manual(values = c("dodgerblue", "hotpink")) +
  facet_wrap(~month)

Coda

You should now be able to finish the homework. The below information is provided for your personal edification

Grammar of graphics

Uniquely describe any plot based on a defined set of information
Leland Wilkinson
Geometric object(s)
- Data
- Mapping
- Statistical transformation
- Position (allows you to shift objects, e.g., spread out overlapping data points)
Facets
Coordinates (coordinate systems other than cartesian, also allows zooming)

Saving plots as new files

ggsave() function saves whatever the last plot you made was.
let’s rerun this code:

ggplot(data = invert,
       mapping = aes(x = body_mass_g,
                     y = taxa,
                     fill = taxa)) +
  geom_boxplot() +
  scale_fill_manual(values = c("dodgerblue", "hotpink")) +
  facet_wrap(~month)

we will now save it into the home directory of our project and name the file:

ggsave("invert_by_month_taxa.jpg")

The type of the file is determined by the file extension

ggsave("invert_by_month_taxa.jpg")

if you want to save it in a specific folder, add it as "folder_name\" in front of the file name
- Make sure you actually have a folder named that before saving!
For example, if you followed the recommended file structure for this course, you should have a homework folder.
to save the last ggplot figure you made as a .png file to your homework folder, enter the following:

ggsave("homework/invert_by_month_taxa.png")

Check your folder to make sure the file is there.
Open it to see how it looks
You may want to delete this from your homework folder so you don’t get confused later.

ENVS475: Experimental Analysis and Design

April, 2024

About

Preliminaries

Getting Started

`ggplot2`

Basics

`ggplot` basic syntax

Plot attributes, mapping and `aes()`

Plot Layers

One Continuous Variable

Histograms

5 bins

Binwidth

Histogram attributes

Boxplots

Boxplot attributes

Density Plots

One Continuous Variable, and Categories

Boxplots

Change fill colors

Manually

Categorical vs. continuous variables

Histograms

Two Continuous Variables

Scatter Plots

map attributes to other columns

Color

Shape

Shape + color

`geom_smooth()`

Facets

Coda

Grammar of graphics

Saving plots as new files

Lab 04b: Graphing using ggplot

ENVS475: Experimental Analysis and Design

April, 2024

About

Preliminaries

Getting Started

ggplot2

Basics

ggplot basic syntax

Plot attributes, mapping and aes()

Plot Layers

One Continuous Variable

Histograms

5 bins

Binwidth

Histogram attributes

Boxplots

Boxplot attributes

Density Plots

One Continuous Variable, and Categories

Boxplots

Change fill colors

Manually

Categorical vs. continuous variables

Histograms

Two Continuous Variables

Scatter Plots

map attributes to other columns

Color

Shape

Shape + color

geom_smooth()

Facets

Coda

Grammar of graphics

Saving plots as new files

`ggplot2`

`ggplot` basic syntax

Plot attributes, mapping and `aes()`

`geom_smooth()`