We now consider in detail one particular type of graphics that is often not covered in standard courses: maps.
Each region on a map is represented as a polygon, which is essentially a shape made up of multiple points connected by lines to form a closed boundary. These polygons are crucial for mapping as they define the spatial boundaries of different geographic areas.
One popular package in R for working with maps is maps. The maps package provides a simple interface to access a variety of map databases, offering detailed map outlines for countries, states, and world regions. When these map databases are loaded, each region (like a country) is broken down into its polygonal components, each described by a series of longitude and latitude points. The ggplot2 package can then take these points and visually render them as polygons, allowing for the addition of aesthetic mappings such as colors to represent different data variables (like population, GDP, or health statistics).1 Before running the code, make sure that the required packages ggplot2 and maps are installed.
We will explore examples of two distinct types of maps: dot maps and choropleths.
A dot map is a map overlaid with dots colored according to some statistic.
A choropleth is a map where each region is colored according to some statistic.
Dot maps
There is a built-in dataset called quakes that contains information on earthquake occurrences, detailing attributes like magnitude, depth, and location coordinates for seismic events in the Fiji area. Our goal is to produce a dot mat to visualize the magnitude and depth of seismic events. We shall approach this in three steps.
data("quakes") # Load the earthquakes data
head(quakes) # Display the first few rows
First, let’s use the map_data() function to retrieve data for the world map. The columns of the retrieved data set store the coordinates (longitude and latitude), along with identifiers that help in grouping and ordering these points to form the shapes of countries or regions on a map.
Now, we are ready to use ggplot to produce the map.
map1= ggplot() +
geom_polygon(data = world_map,
aes(x = long, y = lat, group = group),
fill = "lightgray", color = "black") +
theme_minimal()
map1+labs(x = "Longitude", y = "Latitude")
Simple world map.
Step 2: Zooming in on a region.
Next, let’s use coord_fixed() to zoom in on the Fiji area.
Step 3: Finally. we are ready to overlay the map with dots such that a deeper color indicates greater depth and a larger size indicates greater magnitude.
map2+
geom_point(data = quakes,
aes(x = long, y = lat, size =depth , color = mag),
alpha = 0.5) +
scale_size(range = c(0.1, 2)) +
scale_color_gradient(low = "blue", high = "red") +
labs( x = "Longitude", y = "Latitude",
color = "Magnitude",size="Depth") +
theme_minimal()
Earthquakes around Fiji.
Choropleths
Next, let’s make a choropleth of the United States with each state colored according the the percentage of the population living in urban areas in 1973. We will work with another built-in dataset called USArrests, which contains statistics on arrests per 100,000 residents for assault, murder, and rape in each of the U.S. states in 1973, along with the percentage of the population living in urban areas.
# Get map data for US states (this is in the ggplot package, but you can also download your own lat, long coordinates)
states_map <- map_data("state")
# Merge map data with USArrests data
USArrests$state <- tolower(rownames(USArrests))
arrest_map <- states_map %>%
left_join(USArrests, by = c("region" = "state"))
# Plot with ggplot2
ggplot(arrest_map, aes(long, lat, group = group, fill = UrbanPop)) +
geom_polygon(color = "white", linewidth = 0.3) +
coord_fixed(1.3) +
scale_fill_gradientn(
colors = colorRampPalette(c("lightblue", "darkblue"))(100),
name = "Urban Population (%)"
) +
labs(title = "Urban Population by US State") +
theme_void() +
theme(
legend.position = "right",
plot.title = element_text(hjust = 0.5)
)
Urban population by state
An alternative to making choropleths is using the maps() package, instead of using ggplot().
Working with Categorical Variables
Categorical variables in R are stored in a data type called a factor. It stores both the values of the variable and its unique categories, known as the levels of a factor. Even if they are not explicitly specified, R can determine a factor’s levels. We will discuss how to work with qualitative and ordinal factors through two example datasets.
Renaming Variables and Levels
Let’s first take a look at the iris dataset, which contains 150 total measurements of sepal and petal lengths and widths for three species of Iris flowers. Species is a qualitative variable stored as a factor, so we can print the levels and see that the species included are Setosa, Versicolor, and Virginica. We can make plots to visually observe differences between the species.
ggplot(iris, aes(x = Petal.Length, y = Petal.Width, col = Species)) +
geom_point() +
labs(x = "Petal Length", y = "Petal Width") +
theme_minimal()
Petal width and length for Iris species.
Even though the variable name Species is informative, someone who is unfamiliar with the dataset or flowers may not know that these are referring to Iris species. We can either rename the variable or the levels to clarify exactly what flowers these correspond to. To rename the variable, we modify the vector returned by names() which contains all the column names of iris. Alternatively, tidyverse provides a function rename that gives the same result.
ggplot(iris_copy, aes(x = Petal.Length, y = Petal.Width, col = Species)) +
geom_point() +
labs(x = "Petal Length", y = "Petal Width") +
theme_minimal()
Petal width and length for Iris species.
Creating factors
The next example will show how to create a factor, whether for qualitative or ordinal variables. In some datasets, categorical variables may be stored as another data type and should be converted to a factor first.
Ordinal variables are categorical variables where the categories have an inherent order. One example is the months of a year (January, February, etc.), which are ordered and often represented numerically (1, 2, etc.). In the airquality dataset, there are daily air quality measurements from May to September 1973 in New York. We demonstrate how to encode the ordinal variable Month as a factor.
For an ordinal variable, we want to also specify that the levels are ordered. This allows us to use comparison operators which would not be defined for the unordered factor above.
Finally, let’s rename the levels to be more informative than its original numerical values. Now we can make plots using Month as a categorical variable.
Facets are a powerful tool for visualizing how the relationship between two or
more variables depends on other variables. For instance, let’s say we’re
interested in understanding the relationship between petal length and petal
width for flowers. Depending on the species, the nature of this relationship
might differ, so it can be helpful to have separate plots for each flower type
in our data. The facet_grid method makes this easy:
Note that the above generated horizontally organized subplots. If vertically
organized subplots are desired, then instead of . ~ [variable], you should
use [variable] ~ . as an input to facet_grid.
plot + facet_grid(Species ~ .)
Vertical Subplots
The facet_grid method can also be used to generate subplots for each
combination of two categorical variables. To see this in action, let’s start by
creating a new variable that indicates whether a flower has sepal width greater
than half its sepal length.
The above plot makes it visually clear that the ratio between sepal width and
sepal length doesn’t seem to affect the relationship between petal width and
petal length for any of the three species in the iris data. Depending on your
data, the best method for visualizing specific relationships might involve a
combination of color-coding and facets. An example is below:
plot <- ggplot(iris, aes(x = Petal.Length, y = Petal.Width, col = Species)) +
geom_point() +
labs(x = "Petal Length", y = "Petal Width") +
theme_minimal()
plot + facet_grid(. ~ wide_sepals)
Generally, it is a good idea to use colors consistently throughout your document. For example, if you always display points or lines for setosa in orange, then in any figure, the data related to setosa should be colored orange.
By default, calling facet_grid([Variable 1] ~ [Variable 2]) will create an
grid of subplots, where are the number of categories in
Variable 1 and Variable 2 respectively. To manually control the dimensions
of your facet grid, you can use the facet_wrap() method, which takes a number
of columns ncol as an argument. In the following example, we set the number
of columns to 2:
Below we make a pie chart for species in the Iris dataset. Note that here there is the same number observations for each species, but we do include a pie chart below for illustration purposes. We also include labeling by percentages, just to illustrate the approach. You can modify the plot below for your own purposes.
# Count species
species_counts <- as.data.frame(table(iris$Species))
colnames(species_counts) <- c("Species", "Count")
# Calculate percentages and labels
species_counts$Percent <- round(100 * species_counts$Count / sum(species_counts$Count), 1)
species_counts$Label <- paste0(species_counts$Species, "\n", species_counts$Percent, "%")
# Plot with ggplot
ggplot(species_counts, aes(x = "", y = Count, fill = Species)) +
geom_bar(stat = "identity", width = 1) +
coord_polar("y") + # turns bar into pie
theme_void() + # remove background
geom_text(aes(label = Label),
position = position_stack(vjust = 0.5)) +
labs(title = "Distribution of Iris Species")
Example of Pie Charts
Timeseries
The lubridate package, which is part of tidyverse is designed to handle dates in R. Here is an example:
## date borough district type value
## 1 2005-01-01 Manhattan 09 refuse 3006.0
## 2 2005-01-01 Manhattan 09 paper 276.9
## 3 2005-01-01 Manhattan 09 mgp 180.3
## 4 2005-01-01 Manhattan 09 resident_organics 0.0
## 5 2005-01-01 Manhattan 09 school_organics 0.0
## 6 2005-01-01 Manhattan 09 leaves 0.0
This data set contains information about the amount of waste collected in different locations in New York city. The first column specifies the date of collection. The next two columns specify the borough and district (location). The next column specifies the type of rubbish collection ranging from refuse (general waste) to Christmas trees. Let’s use the waste dataset to look at trends over time across all of New York city. To do this we should group by date and type and use summarize() to add up the amount of waste across all districts.
Let’s first look at the overall proportions of waste across time. Let’s consider a pie-chart for this:
# Summarize total waste by type
waste_summary <- aggregate_waste %>%
group_by(type) %>%
summarise(total_amount = sum(amount)) %>%
ungroup()
# Create pie chart
ggplot(waste_summary, aes(x = "", y = total_amount, fill = type)) +
geom_bar(stat = "identity", width = 1) +
coord_polar("y") +
labs(title = "Total Waste by Type", fill = "Waste Type") +
theme_void()
Pie chart of waste by type
We see that most waste comes from paper, mgp and refuse. Next, we are interested in exploring trends over time, for different types:
# Make sure that date is date type
aggregate_waste <- aggregate_waste %>%
mutate(date = as.Date(date))
# Plot
ggplot(aggregate_waste, aes(x = date, y = amount)) +
geom_line(color = "steelblue") +
facet_wrap(~ type, scales = "free") +
labs(title = "Time Series of Waste by Type",
x = "Date",
y = "Amount of waste") +
theme_minimal()
Time series waste by type
Based on the above, we see some interesting patterns: For example, christmas trees spike at certain times, and school organics are zero during certain times. Let’s calculate the total by year and month, and filter out refuse, mgp and paper:
monthly_waste_filtered %>%
ggplot(aes(x = month, y = amount, color = type)) +
geom_line() +
scale_x_continuous(breaks = (1:12)) +
labs(title = "Waste by Month",
x = "Month",
y = "Amount of waste",
col = "Waste Type")
Waste by month and type
We can see a lot of seasonal effects in the last two graphics. Three that stand out are:
A large number of Christmas trees are collected in January but not in any other month.
Leaves are collected in the fall. There’s a little bit in October, a lot in November and December and again a little bit in January and February.
There’s no organic waste from the schools in July and August! This makes sense as schools are closed for summer.
Let’s say we are interested in seeing whether there are any patter by whether it was the Covid year (2020) or not (previous).
## `summarise()` has grouped output by 'borough', 'month'. You can override using
## the `.groups` argument.
ggplot(refuse_2020_b) +
geom_line(aes(x = month, y = amount, color = is_2020)) +
facet_wrap(~borough) +
labs(x = "Month", y = "Amount", color = "2020?") +
scale_x_continuous(breaks = seq(1, 12))
Comparing waste in covid versus non-covid year
We see that during Covid, a business district such as Manhattan had lower waste. However, Bronx, Brooklyn, Queens and Staten Island had higher waste in the second half of the year.
Plots - a gallery
The codes and examples in this section are done using a standard demo package, with its advantages and disadvantages. Take them simply as a preview of what is possible and the opportunity to learn a bit more about any of these if you are interested, including how to customize them to maximize effectiveness.
demo(graphics)
##
##
## demo(graphics)
## ---- ~~~~~~~~
##
## > # Copyright (C) 1997-2009 The R Core Team
## >
## > require(datasets)
##
## > require(grDevices); require(graphics)
##
## > ## Here is some code which illustrates some of the differences between
## > ## R and S graphics capabilities. Note that colors are generally specified
## > ## by a character string name (taken from the X11 rgb.txt file) and that line
## > ## textures are given similarly. The parameter "bg" sets the background
## > ## parameter for the plot and there is also an "fg" parameter which sets
## > ## the foreground color.
## >
## >
## > x <- stats::rnorm(50)
##
## > opar <- par(bg = "white")
##
## > plot(x, ann = FALSE, type = "n")
##
## > abline(h = 0, col = gray(.90))
##
## > lines(x, col = "green4", lty = "dotted")
##
## > points(x, bg = "limegreen", pch = 21)
##
## > title(main = "Simple Use of Color In a Plot",
## + xlab = "Just a Whisper of a Label",
## + col.main = "blue", col.lab = gray(.8),
## + cex.main = 1.2, cex.lab = 1.0, font.main = 4, font.lab = 3)
##
## > ## A little color wheel. This code just plots equally spaced hues in
## > ## a pie chart. If you have a cheap SVGA monitor (like me) you will
## > ## probably find that numerically equispaced does not mean visually
## > ## equispaced. On my display at home, these colors tend to cluster at
## > ## the RGB primaries. On the other hand on the SGI Indy at work the
## > ## effect is near perfect.
## >
## > par(bg = "gray")
##
## > pie(rep(1,24), col = rainbow(24), radius = 0.9)
##
## > title(main = "A Sample Color Wheel", cex.main = 1.4, font.main = 3)
##
## > title(xlab = "(Use this as a test of monitor linearity)",
## + cex.lab = 0.8, font.lab = 3)
##
## > ## We have already confessed to having these. This is just showing off X11
## > ## color names (and the example (from the postscript manual) is pretty "cute".
## >
## > pie.sales <- c(0.12, 0.3, 0.26, 0.16, 0.04, 0.12)
##
## > names(pie.sales) <- c("Blueberry", "Cherry",
## + "Apple", "Boston Cream", "Other", "Vanilla Cream")
##
## > pie(pie.sales,
## + col = c("purple","violetred1","green3","cornsilk","cyan","white"))
##
## > title(main = "January Pie Sales", cex.main = 1.8, font.main = 1)
##
## > title(xlab = "(Don't try this at home kids)", cex.lab = 0.8, font.lab = 3)
##
## > ## Boxplots: I couldn't resist the capability for filling the "box".
## > ## The use of color seems like a useful addition, it focuses attention
## > ## on the central bulk of the data.
## >
## > par(bg="cornsilk")
##
## > n <- 10
##
## > g <- gl(n, 100, n*100)
##
## > x <- rnorm(n*100) + sqrt(as.numeric(g))
##
## > boxplot(split(x,g), col="lavender", notch=TRUE)
##
## > title(main="Notched Boxplots", xlab="Group", font.main=4, font.lab=1)
##
## > ## An example showing how to fill between curves.
## >
## > par(bg="white")
##
## > n <- 100
##
## > x <- c(0,cumsum(rnorm(n)))
##
## > y <- c(0,cumsum(rnorm(n)))
##
## > xx <- c(0:n, n:0)
##
## > yy <- c(x, rev(y))
##
## > plot(xx, yy, type="n", xlab="Time", ylab="Distance")
##
## > polygon(xx, yy, col="gray")
##
## > title("Distance Between Brownian Motions")
##
## > ## Colored plot margins, axis labels and titles. You do need to be
## > ## careful with these kinds of effects. It's easy to go completely
## > ## over the top and you can end up with your lunch all over the keyboard.
## > ## On the other hand, my market research clients love it.
## >
## > x <- c(0.00, 0.40, 0.86, 0.85, 0.69, 0.48, 0.54, 1.09, 1.11, 1.73, 2.05, 2.02)
##
## > par(bg="lightgray")
##
## > plot(x, type="n", axes=FALSE, ann=FALSE)
##
## > usr <- par("usr")
##
## > rect(usr[1], usr[3], usr[2], usr[4], col="cornsilk", border="black")
##
## > lines(x, col="blue")
##
## > points(x, pch=21, bg="lightcyan", cex=1.25)
##
## > axis(2, col.axis="blue", las=1)
##
## > axis(1, at=1:12, lab=month.abb, col.axis="blue")
##
## > box()
##
## > title(main= "The Level of Interest in R", font.main=4, col.main="red")
##
## > title(xlab= "1996", col.lab="red")
##
## > ## A filled histogram, showing how to change the font used for the
## > ## main title without changing the other annotation.
## >
## > par(bg="cornsilk")
##
## > x <- rnorm(1000)
##
## > hist(x, xlim=range(-4, 4, x), col="lavender", main="")
##
## > title(main="1000 Normal Random Variates", font.main=3)
##
## > ## A scatterplot matrix
## > ## The good old Iris data (yet again)
## >
## > pairs(iris[1:4], main="Edgar Anderson's Iris Data", font.main=4, pch=19)
World Map with Fiji earthquakes data.
