Directory of Visualizations

Author

Jiratchaya Nuanpirom

Published

May 17, 2023

In this chapter, we’ll look at some of the most common plots and charts. We’ll discuss how to interpret and use these visualizations to make data-driven decisions. We’ll also explore how to create these plots and charts using ggplot2 and other graphic libraries.

The following commands will check whether the required R libraries already installed, if not, pacman will install and load the required libraries into your environment.

# Install pacman
if (!require("pacman"))
  install.packages("pacman"
  )

# Use pacman p_load() to install and load other R libraries
pacman::p_load(
1  tidyverse,
2  viridis,
3  ggsci,
4  ComplexHeatmap,
5  RColorBrewer,
6  GGally,
7  kableExtra,
8  ggridges,
9  factoextra
)

1: An integrated data management and visualization library.
2: Color palettes for continuous and discrete variables, color blind friendly.
3: Color palettes based on scientific journals, movies, TV shows.
4: For visualizing heat maps and several plots alongside the heat maps. Provides powerful customization capabilities over the ggplot2’s geom_tile() and base heatmap() functions.
5: Color palettes for continuous and discrete variables.
6: Extension library of ggplot2. Plot complex plots in small steps by simplifying the commands. For exploratory data analysis and correlation analysis.
7: Just for render this document.
8: For ridgeline plot
9: For principal component analysis and visualization

1 Visualizing Amounts

1.1 Bar Plots

The bar plot is the most common way to visualize amounts, i.e., how many things you’ve already counted. Bar plots are versatile and easy to interpret. They are helpful for comparing various groupings of data. They can also be used to compare different values within the same dataset.

In this demo, we will create a data frame with two columns, name and value.

# Create data
data_1 <- data.frame(name = c("A", "B", "C", "D", "E"), 
                   value = c(3, 12, 5, 18, 45))

Example of data_1
name	value
A	3
B	12
C	5
D	18
E	45

1.1.1 Basic bar plot

To create bar plot with ggplot2:

Call ggplot().
Specify the data object.
Make the aesthetics function: use a categorical variable name for X axis and numeric value for Y axis, and fill the bar plot with color by the names of the categories name.
finally call geom_bar(). You have to specify stat="identity" for this kind of data set.
In this case, use labs() to change the axes names, as well as legend name.

ggplot(data_1, aes(x = name, y = value, fill = name)) +
  geom_bar(stat = "identity") +
  labs(x = "New name", y = "New value") +
  scale_fill_npg() +
  theme(legend.position = "none")

1.1.2 Horizontal bar plot

To rotate the plot to another orientation, use coord_flip().

ggplot(data_1, aes(x = name, y = value, fill = name)) +
  geom_bar(stat = "identity") +
  labs(x = "New name", y = "New value") +
  coord_flip() +
  scale_fill_npg() +
  theme(legend.position = "none")

1.1.3 Bar plot with labeled text

To annotate the number in each bar plot, add geom_text to the command. Specify the variable you want to show the text in aes() of geom_text. You can adjust the labels’ position using hjust and vjust to move them in horizontal and vertical directions, respectively.

ggplot(data_1, aes(x = name, y = value, fill = name)) +
  geom_bar(stat = "identity") +
  geom_text(aes(label = value), vjust = -0.2) +
  labs(x = "New name", y = "New value")+
  scale_fill_npg() +
  theme(legend.position = "none")

1.2 Grouped and Stacked Bar Plot

We showed how to visualize quantitative amounts in a single variable in the previous section. Sometimes we’re interested in two categorical variables at once. For example, the mock data frame data_plant that collects data from multiple plant species and multiple conditions.

# create a dataset
set.seed(123456)
data_plant <- data.frame(
  species = c(rep("sorgho", 3),
              rep("poacee", 3),
              rep("banana", 3),
              rep("triticum", 3)),
  condition = rep(c("normal", "stress", "Nitrogen"), 4),
  value = abs(rnorm(12, 0, 15)))

Example of data_plant
species	condition	value
sorgho	normal	12.505998
sorgho	stress	4.140717
sorgho	Nitrogen	5.325028
poacee	normal	1.312311
poacee	stress	33.783836
poacee	Nitrogen	12.516902

1.2.1 Grouped bar plot

In the geom_bar(), use the argument position="dodge" to position the bars next to each other.

ggplot(data_plant, aes(fill = condition, y = value, x = species)) +
  geom_bar(position = "dodge", stat = "identity")+
  scale_fill_npg() +
  theme(legend.position = "none")

1.2.2 Small multiple bar plots

To make multiple bar plots, use facet_wrap(). Each panel will have its name defined by the variable specified. facet_wrap() fixes x and/or y axes scales using the argument scales.

ggplot(data_plant, aes(fill = condition, y = value, x = species)) +
  geom_bar(position = "dodge", stat = "identity") +
  facet_wrap( ~ species, scales = "free_x") +
  scale_fill_npg()

1.2.3 Stacked bar plot

Stack bar plots display subgroups on top of each other. The only thing to change to get this figure is to switch the position argument to stack by adding position = "stack" to the geom_bar() function.

ggplot(data_plant, aes(fill = condition, y = value, x = species)) +
  geom_bar(position = "stack", stat = "identity") +
  scale_fill_npg()

1.2.4 Percent stacked bar plot

Change the position of the percent stacked bar plot to “fill”. We can now study the evolution of each subgroup’s proportion in the overall sample by looking at the percentage of each subgroup.

ggplot(data_plant, aes(fill = condition, y = value, x = species)) +
  geom_bar(position = "fill", stat = "identity") +
  scale_fill_npg()

1.3 Dot Plots

Bars have a limitation that they need to start at zero, so the length is proportional to the amount. Your plot would be vague. We can indicate amounts by placing dots along the x or y axis.

dot plot vs scatter plot

A dot plot is a graph that uses dots to indicate the frequency or amount of a certain value. This makes it easier to compare values, as the dots line up along a continuous axis. A scatter plot, on the other hand, uses points to represent two different variables. This makes it easier to identify the relationship between the two variables.

1.3.1 Simple dot plot

We’ll show you a simple dot plot on US precipitation. Start by loading the data into R and selecting the 30 most precipitated cities in the US.

# Prepare data of Annual Precipitation in US Cities
US_precip <- data.frame(city = names(precip), precipitation = precip)
US_precip <- US_precip %>% 
  top_n(n = 30) %>%
  arrange(desc(precipitation))

Selecting by precipitation

Example of US_precip data
city	precipitation
Mobile	67.0
Miami	59.8
San Juan	59.2
New Orleans	56.8
Juneau	54.7
Jacksonville	54.5
Jackson	49.2
Memphis	49.1
Little Rock	48.5
Atlanta	48.3
Houston	48.2
Columbia	46.4
Nashville	46.0
Atlantic City	45.5
Norfolk	44.7
Hartford	43.4
Louisville	43.1
Providence	42.8
Charlotte	42.7
Richmond	42.6
Boston	42.5
Raleigh	42.5
Baltimore	41.8
Portland	40.8
Charleston	40.8
Wilmington	40.2
New York	40.2
Philadelphia	39.9
Cincinnati	39.0
Washington	38.9

The plot will be a value-ordered dot plot. Basically, ggplot2 will arrange the data alphabetically or ascendingly arranged. It won’t work if you order the data frame before defining ggplot(). Instead, you can arrange the data using reorder(variable_to_order, variable_to_rely_on) during aesthetics mapping as follows.

# Plot value-ordered dot plot
ggplot(US_precip, aes(x = precipitation, y = reorder(city, precipitation))) +
  geom_point(size = 3,
             alpha = 0.8,
             color = "#D51317") +
  labs(x = "Average annual precipitation (in.)", y = NULL) +
  theme_bw()

1.3.2 Time-series

Time can be considered as one of the variables, and it is in an inherent order. Line graphs help us visualize this temporal order. Line graphs can be used to show trends over time, allowing us to compare values at different points in time. They are useful for highlighting changes in data ove

We will use the USPersonalExpenditure data set. This data set consists of United States personal expenditures (in billions of dollars) in the categories; food and tobacco, household operation, medical and health, personal care, and private education for the years 1940, 1945, 1950, 1955 and 1960.

# Prepare data
US_exp <- data.frame(USPersonalExpenditure) %>% 
  rownames_to_column(var = "Category") %>% 
  pivot_longer(!Category) %>% 
  mutate(name = gsub("X", "", name)) %>% 
  rename(year = name)

ggplot(US_exp, aes(x = year, y = value, group = Category, color = Category)) +
  geom_line(linewidth = 0.7, color = "black") +
  geom_point(size = 3) +
  scale_color_frontiers() +
  labs(x = "Year", y = "Amount (billions dollars)")

1.4 Heatmaps

As an alternative to mapping data values onto positions via bars or dots, we can map data values onto colors. Such a figure is called a heat map. Which is a graphical representation of data where the individual values contained in a matrix are represented as colors.

1.4.1 Basic Heatmap with base R

The heatmap() function is natively provided in R. Which required the data in a matrix format, them heatmap() is then run clustering algorithm and visualize the result with dendrogram. We cal use argument scale to normalize matrix to balance the heat colors for easy infer the trend of the data.

# The mtcars dataset:
data <- as.matrix(mtcars)

# Default Heatmap
heatmap(data)

# Use 'scale' to normalize
heatmap(data, scale="column")

1.4.2 Heatmap with `geom_tile()`

We will use USJudgeRatings data set, which reports Lawyers’ Ratings of State Judges in the US Superior Court.

# Prepare data
dt_USJudgeRatings <- USJudgeRatings %>% 
  rownames_to_column(var = "company") %>% 
  pivot_longer(!company)
head(dt_USJudgeRatings)

# A tibble: 6 × 3
  company       name  value
  <chr>         <chr> <dbl>
1 AARONSON,L.H. CONT    5.7
2 AARONSON,L.H. INTG    7.9
3 AARONSON,L.H. DMNR    7.7
4 AARONSON,L.H. DILG    7.3
5 AARONSON,L.H. CFMG    7.1
6 AARONSON,L.H. DECI    7.4

# Plot
ggplot(dt_USJudgeRatings, aes(x = name, y = company, fill = value)) +
  geom_tile() +
  scale_fill_distiller(palette = "Reds") +
  labs(x = "Features", y = NULL)

1.4.3 Heatmap with `Heatmap()` of ComplexHeatmap

With ComplexHeatmap you’re able to visualize associations between different data sources, reveal patterns, and arrange multiple heat maps in a way that’s highly flexible. In this analysis, we’ll use the same data set as above, USJudgeRatings.

# Prepare data
mat_USJudgeRatings <- as.matrix(USJudgeRatings)

# Plot heatmap using ComplexHeatmap
Heatmap(mat_USJudgeRatings,
        row_names_gp = gpar(fontsize = 8),
        column_names_gp = gpar(fontsize = 8),
        column_names_rot = 30)

2 Visualizing Distribution

In a dataset, you might want to know how a variable is distributed. You can look at the distribution of a variable to get a better idea. Visualizations like histograms, boxplots, and density plots help with this. Data patterns and outliers can be identified this way. This can help you understand the data better, and draw conclusions from it. It can also help you decide which type of analysis to use. ### Histogram

A histogram can be easily built with ggplot2’s geom_histogram() function. The input only needs one numerical variable. You might get a warning message during the plot regarding the bin width or stratification as explained in the next section.

2.0.1 Basic histogram

A histogram is created by binning data, or stratifying data into ranges of frequency, so how they look depends on the bin width. Bin widths are usually chosen by default in visualization programs.

users may have to pick the appropriate bin width by hand. The histogram gets too peaky and busy if the bin width is too small. There might be a lot of noise in the data. In contrast, too wide a bin width can blur smaller features in the data distribution, like the dip around age 10.

dt_hist <- data.frame(value = rnorm(100))

ggplot(iris, aes(x = Petal.Length)) +
  geom_histogram(fill = "salmon", color = "black") +
  labs(subtitle = "bin width = 30 (default)")

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

ggplot(iris, aes(x = Petal.Length)) +
  geom_histogram(fill = "salmon", color = "black", bins = 15) +
  labs(subtitle = "bin width = 15")
ggplot(iris, aes(x = Petal.Length)) +
  geom_histogram(fill = "salmon", color = "black", bins = 40) +
  labs(subtitle = "bin width = 40")
ggplot(iris, aes(x = Petal.Length)) +
  geom_histogram(fill = "salmon", color = "black", bins = 5) +
  labs(subtitle = "bin width = 5")

2.1 Density Plot

The density plot shows the underlying probability distribution of the data by drawing a continuous curve. An estimation procedure called kernel density estimation is commonly used to estimate this curve from the data. This estimation procedure is used to estimate the probability of the data falling within a particular range. The resulting curve is then plotted on a graph, allowing for easier comparison and analysis.

2.1.1 Basic density plot

ggplot(iris, aes(x = Petal.Length)) +
  geom_density(fill = "grey80") +
  theme_bw()

2.1.2 Multiple density plots

ggplot(iris, aes(x = Petal.Length, fill = Species)) +
  geom_density(alpha = 0.7) +
  scale_fill_npg()

2.1.3 Facet density plots

ggplot(iris, aes(x = Petal.Length, fill = Species)) +
  geom_density() +
  scale_fill_viridis_d() +
  facet_wrap(. ~ Species) +
  scale_fill_npg()

Scale for fill is already present.
Adding another scale for fill, which will replace the existing scale.

2.2 Box Plot and Violin Plot

A box plot gives a nice summary of one or more numeric variables. A boxplot is composed of several elements:

Anatomy of box plot. Image from https://www.leansigmacorporation.com/box-plot-with-minitab

A median is a line that divides the box into two parts. If the median is 10, it means there are the same number of data points above and below the line.
There are two quartiles at the ends of the box, the upper (Q3) and lower (Q1). A third quartile of 15 indicates that 75% of observations are below that level.
The difference between Quartiles 1 and 3 is called the interquartile range (IQR)
The extreme line shows Q3+1.5 x IQR to Q1-1.5 x IQR (the highest and lowest value excluding outliers). Dots (or other markers) beyond the extreme line shows potential outliers.

However, you should keep in mind that data distribution is hidden behind each box. Normal distributions could look exactly like bimodal distributions. It is recommended to consider a violin plot or a ridgline chart instead.

We will demonstrate on ToothGrowth data set. The data set showed the Effect of Vitamin C on Tooth Growth in Guinea Pigs.

Example of ToothGrowth data set
len	supp	dose
4.2	VC	0.5
11.5	VC	0.5
7.3	VC	0.5
5.8	VC	0.5
6.4	VC	0.5
10.0	VC	0.5
11.2	VC	0.5
11.2	VC	0.5
5.2	VC	0.5
7.0	VC	0.5
16.5	VC	1.0
16.5	VC	1.0
15.2	VC	1.0
17.3	VC	1.0
22.5	VC	1.0
17.3	VC	1.0
13.6	VC	1.0
14.5	VC	1.0
18.8	VC	1.0
15.5	VC	1.0

ggplot(ToothGrowth, 
       aes(x = as.factor(dose), y = len, 
           fill = as.factor(dose), group = as.factor(dose))) +
  geom_boxplot() +
  scale_fill_npg() +
  geom_jitter(color="grey30", size=2, alpha=0.7, width = 0.2) +
  labs(x = "Dose (mg/day)", y = "Tooth length (mm)") +
  theme(legend.position = "none")


ggplot(ToothGrowth, 
       aes(x = as.factor(dose), y = len, 
           fill = as.factor(dose), group = as.factor(dose))) +
  geom_violin() +
  scale_fill_npg() +
  geom_jitter(color="grey30", size=2, alpha=0.7, width = 0.2) +
  labs(x = "Dose (mg/day)", y = "Tooth length (mm)") +
  theme(legend.position = "none")

ggplot(ToothGrowth, 
       aes(x = as.factor(dose), y = len, 
           fill = as.factor(dose), group = as.factor(dose))) +
  geom_violin() +
  scale_fill_npg() +
  geom_boxplot(fill = "white", width = 0.2) +
  labs(x = "Dose (mg/day)", y = "Tooth length (mm)") +
  theme(legend.position = "none")

2.2.1 Ridgeline plot

Previous section showed how to visualize the distribution stack horizontally with histograms and density plots. To expand on this idea, we’ll stagger the distribution plots vertically. Ridgeline plots look like mountain ridgelines, so they’re called that. The ridgeline plot is a good way to show trends in distributions. This example checks the distribution of diamond prices based on their quality.

This graph is made using the ggridges library, which is a ggplot2 extension and compatible with the syntax of the ggplot2. For the X axis, we specify the price column, and for the Y axis, we specify the cut column. By adding 'fill=cut', we can use one color per category and display them separately.

ggplot(diamonds, aes(x = price, y = cut, fill = cut)) +
  geom_density_ridges() +
  theme(legend.position = "none")

Picking joint bandwidth of 458

3 Visualizing Proportions

In many cases, we want to show how something breaks down into parts that each represent a portion of the whole. Among data scientists, pie charts are often maligned because they are ubiquitous in business presentations.

3.1 Pie and Donut Chart

# Create Data
data <- data.frame(
  group=LETTERS[1:5],
  value=c(13,7,9,21,2))

Example of ToothGrowth data set
group	value
A	13
B	7
C	9
D	21
E	2

# Basic piechart
ggplot(data, aes(x = "", y = value, fill = group)) +
  geom_bar(stat = "identity",
           width = 1,
           color = "white") +
  coord_polar("y", start = 0) +
  scale_fill_npg() +
  theme_void() # remove background, grid, numeric labels

To create pie chart with data label

# Piechart with data labels
## Compute the position of labels
data_2 <- data %>% 
  arrange(desc(group)) %>%
  mutate(prop = value / sum(data$value) *100) %>%
  mutate(ypos = cumsum(prop)- 0.5*prop)

Example of ToothGrowth data set
group	value	prop	ypos
E	2	3.846154	1.923077
D	21	40.384615	24.038462
C	9	17.307692	52.884615
B	7	13.461538	68.269231
A	13	25.000000	87.500000

# Plot
ggplot(data_2, aes(x="", y=prop, fill=group)) +
  geom_bar(stat="identity", width=1, color="white") +
  coord_polar("y", start=0) +
  theme_void() + 
  theme(legend.position="none") +
  geom_text(aes(y = ypos, label = group), color = "white", size=6) +
  scale_fill_npg()

For accurate representation of data, alternative plots like percent stacked bar charts or streamlined plots should be considered instead of pie charts. Pie charts can make it hard to compare fractions since the relative sizes are hard to see. So if you want a more accurate representation of the data, you should use another type of chart.

4 Visualizing x-y Relationships

A dataset may contain two or more quantitative variables, and we might be interested in how they relate. In order to plot just the relationship between two variables, such as height and weight, we use a scatter plot. The bubble chart, scatter plot matrix, or correlogram are all good options if we want to show more than two variables at once. Further, principal component analysis is great for high-dimensional datasets, i.e., observing multiple variables at once.

4.1 Scatter Plot

In a scatterplot, two variables are plotted along two axes. It shows how they’re related, eventually revealing a correlation. Here the relationship between Petal width and Petal length in iris data set will demonstrated.

4.1.1 Basic scatter plot

ggplot(iris, aes(x = Petal.Width, y = Petal.Length, color = Species)) +
  geom_point(alpha = 0.8, size = 2.5) +
  labs(x = "Petal width (cm)",
       y = "Petal length (cm)") +
  scale_color_npg()

4.1.2 Facet scatter plot

ggplot(iris, aes(x = Petal.Width, y = Petal.Length, color = Species)) +
  geom_point(alpha = 0.8, size = 2.5) +
  labs(x = "Petal width (cm)",
       y = "Petal length (cm)") +
  scale_color_npg() +
  facet_wrap(. ~ Species)

4.1.3 Scatter plot with trend line

Adding a linear trend to a scatter plot helps the reader identify patterns. The linear trend also helps to identify outliers, as points that are far from the trend are easily visible. It can also be used to make predictions about future values based on current data. geom_smooth() in ggplot2 allows you to add the linear trend and confidence interval around it by using "se = TRUE"

ggplot(mtcars, aes(x = disp, y = mpg)) +
  geom_point(alpha = 0.8, size = 2) +
  geom_smooth(method = "lm", color = "red") +
  labs(x = "Miles/(US) gallon",
       y = "Displacement (cu.in.)")

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

4.2 Correlograms

When we have more than three or four variables, all-against-all scatter plots get a little complicated. Rather than presenting raw data, it’s better to quantify the amount of association between pairs of variables. Correlation coefficients are a common way to do this.

The correlation coefficient r measures how much two variables correlate between -1 and 1. A value of r = 0 means there is no association whatsoever, and a value of 1 or -1 indicates an ideal association. There are two kinds of correlation coefficients: correlated (larger values in one variable coincide with larger values in the other) or anticorrelated (larger values in one variable coincide with smaller values in the other).

Here’s a quick demonstration of all-against-all correlation analysis and a correlogram plotted using the mtcars data set.

# Create data 
data <- data.frame(mtcars)
 
# Check correlation between variables, default uses Pearson
cor_mtcars <- cor(data)

Example of all-against-all correlation coefficient in mtcars data
	mpg	cyl	disp	hp	drat	wt	qsec	vs	am	gear	carb
mpg	1.0000000	-0.8521620	-0.8475514	-0.7761684	0.6811719	-0.8676594	0.4186840	0.6640389	0.5998324	0.4802848	-0.5509251
cyl	-0.8521620	1.0000000	0.9020329	0.8324475	-0.6999381	0.7824958	-0.5912421	-0.8108118	-0.5226070	-0.4926866	0.5269883
disp	-0.8475514	0.9020329	1.0000000	0.7909486	-0.7102139	0.8879799	-0.4336979	-0.7104159	-0.5912270	-0.5555692	0.3949769
hp	-0.7761684	0.8324475	0.7909486	1.0000000	-0.4487591	0.6587479	-0.7082234	-0.7230967	-0.2432043	-0.1257043	0.7498125
drat	0.6811719	-0.6999381	-0.7102139	-0.4487591	1.0000000	-0.7124406	0.0912048	0.4402785	0.7127111	0.6996101	-0.0907898
wt	-0.8676594	0.7824958	0.8879799	0.6587479	-0.7124406	1.0000000	-0.1747159	-0.5549157	-0.6924953	-0.5832870	0.4276059
qsec	0.4186840	-0.5912421	-0.4336979	-0.7082234	0.0912048	-0.1747159	1.0000000	0.7445354	-0.2298609	-0.2126822	-0.6562492
vs	0.6640389	-0.8108118	-0.7104159	-0.7230967	0.4402785	-0.5549157	0.7445354	1.0000000	0.1683451	0.2060233	-0.5696071
am	0.5998324	-0.5226070	-0.5912270	-0.2432043	0.7127111	-0.6924953	-0.2298609	0.1683451	1.0000000	0.7940588	0.0575344
gear	0.4802848	-0.4926866	-0.5555692	-0.1257043	0.6996101	-0.5832870	-0.2126822	0.2060233	0.7940588	1.0000000	0.2740728
carb	-0.5509251	0.5269883	0.3949769	0.7498125	-0.0907898	0.4276059	-0.6562492	-0.5696071	0.0575344	0.2740728	1.0000000

# visualization of correlations
## Squared correlogram
ggcorr(data, method = c("everything", "pearson"))

## Circle correlogram
ggcorr(data, method = c("everything", "pearson"), 
       geom = "circle",
       min_size = 5,
       max_size = 10)

4.3 Dimension Reduction

High-dimensional datasets commonly have multiple correlated variables that convey overlapping information, so dimension reduction relies on that. A smaller number of key dimensions can be reduced without losing data characteristics.

There are many dimension reduction techniques. But, the most widely used method is principal components analysis (PCA) for exploring the data. PCA introduces an additional set of variables (called principal components, PCs) by linear combination of the original variables in the data, standardized to zero mean and unit variance. PCs are chosen to be uncorrelated, and they are ordered so the first component captures the most variation in the data, and each subsequent component captures less and less. In most cases, you can see the key features of data from the first two or three PCs.

We’ll show you how to compute and visualize PCA in R using the prcomp() function and the factoextra package. To show PCA, the factoextra’s built-in data, a Decathlon2.active, will be used.

# Prepare data
data(decathlon2)
decathlon2.active <- decathlon2[1:23, 1:10]

	X100m	Long.jump	Shot.put	High.jump	X400m	X110m.hurdle	Discus	Pole.vault	Javeline	X1500m
SEBRLE	11.04	7.58	14.83	2.07	49.81	14.69	43.75	5.02	63.19	291.7
CLAY	10.76	7.40	14.26	1.86	49.37	14.05	50.72	4.92	60.15	301.5
BERNARD	11.02	7.23	14.25	1.92	48.93	14.99	40.87	5.32	62.77	280.1
YURKOV	11.34	7.09	15.19	2.10	50.42	15.31	46.26	4.72	63.44	276.4
ZSIVOCZKY	11.13	7.30	13.48	2.01	48.62	14.17	45.67	4.42	55.37	268.0
McMULLEN	10.83	7.31	13.76	2.13	49.91	14.38	44.41	4.42	56.37	285.1
MARTINEAU	11.64	6.81	14.57	1.95	50.14	14.93	47.60	4.92	52.33	262.1
HERNU	11.37	7.56	14.41	1.86	51.10	15.06	44.99	4.82	57.19	285.1
BARRAS	11.33	6.97	14.09	1.95	49.48	14.48	42.10	4.72	55.40	282.0
NOOL	11.33	7.27	12.68	1.98	49.20	15.29	37.92	4.62	57.44	266.6

# Calculate principal components
res.pca <- prcomp(decathlon2.active, scale = TRUE)
# Summarize PCA
summary(res.pca)

Importance of components:
                          PC1    PC2    PC3     PC4     PC5     PC6     PC7
Standard deviation     2.0308 1.3559 1.1132 0.90523 0.83759 0.65029 0.55007
Proportion of Variance 0.4124 0.1839 0.1239 0.08194 0.07016 0.04229 0.03026
Cumulative Proportion  0.4124 0.5963 0.7202 0.80213 0.87229 0.91458 0.94483
                           PC8     PC9   PC10
Standard deviation     0.52390 0.39398 0.3492
Proportion of Variance 0.02745 0.01552 0.0122
Cumulative Proportion  0.97228 0.98780 1.0000

Show the percentage of variance explained by each principal component using eigenvalues (scree plot).

# Visualizing scree plot
fviz_eig(res.pca) + theme_grey()

Visualize graph of individuals. Individuals with a similar profile are grouped together.

# Individual PCA
fviz_pca_ind(res.pca,
             col.ind = "cos2", # Color by the quality of representation
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"),
             repel = TRUE     # Avoid text overlapping
             )

Graph of variables. Positive correlated variables point to the same side of the plot. Negative correlated variables point to opposite sides of the graph.

# Variable PCA
fviz_pca_var(res.pca,
             col.var = "contrib", # Color by contributions to the PC
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"),
             repel = TRUE     # Avoid text overlapping
             )

Biplot of individuals and variables

# Biplot
fviz_pca_biplot(res.pca, repel = TRUE,
                col.var = "#FC4E07", # Variables color
                col.ind = "#696969"  # Individuals color
                )

5 References

Fundamentals of Data Visualization: by Claus O. Wilke, 2019
R Graphics Cookbook, 2nd edition. Winston Chang, 2023.
Principal Component Analysis in R: prcomp vs princomp from STHDA, 2017.

1 Visualizing Amounts

1.1 Bar Plots

1.1.1 Basic bar plot

1.1.2 Horizontal bar plot

1.1.3 Bar plot with labeled text

1.2 Grouped and Stacked Bar Plot

1.2.1 Grouped bar plot

1.2.2 Small multiple bar plots

1.2.3 Stacked bar plot

1.2.4 Percent stacked bar plot

1.3 Dot Plots

1.3.1 Simple dot plot

1.3.2 Time-series

1.4 Heatmaps

1.4.1 Basic Heatmap with base R

1.4.2 Heatmap with geom_tile()

1.4.3 Heatmap with Heatmap() of ComplexHeatmap

2 Visualizing Distribution

2.0.1 Basic histogram

2.1 Density Plot

2.1.1 Basic density plot

2.1.2 Multiple density plots

2.1.3 Facet density plots

2.2 Box Plot and Violin Plot

2.2.1 Ridgeline plot

3 Visualizing Proportions

3.1 Pie and Donut Chart

4 Visualizing x-y Relationships

4.1 Scatter Plot

4.1.1 Basic scatter plot

4.1.2 Facet scatter plot

4.1.3 Scatter plot with trend line

4.2 Correlograms

4.3 Dimension Reduction

5 References

1.4.2 Heatmap with `geom_tile()`

1.4.3 Heatmap with `Heatmap()` of ComplexHeatmap