Complete Guide to Coloring Scatter Plots by Factor Variables in R

Introduction

In data visualization, color-coding data points based on categorical variables is a common and effective approach. R provides multiple implementation methods, with the base graphics system offering the most straightforward solution. This article uses the classic iris dataset to thoroughly explore how to color scatter plots based on factor variables while ensuring proper legend display.

Basic Implementation Method

Using R's base graphics system, color coding based on factor variables can be achieved through simple code. The core principle lies in R's automatic conversion of factor variables to integer sequences, which are then mapped to the default color palette.

data <- iris
plot(data$Sepal.Length, data$Sepal.Width, col=data$Species)

In this code, the col=data$Species parameter implements color mapping. Since Species is a factor variable with three levels (setosa, versicolor, virginica), R automatically converts it to numerical values 1, 2, 3, corresponding to the first three colors in the palette.

Necessity of Adding Legends

While color coding is implemented, the absence of a legend makes it impossible for readers to determine which color corresponds to which specific category. To address this issue, the legend function must be used to add explanatory legends.

legend(7, 4.3, unique(data$Species), col=1:length(unique(data$Species)), pch=1)

This code adds a legend at coordinates (7,4.3), displaying the colors and symbols corresponding to each species. unique(data$Species) ensures only unique factor levels appear in the legend, col=1:length(unique(data$Species)) specifies the color sequence, and pch=1 sets the point shape to circles.

Detailed Explanation of Color Mapping Mechanism

Understanding R's color mapping mechanism is crucial for mastering data visualization. When using factor variables as color parameters, R performs the following conversion process:

# Examine internal representation of factor variables
str(iris$Species)
# View default palette
palette()

By default, R uses the color sequence defined by the palette() function. The first factor level maps to "black", the second to "red", the third to "green3", and so on. This mapping relationship can be overridden by explicitly specifying colors.

Advanced Color Control

Beyond using default colors, custom color schemes can be defined to meet specific requirements. R provides various color generation functions such as rainbow(), heat.colors(), terrain.colors(), etc.

# Using custom colors
custom_colors <- c("red", "blue", "green")
plot(iris$Sepal.Length, iris$Sepal.Width, 
     col=custom_colors[as.numeric(iris$Species)])
legend("topright", legend=levels(iris$Species), 
       col=custom_colors, pch=1)

Comparison of Alternative Approaches

While the base graphics system is powerful, R's ecosystem includes other excellent visualization packages. ggplot2 offers more concise syntax and more aesthetically pleasing default styles:

library(ggplot2)
qplot(Sepal.Length, Sepal.Width, data=iris, color=Species)

ggplot2 automatically handles color mapping and legend generation, significantly simplifying code complexity. The lattice package provides similar functionality:

library(lattice)
xyplot(Sepal.Width ~ Sepal.Length, group=Species, data=iris,
       auto.key=list(space="right"))

Practical Application Considerations

In practical data analysis, several important factors must be considered. First, when dealing with numerous factor levels, ensure the color scheme provides good discriminability. Second, for colorblind-friendly visualizations, avoid red-green combinations. Finally, consider aesthetic effects and readability when preparing visualizations for publication or presentation.

# Generate well-discriminated color schemes
n_colors <- nlevels(iris$Species)
better_colors <- rainbow(n_colors)
plot(iris$Sepal.Length, iris$Sepal.Width, 
     col=better_colors[as.numeric(iris$Species)])
legend("topright", legend=levels(iris$Species), 
       col=better_colors, pch=1)

Conclusion

Coloring scatter plots based on factor variables is a fundamental skill in R data visualization. Through proper use of base graphics system's plot and legend functions, combined with appropriate color mapping strategies, information-rich and aesthetically pleasing visualization results can be created. For more complex application scenarios, advanced packages like ggplot2 and lattice provide additional functionality and better default styles.