Mastering Text Size Adjustment in R Plots

Introduction

Creating visualizations is a crucial aspect of data analysis and interpretation, especially when presenting findings to an audience. In the R programming language, adjusting text size within plots enhances readability and tailors your visuals to specific presentation needs. This tutorial will guide you through various techniques for modifying font sizes in R plots, focusing on titles, axis labels, legends, and annotations.

Understanding Font Size Parameters

R offers several parameters that allow customization of text appearance:

cex: A universal scaling factor for text size. It affects most text elements but may not work with all plotting functions directly.
cex.lab: Scales the font size of axis labels (xlab and ylab).
cex.axis: Adjusts the font size of numbers on axes.
cex.main: Modifies the main title’s font size.
cex.sub: Changes the subtitle text size.

Basic Example with Histogram

Consider a simple histogram using normally distributed data:

x <- rnorm(100)
hist(x, xlab="Variable Label", ylab="Density", main="Title of Plot")

To increase font sizes for various elements, you can apply the cex family parameters directly within your plotting function:

hist(x, 
     xlab="Variable Label", 
     ylab="Density", 
     main="Title of Plot",
     cex.lab=1.5,  # Scale axis labels
     cex.axis=1.5, # Scale numbers on the axes
     cex.main=1.5, # Scale the main title
     cex.sub=1.5   # Scale subtitles if any
)

Using `par()` for Global Settings

You can also set these parameters globally using par(), which affects subsequent plots unless changed back:

# Set global text size adjustments
par(cex.lab=0.8, cex.axis=0.9)

# Plotting after parameter adjustment
hist(x, xlab="Variable Label", ylab="Density", main="Title of Plot")

# Reset to default settings
par(cex.lab=1.0, cex.axis=1.0)

Special Considerations

Function-Specific Behavior: Not all functions respect the cex parameter. For instance, in some cases, like hist(), direct use of cex might not affect text size as expected.

Plot Type Variability: In plots like hierarchical cluster dendrograms (agnes() from the cluster package), cex can effectively alter text size:

library(cluster)
data(votes.repub)
agn1 <- agnes(votes.repub, metric = "manhattan", stand = TRUE)
plot(agn1, cex=0.5)  # Scales down the font size

Using mtext() for Labels: When direct label setting within a function doesn’t work as expected, mtext() can be an alternative to customize text:
```
hist(x)
mtext("Custom X-axis Label", side=1, line=2, cex=0.8)  # Use cex for font scaling
```
PDF Output Adjustments: For PDF output, the pointsize parameter in pdf() can control default text size:
```
pdf("plot.pdf", pointsize=12)
hist(x)
dev.off()
```

Best Practices

Consistency Across Plots: Use consistent scaling factors (cex) across multiple plots to maintain uniformity.
Resetting Parameters: Always remember to reset par() settings after customizing them to avoid unintended effects on subsequent plots.
Experimentation: Different plotting functions may react differently to text size parameters, so testing is key.

Conclusion

Mastering the manipulation of text sizes in R plots enables you to create clear and professional visualizations tailored for various audiences. Whether through direct function arguments or global settings via par(), understanding these tools empowers you to effectively communicate your data insights.