How to Use geom_smooth in ggplot2 for Trend Lines

Use geom_smooth() in ggplot2 to add a smooth trend line or curve to your plot. It fits a model like LOESS or linear regression by default and shows confidence intervals. Simply add geom_smooth() to your ggplot object with optional parameters to customize the smoothing method and appearance.

Syntax

The basic syntax of geom_smooth() is:

• mapping: aesthetic mappings like aes(x, y).
• method: smoothing method, e.g., "loess" (default for small data) or "lm" for linear model.
• se: logical, whether to display confidence interval (default TRUE).
• formula: formula for the model, e.g., y ~ x.
• color, fill: colors for line and confidence band.

Example syntax:

geom_smooth(mapping = aes(x, y), method = "loess", se = TRUE)

r

geom_smooth(mapping = aes(x, y), method = "loess", se = TRUE)

Example

This example shows how to add a smooth trend line to a scatter plot of mpg dataset with geom_smooth(). It uses the default LOESS method and shows the confidence interval.

r

library(ggplot2)

# Scatter plot with smooth trend line
p <- ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_point() +
  geom_smooth() +
  labs(title = "Engine Displacement vs Highway MPG with Smooth Line")

print(p)

Output

[A scatter plot with points and a smooth curve with shaded confidence band]

Common Pitfalls

Common mistakes when using geom_smooth() include:

• Not specifying method when data is large, causing slow LOESS computation.
• Forgetting to set se = FALSE if you don't want the confidence band.
• Using geom_smooth() without aes() mappings, which can cause errors.
• Passing incorrect formula syntax for custom models.

Example of wrong and right usage:

r

# Wrong: No aes mapping
# geom_smooth() # may cause error if no x and y defined

# Right: Proper aes mapping and method
geom_smooth(aes(x = displ, y = hwy), method = "lm", se = FALSE)

Quick Reference

Parameter	Description	Default
mapping	Aesthetic mappings like x and y	Required
method	Smoothing method: "loess", "lm", etc.	"loess" for small data
se	Show confidence interval band	TRUE
formula	Model formula, e.g. y ~ x	Depends on method
color	Line color	ggplot default
fill	Confidence band fill color	ggplot default

Key Takeaways

Add smooth trend lines easily with geom_smooth() in ggplot2.

Specify method like "lm" for linear or "loess" for local smoothing.

Use se = FALSE to hide confidence intervals if not needed.

Always map x and y aesthetics inside aes() for geom_smooth().

Large datasets may require method = "lm" for faster smoothing.