R ggplot2 - Scatter Plot with Examples

Key Insights

ggplot2 uses a layered grammar of graphics approach where you build plots by adding geometric objects, scales, and themes to a base ggplot object
Scatter plots in ggplot2 are created using geom_point(), with extensive customization options for aesthetics like color, size, shape, and transparency
Advanced scatter plot techniques include faceting, trend lines, marginal distributions, and interactive annotations that transform basic visualizations into analytical tools

Basic Scatter Plot Construction

The fundamental ggplot2 scatter plot requires a dataset, aesthetic mappings, and a point geometry layer. Here’s the minimal implementation:

library(ggplot2)

# Create sample data
set.seed(123)
df <- data.frame(
  x = rnorm(100, mean = 50, sd = 10),
  y = rnorm(100, mean = 50, sd = 10)
)

# Basic scatter plot
ggplot(df, aes(x = x, y = y)) +
  geom_point()

The ggplot() function initializes the plot object with data and aesthetic mappings. The aes() function maps variables to visual properties. The + operator adds layers—in this case, geom_point() renders the scatter plot.

Customizing Point Aesthetics

Control point appearance through size, color, shape, and transparency:

# Fixed aesthetics (applied to all points)
ggplot(df, aes(x = x, y = y)) +
  geom_point(color = "steelblue", size = 3, shape = 17, alpha = 0.6)

# Variable aesthetics (mapped to data)
df$category <- sample(c("A", "B", "C"), 100, replace = TRUE)
df$value <- runif(100, 1, 10)

ggplot(df, aes(x = x, y = y, color = category, size = value)) +
  geom_point(alpha = 0.7) +
  scale_size_continuous(range = c(1, 8))

Shape codes: 0-25 represent different shapes. Common choices: 16 (filled circle), 17 (filled triangle), 15 (filled square). Alpha values range from 0 (transparent) to 1 (opaque).

Color Scales and Palettes

ggplot2 provides multiple color scale functions for different data types:

# Discrete colors
ggplot(df, aes(x = x, y = y, color = category)) +
  geom_point(size = 3) +
  scale_color_manual(values = c("A" = "#E41A1C", "B" = "#377EB8", "C" = "#4DAF4A"))

# Using ColorBrewer palettes
ggplot(df, aes(x = x, y = y, color = category)) +
  geom_point(size = 3) +
  scale_color_brewer(palette = "Set1")

# Continuous color gradient
ggplot(df, aes(x = x, y = y, color = value)) +
  geom_point(size = 3) +
  scale_color_gradient(low = "yellow", high = "red")

# Viridis color scales (perceptually uniform)
ggplot(df, aes(x = x, y = y, color = value)) +
  geom_point(size = 3) +
  scale_color_viridis_c(option = "plasma")

Adding Trend Lines and Statistical Summaries

Overlay regression lines, smoothers, or confidence intervals:

# Linear regression line
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_smooth(method = "lm", se = TRUE, color = "red")

# LOESS smoother
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_smooth(method = "loess", span = 0.3)

# Multiple regression lines by group
ggplot(df, aes(x = x, y = y, color = category)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE)

# Custom formula
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_smooth(method = "lm", formula = y ~ poly(x, 2), se = TRUE)

The se parameter controls confidence interval display. The span parameter in LOESS controls smoothing degree (smaller = less smooth).

Faceting for Multi-Panel Plots

Split data into subplots based on categorical variables:

# Create richer dataset
df$group1 <- sample(c("Group1", "Group2"), 100, replace = TRUE)
df$group2 <- sample(c("Type_X", "Type_Y"), 100, replace = TRUE)

# Facet by one variable
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  facet_wrap(~ category, ncol = 2)

# Facet by two variables (grid)
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  facet_grid(group1 ~ group2)

# Free scales for each facet
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  facet_wrap(~ category, scales = "free")

Use scales = "free" when subgroups have different ranges. facet_grid() creates a matrix layout, while facet_wrap() wraps panels into a rectangular layout.

Handling Overplotting

When points overlap, use transparency, jittering, or 2D density representations:

# Create overlapping data
df_overlap <- data.frame(
  x = rnorm(5000, mean = 50, sd = 5),
  y = rnorm(5000, mean = 50, sd = 5)
)

# Alpha transparency
ggplot(df_overlap, aes(x = x, y = y)) +
  geom_point(alpha = 0.1)

# Jittering
ggplot(df_overlap, aes(x = x, y = y)) +
  geom_jitter(width = 0.5, height = 0.5, alpha = 0.3)

# 2D density contours
ggplot(df_overlap, aes(x = x, y = y)) +
  geom_density_2d() +
  geom_point(alpha = 0.2)

# Hexagonal binning
ggplot(df_overlap, aes(x = x, y = y)) +
  geom_hex(bins = 30) +
  scale_fill_viridis_c()

Annotations and Labels

Add text, arrows, and reference lines:

# Text labels for specific points
df_labeled <- df[1:5, ]
df_labeled$label <- paste("Point", 1:5)

ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_text(data = df_labeled, aes(label = label), 
            vjust = -0.5, hjust = 0.5, size = 3)

# Using ggrepel to avoid label overlap
library(ggrepel)
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_text_repel(data = df_labeled, aes(label = label))

# Reference lines
ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  geom_hline(yintercept = mean(df$y), linetype = "dashed", color = "red") +
  geom_vline(xintercept = mean(df$x), linetype = "dashed", color = "blue") +
  annotate("rect", xmin = 40, xmax = 60, ymin = 40, ymax = 60, 
           alpha = 0.2, fill = "yellow")

Theme Customization

Control non-data plot elements:

ggplot(df, aes(x = x, y = y, color = category)) +
  geom_point(size = 3) +
  labs(
    title = "Scatter Plot Analysis",
    subtitle = "Sample data with categorical grouping",
    x = "X Variable (units)",
    y = "Y Variable (units)",
    color = "Category"
  ) +
  theme_minimal() +
  theme(
    plot.title = element_text(size = 16, face = "bold"),
    legend.position = "bottom",
    panel.grid.minor = element_blank(),
    axis.text = element_text(size = 10)
  )

Built-in themes: theme_minimal(), theme_classic(), theme_bw(), theme_dark(). The theme() function provides granular control over every visual element.

Real-World Example: Correlation Analysis

Complete example analyzing the mtcars dataset:

library(ggplot2)
library(dplyr)

# Prepare data
mtcars_prep <- mtcars %>%
  mutate(
    cyl_factor = factor(cyl),
    am_label = factor(am, labels = c("Automatic", "Manual"))
  )

# Comprehensive scatter plot
ggplot(mtcars_prep, aes(x = wt, y = mpg)) +
  geom_point(aes(color = cyl_factor, size = hp), alpha = 0.7) +
  geom_smooth(method = "lm", se = TRUE, color = "black", linetype = "dashed") +
  facet_wrap(~ am_label) +
  scale_color_manual(
    values = c("4" = "#2E86AB", "6" = "#A23B72", "8" = "#F18F01"),
    name = "Cylinders"
  ) +
  scale_size_continuous(range = c(3, 10), name = "Horsepower") +
  labs(
    title = "Fuel Efficiency vs Vehicle Weight",
    subtitle = "Grouped by transmission type and engine configuration",
    x = "Weight (1000 lbs)",
    y = "Miles per Gallon"
  ) +
  theme_bw() +
  theme(
    legend.position = "right",
    strip.background = element_rect(fill = "lightblue"),
    plot.title = element_text(hjust = 0.5, face = "bold")
  )

This example demonstrates multiple techniques: faceting by transmission type, color-coding by cylinder count, sizing by horsepower, and adding regression lines to identify trends within each transmission category.