Eric Campbell 5 min read Data Science

How to Create a Bar Chart in ggplot2

Bar charts are the workhorse of data visualization. They excel at comparing quantities across categories, showing distributions, and highlighting differences between groups. When you need to answer...

Key Insights

  • Use geom_bar() for counting observations and geom_col() when you already have aggregated values—understanding this distinction prevents common data transformation mistakes
  • The position parameter ("dodge", "stack", "fill") fundamentally changes how multiple categories display, with grouped bars best for direct comparisons and stacked bars for part-to-whole relationships
  • Professional bar charts require deliberate aesthetic choices: proper color palettes, readable labels, and strategic use of geom_text() for data labels that enhance rather than clutter

Introduction to ggplot2 Bar Charts

Bar charts are the workhorse of data visualization. They excel at comparing quantities across categories, showing distributions, and highlighting differences between groups. When you need to answer “which category has the most?” or “how do these groups compare?”, bar charts deliver clarity.

The ggplot2 package implements Leland Wilkinson’s grammar of graphics, treating visualizations as layered components rather than pre-packaged chart types. This approach gives you precise control over every element while maintaining clean, readable code. For bar charts specifically, you’ll work with geometric objects (geom_bar() or geom_col()), aesthetic mappings, position adjustments, and coordinate systems.

The key to effective ggplot2 bar charts is understanding that you’re not just calling a “make bar chart” function—you’re building a visualization by combining data, aesthetics, and geometric representations.

Basic Bar Chart Setup

The most common confusion with ggplot2 bar charts is choosing between geom_bar() and geom_col(). Here’s the rule: geom_bar() counts your data automatically, while geom_col() uses values you’ve already calculated.

Use geom_bar() when you have raw categorical data:

library(ggplot2)
library(dplyr)

# Sample dataset: customer purchases by product category
purchases <- data.frame(
  category = c("Electronics", "Clothing", "Electronics", "Home", 
               "Clothing", "Electronics", "Home", "Clothing", 
               "Electronics", "Home")
)

# geom_bar() counts occurrences automatically
ggplot(purchases, aes(x = category)) +
  geom_bar()

This creates a bar chart showing the count of purchases in each category. The stat = "count" happens behind the scenes.

Use geom_col() when you’ve pre-aggregated your data:

# Pre-computed sales data
sales_summary <- data.frame(
  category = c("Electronics", "Clothing", "Home"),
  revenue = c(45000, 32000, 28000)
)

# geom_col() uses the values directly
ggplot(sales_summary, aes(x = category, y = revenue)) +
  geom_col()

Notice that geom_col() requires both x and y aesthetics, while geom_bar() only needs x. This distinction saves you from unnecessary data manipulation.

Customizing Bar Appearance

Default gray bars rarely communicate effectively. Color, width, and borders transform basic charts into professional visualizations.

The fill parameter controls interior color, while color controls the border:

ggplot(sales_summary, aes(x = category, y = revenue)) +
  geom_col(fill = "#2E86AB", color = "#06141B", linewidth = 0.8)

For category-specific colors, map fill to a variable inside aes():

ggplot(sales_summary, aes(x = category, y = revenue, fill = category)) +
  geom_col() +
  scale_fill_manual(values = c("Electronics" = "#E63946",
                                "Clothing" = "#F1FAEE", 
                                "Home" = "#A8DADC"))

Adjust bar width to create spacing or emphasis:

ggplot(sales_summary, aes(x = category, y = revenue)) +
  geom_col(width = 0.6, fill = "#457B9D", color = "black", linewidth = 1.2) +
  theme_minimal()

Narrower bars (width < 1) create white space that improves readability, especially with many categories. Thicker borders draw attention but can overwhelm—use them sparingly.

Grouped and Stacked Bar Charts

Comparing multiple variables requires grouped or stacked bars. Your choice depends on whether you’re emphasizing individual values or part-to-whole relationships.

For grouped bars, use position = "dodge":

# Quarterly sales by category
quarterly_sales <- data.frame(
  category = rep(c("Electronics", "Clothing", "Home"), each = 4),
  quarter = rep(c("Q1", "Q2", "Q3", "Q4"), 3),
  revenue = c(12000, 15000, 10000, 8000,   # Electronics
              8000, 9000, 7500, 7500,       # Clothing
              7000, 8000, 6500, 6500)       # Home
)

ggplot(quarterly_sales, aes(x = category, y = revenue, fill = quarter)) +
  geom_col(position = "dodge") +
  scale_fill_brewer(palette = "Set2")

Grouped bars let you compare quarters within each category and categories within each quarter. They work best with 2-4 groups per category.

For stacked bars showing composition:

ggplot(quarterly_sales, aes(x = category, y = revenue, fill = quarter)) +
  geom_col(position = "stack") +
  scale_fill_brewer(palette = "Blues")

Stacked bars show total revenue per category plus the quarterly breakdown. For proportional stacking (100% bars), use position = "fill":

ggplot(quarterly_sales, aes(x = category, y = revenue, fill = quarter)) +
  geom_col(position = "fill") +
  scale_y_continuous(labels = scales::percent) +
  labs(y = "Percentage of Annual Revenue")

This reveals each quarter’s contribution regardless of total category size.

Horizontal Bars and Axis Formatting

Horizontal bars improve readability for long category names and create a natural reading flow for rankings.

Use coord_flip() to rotate any bar chart:

# Product rankings
products <- data.frame(
  product = c("Wireless Headphones", "Smart Watch", 
              "Bluetooth Speaker", "Laptop Stand", "USB-C Hub"),
  units_sold = c(1250, 980, 875, 720, 650)
)

ggplot(products, aes(x = reorder(product, units_sold), y = units_sold)) +
  geom_col(fill = "#06141B") +
  coord_flip() +
  labs(x = NULL, y = "Units Sold", title = "Top 5 Products by Sales Volume")

The reorder() function sorts categories by value, creating an ordered ranking. Setting x = NULL removes the redundant axis label.

Customize axis formatting for clarity:

ggplot(sales_summary, aes(x = category, y = revenue)) +
  geom_col(fill = "#2A9D8F") +
  scale_y_continuous(
    labels = scales::dollar_format(prefix = "$", suffix = "K", scale = 0.001),
    breaks = seq(0, 50000, 10000),
    expand = c(0, 0)
  ) +
  labs(
    title = "Revenue by Product Category",
    subtitle = "Fiscal Year 2024",
    x = "Product Category",
    y = "Revenue"
  ) +
  theme_minimal()

The expand = c(0, 0) removes default padding, making bars touch the axis—a cleaner look for bar charts.

Adding Labels and Annotations

Data labels eliminate guesswork. Place them strategically to enhance, not clutter.

Add values above bars with geom_text():

ggplot(sales_summary, aes(x = category, y = revenue)) +
  geom_col(fill = "#E76F51") +
  geom_text(aes(label = scales::dollar(revenue, scale = 0.001, suffix = "K")),
            vjust = -0.5, size = 4, fontface = "bold") +
  scale_y_continuous(expand = expansion(mult = c(0, 0.1)))

The vjust = -0.5 positions labels above bars, while the y-axis expansion creates headroom.

For stacked bars, position labels inside segments:

ggplot(quarterly_sales, aes(x = category, y = revenue, fill = quarter)) +
  geom_col() +
  geom_text(aes(label = scales::dollar(revenue, scale = 0.001, suffix = "K")),
            position = position_stack(vjust = 0.5),
            color = "white", fontface = "bold", size = 3.5)

With position_stack(vjust = 0.5), labels center within their segments. White text ensures visibility against colored backgrounds.

Theme and Polish

Professional visualizations require cohesive styling. Combine theme elements, color palettes, and typography for polish.

Here’s a complete, production-ready bar chart:

library(scales)

ggplot(quarterly_sales, aes(x = reorder(category, -revenue, sum), 
                             y = revenue, fill = quarter)) +
  geom_col(position = "dodge", width = 0.7) +
  geom_text(aes(label = dollar(revenue, scale = 0.001, suffix = "K")),
            position = position_dodge(width = 0.7),
            vjust = -0.5, size = 3, fontface = "bold") +
  scale_fill_manual(
    values = c("Q1" = "#264653", "Q2" = "#2A9D8F", 
               "Q3" = "#E9C46A", "Q4" = "#F4A261"),
    name = "Quarter"
  ) +
  scale_y_continuous(
    labels = dollar_format(scale = 0.001, suffix = "K"),
    expand = expansion(mult = c(0, 0.15))
  ) +
  labs(
    title = "Quarterly Revenue Performance by Category",
    subtitle = "All figures in thousands (USD)",
    x = NULL,
    y = "Revenue"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    plot.title = element_text(face = "bold", size = 16, margin = margin(b = 5)),
    plot.subtitle = element_text(color = "gray40", margin = margin(b = 15)),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    legend.position = "top",
    axis.text = element_text(color = "gray20"),
    axis.title = element_text(face = "bold")
  )

This example demonstrates professional practices: ordered categories for readability, a cohesive color palette, adequate white space, removed unnecessary grid lines, and clear hierarchy in text elements.

Bar charts in ggplot2 reward thoughtful construction. Master the fundamentals—choosing the right geom, controlling position, and styling deliberately—and you’ll create visualizations that communicate clearly and look professional.

Liked this? There's more.

Every week: one practical technique, explained simply, with code you can use immediately.