Day 3: Data Visualization with ggplot2 | R Programming in 5 Days

The Grammar of Graphics

ggplot2 maps data variables (aesthetics: x, y, color, size, shape) to geometric objects (geoms: points, lines, bars, boxes). Every plot starts with ggplot(data, aes(x=var, y=var)) then adds layers with +. geom_point() makes scatter plots; geom_line() line charts; geom_histogram() histograms; geom_col()/geom_bar() bar charts; geom_boxplot() box plots. Facets (facet_wrap, facet_grid) create small multiples.

Scales, Themes, and Labels

Scales control how data maps to visual properties: scale_color_brewer() uses ColorBrewer palettes (designed for clarity and accessibility), scale_x_log10() transforms axes, scale_fill_manual() sets custom colors. Themes control non-data elements: theme_minimal(), theme_classic(), theme_bw(). Labs(title, x, y, color) sets labels. theme() customizes individual elements: theme(legend.position='bottom').

Advanced ggplot2: Statistical Layers

stat_smooth() adds regression lines with confidence intervals. geom_violin() shows distribution shape. geom_density_2d() shows 2D density contours. geom_tile() makes heat maps. coord_flip() rotates axes for horizontal bar charts. patchwork library combines multiple plots. ggsave() exports to PDF, PNG, or SVG at any resolution — R produces print-ready graphics with no post-processing.

library(ggplot2)
library(dplyr)

# Scatter plot with regression line
ggplot(mtcars, aes(x = wt, y = mpg, color = factor(cyl))) +
  geom_point(size = 3, alpha = 0.8) +
  geom_smooth(method = 'lm', se = TRUE, aes(group = 1),
              color = 'black', linetype = 'dashed') +
  scale_color_brewer(palette = 'Set1', name = 'Cylinders') +
  labs(
    title    = 'Car Weight vs Fuel Efficiency',
    subtitle = 'Data: mtcars | Dashed line: linear fit',
    x        = 'Weight (1000 lbs)',
    y        = 'Miles per Gallon'
  ) +
  theme_minimal(base_size = 13)

# Save high-resolution
ggsave('mpg_plot.pdf', width = 8, height = 5, dpi = 300)

# Faceted bar chart
ggplot(diamonds, aes(x = cut, fill = clarity)) +
  geom_bar(position = 'fill') +
  facet_wrap(~color) +
  coord_flip() +
  labs(y = 'Proportion', title = 'Diamond Cut by Clarity and Color') +
  theme_bw()

💡

Save all ggplot2 graphics with ggsave() as PDF or SVG for scalable vector output. PNG at 300 DPI is fine for presentations. Never screenshot plots from the RStudio Viewer — the resolution will be poor.

📝 Day 3 Exercise

Build a Complete Analysis Dashboard

Load the gapminder package: install.packages('gapminder'); library(gapminder)
Create a scatter plot of GDP per capita vs life expectancy, colored by continent
Add stat_smooth(method='loess') to show the trend
Use facet_wrap(~year) to create small multiples for 5 different years
Export to a PDF with ggsave() at 300 DPI

Day 3 Summary

ggplot2 builds plots by layering data, aesthetics, and geometric objects
Aesthetics (aes): x, y, color, fill, size, shape, alpha map data to visuals
Geoms: geom_point, geom_line, geom_col, geom_histogram, geom_boxplot
Facets create small multiples: facet_wrap (1D) and facet_grid (2D)
ggsave() exports print-ready PDF/PNG/SVG at any resolution

Challenge

Create a complete ggplot2 report with 4 publication-quality plots analyzing a dataset of your choice. Each plot should use a different geom type, have proper titles/axis labels, and use a colorblind-accessible palette.

← Previous

Day 2: Data Manipulation with dplyr & tidyr

Day 4: Statistical Analysis & Modeling

→