ggplot2 Cheat Sheet
Reference for ggplot2's grammar of graphics, common geoms, faceting, and theming used to build layered statistical visualizations in R.
2 PagesIntermediateFeb 28, 2026
Grammar of Graphics Basics
Map data to aesthetics and add a geometry layer.
r
library(ggplot2)ggplot(data = mpg, aes(x = displ, y = hwy, color = class)) + geom_point(size = 2, alpha = 0.7) + labs(title = "Engine Size vs Highway MPG", x = "Displacement (L)", y = "Highway MPG") + theme_minimal()
Common Geoms
The most frequently used chart types.
r
# Bar chartggplot(mpg, aes(x = class)) + geom_bar()# Line chartggplot(economics, aes(x = date, y = unemploy)) + geom_line()# Boxplotggplot(mpg, aes(x = class, y = hwy)) + geom_boxplot()# Histogramggplot(mpg, aes(x = hwy)) + geom_histogram(binwidth = 2)# Smoothed trend lineggplot(mpg, aes(x = displ, y = hwy)) + geom_point() + geom_smooth(method = "lm", se = TRUE)
Faceting & Themes
Build small multiples and control non-data plot elements.
r
ggplot(mpg, aes(x = displ, y = hwy)) + geom_point() + facet_wrap(~ class, ncol = 3) + # small multiples by category theme_bw() + theme(legend.position = "bottom")ggsave("plot.png", width = 8, height = 5, dpi = 300)
Key Concepts
The layered vocabulary behind every ggplot2 chart.
- ggplot()- Initializes a plot object with a dataset and default aesthetic mappings
- aes()- Maps data columns to visual properties (x, y, color, fill, size, shape)
- geom_*- Geometric layer that determines the chart type: geom_point, geom_bar, geom_line, etc.
- facet_wrap/facet_grid- Splits the plot into a grid of subplots (small multiples) by a categorical variable
- scale_*- Controls axis/legend mapping, e.g. scale_color_manual() for custom colors
- theme()- Adjusts non-data plot elements: fonts, gridlines, legend position, background
- stat vs geom- Every geom has a default stat (e.g. geom_bar uses stat_count) that transforms data before drawing
Custom Scales & Color Palettes
Control color mapping, manual palettes, and axis transforms.
r
# Continuous color scale from the viridis family (colorblind-safe)ggplot(mpg, aes(displ, hwy, color = cty)) + geom_point() + scale_color_viridis_c(option = "plasma")# Manual discrete palette with a named vectorggplot(mpg, aes(class, fill = drv)) + geom_bar() + scale_fill_manual(values = c("4" = "#1b9e77", "f" = "#d95f02", "r" = "#7570b3"))# Log-scaled axis with formatted labelsggplot(diamonds, aes(carat, price)) + geom_point(alpha = 0.1) + scale_x_log10() + scale_y_continuous(labels = scales::dollar)
Building & Setting a Custom Theme
Compose a reusable theme object and apply it session-wide.
r
library(ggplot2)my_theme <- theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), panel.grid.minor = element_blank(), legend.position = "top", strip.background = element_rect(fill = "grey20"), strip.text = element_text(color = "white") )theme_set(my_theme) # applies to every plot for the rest of the sessionggplot(mpg, aes(displ, hwy)) + geom_point() + labs(title = "Custom Theme")
Coordinate Systems & Annotations
Reshape the plotting space and add fixed, non-data reference elements.
r
# Flip axes without swapping the aes mappingsggplot(mpg, aes(x = class, y = hwy)) + geom_boxplot() + coord_flip()# Polar coordinates turn a stacked bar into a pie/donutggplot(mpg, aes(x = "", fill = class)) + geom_bar(width = 1) + coord_polar(theta = "y")# Free-text annotation at fixed data coordinates (not tied to a geom)ggplot(mpg, aes(displ, hwy)) + geom_point() + annotate("text", x = 6, y = 40, label = "outlier region", color = "red") + annotate("rect", xmin = 5, xmax = 7, ymin = 35, ymax = 45, alpha = 0.1)
stat_summary(), Ribbons & Multi-Plot Composition
Summarize on the fly, draw confidence bands, and combine plots with patchwork.
r
# Summarize on the fly without pre-aggregating the data frameggplot(mpg, aes(class, hwy)) + geom_jitter(width = 0.1, alpha = 0.4) + stat_summary(fun = mean, geom = "point", color = "red", size = 3) + stat_summary(fun.data = mean_se, geom = "errorbar", width = 0.2)# Confidence-style bands with geom_ribbonggplot(economics, aes(date, unemploy)) + geom_ribbon(aes(ymin = unemploy * 0.95, ymax = unemploy * 1.05), fill = "grey80") + geom_line()# Compose multiple ggplots into one figurelibrary(patchwork)p1 <- ggplot(mpg, aes(displ, hwy)) + geom_point()p2 <- ggplot(mpg, aes(class)) + geom_bar()p1 + p2 + plot_layout(ncol = 2)
Advanced ggplot2 Concepts
Deeper machinery and ecosystem packages beyond the core grammar.
- Position adjustments- position_dodge(), position_stack(), and position_jitter() control how overlapping geoms are arranged
- guides()- Fine-grained legend control, e.g. guides(color = guide_legend(nrow = 2))
- ggplot_build()- Inspects the computed data/render tree behind a plot object, useful for debugging stats
- ggproto- The OOP system ggplot2 extensions use to define new Geoms/Stats/Positions
- Extension ecosystem- patchwork (composition), gganimate (animation), ggrepel (non-overlapping labels), plotly::ggplotly() (interactivity)
- Discrete vs continuous defaults- ggplot2 infers scale type from the aesthetic's data type; override with scale_*_continuous()/scale_*_discrete()
- Layer-specific data/aes- Each geom_*() layer can override the plot-level data/aes, e.g. geom_point(data = subset_df)
Pro Tip
Build ggplot2 charts by adding layers incrementally with + and re-running after each addition - since the grammar of graphics is layer-based, this makes it easy to see exactly which layer introduced an unexpected change.
Was this cheat sheet helpful?
Explore Topics
#Ggplot2#Ggplot2CheatSheet#DataScience#Intermediate#GrammarOfGraphicsBasics#CommonGeoms#FacetingThemes#KeyConcepts#MachineLearning#CheatSheet#SkillVeris
