`plot2`

ggplot2, without the boilerplate.

plot2 is a single-function wrapper around ggplot2 that handles the routine work — grouping, aggregating, sorting, facetting, labelling — so you do not have to. It returns a standard ggplot object, so every ggplot2 extension and layer works as usual.

Install from CRAN or r-universe:

install.packages("plot2", repos = c("https://cran.r-project.org",
                                    "https://msberends.r-universe.dev"))

The same chart, two ways

A sorted column chart of vehicle counts per class, with data labels. The dataset mpg is bundled in ggplot2.

# ggplot2 + dplyr + forcats
library(ggplot2)
library(dplyr)
library(forcats)

mpg |>
  count(class) |>
  mutate(class = fct_reorder(class, n)) |>
  ggplot(aes(x = class, y = n)) +
  geom_col(width = 0.6) +
  geom_text(aes(label = n), vjust = -0.5, size = 3.5) +
  scale_y_continuous(expand = expansion(mult = c(0, 0.15))) +
  labs(x = "Vehicle class", y = "Count") +
  theme_minimal()

# plot2
library(plot2)

mpg |> plot2(x = class, y = n(), x.sort = "freq-asc")

Both produce the same chart. The difference is in how much you write.

What you stop writing

Task	ggplot2	plot2
Aggregate then plot	`group_by()` + `summarise()` + `ggplot()` + `geom_col()`	`plot2(x = a, y = mean(b))`
Sort bars	`mutate(fct_reorder(...))`	`x.sort = "freq-desc"`
Group by colour	`aes(fill = var)` for bars, `aes(colour = var)` for points, each needing its own `scale_*`	`category = var`
Set a colour palette	`scale_fill_manual(values = ...)` or `scale_colour_manual(values = ...)` depending on geom	`colour = c(...)` in all cases
Facet	`+ facet_wrap(~var)`	`facet = var`
Data labels	`+ geom_text(aes(label = ...), vjust = ...)`	automatic on discrete axes
Stacked / filled bars	`geom_col(position = "stack")` + scale	`stacked = TRUE` / `stacked_fill = TRUE`
Clean axis expansion	`scale_y_continuous(expand = expansion(...))`	built-in default

Examples

All examples below use datasets bundled in base R or ggplot2.

The four arguments: x, y, category, facet

plot2 is built around four named arguments. That is all you need for most plots:

# x and y: plot2 picks the type. Two numerics → scatter.
iris |> plot2(x = Sepal.Width, y = Sepal.Length)

# category: colours the groups — works the same regardless of geom type.
# In ggplot2 you would use aes(fill = ...) for bars and aes(colour = ...) for
# points, each requiring its own scale_*. Here it is always category =.
iris |> plot2(x = Sepal.Width, y = Sepal.Length, category = Species)

# Categorical x → column with data labels. Inline aggregation in y.
iris |> plot2(x = Species, y = mean(Sepal.Length))

# facet: splits into panels. Fourth argument, same idea.
mpg |> plot2(x = class, y = mean(hwy), category = drv, facet = year)

Sorting

mpg |> plot2(x = manufacturer, y = n(), x.sort = "freq-desc")

Colour

colour = sets the palette for any plot type — no need to choose between scale_fill_* and scale_colour_*:

iris |> plot2(x = Sepal.Width, y = Sepal.Length, category = Species,
              colour = "viridis")
iris |> plot2(x = Sepal.Width, y = Sepal.Length, category = Species,
              colour = c(setosa = "#3F681C", versicolor = "#375E97", virginica = "#FFBB00"))

When you need to distinguish the stroke from the fill (e.g. bars with a coloured border), colour controls the stroke and colour_fill controls the interior:

iris |> plot2(x = Species, y = mean(Sepal.Length),
              colour = "black",
              colour_fill = c(setosa = "#3F681C", versicolor = "#375E97", virginica = "#FFBB00"))

Inline aggregations

Any summarising function works directly in the y argument — no group_by() or summarise() needed:

mpg |> plot2(x = class, y = mean(hwy))
mpg |> plot2(x = class, y = median(cty))
mpg |> plot2(x = class, y = n())
mpg |> plot2(x = class, y = n(), category = drv, stacked_fill = TRUE)

Titles accept the same inline expressions:

mpg |> plot2(
  x = class, y = n(),
  title = paste("Based on", n(), "vehicles from", n_distinct(manufacturer), "manufacturers"),
  caption = paste("Highway mpg ranges from", paste(range(hwy), collapse = " to "))
)

Special plot types

# Dumbbell: compare two groups side by side
mpg |> plot2(x = manufacturer, y = mean(hwy), category = drv,
             type = "dumbbell",
             x.sort = "freq-desc")

# Histogram with automatic bin width
diamonds |> plot2(x = price, type = "hist")

# Stacked column
mpg |> plot2(x = class, y = n(), category = drv, stacked = TRUE)

Sankey / alluvial diagram

Titanic |>
  plot2(x = c(Age, Class, Survived),
        category = Sex,
        type = "sankey")

Correlation matrix

mtcars |>
  cor() |>
  plot2(datalabels = TRUE,
        colour = c("steelblue", "white", "firebrick"),
        title = "Correlation matrix of `mtcars`")

Full ggplot2 compatibility

plot2 returns a standard ggplot object. Any ggplot2 layer, scale, or extension can be appended:

mpg |>
  plot2(x = class, y = mean(hwy), x.sort = "freq-desc") +
  geom_hline(yintercept = mean(mpg$hwy), linetype = "dashed") +
  labs(caption = "Dashed line: overall mean")

Going further

plot2 supports secondary y-axes, Google and system fonts, geographic plots via sf objects, regression model objects, viridis and custom colour palettes, and an interactive Shiny-based plot builder.

The package ships with admitted_patients, a synthetic hospital dataset used throughout the full vignette. For an overview of every supported plot type, see the supported types reference.

# Build any plot interactively — all plot2 arguments available in the UI
iris |> create_interactively()

Ways to call plot2()

Like base plot(), all input styles work:

# Formula
plot2(hp ~ mpg, data = mtcars)

# Pipe
mtcars |> plot2(mpg, hp)

# Named arguments
plot2(mtcars, x = mpg, y = hp)

# Positional
plot2(mtcars, mpg, hp)

Getting involved

Issues, feature requests, and pull requests are welcome at https://github.com/msberends/plot2. Familiarity with ggplot2 and the tidyverse is especially useful as the package continues to develop.

Licence

GPL-2. See the LICENSE file for details.