The objective of ggrefine is to provide complete themes for publication-quality ‘ggplot2’ visualisation. Functions are provided to modify these based on the positional axis scales and orientation of a particular plot.
Installation
Install from CRAN, or development version from GitHub.
install.packages("ggrefine")
pak::pak("davidhodge931/ggrefine")Theme
The themes are built to work with the refine functions in that they have all axis and panel grid elements.
They can also be customised easily.
The theme_ggplot2 function has a smart panel_grid_colour default that is derived from the panel_background_fill.
library(ggplot2)
library(ggrefine)
set_theme(theme_lights())
update_panel_size(heights = unit(5, "cm"), widths = unit(7.5, "cm"))
p <- mpg |>
ggplot() +
aes(x = hwy) +
geom_histogram(
stat = "bin",
) +
scale_y_zero() +
scale_fill_colourblend_discrete() +
refine_axis_grid(discrete = "none")
p + labs(title = "theme_lights")
#> `stat_bin()` using `bins = 30`. Pick better value `binwidth`.
update_greys()
p + labs(title = "theme_greys")
#> `stat_bin()` using `bins = 30`. Pick better value `binwidth`.
update_darks()
p + labs(title = "theme_darks")
#> `stat_bin()` using `bins = 30`. Pick better value `binwidth`.
Scales and Aesthetics
The package provides dynamic color and fill aesthetic and scale helpers that evaluate aesthetics late to provide automatic colour or fill properties.
These can be:
- a blend of fill (or colour)
- a contrast of fill (intended for text geoms)
- a contrast of the panel.background fill (intended for text geoms).
Note aes_colourcontrast requires the ggrefine function to be used instead of ggplot2::ggplot to work where no fill aesthetic is mapped. As such, it can be useful to always use ggrefine followed by ggplot2::aes to avoid thinking about this.
penguins |>
ggrefine() +
aes(x = flipper_len, y = body_mass, fill = species) +
geom_point() +
scale_fill_colourblend_discrete() +
refine_axis_grid(discrete = "none")
#> Warning: Removed 2 rows containing missing values or values outside the scale range
#> (`geom_point()`).
p <- mpg |>
dplyr::count(class) |>
dplyr::mutate(class = forcats::fct_reorder(class, n)) |>
ggrefine() +
aes(x = n, y = class, label = n) +
geom_col(width = 0.7) +
scale_x_zero() +
scale_fill_colourblend_discrete() +
refine_axis_grid(discrete = "y")
p +
geom_text(aes_colourcontrast(), hjust = 1.25)
update_palette(fixed = jumble::navy)
p +
geom_text(aes_colourcontrast(), hjust = 1.25) 
update_palette(discrete = jumble::jumble)
p +
aes(fill = class) +
geom_text(aes_colourcontrast(), hjust = 1.25) +
refine_legend(position = "top") 
p +
geom_text(aes(x = n + (max(n) * 0.05), !!!aes_colourpanel()))
update_darks()
p +
geom_text(aes(x = n + (max(n) * 0.05), !!!aes_colourpanel()))
