Multivariate scales for ggplot2, written by Claus O. Wilke
This package can be installed from github. It requires the development version of the colorspace package. It also requires ggplot2 3.0.0, which was released on July 3, 2018.
install.packages("colorspace", repos = "http://R-Forge.R-project.org")
devtools::install_github("clauswilke/multiscales")
This is an experimental package. Use at your own risk. API is not stable. No user support provided.
Visualizing the lead/lag of Clinton vs. Trump in the 2016 presidential election jointly with the uncertainty of the lead/lag estimates. This visualization shows that for many states the outcome was difficult to predict. (Example taken from Correll et al., Value-Suppressing Uncertainty Palettes, CHI 2018.)
library(ggplot2)
library(multiscales)
colors <- scales::colour_ramp(
colors = c(red = "#AC202F", purple = "#740280", blue = "#2265A3")
)((0:7)/7)
ggplot(US_polling) +
geom_sf(aes(fill = zip(Clinton_lead, moe_normalized)), color = "gray30", size = 0.2) +
coord_sf(datum = NA) +
bivariate_scale("fill",
pal_vsup(values = colors, max_desat = 0.8, pow_desat = 0.2, max_light = 0.7, pow_light = 1),
name = c("Clinton lead", "uncertainty"),
limits = list(c(-40, 40), c(0, 1)),
breaks = list(c(-40, -20, 0, 20, 40), c(0, 0.25, 0.50, 0.75, 1.)),
labels = list(waiver(), scales::percent),
guide = "colourfan"
) +
theme_void() +
theme(
legend.key.size = grid::unit(0.8, "cm"),
legend.title.align = 0.5,
plot.margin = margin(5.5, 20, 5.5, 5.5)
)
For comparison, the same plot with a univariate color scale. Not it appears that in every state we know who is in the lead.
ggplot(US_polling) +
geom_sf(aes(fill = Clinton_lead), color = "gray30", size = 0.2) +
coord_sf(datum = NA) +
scale_fill_gradient2(
low = "#AC202F", mid = "#740280", high = "#2265A3",
name = c("Clinton lead", "uncertainty"),
limits = c(-40, 40),
breaks = c(-40, -20, 0, 20, 40),
guide = guide_colorbar(
direction = "horizontal",
label.position = "bottom",
title.position = "top",
barwidth = grid::unit(4.0, "cm")
)
) +
theme_void() +
theme(
legend.title.align = 0.5,
legend.key.size = grid::unit(0.8, "cm"),
plot.margin = margin(5.5, 20, 5.5, 5.5)
)