In taylordunn/dgitheme: DGI Clinical Theme

options(width = 150)
knitr::opts_chunk$set(
    message = TRUE,
    fig.width = 11
)
library(tidyverse)
library(gt) # tables
library(here) # relative file paths
library(patchwork) # combining ggplots
library(glue) # combining strings and data

# Set ggplot2 theme and defaults
theme_set(cowplot::theme_cowplot() + cowplot::background_grid(major = "xy"))
ggp <- function(...) ggplot(...) +
  scale_color_brewer(palette = "Set1") +
  scale_fill_brewer(palette = "Set1")

# Alternatively, use DGI Clinical themes and fonts
library(dgitheme)
# This command must be run once to register all fonts installed on Windows
# extrafont::font_import()
# This command must be run once in each R session to register fonts
extrafont::loadfonts(device = "win")

# DGI fonts
theme_update(text = element_text(family = "Lato"),
             plot.title = element_text(family = "Bebas Neue", size = 25))
# DGI colors
ggp <- function(...) ggplot(...) +
  scale_color_dgi() +
  scale_fill_dgi()

Plots

Below is a comparison of

(A) the default ggplot2 qualitative color palette,

(B) the "Set1" palette from scale_fill_brewer(),

(D) the "triadic" palette from dgitheme::scale_fill_dgi().

Also demonstrates how to use the patchwork package to arrange plots.

p1 <- ggplot(iris, aes(Sepal.Length, fill = Species)) +
  geom_density() +
  scale_y_continuous(NULL, expand = c(0, 0)) +
  labs(title = "Default ggplot2 palette") +
  theme(legend.position = c(0.7, 0.7))
p2 <- p1 +
  scale_fill_brewer(palette = "Set1") +
  labs(title = "Brewer Set1 palette")
p3 <- p1 +
  scale_fill_dgi() +
  labs(title = "DGI Clinical main palette")
p4 <- p1 +
  scale_fill_dgi("triadic") +
  labs(title = "DGI Clinical triadic palette")
p1 + p2 + p3 + p4 +
  plot_layout(nrow = 2) +
  plot_annotation(tag_levels = "A")

When plotting ordered factors, some options include:

(A) the ggplot2 default virdis palette, (B) the default sequential color palette ("Blues") from scale_fill_brewer(), (C) the "main" palette (reversed) from dgitheme::scale_fill_dgi(), and (D) the "teal grey" palette (reversed) from dgitheme::scale_fill_dgi().

p1 <- diamonds %>%
  ggplot(aes(clarity, fill = cut)) +
  geom_bar() +
  scale_y_continuous(expand = c(0, 0)) +
  labs(title = "Default ggplot2 palette") +
  theme(legend.position = c(0.7, 0.7))
p2 <- p1 +
  scale_fill_brewer() +
  labs(title = "Brewer Blues palette")
p3 <- p1 +
  scale_fill_dgi(reverse = TRUE) +
  labs(title = "DGI Clinical main palette")
p4 <- p1 +
  scale_fill_dgi("teal grey", reverse = TRUE) +
  labs(title = "DGI Clinical teal grey palette")
p1 + p2 + p3 + p4 +
  plot_layout(nrow = 2) +
  plot_annotation(tag_levels = "A")

Text

From the options(width = 150) argument at setup, code output should fill the width of the page.

glimpse(mpg)

Also, printing of dataframes will default to a paged format to allow full exploration:

mpg

Text can be **bold** or *italic*.

Mathematical expressions can be rendered using LaTeX syntax: $\sigma^2 = \sum \frac{(x_i - \mu)^2}{N}$, or on a separate line as:

$$\chi^2 = \frac{(n-1) s^2}{\sigma^2}$$

Tables

Tables are best rendered using the gt() package, which has extensive documentation and examples.

Here is an example making use of the gt::data_color() function to add dgitheme colors corresponding to data:

pizzaplace %>%
  filter(
    type %in% c("chicken", "supreme")) %>%
  group_by(type, size) %>%
  summarize(
    sold = n(),
    income = sum(price)
  ) %>%
  gt(rowname_col = "size") %>%
  data_color(
    columns = vars(sold, income),
    colors = scales::col_numeric(
      palette = dgitheme::dgi_pal("teal white", reverse = T)(2),
      domain = NULL)
  )

countrypops %>%
  filter(country_name == "Mongolia") %>%
  select(-contains("code")) %>%
  tail(10) %>%
  gt() %>%
  data_color(
    columns = vars(population),
    colors = scales::col_numeric(
      palette = dgitheme::dgi_pal("complementary")(2),
      domain = c(2.6E6, 3.1E6))
  )