words_per_constitution_article.R

library(tidyverse)
library(gutenbergr)

constitution_raw <- gutenberg_download(5)

constitution <- constitution_raw |> 
  slice(36:546) |> 
  filter(text != "") |> 
  mutate(text_lc = str_to_lower(text)) |> 
  mutate(is_article = str_starts(text_lc, "article")) |> 
  mutate(
    article_number = cumsum(is_article),
    article_number = ifelse(article_number == 0, "Preamble", as.character(as.roman(article_number))),
    article_number = fct_inorder(article_number)
  )

words_per_article <- constitution |> 
  mutate(words = stringi::stri_count_words(text_lc)) |> 
  group_by(article_number) |> 
  summarize(total_words = sum(words)) |> 
  mutate(
    prop = total_words / sum(total_words),
    prop_nice = scales::label_percent(accuracy = 0.1)(prop),
    total_nice = scales::label_comma()(total_words)
  ) |> 
  mutate(label = paste0(total_nice, "\n(", prop_nice, ")"))
    
ggplot(words_per_article, aes(x = article_number, y = total_words)) +
  geom_col(aes(fill = article_number == "I")) +
  geom_text(
    aes(label = label), nudge_y = 150, 
    size = 7, size.unit = "pt", family = "Public Sans", fontface = "bold"
  ) +
  scale_y_continuous(labels = scales::label_comma()) +
  # Colors from the USWDS
  # https://designsystem.digital.gov/utilities/color/#basic-palette-2
  scale_fill_manual(values = c("#0076d6", "#e52207"), guide = "none") +
  labs(
    x = NULL,
    y = "Total words",
    title = "More than half of the\nUS Constitution is in Article I"
  ) +
  theme_minimal(base_family = "Public Sans") +
  theme(
    panel.grid.minor.y = element_blank(),
    panel.grid.major.x = element_blank(),
    plot.title = element_text(face = "bold", hjust = 0.5)
  )

ggsave("~/Desktop/words.png", width = 5, height = 3.5, units = "in", dpi = 300, bg = "white")