pyramids-2024.R

library(tidyverse)
library(tidycensus)

# Get the data
pyramid_data <- get_estimates(
  geography = "state",
  product = "characteristics",
  breakdown = c("AGEGROUP", "SEX"),
  breakdown_labels = TRUE,
  vintage = 2024,
  year = 2024
) %>%
  filter(AGEGROUP != "All ages", SEX != "Both sexes", NAME %in% c("Maine", "Utah")) %>%
  group_by(NAME) %>%
  mutate(
    value_pct = value / sum(value) * 100,
    value_pct = ifelse(SEX == "Male", -value_pct, value_pct)
  ) %>%
  ungroup()

# Create enhanced population pyramids
ggplot(pyramid_data, aes(x = value_pct, y = AGEGROUP, fill = SEX)) +
  geom_col(alpha = 0.9, width = 0.85) +
  geom_vline(xintercept = 0, color = "gray40", linewidth = 0.5) +
  scale_x_continuous(
    labels = function(x) paste0(abs(x), "%"),
    expand = expansion(mult = c(0.05, 0.05))
  ) +
  scale_y_discrete(
    labels = function(x) str_remove_all(x, "Age\\s|\\syears") 
  ) +
  scale_fill_manual(
    values = c("Female" = "#A23B72", "Male" = "#2E86AB"),
    breaks = c("Male", "Female")  # This controls the legend order
  ) +
  facet_wrap(~NAME, scales = "free_x") +
  labs(
    title = "The Oldest (Maine) and Youngest (Utah) States in the U.S.",
    subtitle = "Population distribution by age and sex, 2024",
    x = "Percentage of State Population",
    y = "Age Group",
    caption = "Source: U.S. Census Bureau, Vintage 2024 Population Estimates | tidycensus R package"
  ) +
  theme_minimal(base_size = 11) +
  theme(
    # Title and text styling
    plot.title = element_text(face = "bold", size = 18, hjust = 0.5, 
                              margin = margin(b = 5)),
    plot.subtitle = element_text(color = "gray50", size = 13, hjust = 0.5,
                                 margin = margin(b = 15)),
    plot.caption = element_text(color = "gray60", size = 9, hjust = 1,
                                margin = margin(t = 15)),
    
    # Facet styling
    strip.text = element_text(size = 14, face = "bold", color = "gray20"),
    strip.background = element_rect(fill = "gray95", color = NA),
    
    # Axis styling
    axis.title.x = element_text(size = 11, margin = margin(t = 10)),
    axis.title.y = element_text(size = 11, margin = margin(r = 10)),
    axis.text.y = element_text(size = 10, color = "gray30"),
    axis.text.x = element_text(size = 10, color = "gray30"),
    
    # Grid styling
    panel.grid.major.y = element_line(color = "gray90", linewidth = 0.3),
    panel.grid.minor.y = element_blank(),
    panel.grid.major.x = element_line(color = "gray85", linewidth = 0.3),
    panel.grid.minor.x = element_blank(),
    
    # Panel styling
    panel.spacing = unit(1.5, "lines"),
    
    # Legend styling - moved to bottom with horizontal layout
    legend.position = "bottom",
    legend.direction = "horizontal",
    legend.title = element_text(size = 10, margin = margin(r = 3)),  # Reduced spacing
    legend.text = element_text(size = 10),
    legend.key.width = unit(1.5, "cm"),
    legend.key.height = unit(0.3, "cm"),
    legend.margin = margin(t = 5, b = 0),
    legend.spacing.x = unit(0.3, "cm"),
    
    # Overall plot margins
    plot.margin = margin(20, 20, 10, 20)
  )