---

title: “final project” author: “Joshua Hawk” date: “2026-05-01” output: html_document

## `height` was translated to `width`.

## function (data = NULL, mapping = aes(), ..., environment = parent.frame()) 
## {
##     UseMethod("ggplot")
## }
## <bytecode: 0x5c5c86beb640>
## <environment: namespace:ggplot2>

## function (..., orientation = "y") 
## {
##     deprecate_soft0("4.0.0", "geom_errorbarh()", "geom_errorbar(orientation = \"y\")", 
##         id = "no-more-errorbarh")
##     geom_errorbar(..., orientation = orientation)
## }
## <bytecode: 0x5c5c8c502b78>
## <environment: namespace:ggplot2>

# Install if needed:
# install.packages(c("tidycensus", "tigris", "sf", "dplyr", "ggplot2", "stringr"))

library(tidycensus)
library(tigris)
library(sf)
library(dplyr)
library(ggplot2)
library(stringr)

options(tigris_use_cache = TRUE)
theme_set(theme_minimal())

# ----------------------------------------------------------
# Step 3: Transmit Census API key (uncomment and paste yours)
# ----------------------------------------------------------

census_api_key("d18305720219d4680a63fc284f2f8c8f5bf8bc79")

#------------------------------------------------------------
# 1. Helper: get percent white + MOE for a geography/year
#------------------------------------------------------------

get_pct_white <- function(geo, year, state_fips = "47") {
  # B02001_001 = total population
  # B02001_002 = white alone
  dat <- get_acs(
    geography = geo,
    state     = state_fips,
    variables = c(total = "B02001_001", white = "B02001_002"),
    year      = year,
    survey    = "acs1",
    geometry  = FALSE,
    output    = "wide"
  )
  
  dat %>%
    mutate(
      pct_white      = 100 * whiteE / totalE,
      pct_white_moe  = moe_prop(whiteE, totalE, whiteM, totalM) * 100
    )
}

#------------------------------------------------------------
# 2. 2021 TN U.S. House districts map (percent white)
#------------------------------------------------------------

# Data
tn_cd_2021 <- get_pct_white("congressional district", 2021)

# Geometry for 2021 districts
cd_shapes_2021 <- congressional_districts(year = 2021, cb = TRUE) %>%
  st_transform(4326) %>%
  filter(STATEFP == "47") %>%
  left_join(tn_cd_2021, by = c("GEOID" = "GEOID"))

# County borders for overlay
tn_counties <- counties(state = "47", year = 2021, cb = TRUE) %>%
  st_transform(4326)

# Breaks to match your legend (adjust if needed)
breaks_2021 <- c(20.50, 67.36, 78.26, 79.78, 85.56, 89.70)

cd_shapes_2021 <- cd_shapes_2021 %>%
  mutate(
    pct_cat = cut(
      pct_white,
      breaks = breaks_2021,
      include.lowest = TRUE,
      right = TRUE
    )
  )

gg_2021 <- ggplot() +
  geom_sf(data = cd_shapes_2021, aes(fill = pct_cat), color = "grey40", size = 0.3) +
  geom_sf(data = tn_counties, fill = NA, color = "white", size = 0.2) +
  scale_fill_brewer(
    name   = "Estimated percent white",
    palette = "Blues",
    drop   = FALSE
  ) +
  labs(
    title = "U.S. House districts in Tennessee, 2021, shaded by estimated percent white",
    subtitle = "ACS 1-year estimates; county borders shown"
  ) +
  theme(
    legend.position = "right",
    panel.grid.major = element_blank()
  )

gg_2021

#------------------------------------------------------------
# 3. 2022 TN U.S. House districts map (percent white)
#------------------------------------------------------------

tn_cd_2022 <- get_pct_white("congressional district", 2022)

cd_shapes_2022 <- congressional_districts(year = 2022, cb = TRUE) %>%
  st_transform(4326) %>%
  filter(STATEFP == "47") %>%
  left_join(tn_cd_2022, by = c("GEOID" = "GEOID"))

tn_counties_2022 <- counties(state = "47", year = 2022, cb = TRUE) %>%
  st_transform(4326)

breaks_2022 <- c(26.30, 72.28, 74.52, 79.40, 81.72, 89.70)

cd_shapes_2022 <- cd_shapes_2022 %>%
  mutate(
    pct_cat = cut(
      pct_white,
      breaks = breaks_2022,
      include.lowest = TRUE,
      right = TRUE
    )
  )

gg_2022 <- ggplot() +
  geom_sf(data = cd_shapes_2022, aes(fill = pct_cat), color = "grey40", size = 0.3) +
  geom_sf(data = tn_counties_2022, fill = NA, color = "white", size = 0.2) +
  scale_fill_brewer(
    name   = "Estimated percent white",
    palette = "Blues",
    drop   = FALSE
  ) +
  labs(
    title = "U.S. House districts in Tennessee, 2022, shaded by estimated percent white",
    subtitle = "ACS 1-year estimates; county borders shown"
  ) +
  theme(
    legend.position = "right",
    panel.grid.major = element_blank()
  )

gg_2022

#------------------------------------------------------------
# 4. “Estimates by area” dotplot for TN congressional districts
#    (similar to your 116th Congress chart)
#------------------------------------------------------------

# Use 2018 (116th Congress) as an example; change year if needed
tn_cd_2018 <- get_pct_white("congressional district", 2018) %>%
  mutate(
    name_clean = str_replace(NAME, ", Tennessee", ""),
    name_clean = str_replace(name_clean, "Congressional District ", "CD ")
  )

gg_dot <- ggplot(tn_cd_2018,
                 aes(x = pct_white, y = reorder(name_clean, pct_white))) +
  geom_errorbarh(
    aes(
      xmin = pct_white - pct_white_moe,
      xmax = pct_white + pct_white_moe
    ),
    height = 0.2,
    color = "grey40"
  ) +
  geom_point(size = 2.5, color = "navy") +
  labs(
    x = "ACS estimate (percent white, with MOE)",
    y = NULL,
    title = "Estimates by area",
    subtitle = "Congressional districts in Tennessee (116th Congress)"
  ) +
  theme(
    panel.grid.major.y = element_blank()
  )

gg_dot