get_lbk_data

rm(list=ls())

library(dplyr)
library(ggplot2)
library(lmtest)
library(multiwayvcov)
library(tmap)
library(tigris)

library(tidyverse)
library(maps)
library(sf)
library(data.table)

high <- read.csv("LBK_high.csv")
low <- read.csv("LBK_low.csv")

low$HouseAge <- low$HouseAge-1

high$concat <- paste0(high$Price, high$SquareFoot)
low$concat <- paste0(low$Price, low$SquareFoot)

length(unique(high$concat))
length(unique(low$concat))

drops <- c("Price","SquareFoot", "HouseAge", "Lot", "Garage", "Env", "Distance")
low <- low[ , !(names(low) %in% drops)]

data <- merge(high, low, by="concat")

#mapping

library(tidycensus)
options(tigris_use_cache = TRUE)

lbk_race <- get_decennial(
  geography = "block group",
  state = "TX",
  county = "Lubbock",
  variables = c(Black = "H006003"),
  summary_var = "P005001",
  year = 2010,
  geometry = TRUE) %>%
  mutate(percent = 100 * (value / summary_value))

lbk_black <- filter(lbk_race, 
                       variable == "Black")

current.mode <- tmap_mode("plot")
tmap_mode("view")

#empty map
tm_shape(lbk_black) + 
  tm_polygons()

#map w/ black %
tm_shape(lbk_black) + 
  tm_polygons(col = "percent")

#map w/ black % but more colorful
tm_shape(lbk_black) + 
  tm_polygons(col = "percent",
              style = "quantile",
              n = 5,
              palette = "Purples",
              title = "2020 US Census") + 
  tm_layout(title = "Percent Black\nby Census tract",
            frame = FALSE,
            legend.outside = TRUE)


#cropping area outside atlanta
df <- sf::st_as_sf(data, coords = c("Longitude","Latitude"))

#lbk_black = sf::st_crop(lbk_black, xmin=-101.90, xmax=-102.30, ymin=31.85, ymax=33.60)

#map w/ black % and tax/salesprice
tm_shape(lbk_black) + 
  tm_polygons(col = "percent",
              style = "quantile",
              n = 5,
              palette = "Blues",
              title = "Percent Black\nby Census tract") + 
  tm_shape(df) + tm_bubbles(size="Price", col="Price", palette = "Reds", style = "quantile", legend.size.show = FALSE,
                            border.lwd = 0.1, border.alpha = 0.1, border.col = "black",
                            title.col = "Tax/Sales Price", alpha = 0.5)

#acs data

#fast

vt1 <- get_acs(
  geography = "tract", #block group
  state = "TX",
  county = "Lubbock",
  variables = c(hh_number="DP02_0001", hh_medianincome = "B19013_001",
                hh_poverty="B17017_002", tract_pop="B01003_001", 
                hh_medianval="B25097_001", white="B02001_002", 
                black="B02001_003", asian="B02001_005", hispanic="B01001I_001",
                b_hs="B06009_002", 
                b_25_1="B15002_012", b_25_2="B15002_013", 
                b_25_3="B15002_029", b_25_4="B15002_030",
                associate="B06009_004", bachelor="B06009_005", 
                a_bachelor="B06009_006"),
  year = 2010,
  geometry = TRUE)

vt1_dt <- as.data.table(vt1)
data_wide_vt1 <- dcast(vt1_dt, GEOID ~ variable, fun.aggregate = mean,
                       value.var=c("estimate"))

data_wide_vt1$pov_rate <- (data_wide_vt1$hh_poverty/data_wide_vt1$hh_number)*100
data_wide_vt1$hh_medianincome000 <- data_wide_vt1$hh_medianincome/1000
data_wide_vt1$hh_medianval000 <- data_wide_vt1$hh_medianval/1000

data_wide_vt1$below_hs <- (data_wide_vt1$b_hs/data_wide_vt1$tract_pop)*100
data_wide_vt1$no_degree <- ((data_wide_vt1$b_25_1 + data_wide_vt1$b_25_2 +
                               data_wide_vt1$b_25_3 + data_wide_vt1$b_25_4)/data_wide_vt1$tract_pop)*100
data_wide_vt1$some_college <- (data_wide_vt1$associate/data_wide_vt1$tract_pop)*100
data_wide_vt1$grad <- ((data_wide_vt1$bachelor + data_wide_vt1$a_bachelor)/data_wide_vt1$tract_pop)*100

data_wide_vt1$white_pop <- (data_wide_vt1$white/data_wide_vt1$tract_pop)*100
data_wide_vt1$black_pop <- (data_wide_vt1$black/data_wide_vt1$tract_pop)*100
data_wide_vt1$asian_pop <- (data_wide_vt1$asian/data_wide_vt1$tract_pop)*100
data_wide_vt1$hispanic_pop <- (data_wide_vt1$hispanic/data_wide_vt1$tract_pop)*100

data_wide_vt1 <- select(data_wide_vt1, GEOID, hh_medianincome000, pov_rate,
                        hh_medianval000, below_hs, grad, white_pop, black_pop)

vt1 <- get_acs(
  geography = "tract", #block group
  state = "TX",
  county = "Lubbock",
  variables = c(hh_number="DP02_0001"),
  year = 2010,
  geometry = TRUE)

vt1$NAME <- NULL
vt1$variable <- NULL
vt1$estimate <- NULL
vt1$moe <- NULL

data_wide_vt1$row <- as.numeric(rownames(data_wide_vt1))

#boundary <- st_bbox(col_income)

#data <- subset(data, hlong >= boundary[1] & hlong <= boundary[3] & hlat >= boundary[2] & hlat <= boundary[4])

coord <- data.frame(lat = c(data$Latitude), long = c(data$Longitude)) %>%
  st_as_sf(coords = c("long", "lat"),
           crs = st_crs(vt1))

coord$row <- as.numeric(st_within(coord, vt1))

data <- cbind(data, coord)
data$geometry <- NULL

new_data <- merge(data, data_wide_vt1, by="row", all.x = TRUE)

#new_data <- merge(new_data, vt1, by="GEOID", all.x = TRUE)
#new_data <- st_as_sf(new_data)

#from decennial

lbk_race <- get_decennial(
  geography = "block group",
  state = "TX",
  county = "Lubbock",
  variables = c(Black = "H006003"),
  summary_var = "P005001",
  year = 2010,
  geometry = TRUE) %>%
  mutate(percent = 100 * (value / summary_value))

lbk_black <- filter(lbk_race, 
                       variable == "Black")

lbk_black$NAME <- NULL
lbk_black$variable <- NULL
lbk_black$value <- NULL
lbk_black$summary_value <- NULL

lbk_black$row <- as.numeric(rownames(lbk_black))

#boundary <- st_bbox(col_income)

#data <- subset(data, hlong >= boundary[1] & hlong <= boundary[3] & hlat >= boundary[2] & hlat <= boundary[4])

coord <- data.frame(lat = c(data$Latitude), long = c(data$Longitude)) %>%
  st_as_sf(coords = c("long", "lat"),
           crs = st_crs(lbk_black))

coord$row <- as.numeric(st_within(coord, lbk_black))

data <- cbind(data, coord)
data$geometry <- NULL

new_data <- merge(data, lbk_black, by="row", all.x = TRUE)

model1 <- lm(Price ~ SquareFoot + HouseAge + Age + Gender + HHSize + Educ + percent, data=new_data)
summary(model1)