his.R

pop <- read.csv("../downloader/data/pop_muns.csv") %>% filter(date == '2010-06-01 00:00:00')

mex <- injury.intent %>%
  filter(year_reg == 2016 & intent %in% c("Homicide", "Legal Intervention")) %>%
  group_by(state_occur_death, mun_occur_death) %>%
  rename(state_code = state_occur_death, mun_code = mun_occur_death) %>%
  summarise(count = n()) %>%
  left_join(read.csv("../downloader/data/pop_muns.csv") %>% filter(date == '2016-06-01 00:00:00'), 
            by = c("state_code", "mun_code")) %>%
  mutate(rate = count / population * 10^5) %>%
  mutate(country = 'México') %>%
  left_join(read.csv("../downloader/data/municipio_names.csv", fileEncoding="windows-1252"), 
            by = c("state_code", "mun_code")) %>%
  left_join(read.csv("../downloader/data/state_codes.csv", fileEncoding="windows-1252"), 
            by = c("state_code")) %>%
  mutate(County = stringi::stri_trans_totitle((str_c(MUNICIPIO, ", ", state)))) %>%
  ungroup() %>%
  dplyr::select(population, rate, County, country) %>%
  filter(population > 10^5)

#https://wonder.cdc.gov/controller/saved/D76/D27F575
us <- read.csv("~/Downloads/Hispanics-County.txt", sep = "\t", stringsAsFactors = FALSE) %>%
  rename(population = Population, rate = Crude.Rate ) %>%
  mutate(country = "USA") %>%
  dplyr::select(population, rate, County, country) %>%
  mutate(rate = as.numeric(rate)) %>%
  na.omit() %>%
  filter(population > 10^5)

ggplot(bind_rows(us, mex) %>% arrange(-rate) %>% head(150),
       aes(reorder(County, rate), rate, fill = country)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  xlab("County") +
  theme_bw() +
  ggtitle('2016 Homicides and Legal Intervention Deaths by County\nin the US (only Hispanics) and Mexico (Everyone)',
          subtitle = 'Only includes counties with more than 100,000 persons or 100,000 White Hispanics\nData Sources: INEGI and CDC Wonder')
ggsave("test.png", dpi = 100, width = 10, height = 18)