Preprocessing

Filter only Montana schools, remove non-plausible mmr values, and select distinct cases

mt_meas<-measles %>% 
  filter(state == "Montana") %>% 
  filter(mmr > 0) %>% 
  distinct(index,state,name,county,enroll,mmr, .keep_all = T)

mt_meas %>% 
  head(8)
## # A tibble: 8 x 16
##   index state year  name  type  city  county district enroll   mmr overall xrel 
##   <dbl> <chr> <chr> <chr> <chr> <chr> <chr>  <lgl>     <dbl> <dbl>   <dbl> <lgl>
## 1     1 Mont… 2017… Acad… <NA>  <NA>  Silve… NA           64   100      -1 NA   
## 2     2 Mont… 2017… Anna… <NA>  <NA>  Glaci… NA          105   100      -1 NA   
## 3     3 Mont… 2017… Ashl… <NA>  <NA>  Roseb… NA           75   100      -1 NA   
## 4     4 Mont… 2017… Aton… <NA>  <NA>  Yello… NA           26   100      -1 NA   
## 5     5 Mont… 2017… Big … <NA>  <NA>  Casca… NA           38   100      -1 NA   
## 6     6 Mont… 2017… Blac… <NA>  <NA>  Glaci… NA           76   100      -1 NA   
## 7     7 Mont… 2017… Chil… <NA>  <NA>  Misso… NA          318   100      -1 NA   
## 8     8 Mont… 2017… Crow… <NA>  <NA>  Big H… NA          315   100      -1 NA   
## # … with 4 more variables: xmed <dbl>, xper <dbl>, lat <dbl>, lng <dbl>

Load in spatial datasets

# load in world and state sf
world <- ne_countries(scale = "medium", returnclass = "sf")
states <- st_as_sf(map("state", plot = FALSE, fill = TRUE))

# load in county data
load(url("https://github.com/mgimond/ES218/blob/gh-pages/Data/counties48.RData?raw=true"))

County-level processing:

  • Seperate ID into state and county names

  • Filter only Montana data

  • Change the case of “county” so that it matches measles data formatting

mtcounties<-cnty %>% 
  separate(col=ID, into = c("state","county"),sep = ",", remove = F) %>%    
  filter(state=="montana") %>%                                       
  mutate(county = snakecase::to_upper_camel_case(county))   

#mtcounties %>% 
 # head(8)

Make a basic map to explore distribution of schools/school sizes throughout the state

Summarize data by county

# summarize measles data by county
mt1<-mt_meas %>% 
  filter(name != "West Valley School Prek-6") %>%    # extreme outlier, something wrong with this datapoint
  group_by(county) %>% 
  mutate(xmt = (xrel + xmed + xper)) %>% 
  mutate(totalvacc = enroll*(mmr/100)) %>% 
  mutate(unvacc = enroll-totalvacc) %>% 
  summarise(mean_mmr=mean(mmr,na.rm = T), sum_exmt = sum(xrel,na.rm = T), total_unvacc = sum(unvacc,na.rm = T)) 

# join summarized measles data to county spatial data
mt2<-mt1 %>%
  mutate(round_unvacc = paste(round(mean_mmr,0),"%",sep = "")) %>% 
  left_join(mtcounties, by = "county") 

# create wesanderson palette 
pal <- wesanderson::wes_palette("Zissou1", 100, type = "continuous")

# create plot of mean mmr vaccination rate by county
mmrplot<-ggplot(data = world) + geom_sf() + geom_sf(data = states) + geom_sf(data=mtcounties) +
  geom_sf(data=mt2$geometry, aes(fill=mt2$mean_mmr)) +
  coord_sf(xlim = c(min(mt_meas$lng,na.rm = T)-1,max(mt_meas$lng,na.rm = T)+1),
           ylim = c(min(mt_meas$lat,na.rm = T)-1, max(mt_meas$lat,na.rm = T)+1),expand = FALSE) +
  LGCtheme + scale_x_continuous(breaks=seq(-116,-104,4)) + 
  scale_y_continuous(breaks=seq(43,49,3)) + 
  scale_fill_gradientn(colours = rev(pal)) +
  labs(fill="", x="",y="") + 
  theme(legend.position = c(0.1, 0.2), legend.background = element_blank(), legend.title = element_text(size=10))


# create plot of number unvaccinated by county
unvaccplot<-ggplot(data = world) + geom_sf() + geom_sf(data = states) + geom_sf(data=mtcounties) +
  geom_sf(data=mt2$geometry, aes(fill=(mt2$total_unvacc))) +
  coord_sf(xlim = c(min(mt_meas$lng,na.rm = T)-1,max(mt_meas$lng,na.rm = T)+1),
           ylim = c(min(mt_meas$lat,na.rm = T)-1, max(mt_meas$lat,na.rm = T)+1),expand = FALSE) +
  LGCtheme + scale_x_continuous(breaks=seq(-116,-104,4)) + 
  scale_y_continuous(breaks=seq(43,49,3)) + 
  scale_fill_gradientn(colours = pal) +
  labs(fill="", x="",y="") + 
  theme(legend.position = c(0.1, 0.2), legend.background = element_blank(), legend.title = element_text(size=10))

County choropleth maps

Mean MMR vaccination rate (percent vaccinated)

Counties with most unvaccinated students (top 10 counties)

County Mean MMR Vacc Total Exemptions Total Unvacc
Flathead 91.15812 0 721.0298
Gallatin 90.26943 0 428.0193
Ravalli 79.99909 0 405.9783
Hill 87.55222 0 325.9764
Missoula 95.30806 0 284.9955
Yellowstone 96.95689 0 278.9469
Lake 89.31176 0 255.0102
Toole 73.20500 0 249.9804
Cascade 96.88862 0 158.9801
Lincoln 82.93333 0 151.0146

Number of unvaccinated students by county (n students not vaccinated)