Download the script

Copying and pasting syntax from a browser can cause problems. To avoid this issue, please download a script with all of the needed code presented in this walkthrough.

Prior Things

vars_of_interest <- c(median_income = "B19013_001")

National Level Data

national_data <- get_acs(geography = "county",
                         variables = vars_of_interest,
                         year = 2016,
                         survey = "acs5",
                         shift_geo = TRUE,
                         geometry = TRUE) %>%
                         rename(`Median Income` = estimate)

Make sure to check it

datatable(national_data)  

Let’s now create a palette

wes_palette("Darjeeling2")

make it continuous

darj2 <- wes_palette("Darjeeling2",
                      max(national_data$`Median Income`, 
                      na.rm = TRUE), 
                      type = "continuous")

and plot it for the entire US

ggplot() + 
  geom_sf(data = national_data,
          mapping = aes(fill = `Median Income`),
          color = NA) + 
  coord_sf(datum = NA) +
  scale_fill_gradientn(colors = darj2)  + 
  theme_minimal() +
  theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5)) +
  ggtitle(stringr::str_wrap("2016 ACS Median Income for the United States", 35))

Regional Level Data

Here is a US Census PDF of what the regions are. Now Region 1 can be defined by

region1 = c("CT", "ME", "MA", "NH", "RI", "VT", "NJ", "NY", "PA")

We get acs data and geographies for our variable at the

• tract level

multistate_tract_data <- get_acs(geography = "tract",
                                 variables = vars_of_interest,
                                 state = region1,
                                 year = 2016,
                                 survey = "acs5",
                                 geometry = TRUE) %>%
                          rename(`Median Income` = estimate)

Make sure to check it

datatable(multistate_tract_data)  

• state level

multistate_state_data <- get_acs(geography = "state",
                                 variables = vars_of_interest,
                                 state = region1,
                                 year = 2016,
                                 survey = "acs5",
                                 geometry = TRUE) %>%
                          rename(`Median Income` = estimate)

Make sure to check it

datatable(multistate_state_data)  

Putting geometry = TRUE brings in the geographic information (aka shapefiles) so make sure that is on, though if you only need the data and not a map, save yourself some time and turn it off. Geographic data is relatively dense and can take some time depending on how much you need.

Optionally let’s create a palette,

wes_palette("Darjeeling1")

make it continuous

darj1 <- wes_palette("Darjeeling1",
                     max(multistate_tract_data$`Median Income`, 
                         na.rm = TRUE), 
                     type = "continuous")

and plot it. This is tract level data so you’ll have to wait a little bit

ggplot() + 
  geom_sf(data = multistate_tract_data,
                   mapping = aes(fill = `Median Income`),
                   color = NA) + 
  scale_fill_gradientn(colors = darj1) +
  theme_map() 

Let’s give it a go with state borders so its a bit more familiar and with a formatted title and some finagling of the legend

ggplot() + 
  geom_sf(data = multistate_tract_data,
                   mapping = aes(fill = `Median Income`),
                   color = NA) + 
  geom_sf(data = multistate_state_data,
                   color = "#f7f7f7",
                   fill = NA) + 
  scale_fill_gradientn(colors = darj1,
                       label = scales::dollar) +
  theme_map() +
  theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5),
        legend.position = c(0.8,0.1),
        legend.title = element_text(size = 12,
                                    face = "bold",
                                    color = "#3392a4",
                                    hjust = 0.5,
                                    margin = margin(b = 10)),
        legend.text = element_text(size = 10,
                                   face = "bold",
                                   color = "#2d8292")) +
  ggtitle("2016 ACS Median Income for Census Region 1")

State Level Data

We get acs data and geographies for our variable using

state_data <- get_acs(geography = "state",
                      variables = vars_of_interest,
                      year = 2016,
                      survey = "acs5",
                      shift_geo = TRUE,
                      geometry = TRUE) %>%
                      rename(`Median Income` = estimate)

Make sure to check it

datatable(state_data)  

Optionally let’s create a palette,

wes_palette("Zissou1")

make it continuous

ziss1 <- wes_palette("Zissou1",
                     max(state_data$`Median Income`, 
                         na.rm = TRUE), 
                     type = "continuous")

and plot it for the entire US

ggplot() + 
  geom_sf(data = state_data,
          mapping = aes(fill = `Median Income`),
          color = NA) + 
  scale_fill_gradientn(colors = ziss1) +
  theme_map() +
  theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5)) +
  ggtitle("2016 ACS Median Income for All US States")

or just the lower 48 states

lower_48 <- state_data %>%
            filter(GEOID < 60) %>%
            filter(!NAME == "Hawaii") %>%
            filter(!NAME == "Alaska")

ggplot() + 
 geom_sf(data = lower_48,
         mapping = aes(fill = `Median Income`),
         color = NA) + 
 scale_fill_gradientn(colors = ziss1) +
 theme_map() +
    coord_sf(crs = 5070, 
             datum = NA) +
 theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5)) +
  ggtitle(stringr::str_wrap("2016 ACS Median Income for the Contiguous US States", 35))

County Level Data

Tract level data is sometimes tricky to get, in that not all data sets with geographies have them. Sticking to the widespread surveys such as the decennial census and acs will always work, while others are up in the air.

kansas <- get_acs(state = "KS", 
                  geography = "county", 
                  variables = vars_of_interest,
                  geometry = TRUE) %>%
                  rename(`Median Income` = estimate)

datatable(kansas)

wes_palette("FantasticFox1")

make it continuous

ffox <- wes_palette("FantasticFox1",
                     max(kansas$`Median Income`, 
                         na.rm = TRUE), 
                     type = "continuous")

and plot it!

kansas %>%
  ggplot(aes(fill = `Median Income`)) + 
  geom_sf(color = NA) + 
  scale_fill_gradientn(colors = ffox,
                       label = scales::dollar) +
  theme_map() +
  theme(legend.position = "bottom",
        legend.direction = "horizontal",
        legend.title.align=0.5) +   
  guides(fill = guide_legend(title.position="top")) +
 theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5)) +
  ggtitle("2016 ACS Median Income for the Counties in Kansas")

Tract Level Data

Tract level data is sometimes tricky to get, in that not all data sets with geographies have them. Sticking to the widespread surveys such as the decennial census and acs will always work, while others are up in the air.

doug <- get_acs(state = "KS", 
                county = "Douglas", 
                geography = "tract", 
                variables = vars_of_interest,
                geometry = TRUE) %>%
                rename(`Median Income` = estimate)

datatable(doug)

wes_palette("Cavalcanti1")

make it continuous

caval <- wes_palette("Cavalcanti1",
                     max(doug$`Median Income`, 
                         na.rm = TRUE), 
                     type = "continuous")

and plot it!

doug %>%
  ggplot(aes(fill = `Median Income`)) + 
  geom_sf(color = NA) + 
  scale_fill_gradientn(colors = caval,
                       label = scales::dollar) +
  theme_map() +
  theme(legend.position = "bottom",
        legend.key.width = unit(3, "line")) +   
  guides(fill = guide_legend(title.position="top", 
                              title.hjust =0.5))  +
 theme(plot.title = element_text(size = 20,
                                  face = "bold",
                                  color = "#005b5b",
                                  hjust = 0.5)) +
  ggtitle(stringr::str_wrap("2016 ACS Median Income for the Census Tracts in Kansas", 35))