Smoking Choropleths

1. This code gathers Google search result data for “tobacco” and organizes it by region. It also gathers Census data in order

tobacco <- trendy("tobacco", geo = "US", from = "2019-01-01", to = "2020-01-01")

tob_states <- tobacco %>%
  get_interest_region()

states_leaflet <- get_acs(geography = "state",       
                  variables = "B19013_001",         
                  geometry = TRUE)                   
                  shift_geo = T         

This code reads combines the Google tobacco data and CDC data on smoking.

smoking <- read_csv("Cig_smoking_percent.csv")

tob_data <- tob_states %>% 
  mutate(State = state2abbr(location)) %>% 
  inner_join(smoking)

This code maps the combined Google search and CDC data.

tob_colors <- colorNumeric(palette = "viridis", domain = tob_data$Cig_percent)

states_leaflet %>% 
  rename(location = NAME) %>% 
  inner_join(tob_data) %>% 
  leaflet() %>% 
  addTiles() %>%
  addPolygons(weight = 1,
              fillColor = ~tob_colors(Cig_percent), 
              label = ~paste0(location, ", Percent of Smokers = ", Cig_percent),
              highlight = highlightOptions(weight = 2)) %>% 
  setView(-95, 40, zoom = 4) %>% 
  addLegend(pal = tob_colors, values = ~Cig_percent)

Joining, by = "location"
sf layer has inconsistent datum (+proj=longlat +datum=NAD83 +no_defs).
Need '+proj=longlat +datum=WGS84'

2. This code gets the Google search result data for “cigarettes” and organizes it by region. It also gathers Census data so the google data is able to be mapped.

cigarettes <- trendy("cigarettes", geo = "US", from = "2019-01-01", to = "2020-01-01")

cig_states <- cigarettes %>%
  get_interest_region()

states_leaflet <- get_acs(geography = "state",       
                  variables = "B19013_001",         
                  geometry = TRUE)                   
                  shift_geo = T                    

This code maps the Google data from above by hits so you can see which states search “cigarettes” the most.

cig_colors <- colorNumeric(palette = "viridis", domain = cig_states$hits)

states_leaflet %>% 
  rename(location = NAME) %>% 
  inner_join(cig_states) %>% 
  leaflet() %>% 
  addTiles() %>%
  addPolygons(weight = 1,  fillColor = ~cig_colors(hits), label = ~paste0(location, ", Search volume = ", hits), highlight = highlightOptions(weight = 2)) %>% 
  setView(-95, 40, zoom = 4) %>% 
  addLegend(pal = cig_colors, values = ~hits)

Joining, by = "location"
sf layer has inconsistent datum (+proj=longlat +datum=NAD83 +no_defs).
Need '+proj=longlat +datum=WGS84'

3. This code joins the Google search data from above and joins it to CDC data on smoking.

smoking <- read_csv("Cig_smoking_percent.csv")

cig_data <- cig_states %>% 
  mutate(State = state2abbr(location)) %>% 
  inner_join(smoking)

This code creates a model between the CDC smoking data and the cigarette data.

cig_model <- lm(Cig_percent ~ hits, data = cig_data)

glance(cig_model)

tidy(cig_model)

This code creates a scatter plot of the CDC and Google search data so that you can see the relationship between them.

cig_data %>% 
  drop_na() %>% 
  plot_ly(x = ~hits, 
          y = ~Cig_percent,
          hoverinfo = "text", 
          text = ~paste("State: ", location, "<br>", "'Cigarettes' search rate: ", hits, "<br>", "Percent of smokers: ", Cig_percent)) %>% 
  add_markers(showlegend = F) %>% 
  add_lines(y = ~fitted(cig_model)) %>% 
  layout(title = "Relationship Between Google Searches for 'cigarettes' and Percent of Smokers, by State",
   xaxis = list(title = "Google search volume for 'cigarettes'"),
   yaxis = list(title = "State Percent of Smokers, per capita"))

4. This code gathers Google search results for “vaping”, organizes it by region, and gathers Census data so it can be mapped.

vaping <- trendy("vaping", geo = "US", from = "2019-01-01", to = "2020-01-01")

vape_states <- vaping %>%
  get_interest_region()

states_leaflet <- get_acs(geography = "state",       
                  variables = "B19013_001",         
                  geometry = TRUE)                   
                  shift_geo = T                    

This code maps the Google data from above by hits so you can see which states search “vaping” the most.

vape_colors <- colorNumeric(palette = "viridis", domain = vape_states$hits)

states_leaflet %>% 
  rename(location = NAME) %>% 
  inner_join(vape_states) %>% 
  leaflet() %>% 
  addTiles() %>%
  addPolygons(weight = 1,  fillColor = ~vape_colors(hits), label = ~paste0(location, ", Search volume = ", hits), highlight = highlightOptions(weight = 2)) %>% 
  setView(-95, 40, zoom = 4) %>% 
  addLegend(pal = vape_colors, values = ~hits)

Joining, by = "location"
sf layer has inconsistent datum (+proj=longlat +datum=NAD83 +no_defs).
Need '+proj=longlat +datum=WGS84'

5. Do the linear model and plotly graph between your second term and the CDC data.

smoking <- read_csv("Cig_smoking_percent.csv")

vape_data <- vape_states %>% 
  mutate(State = state2abbr(location)) %>% 
  inner_join(smoking)

This code creates a model between the CDC smoking data and the vaping data.

vape_model <- lm(Cig_percent ~ hits, data = vape_data)

glance(cig_model)

tidy(cig_model)

This code creates a scatter plot of the CDC and Google search data so that you can see the relationship between them.

vape_data %>% 
  drop_na() %>% 
  plot_ly(x = ~hits, 
          y = ~Cig_percent,
          hoverinfo = "text", 
          text = ~paste("State: ", location, "<br>", "Vaping' search rate: ", hits, "<br>", "Percent of smokers: ", Cig_percent)) %>% 
  add_markers(showlegend = F) %>% 
  add_lines(y = ~fitted(vape_model)) %>% 
  layout(title = "Relationship Between Google Searches for 'vaping' and Percent of Smokers, by State",
   xaxis = list(title = "Google search volume for 'vaping'"),
   yaxis = list(title = "State Percent of Smokers, per capita"))