Seattle Comprehensive Plan Map - Part II

Load data

We will load three different data sets to conduct our analysis: 1. State data from the U.S. Census 2. Municipal boundary data from Washington state 3. Seattle planning data on progress aggregated from multiple sources (includes geometry)

A listing of the available non-census data sets (WA / Seattle data sets) can be found here.

NOTE: We will want to ensure all of our data our in compatible data types and projections. This requires setting our options for tigris package at the outset AND aligning the crs data with sf::st_crs function

#Note: Add results="hide" to avoid printing download progress of tigris data...

# Option setting mentioned above to convert tigris to sf
options(tigris_class = "sf")

# URLS for data: 
muni_boundary <- "https://opendata.arcgis.com/datasets/d508083ebd7d444b9997639af845937d_1.geojson"
url_vlg_boundary <- "https://opendata.arcgis.com/datasets/9267e111804b4cc7b44bc73c673e6bda_0.geojson"


# Load data into environment

## 1. WA State Tract Level Geometry using Tigris (no link required)
d_wa_tract <- tracts(state = "WA", cb = T)%>%suppressMessages()

## 2. Load municipal boundaries for Seattle, note we set the crs = to our tigris tract dataset

d_sea_muni <- geojson_sf(muni_boundary)%>%
  filter(., CITYNAME == "Seattle")%>%
  st_transform(., st_crs(d_wa_tract), "+no_defs +datum=WGS84 +proj=longlat")


## 3. Finally we get our dataset from City of Seattle with the planning measures

d_vlg_designations <- geojson_sf(url_vlg_boundary)%>%
  st_transform(., st_crs(d_wa_tract), "+no_defs +datum=WGS84 +proj=longlat")

Clip Datasets and Explore

Next we will take the 3 different data sets and clip the geographies to explore what they look like and which one process is the most accurate.

1. View what the WA Tract data looks like from U.S. Census, we’ll see we only want a subset of the data
2. Simple Clip: Clip the tract data based on the boundaries set out in the municipal data set
3. Advanced Clip: Use sf::st_within to get Washington tracts within bounding box, then use true v. false (index) logic to subset the observations within the tract dataset.

# 2. Simple clip
d_clip1 <- d_wa_tract[d_sea_muni,]

# 3. Advanced Clip

d_inter <- st_within(d_wa_tract, d_sea_muni)
d_idx <- map_lgl(d_inter, function(x) {
  if (length(x) == 1) {
    return(TRUE)
  } else {
    return(FALSE)
  }
})

d_clip_adv2 <- d_wa_tract[d_idx,]


# Compare maps #1 WA State Tract; #2 Simple Clip; #3 Adv. Clip

ggplot(d_wa_tract)+
  geom_sf()

ggplot()+
  geom_sf(data = d_clip1)+
  geom_sf(data = d_vlg_designations, color = "darkred")

ggplot()+
  geom_sf(data = d_clip_adv2)+
  geom_sf(data = d_vlg_designations, color = "darkred")

We see that the advanced clipping process creates a more tailored dataset that is a more accurate and skinny outline of the city of Seattle. Moreover, we see the boundary projections from our city planning dataset are more compatible with the clipping. Therefore, we’ll proceed with that data set for our mapping.

Map Dataset

In order to visualize our data we’ll plot select features on our interactive map. We will utilize the leaflet package, which allows for us to add layers to the plot for the results we desire.

1. Set bounding box or boundaries of the plot area so our focus is only on Seattle
2. Set a color palette for a specific value within the data set
3. Create popup text that we will pass to our map to appear on selecting a colored geography
4. Create the plot with each of the layers

# Gather bounding box details from our advanced clip go data
box_max <- d_clip_adv2%>%
  st_bbox()%>%
  as.vector()

# Create a color palette based off of the type of Seattle designated geographies 

col_pal <- colorFactor("Blues", domain = d_vlg_designations$TYPE_NAME, reverse = T)

# Build the text of our pop-up when we select a given geography
# This utilizes htmltools

popup <- paste('<b>Name:</b>', d_vlg_designations$UV_NAME, '<br>',
               '<b>Type:</b>', d_vlg_designations$TYPE_NAME,'<br>',
               '<b>Equity Category:</b>', d_vlg_designations$EQUITY_CATEGORY, '<br>', 
               '<b>Hunits Built Since 2015: </b>', 
                prettyNum(d_vlg_designations$HU_BUILT_SINCE_15, big.mark=","), '<br>',
               '<b>Hunits Target Total 2015-35:</b>', 
                prettyNum(d_vlg_designations$HU_TARGET_15_35, big.mark=","), '<br>', 
               '<b>Hunits Remaining:</b>', 
                prettyNum(d_vlg_designations$REMAIN_HU_TARGET, big.mark=","), '<br>')

m <- leaflet(d_vlg_designations)%>%
  addProviderTiles('CartoDB.DarkMatter')%>%
  fitBounds(box_max[1], box_max[2], box_max[3], box_max[4])%>%
  addPolygons(data = d_clip_adv2, stroke = 0.7, weight = 0.5, color = "white", 
              opacity = 0.8,
              fillOpacity = 0.05)%>%
  addPolygons(data = d_vlg_designations, 
              color = ~col_pal(TYPE_NAME), 
              stroke = F, 
              fillOpacity = 0.85,
              popup = ~popup,
              highlightOptions = highlightOptions(
                stroke = 0.5,
                weight = 1,
                color = "white", 
                bringToFront = T)
  )%>%
  addLegend(map = ., pal = col_pal, 
            values = ~TYPE_NAME, position = "bottomright", 
            title = "Region Designation")

m