Creating Howard County Precinct Cartograms Based on 2014 Registered Voters, Part 3

Introduction

This is part 3 of a three-part series; see also part 1 and part 2.

In doing analyses of precinct-level voting data from the Howard County 2014 general election (using data from the Maryland State Board of Elections and the Howard County Board of Elections), one problem is that maps drawn using that data are inherently misleading: Precincts in western Howard County are much larger than precincts in Columbia and eastern Howard County, so they visually dominate the maps.

One way to reduce this effect is to create cartograms, maps in which the visual sizes of the geographic subdivisions are based not on their actual geographic area but on some other variable associated with them. In particular, for political maps of Howard County I would like to display precincts sized according to the number of registered voters in each precinct.

Creating a voter-based cartogram of Howard County precincts requires distorting the precinct boundaries so as to change the area of each precinct while still preserving the overall shape of the precinct as much as possible, and also preserving its relationship to its neighboring precincts. This requires some relatively sophisticated mathematics, and unfortunately there is no existing R package that can do it well. Instead I use scapetoad, a Java-based application available for Microsoft Windows, Mac OS X, and Linux.

In part 1 of this series I described the overall process of preparing Howard County map data for use with the scapetoad application. In part 2 I described the process of running the scapetoad application to produce the desired cartogram(s) and checking the results. In this document I present examples of the completed cartograms, including overlaying Howard County council district, Maryland state legislative district, and US congressional district boundaries on the base precinct-level cartogram. I present an unmodified precinct map as a comparison.

Loading libraries

I use the R statistical package run from the RStudio development environment, along with the dplyr package for data manipulation, the ggplot2 package for plotting, and the sp and rgdal packages to read mapping data.

library("dplyr", warn.conflicts = FALSE)
library("ggplot2")
library("sp")
library("rgdal")

## rgdal: version: 0.9-1, (SVN revision 518)
## Geospatial Data Abstraction Library extensions to R successfully loaded
## Loaded GDAL runtime: GDAL 1.11.1, released 2014/09/24
## Path to GDAL shared files: /Library/Frameworks/GDAL.framework/Versions/1.11/Resources/gdal
## Loaded PROJ.4 runtime: Rel. 4.8.0, 6 March 2012, [PJ_VERSION: 480]
## Path to PROJ.4 shared files: (autodetected)

The rgdal package also requires installing the GDAL mapping library on the underlying operating system.

Overall approach

In this document I perform the following steps:

• Download and load all the necessary cartogram shapefiles, along with an unmodified precinct map for comparison.
• Transform the cartogram shapefiles and the unmodified precinct map for use with the ggplot() function.
• Determine where to place labels for individual map areas (precincts and districts).
• Plot the unmodified precinct map and the base precinct cartogram showing the precincts with labels.
• Plot each of the district cartograms in turn as a layer with labels on top of the precinct-level cartogram.

Downloading the cartogram shapefiles

I first download an unmodified precinct map as published on data.howardcountymd.gov.

download.file("https://data.howardcountymd.gov/geoserver/ows?service=WFS&version=1.0.0&request=GetFeature&typeName=general:Voting_Precincts&outputFormat=shape-zip",
              "Voting_Precincts.zip",
              method = "curl")

I then download all of the cartogram shapefiles as published to the datasets section of my hocodata GitHub repository.

url <- "https://raw.githubusercontent.com/frankhecker/hocodata/master/datasets/"
files <- c("Voting_Precincts_Cartogram.zip",
           "Council_Districts_Cartogram.zip",
           "Legislative_Districts_Cartogram.zip",
           "Congressional_Districts_Cartogram.zip")
for (f in files) {
    download.file(paste(url, f, sep = ""), f, method = "curl")
}

Since the boundary data is in .zip files I need to unzip the files to extract the actual shapefiles.

unzip("Voting_Precincts.zip", overwrite = TRUE)
for (f in files) {
    unzip(f, overwrite = TRUE)
}

Finally I read in the boundary shapefile data.

unmodified_map <- readOGR(dsn = ".",
                          layer = "Voting_PrecinctsPolygon")

## OGR data source with driver: ESRI Shapefile 
## Source: ".", layer: "Voting_PrecinctsPolygon"
## with 118 features and 9 fields
## Feature type: wkbPolygon with 2 dimensions

precinct_map <- readOGR(dsn = ".",
                        layer = "Voting_Precincts_CartogramPolygon")

## OGR data source with driver: ESRI Shapefile 
## Source: ".", layer: "Voting_Precincts_CartogramPolygon"
## with 118 features and 12 fields
## Feature type: wkbPolygon with 2 dimensions

council_map <- readOGR(dsn = ".",
                       layer = "Council_Districts_CartogramPolygon")

## OGR data source with driver: ESRI Shapefile 
## Source: ".", layer: "Council_Districts_CartogramPolygon"
## with 5 features and 5 fields
## Feature type: wkbPolygon with 2 dimensions

legis_map <- readOGR(dsn = ".",
                           layer = "Legislative_Districts_CartogramPolygon")

## OGR data source with driver: ESRI Shapefile 
## Source: ".", layer: "Legislative_Districts_CartogramPolygon"
## with 4 features and 5 fields
## Feature type: wkbPolygon with 2 dimensions

cong_map <- readOGR(dsn = ".",
                             layer = "Congressional_Districts_CartogramPolygon")

## OGR data source with driver: ESRI Shapefile 
## Source: ".", layer: "Congressional_Districts_CartogramPolygon"
## with 3 features and 2 fields
## Feature type: wkbPolygon with 2 dimensions

Converting maps for use with ggplot()

Before I plot the maps I have to convert the spatial data to normal data frames usable with the ggplot() function.

# Define a function to convert spatial data into a ggplot-compatible data frame.
convert_map <- function(map) {
    temp_map <- map  # Avoid modifying original map in the next step
    temp_map@data$id <- rownames(temp_map@data)
    points <- fortify(temp_map, region = "id")
    df <- full_join(points, temp_map@data, by = "id")
    return(df)
}
# Convert the spatial data.
unmodified_df <- convert_map(unmodified_map)
precinct_df <- convert_map(precinct_map)
council_df <- convert_map(council_map)
legis_df <- convert_map(legis_map)
cong_df <- convert_map(cong_map)

Labeling map areas

Since I want to label the various map areas I next compute the centroids of the precincts and districts in order to position the labels on the map.

# Unmodified precincts
unmodified_centers = coordinates(unmodified_map)
unmodified_centers_df <- as.data.frame(unmodified_centers)
names(unmodified_centers_df) <- c("long", "lat")
unmodified_centers_df$Precinct = as.character(unmodified_map@data$PRECINCT20)
# Precincts
precinct_centers = coordinates(precinct_map)
precinct_centers_df <- as.data.frame(precinct_centers)
names(precinct_centers_df) <- c("long", "lat")
precinct_centers_df$Precinct = as.character(precinct_map@data$PRECINCT20)
# Howard County council districts
council_centers = coordinates(council_map)
council_centers_df <- as.data.frame(council_centers)
names(council_centers_df) <- c("long", "lat")
council_centers_df$District = as.character(council_map@data$DISTRICT20)
# Maryland state legislative districts
legis_centers = coordinates(legis_map)
legis_centers_df <- as.data.frame(legis_centers)
names(legis_centers_df) <- c("long", "lat")
legis_centers_df$District = as.character(legis_map@data$DISTRICT)
# US congressional districts
cong_centers = coordinates(cong_map)
cong_centers_df <- as.data.frame(cong_centers)
names(cong_centers_df) <- c("long", "lat")
cong_centers_df$District = as.character(cong_map@data$Congress_D)

The cartogram’s shapes for Howard County council districts 1 and 5 are so distorted that the centroids are almost outside the district boundaries. I therefore tweak the locations for the labels for those districts so that they’ll appear in more suitable locations, moving the label for district 1 a bit north and the label for district 5 a bit west.

# District 1
council_centers_df$lat[1] <- council_centers_df$lat[1] + 9000
# District 5
council_centers_df$long[5] <- council_centers_df$long[5] - 9000

I also tweak the locations for the labels for Maryland state legislative districts 12, 9A, and 9B so that they’ll appear in more suitable locations.

# District 12
legis_centers_df$long[1] <- legis_centers_df$long[1] + 17400
legis_centers_df$lat[1] <- legis_centers_df$lat[1] - 3000
# District 9A
legis_centers_df$lat[4] <- legis_centers_df$lat[4] + 8400
# District 9B
legis_centers_df$long[3] <- legis_centers_df$long[3] + 9000

Finally I tweak the location for the label for US congressional district 2 so that it will appear in a more suitable location.

# District 2
#cong_centers_df$long[1] <- cong_centers_df$long[1] + 17400
cong_centers_df$lat[1] <- cong_centers_df$lat[1] - 6000

Plotting the cartograms

I first plot the unmodified precinct map, as follows:

• Plot the unmodified map for the precincts, outlining their boundaries in gray.
• Plot the text labels for the map areas in black at an appropriate size.
• Force the x and y coordinate scales to be equal to avoid further distortion.
• Remove all extraneous plot elements, leaving only the map itself.

I don’t make any effort to tweak the locations of the labels for the precincts, so labels will appear out of place in cases where the precincts have odd shapes and the precinct centroids are close to or even outside the boundaries of the precincts themselves.

g <- ggplot() +
    geom_polygon(data = unmodified_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "grey") +
    geom_text(data = unmodified_centers_df,
              aes(x = long, y = lat, label = Precinct),
              size = 2,
              colour = "black",
              show_guide = FALSE) +
    coord_equal() +
    theme(axis.title = element_blank(),
          axis.text = element_blank(),
          axis.ticks = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
          panel.background = element_blank())
print(g)

I next plot the precinct cartogram, following the same procedure as for the unmodified precinct map. As with the unmodified map, because I don’t modify the locations of the text labels some show up in odd places; the problem is made worse for the cartogram because the precincts’ shapes are distorted.

g <- ggplot() +
    geom_polygon(data = precinct_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "grey") +
    geom_text(data = precinct_centers_df,
              aes(x = long, y = lat, label = Precinct),
              size = 2,
              colour = "black",
              show_guide = FALSE) +
    coord_equal() +
    theme(axis.title = element_blank(),
          axis.text = element_blank(),
          axis.ticks = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
          panel.background = element_blank())
print(g)

I next plot the Howard County council districts overlaid on the precinct cartogram. The procedure is the same as in the previous plots, except that I add a second layer to the plot showing the council boundaries and I label the council districts instead of the precincts.

g <- ggplot() +
    geom_polygon(data = precinct_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "grey") +
    geom_polygon(data = council_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "black") +
    geom_text(data = council_centers_df,
              aes(x = long, y = lat, label = District),
              size = 7,
              colour = "black",
              show_guide = FALSE) +
    coord_equal() +
    theme(axis.title = element_blank(),
          axis.text = element_blank(),
          axis.ticks = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
          panel.background = element_blank())
print(g)

I plot the Maryland state legislative districts overlaid on the precinct cartogram.

g <- ggplot() +
    geom_polygon(data = precinct_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "grey") +
    geom_polygon(data = legis_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "black") +
    geom_text(data = legis_centers_df,
              aes(x = long, y = lat, label = District),
              size = 7,
              colour = "black",
              show_guide = FALSE) +
    coord_equal() +
    theme(axis.title = element_blank(),
          axis.text = element_blank(),
          axis.ticks = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
          panel.background = element_blank())
print(g)

Finally, I plot the US congressional districts overlaid on the precinct cartogram.

g <- ggplot() +
    geom_polygon(data = precinct_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "grey") +
    geom_polygon(data = cong_df,
                 aes(x = long, y = lat, group = group),
                 fill = NA,
                 colour = "black") +
    geom_text(data = cong_centers_df,
              aes(x = long, y = lat, label = District),
              size = 7,
              colour = "black",
              show_guide = FALSE) +
    coord_equal() +
    theme(axis.title = element_blank(),
          axis.text = element_blank(),
          axis.ticks = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
          panel.background = element_blank())
print(g)

Appendix

I used the following R environment in creating this document:

sessionInfo()

## R version 3.1.2 (2014-10-31)
## Platform: x86_64-apple-darwin13.4.0 (64-bit)
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] maptools_0.8-34 rgdal_0.9-1     sp_1.0-17       ggplot2_1.0.0  
## [5] dplyr_0.4.0     RCurl_1.95-4.3  bitops_1.0-6   
## 
## loaded via a namespace (and not attached):
##  [1] assertthat_0.1   colorspace_1.2-4 DBI_0.3.1        digest_0.6.4    
##  [5] evaluate_0.5.5   foreign_0.8-61   formatR_1.0      grid_3.1.2      
##  [9] gtable_0.1.2     htmltools_0.2.6  knitr_1.7        labeling_0.3    
## [13] lattice_0.20-29  magrittr_1.0.1   MASS_7.3-35      munsell_0.4.2   
## [17] parallel_3.1.2   plyr_1.8.1       proto_0.3-10     Rcpp_0.11.3     
## [21] reshape2_1.4     rgeos_0.3-8      rmarkdown_0.5.1  scales_0.2.4    
## [25] stringr_0.6.2    tools_3.1.2      yaml_2.1.13

The underlying GDAL library for the rgdal packages is from the KyngChaos GDAL Complete distribution version 1.11 for Mac OS X.

You can find the source code for this document and others at my HoCoData repository on GitHub. This document and its source code are available for unrestricted use, distribution and modification under the terms of the Creative Commons CC0 1.0 Universal (CC0 1.0) Public Domain Dedication. Stated more simply, you’re free to do whatever you’d like with it.