Creating Route Maps in R with ggplot2

In this exercise, we outline how to gererate a “route map” - similar to one you might see published by an airline carrier. But instead of using flight routes, the map will show the “migration” of family members.

Initial Steps

We begin by loading the ggplot2 and ggrepel packages. The ggrepel package is loaded so that in case we wish to later add labels to the map, we can aesthetically do so.

library(ggplot2)
library(ggrepel)

We then load the data into r using the read.csv function.

migrate <- read.csv("addresses2.csv", stringsAsFactors = FALSE)

In order to construct the map, we will need longitude and latitude coordinates for all points which we wish to chart. The coordinates are already included in the migrate dataframe, so no further processessing is needed:

head(migrate,3)

##              Name genLevel                          From.Address     flat
## 1 Walter T Herget        2 335 Buena Vista Ave, Pekin, IL 61554  40.56595
## 2    James Herget        2 335 Buena Vista Ave, Pekin, IL 61554  40.56595
## 3 Robert L Herget        2 335 Buena Vista Ave, Pekin, IL 61554  40.56595
##        flon                                    To.Address     tlat
## 1 -89.64698         2255 W Berry Ave, Littleton, CO 80120 39.61802
## 2 -89.64698 2700 Town Center Blvd N, Sugar Land, TX 77479 29.59531
## 3 -89.64698             438 S Gladstone, Aurora, IL 60506 41.75254
##         tlon
## 1 -105.01363
## 2  -95.62172
## 3  -88.34685

Note that for each “route”, there must be 4 values: origin (1) longitude and (2) latitude coordinates, and destination (3)longitude and (4) latitude coordinates.

Generate Blank Map

The following command generates a blank map of the US that includes state borders:

#other map arguments include "world", "usa" and "county"
usMap <- borders("state", colour="grey", fill="white")

Generate US Migration Map Now we draw the map to include the migration routes.

ggplot() + usMap

allUSA <- ggplot() + usMap +
  geom_curve(data=migrate,
             aes(x=flon, y=flat, xend=tlon, yend=tlat),
             col="#00008b",
             size=.5,
             curvature=0.2) +
  geom_point(data=migrate,
             aes(x=flon, y=flat), 
             colour="blue",
             size=1.5) +
  geom_point(data=migrate,
             aes(x=tlon, y=tlat), 
             colour="blue") +
  theme(axis.line=element_blank(),
        axis.text.x=element_blank(),
        axis.text.y=element_blank(),
        axis.title.x=element_blank(),
        axis.title.y=element_blank(),
        axis.ticks=element_blank(),
        plot.title=element_text(hjust=0.5, size=12)) +
  ggtitle("Migration Map: Three Generations of Adult Decendants of Walter F. Herget, Pekin, IL")

allUSA

Zooming In

Much of the famial migration is concentrated in the midwest, in particular, Illinois and Indiana. The following code “zooms-in” by setting x and y coordinate limits:

midwest <- ggplot() + usMap +
  geom_curve(data=migrate,
             aes(x=flon, y=flat, xend=tlon, yend=tlat),
             col="#00008b",
             size=.5,
             curvature=0.2) +
  geom_point(data=migrate,
             aes(x=flon, y=flat), 
             colour="blue",
             size=1.5) +
  geom_point(data=migrate,
             aes(x=tlon, y=tlat), 
             colour="blue") +
    theme(axis.line=element_blank(),
          axis.text.x=element_blank(),
          axis.text.y=element_blank(),
          axis.title.x=element_blank(),
          axis.title.y=element_blank(),
          axis.ticks=element_blank(),
          plot.title=element_text(hjust=0.5, size=12)) +
  coord_cartesian(ylim=c(36.5, 42.5), xlim=c(-92, -82)) +
  ggtitle("Migration Focus: Illinois, Indiana, Ohio")

midwest

Creating Route Maps in R with ggplot2

Robert B. Herget