Everything you wanted to know about starting to make The Facebook Plot…
author: Alex Adler date: 2/6/14
The Facebook IPO Plot illustrates ggplot’s prowess
Connecting friends across cities by Facebook (credit: Paul Butler, Visualizing Friendships)
What’s so attractive about this plot?
- Great Circle Arcs
- It uses ggplot2
- Cool Coloring
- No mapping package
Great Circle Arcs
“A great circle is the largest circle that can be drawn on any given sphere…. For any two points on the surface of a sphere there is a unique great circle through the two points.”
This can be done with the geosphere library and in particular gcIntermediate()
library(ggmap)## Loading required package: ggplot2library(geosphere)## Loading required package: spnyc<-geocode('New York City')## Information from URL : http://maps.googleapis.com/maps/api/geocode/json?address=New+York+City&sensor=false
## Google Maps API Terms of Service : http://developers.google.com/maps/termssfo<-geocode('San Francisco')## Information from URL : http://maps.googleapis.com/maps/api/geocode/json?address=San+Francisco&sensor=false
## Google Maps API Terms of Service : http://developers.google.com/maps/termsinter1<-gcIntermediate(nyc,sfo,n=50,addStartEnd=TRUE)http://flowingdata.com/2011/05/11/how-to-map-connections-with-great-circles/
Great Circle Arcs
“Get intermediate points on a great circle inbetween the two points used to define the great circle.”
## lon lat
## lon -74.00594 40.71278
## -74.94872 40.85623
## -75.89549 40.99197
## -76.84604 41.11993
## -77.80020 41.24005
## -78.75776 41.35226
## -79.71850 41.45651
## -80.68222 41.55274
## -81.64869 41.64090
## -82.61769 41.72094
## -83.58898 41.79283
## -84.56233 41.85651
## -85.53749 41.91197
## -86.51422 41.95915
## -87.49227 41.99805
## -88.47139 42.02864
## -89.45132 42.05090
## -90.43182 42.06482
## -91.41261 42.07039
## -92.39345 42.06760
## -93.37407 42.05647
## -94.35422 42.03700
## -95.33364 42.00919
## -96.31207 41.97306
## -97.28927 41.92864
## -98.26499 41.87594
## -99.23897 41.81500
## -100.21097 41.74585
## -101.18076 41.66852
## -102.14810 41.58306
## -103.11276 41.48952
## -104.07452 41.38794
## -105.03316 41.27838
## -105.98848 41.16089
## -106.94026 41.03554
## -107.88831 40.90240
## -108.83244 40.76151
## -109.77247 40.61297
## -110.70822 40.45684
## -111.63953 40.29321
## -112.56624 40.12214
## -113.48820 39.94372
## -114.40527 39.75804
## -115.31730 39.56518
## -116.22419 39.36524
## -117.12581 39.15829
## -118.02205 38.94444
## -118.91282 38.72378
## -119.79802 38.49640
## -120.67757 38.26240
## -121.55139 38.02188
## lon -122.41942 37.77493Great Circle Lines: An Example
For now, let’s use a map because we’re only dealing with two points.
library(maps)
USMap<-map_data("usa")
xlim <- c(-171.738281, -56.601563)
ylim <- c(12.039321, 71.856229)
USMap<-map("world", col="#f2f2f2", fill=TRUE, bg="white", lwd=0.05, xlim=xlim, ylim=ylim)
lines(inter1)
But that’s not nearly cool enough…
We need more points: Flying the Friendly Skies
airports <- read.csv("http://datasets.flowingdata.com/tuts/maparcs/airports.csv", header=TRUE)
flights <- read.csv("http://datasets.flowingdata.com/tuts/maparcs/flights.csv", header=TRUE, as.is=TRUE)
major<-c("ATL","LAX","ORD","DFW","DEN","JFK","SFO")
trips<-flights[flights$airport1 %in% major,]Clean and arrange the data to get start and end points
##
## Attaching package: 'dplyr'
##
## The following object is masked from 'package:stats':
##
## filter
##
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union## Warning: joining factor and character vector, coercing into character
## vector## Warning: joining factor and character vector, coercing into character
## vectorNow, compile all the data into a data frame for plotting
routes<-NULL
for(i in 1:nrow(trips)){
gcirc<-as.data.frame(gcIntermediate(
c(trips$lon.origin[i],trips$lat.origin[i]),
c(trips$lon.dest[i],trips$lat.dest[i]),
n=50,breakAtDateLine=F,addStartEnd=TRUE))
gcirc$pathID<-i # allowing for group plotting in ggplot
gcirc$iata.origin<-trips$airport1[i]
routes<<-rbind(routes,gcirc)
}
head(routes,n=10)## lon lat pathID iata.origin
## 1 -97.03720 32.89595 1 DFW
## 2 -96.35066 32.68628 1 DFW
## 3 -95.66736 32.47288 1 DFW
## 4 -94.98733 32.25582 1 DFW
## 5 -94.31056 32.03513 1 DFW
## 6 -93.63708 31.81086 1 DFW
## 7 -92.96688 31.58306 1 DFW
## 8 -92.29997 31.35178 1 DFW
## 9 -91.63637 31.11707 1 DFW
## 10 -90.97606 30.87897 1 DFWA glimpse at the data frame to be plotted
head(routes,n=10)## lon lat pathID iata.origin
## 1 -97.03720 32.89595 1 DFW
## 2 -96.35066 32.68628 1 DFW
## 3 -95.66736 32.47288 1 DFW
## 4 -94.98733 32.25582 1 DFW
## 5 -94.31056 32.03513 1 DFW
## 6 -93.63708 31.81086 1 DFW
## 7 -92.96688 31.58306 1 DFW
## 8 -92.29997 31.35178 1 DFW
## 9 -91.63637 31.11707 1 DFW
## 10 -90.97606 30.87897 1 DFWggplot options to finish it all up
library(ggplot2)
library(grid)
p<-ggplot()+
xlim(-130,-60)+ylim(20,56)+
theme(plot.margin = unit(c(-1, -1, -1, -1), "cm"),
panel.background=element_blank(),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
plot.background=element_rect(fill="#3e4045"),
axis.line=element_blank(),
axis.text.x=element_blank(),
axis.text.y=element_blank(),
axis.ticks=element_blank(),
axis.title.x=element_blank(),
axis.title.y=element_blank(),legend.position="none"
)
airports<-geom_point(data=airports[airports$country=="USA",],aes(x=long,y=lat),col='#3d838a',size=0.7)
arcs<-geom_path(data=as.data.frame(routes), aes(x=lon, y=lat,group=pathID,color=iata.origin),alpha=0.2,size=0.5)Finally something a little nicer looking
## Warning: Removed 303 rows containing missing values (geom_point).## Warning: Removed 906 rows containing missing values (geom_path).
PDF export provides even more options
Paths to illustrator: Adding glows and changing visual aspects in a post-processing step
