How to plot animated maps with gganimate

Intro

Leipzig is the largest city in the federal state of Saxony, Germany. Since about 2009, Leipzig is one of the fastest growing German cities. Between 2006 and 2016, the population has grown from about 506.000 to 580.000 inhabitants.The largest part of the population growth can be attributed to inward migration.

In this blog post, I show how to create an animated map of Leipzig visualizing the migration balances for each of Leipzig's 63 districts (“Ortsteile”).

The data

The data I will visualize in this blog post are provided by the local Statistical Bureau. They can be downloaded as .csv file from the following website. The shapefile I use, is not publicly available and can be purchased for about 25 Euro from the Statistical Bureau.

R packages

The following code will install load and / or install the R packages required for this blog post.

if (!require("pacman")) install.packages("pacman")
pacman::p_load(readr, rgdal, ggplot2, reshape2)
pacman::p_load_current_gh("dgrtwo/gganimate")

While the packages readr and rgdal are required for importing the .csv and the shapefile, ggplot2 and gganimate are needed to plot and animate the data. The reshape2 package is required for transforming data frames from wide to long format.

After importing the csv file containing the migration data, we add a key variable (id) to identify each of Leipzig's 63 districts. Using the melt-function of the reshape2 package, we reformat the data frame from wide into long form.

# import csv file
df.migrat <- read_csv2("./Data_LE/Bevölkerungsbewegung_Wanderungen_Wanderungssaldo.csv",
                      col_types = cols(`2015` = col_number()),
                      n_max = 63,
                      locale = locale(encoding = "latin1"),
                      trim_ws = TRUE)
# add id variable
df.migrat$id <- c(1:63)
# drop variables not needed
df.migrat <- df.migrat[,c(1,18,2:17)]
# change name of first variable
colnames(df.migrat)[1] <- 'district'
# convert from wide to long
df.migrat <- reshape2::melt(df.migrat, id.vars=c("district","id"))
# change name of third variable
colnames(df.migrat)[3] <- 'year'

With the next code snippet, we import the shapefile using the readOGR-function of the rgdal package. The object we receive (sh.file) is of class SpatialPolygonsDataFrame. After adding a key variable, we transform this object into an ordinary data frame using the fortify-function of the ggplot2 package. Furthermore, we join both data frames by the key variable id. Before we plot the data, we create a new variable (co) categorizing the migration balance variable (value).

# import shapefile
sh.file <- readOGR("ot.shp", layer="ot")
# add id variable
sh.file$id <- c(1:63)
# Create a data frame
sh.file <- fortify(sh.file, region = 'id')
# Merge data frames by id variable
mydata <- merge(sh.file, df.migrat, by='id')
# create a categorized variable
mydata$co <- ifelse(mydata$value >= 1000, 4,
                   ifelse(mydata$value >= 0, 3,
                                ifelse(mydata$value >= -1000, 2, 1)))
# define levels and labels
mydata$co  <- factor(mydata$co, levels = c(4:1),
                     labels = c('> 1000', '0-1000', '-1000-0', '< -1000'))

Moreover, we define a theme suitable for plotting maps with ggpot2. The following theme is a customized version of a theme I found on Stack Overflow.

theme_opts <- list(theme(panel.grid.minor = element_blank(),
                         panel.grid.major = element_blank(),
                         panel.background = element_blank(),
                         plot.background = element_blank(),
                         panel.border = element_blank(),
                         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="right",
                         plot.title = element_text(size=16)))

Plotting the data

Finally, we can plot the data. The first plot (static) visualizes the migration data for the whole period of time (2000–2015).

ggplot(mydata, aes(x = long, y = lat, group = group, fill=value)) +
  theme_opts + 
  coord_equal() +
  geom_polygon(color = 'grey') +
  labs(title = "Migration balance between 2000 and 2015",
       subtitle = "Static Map",
       caption = 'Data source: Stadt Leipzig, Amt für Statistik und Wahlen, 2015') +
  scale_fill_distiller(name='Migration\nbalance', direction = 1, palette='RdYlGn')

plot of chunk map

For creating an animated plot, we add the frame argument stemming from the gganimate package. With interval = 5 we define the time (in seconds) between the “slides”.

p <- ggplot(mydata, aes(x = long, y = lat, group = group, fill=value, frame=year)) +
  theme_opts + 
  coord_equal() +
  geom_polygon(color = 'grey') +
  labs(title = "Migration balance in year: ",
       subtitle = "Animated Map",
       caption = 'Data source: Stadt Leipzig, Amt für Statistik und Wahlen, 2015') +
  scale_fill_distiller(name='Migration\nbalance', direction = 1, palette='RdYlGn')

gg_animate(p, interval = 5) 

Advertisements

About norbert

I am post doc at the Department of Medical Psychology and Sociology, Leipzig University (GER), with degrees in sociology (MA) and public health (MPH).
This entry was posted in Visualizing Data and tagged , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s