Spatial Data Analysis

R has a range of packages which provide functionality for handling spatial data and performing complex spatial analysis operations. Spatial data analysis is no longer reserved just for experts with expensive hardware and software. R has impressive geographic capabilities which anyone with the desire for exploration of geospatial data can use.

Before we get into geospatial mapping in R, let us introduce some basic ideas about spatial data.

Spatial data comprises of:

• coordinates of the object
• the coordinates relate to a physical location on Earth: coordinate reference system (CRS)

There are two types of CRS:

1. Geographical

• specific locations on the surface are defined by a three dimensional model of the Earth
• longitude and latitude

2. Projected

• A conversion of the three-dimensional grid onto a two-dimensional plane

According to the storing technique, spatial data is one of two types:

1. Raster data: composed of grid cells identified by row and column. The whole geographic area is divided into groups of individual cells, which represent an image (satellite images).

2. Vector data: composed of points, polylines, and polygons. For example hospitals, houses etc. are represented by points, while rivers, roads, etc., are represented by polylines. Villages and towns are represented by polygons.

When doing spatial analysis you will also deal with attribute data, which contains the relevant information about the spatial data. The analysis will be based on attribute data attached to geospatial data. Such data could be:

• nominal
• ordinal
• interval data
• ratio data

ie. attribute or measured type of data.

Static Maps: Shape File

R has impressive geographic capabilities and can handle different kinds of spatial data file formats including geojson and KML. We will illustrate the use of the ESRI Shapefile format, which stores non-topological geometry and attribute information for the spatial features in a data set. A shapefile consists minimally of a main file, an index file, and a dBASE table.

• .shp - lists shape and vertices
• .shx - has index with offsets
• .dbf - relationship file between geometry and attributes (data)

To import an ESRI shapefile into R correctly, all three files must be present in the directory and named the same (except for the file extension).

Let us start by reading a shape file of Serbian districts’ boundaries available from GADM maps and data.

First, download the files we will need from the following GitHub repository: https://github.com/TanjaKec/gadm36_SRB_shp. Make sure you save the file into your R-Project working directory before you ask R to execute the following

library(ggplot2)
library(sf)

#pointed to the shape file
eu <- "eu_countries_simplified.shp"

#used the st_read() function to import it
eu <- st_read(eu)

# plot the map
ggplot(eu) + 
  geom_sf()

This looks good 😃 Next, we want to add some data to it and do some ‘colouring’. We will incorporate information about `Cumulative number for 14 days of COVID-19 cases per 100000 on 14-Novov-202’ given in “covid_EU.csv” file.

library(dplyr)
covid_EU <- read.csv("covid_EU.sc")


# tidy up
# Make the country names correspond to ecdc data 
bound$country <- gsub(" ", "_", bound$country)
bound <- bound %>% 
  mutate(country = fct_recode(country,
                              "Czechia" = "Czech_Republic",
                              "North_Macedonia" = "Macedonia"))

# join data from the two data frames  
my_map <- left_join(bound, covid_EU,
                    by = c("country" = "country"))

# plot the choropleth
ggplot(my_map) +
  geom_sf(aes(fill = Fx14dper100K)) +
  scale_fill_distiller(direction = 1, name = "Fx14per100K") +
  scale_fill_viridis_c(option = "magma", begin = 0.1) +
  labs(title="Cumulative number for 14 days of COVID-19 cases per 100000", caption="Source: ecdc")

Interactive Maps: leaflet

Leaflet is one of the most popular open-source JavaScript libraries for interactive maps that was integrated into R, by a team of people from RStudio. We will illustrate how easy it is to plot a location map using the leaflet package, but you can learn more about this package from the Leaflet for R website.

The following map pinpoints the Libraries accross Leeds using data available from the U.K. open data portal data.gov.uk

## If you don't have leaflet installed yet, uncomment and run the line below
#install.packages("leaflet")
library(leaflet)
suppressPackageStartupMessages(library(dplyr))

#Leeds libraries available from #https://data.gov.uk/dataset/4853a980-0c6d-4d83-9daa-d0441b21f6e1/leeds-libraries
mydata <- read.csv('https://datamillnorth.org/download/leeds-libraries/cfc9f345-4916-4b74-aa9e-cecc78db9075/library%2520locations.csv', 
                   header=T, 
                   na.strings=c("","NA"),
                   stringsAsFactor = FALSE,
                   fileEncoding = "latin1") 
# define the corners of the map
minlat <- min(mydata$Latitude)
maxlat <- max(mydata$Latitude)
minlng <- min(mydata$Longitude)
maxlng <- max(mydata$Longitude)

# Initialize and assign libs as a leaflet object
libs <- mydata %>% 
  leaflet() %>% 
  # add tiles to the leaflet object
  addTiles() %>%
  # setting the corners/centre of the map
  fitBounds(~minlng, ~minlat, ~maxlng, ~maxlat) %>% 
  # add the markers with the popups 
  addCircles(lng = ~Longitude, lat = ~Latitude,
             radius = 150, weight = 10, color = "black",
             fillColor = "red", fillOpacity = 0.7,  
             popup = ~paste("<b>", Address.Line.1, "<br>", Postcode , "<br> tel:", Telephone, "<br> email:", Email, "<br>"))
#print the map
libs

Here is an interesting website which introduces all kinds of possibilities when creating maps with ‘leaflet’. You should check it out and explore further for yourself.

YOUR TURN 👇

Practise by doing the following set of exercises:

1. Read a shape file of Serbian districts’ boundaries available from GADM maps and data and plot population per square kilometre available form ‘Serbian_Pop.csv’ used in the exercise.

2. Create a map pointing out your favourite hang-out places using the leaflet package.

useful links:

R for the analysis of spatial or spatio-temporal data (blog)

Geocomputation with R: a book on geographic data analysis, visualisation and modelling.

GADM maps and data

Creating Interactive Spatial Maps in R Using Leaflet

Thematic Mapping with tmap

© 2020 Tatjana Kecojevic