Automatic Repair of Spatial Polygons

Overview

The prepr R R package is used to repair broken spatial polygons. To achieve this, the package uses the constrained triangulation approach implemented in the prepair C++ library (Ledoux et al. 2014). Specifically, the st_prepair() function is used to repair polygons. For more information, please see the package vignette.

Installation

The latest version can be installed from the online source code repository using the following R code.

if (!require(remotes)) install.packages("remotes")
remotes::install_github("prioritizr/prepr")

Windows

The Rtools software needs to be installed to install the prepr R package from source. This software provides system requirements from rwinlib.

Ubuntu

The gmp, mpfr, and gdal libraries need to be installed to install the prepr R package. For recent versions of Ubuntu (18.04 and later), these libraries are available through official repositories. They can be installed using the following system commands:

sudo add-apt-repository ppa:ubuntugis/ubuntugis-unstable
sudo apt-get -y update
sudo apt-get install \
  libudunits2-dev libgdal-dev libgeos-dev libproj-dev \
  libsqlite0-dev libgmp3-dev libmpfr-dev

Linux

For Unix-alikes, gmp (>= 4.2.3), mpfr (>= 3.0.0), and gdal (>= 2.0.1) are required.

macOS

The pkg-config, gmp, mpfr, and gdal libraries are required. The easiest way to install these libraries is using HomeBrew. After installing HomeBrew, these libraries can be installed using the following commands in the system terminal:

brew install pkg-config
brew install gmp
brew install mpfr
brew install gdal

Citation

We recommend citing the methodology that underpins the prepr R package. To cite this work, please use:

Ledoux H, Arroyo Ohori K, and Meijers M (2014) A triangulation-based approach to automatically repair GIS polygons. Computers & Geosciences 66:121–131.

Usage

Here we provide a short example showing how the prepr package can be used to repair broken polygon geometries. To start with, we will load the package. We will also load the sf package for working with spatial data.

# load packages
library(sf)    # package for working with spatial data
library(prepr) # package for repairing spatial geometries

Next, we will create a broken spatial polygon.

# create broken polygon
x <- st_as_sfc("POLYGON((0 0, 0 10, 10 0, 10 10, 0 0))")

# verify that polygon is broken, and display reason why it's broken
st_is_valid(x, reason = TRUE)

## [1] "Self-intersection[5 5]"

Now we will use the st_prepair() function to repair the broken polygon.

# repair the polygon
y <- st_prepair(x)

# verify that repaired polygon is NOT broken
st_is_valid(y)

## [1] TRUE

Finally, we can compare the two polygons. To achieve this, let’s plot both the two polygons (shown in gray) and also plot the vertices for the two polygons (shown in red).

# visualize the polygons
par(mfrow = c(1, 2), mar = c(0, 0, 2.5, 0))
plot(x, main = "original", col = "#999999")
points(st_coordinates(st_cast(x, "POINT")), col = "#e41a1c", pch = 16, cex = 2)
plot(y, main = "repaired", col = "#999999")
points(st_coordinates(st_cast(y, "POINT")), col = "#e41a1c", pch = 16, cex = 2)

Although we can see that both of the polygons (shown in gray) appear visually identical, the repaired polygon (y) has an additional vertex (red point) in the center of the ‘bowtie’. This extra vertex is important to maintain the validity fo the spatial polygon. For more information and additional examples, see the package vignette.

Getting help

If you have any questions about using the prepr R package or suggestions for improving it, please file an issue at the package’s online code repository.