afriadmin

african administrative boundaries

To make administrative boundary polygons for Africa easily accessible from R.

A part of the afrimapr project.

In early development, will change, contact Andy South with questions.

There are various sources of administrative boundaries available on the web, and existing R packages for accessing them.

This package will improve access to boundaries for Africa.

It started by looking at African admin boundaries at all levels from gadm.org.

Next step is to look at geoboundaries which seem to be more current and comprehensive - via rgeoboundaries.

As a first step I've saved all the full resolution gadm files (~ 88MB) in the package to save having to download them each time. Later we may choose to save just lower resolution versions. CRAN policy "As a general rule, neither data nor documentation should exceed 5MB ... Where a large amount of data is required (even after compression), consideration should be given to a separate data-only package which can be updated only rarely". This is also just a temporary solution, given that the GADM licence conditions indicate that they cannot be re-distributed without permission.

I'll develop interface functions that are intended to stay similar, what is happening behind may change.

I'm also looking into accessing boundaries from HDX (Humanitarian Data eXchange), which are more up-to-date but currently are trickier to reach for all countries in an automated way.

Install afriadmin

Install the development version from GitHub using devtools.

# install.packages("devtools") # if not already installed

devtools::install_github("afrimapr/afriadmin")

First Usage

library(afriadmin)
library(sf)

## single admin levels

# with interactive map
sfken <- afriadmin("kenya",level='max', plot='mapview')

# static map
sfeth <- afriadmin("ethiopia",level='max', plot='sf')

afriadmin("Angola",level=2)


## multiple admin levels - experimental

# single static map
afplotadmin("malawi")

# facetted using tmap
# returns a single object containing all admin levels
sfall <- alllevels("angola")


# todo move the readme to an Rmd to allow plots

Use cases

A. I'm a developer in country X, I want to know what options I have for obtaining boundary data at different administrative levels for my country.

1. Single country maps. We want to target in-country use. People in Africa making maps of Africa. Maybe there will be a greater utility for tools that make it easy to make maps of a single country rather than of multiple countries ?

2. low resolution boundaries just for display at country level

3. higher resolution boundaries to allow zooming in. (e.g. what detail is needed in a polygon boundary file to be able to create zonal statistics for a raster file of a given cell size ?)