An R Shiny application to visually explore DfT traffic statistics for the whole of GB from 2000-2023, including count points, regional and local authority area summaries, and flows on major road links.
The application is currently hosted on a virtual server provided by iSolutions at the University of Southampton. It can be accessed at https://trg-apps.soton.ac.uk/gbtraffic.
An R script is available to automatically download and process the DfT data. This script writes the processed data to a set of files in the dev/dft_csv folder. These files are later imported into the database (see below). Note that there may be changes to the DfT data in future years that require this script to be amended, for example the addition of new local authority areas.
This R Shiny application requires a backend PostgreSQL/PostGIS spatial database which contains imported data from DfT Road Traffic Statistics. Additional tables and views are derived from this data.
Steps required to create the database:
- Create a spatial database.
- Create a schema within the database called
traffic_gb. - Create a database user/role called
gbtraffic. - An SQL script is available to create the database tables, import the DfT data, create necessary views, materialized views, and indexes, and to set permissions for the
gbtrafficuser.
The application is provided as an R package. The required release of the R package should be uploaded to the Linux machine running Shiny Server as a tar.gz file.
Edit the file /etc/R/Renviron.site and add the following environment variables with the correct values for the PostgreSQL database, user and password:
GBTRAFFIC_PG_HOST = 'localhost'
GBTRAFFIC_PG_PORT = 5432
GBTRAFFIC_PG_DB = 'spatial'
GBTRAFFIC_PG_USER = 'gbtraffic'
GBTRAFFIC_PG_PASSWORD = 'blahblah'
STADIA_API_KEY = 'blahblah'
In the Linux terminal install the gbtraffic R package (uploaded earlier) as root (amend path as required):
sudo su - -c "R -e \"install.packages(c('/home/blah/file.tar.gz'), repos = NULL)\""Or, alternatively:
sudo -i
Rscript -e 'install.packages("/home/blah/file.tar.gz", repos = NULL)'A symlink needs to be created in the /srv/shiny-server/ folder to the folder containing the app.R file. This file is in the folder /inst/shinyApp in the package prior to installation. After installation the contents of /inst are moved to top level of the package folder:
cd /srv/shiny-server/
sudo ln -s /usr/local/lib/R/site-library/gbtraffic/shinyApp ./trafficR
The above can be implemented fairly easily with a local installation of R and PostgreSQL. For a hosted solution expertise in setting up a virtual machine with R, Shiny Server and PostgreSQL/PostGIS installed will be required. It is recommended to use the r2u tool to install the required CRAN packages. This is fast and automatically resolves all OS dependencies.
