The AccessR tool set provides a simple set of NMTK tools for preparing an accessibility map and conducting simple isochrone accessibility analyses from a set of points.
This tool can be installed under the name "AccessR" in the NMTK's NMTK_apps subfolder, with a corresponding entry in local_settings.py to enable the application. There are system prerequisites detailed below in in the Installation section.
The tool works in three steps.
The first step takes an input polygon geographic file and rasterizes it to establish a study area with a base level of accessibility (the units are arbitrary). The tool can use either feature values or a constant value to provide the base level of impedance in each cell. Any value used for the baseline should be greater than zero; a zero value is treated as inaccessible. This step generates a basic raster file called an "accessibility map".
The second step accepts a spatial layer (areas, lines or points) and overlays those features onto an existing accessibility map. The accessibility map can come from the first step, or from a previous iteration of this second step, which can be conducted as many times as necessary to include different types of facilities or barriers. The accessibility value of each feature (or a constant value applied to all features) is used to update the provided accessibility map, and new accessibility map is generated; that number should be positive, non-zero value. Three "styles" of update are provided: "Barriers" are areas that will be considered completely untraversable or off-limits. Areas occupied by barriers will be considered outside the analysis area. "Obstacles" simply reduce the base accessibility by the accessibility value but still allow passage. "Facilities" increase the base accessibility.
The third step accepts an accessibility map from one of the previous steps, plus a file of points for which isochrones are computed using the accessibility map.
Installation
The AccessR accessibility analysis functions are designed as a tool for the Nonmotorized Toolkit (https://github.com/chander/nmtk). To use this code, you'll need a working installation of the Nonmotorized Toolkit first.
To add these tools to the Nonmotorized Toolkit, the entire set of repository code should be placed in a subfolder of the NMTK_apps folder of the NMTK called "AccessR", and the "AccessR" tool should be added to the list of available tools in local_settings.py (see the NMTK documentation for details). The tool will be enabled when the deploy.sh script is run (see below).
As the name suggests, the tool uses the R Statistical Environment to perform its computations. So you'll need to install a recent version of R from one of the CRAN mirrors (see http://cran.r-project.org). For Ubuntu, you can look here: https://cran.r-project.org/bin/linux/ubuntu/
On Ubuntu 14.04 ("trusty") you can do the following to connect a R repository (replacing "<my.favorite.cran.mirror>" with one of the mirrors you can find at https://cran.r-project.org/mirrors.html:
sudo echo "deb https://<my.favorite.cran.mirror>/bin/linux/ubuntu trusty/" > /etc/apt/sources.list.d/cran.list
The repository is secured by a specific signature key. You need to install the key by doing this:
sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys E084DAB9
Then you can run these commands to install R itself, plus the development environment for compiliing packages from source:
sudo apt-get update # to read the repository
sudo apt-get install r-base r-base-dev
You will need several R packages (specifically: sp, raster, rgdal, gdistance). You will also need Rserve to allow the NMTK environment to run R commands. Installation of Rserve is easy, see the Rserve documentation: https://www.rforge.net/Rserve/doc.html
It is most convenient to install these packages from within R using the function 'install.packages':
sudo R # run R as an administrator
install.packages(c("sp","raster","rgdal","gdistance","Rserve","RSclient"))
quit()
For accessing Rserve from the Python environment, the pyRserve package is used. Activate the NMTK python environment from the NMTK root folder, then install pyRserve using 'pip install -r requirements.txt' Note that pyRserve requires numpy, which the base NMTK installs. The complete sequence for installing pyRserve:
source ../../venv/bin/activate
pip install -r requirements.txt
The Rserve configuration file (/etc/Rserv.conf) needs to be adjusted with consistent permissions for the web server and to set up "out-of-band" (oob) messages to pass status updates from within R itself. The Rserv.conf file should contain these instructions:
# Configuration for RServe
control enable
# Change uid/gid to www-data
su server
gid 33
uid 33
# Enable Out-of-Band Messages
oob enable
eval require(Rserve)
There are two shell script files available to start and stop Rserve, and there will also soon be an init.d script that is automatically installed to stop and restart Rserve on system startup and shutdown.
bash startRserve.sh
bash stopRserve.sh
Finally, the included "deploy.sh" script will set everything going once you have all the pieces installed and have added` "AccessR" to the NMTK's list of tools in local_settings.py. It calls the root NMTK deployment script to reload the NMTK environment and it also starts Rserve. Like the root NMTK deployment script, it supports an option "-c" which will clear log files.
sudo bash deploy.sh # -c optional to clear log files