This is the home of Geo Open Accessibility Tool (GOAT)
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
app Update buildings_residential.sql Dec 18, 2018
data_samples Fixed Shapefiles Nov 24, 2018
.gitattributes Changed GitAttribute Nov 24, 2018
.gitignore Removed node_modules folder from node project and downgraded event-st… Nov 27, 2018
LICENSE Synced with Master Branch Nov 20, 2018
README.md Update README.md Dec 17, 2018
Vagrantfile Synced with Master Branch Nov 20, 2018

README.md

Geo Open Accessibility Tool (GOAT)

About

This is the experimental release of Geo Open Accessibility Tool (GOAT). GOAT is meant to be an open source, interactive, flexible and useful web-tool for accessibility planning. Its main data source is OpenStreetMap, however other data sources can be used as well. The backbone of the application is a PostgreSQL/PostGIS database.

Software and Libraries used

The following software, languages and libraries are currently used:

PostgresSQL 10

PostGIS 2.4

Pl/PgSQL

Geoserver 2.13.2

Python3

Shell

Openlayers 5

Parcel

NodeJS 8.12.0

Virtual Box 5.2.18

Vagrant 2.1.2

Docker

Setup GOAT

For the developed Vagrant and Docker are used. It is recommended to use Git for fetching the project and if you are on Windows Git Bash is also a nice alternative to the windows command prompt. In order to start and customize GOAT for your study area, you have to follow these steps:

1. Get a copy of GOAT

git clone https://github.com/EPajares/goat.git (run on your host)

or copy as zip-folder

2. Install Virtualbox

https://www.virtualbox.org/

3. Install Vagrant

It was only tested with the version mentioned above. Accordingly if you want to avoid unexpected issues, stick with that version.

https://www.vagrantup.com/

4. Prepare your data

Put all your data into the app/data folder!

If you want to disaggregate population data

You need a shapefile with administrative boundaries and a column with the number of inhabitants in this administrative unit, it works for any spatial resolution. The column has to be named “sum_pop” and has to be saved as integer. (name the file: study_area.shp) As the population data is used for population disaggregation, data on higher resolution will give you a more accurate disaggregation.

Optional: In the case you have custom landuse data you can place the data as shapefile (name the file: landuse.shp) into your data folder. The table has to include a column named "landuse". You can define in the table variable_container, which landuse category you want to exclude from the population disaggregation. For instance you can exclude graveyards or farmland and as consequences houses standing on these landuse categories are marked as not uninhabited.

If you already have population data on a high-resolution

Just place a shapefile called population.shp into your data folder. The geometry type has to be point and the number of residents have to be saved as integer into a column called "population".

5. Define your bounding box and the OSM-Downloadlink

Open the file secret.js in the app/config folder. Customize the DOWNLOAD_LINK and define your BOUNDING_BOX.

6. Setup GOAT

6.1. Start Vagrant

Open a command window and go into the project folder. Run the command:

vagrant up (run on your host)

For more Vagrant commands checkout:

https://gist.github.com/wpscholar/a49594e2e2b918f4d0c4

6.2. Install the necessary software

vagrant ssh (run on your host)

sudo bash app/installation/install_software.sh (run on your VM)

This script can take a while as it installs quite some software on your VM. If you want to check what is installed exactly you can view the install_software.sh script.

6.3. Fill your database

sudo bash app/installation/setup_goat.sh (run on your VM)

UPDATE data

In case you want to UPDATE all your data you can simply run the following from your project directory.

sudo /etc/init.d/postgresql restart (run on your VM)

sudo bash app/installation/setup_goat.sh (run on your VM)

!!Note this will drop your database and create a new database.!!

7. Start Geoserver

cd ~/app/geoserver (run on your VM)

sudo bash install_geoserver.sh (run on your VM)

Geoserver is running inside docker, which itself is inside your VM. You can check if Geoserver is up and running by typing http://localhost:8080/geoserver/index.html into your browser. The default password for your Geoserver instance is:

User: admin Password : geoserver

8. View GOAT in the browser

The front-end is bundled using parcel. At the moment it is recommended to run parcel on your host. For this you need to have NodeJS installed on your host:

https://nodejs.org/en/

In order to start the bundling go to the front-end directory and run:

npm install (run on your host)

npm start (run on your host)

If all steps were successful you will be able to use GOAT by typing the following into your browser:

http://localhost:8585

You can also run parcel on your VM, however you have to open port 9090 and port 8585. This can be done on your Vagrantfile.

9. Optional: Pre-calculate accessibility heat-map

GOAT allows you to use pre-calculated matrices that are used to visualize the dynamic heatmap. In order to start the pre-calculation you currently have to start the script manually with the following command:

python3 ~/app/data_preparation/Python/precalculate_grid_thematic.py (run on your VM)

Depending on the size of your study area the calculation can take a bit.

You can also set different grid_sizes in the script.

10. Connect to your database

You can connect to the PostgreSQL database with the following default credentials:

Change your credentials especially if you want to run GOAT in production

Host: localhost

User: goat

Database: goat

Password: earlmanigault

Port: 65432

Common Issues

If you use Windows as your host OS it might happen that you have issue when executing the shell scripts. Due to the different ways Unix-like systems and Windows are dealing with line endings. You will get an warning like "\r command not found". This especially happens if you open the shell scripts or the config.js with Wordpad or the text editor coming with Windows. You can avoid this by using an editor like Visual Studio Code. In the case you still face the issue you can convert the shell scripts and the secret.js with a tool like dos2unix.

You potentially have to run all the following commands:

sudo apt install dos2unix (run on your VM)

dos2unix ~/app/installation/install_software.sh (run on your VM)

dos2unix ~/app/installation/setup_goat.sh (run on your VM)

dos2unix ~/app/config/secrets.js (run on your VM)

dos2unix ~/app/geoserver/install_geoserver.sh (run on your VM)