RADIAN: Synthetic Spatial Data in Python

Overview

RADIAN (RAnDom spatIal dAta geNerator) is a Python-based tool to generate synthetic geographic datasets for classroom and teaching environments. RADIAN utilizes a unique voronoi-based buffering system in order to replicate the radial nature of many real-world spatial datasets. Given a polygon in GeoJSON format and the relevent JSON parameter file, RADIAN can export synthetic datasets in GeoJSON and postgreSQL formats.

This readme will briefly outline the use of the tool and some examples of what it is capable of. For a more in-depth exploration of the underlying algorithms, please consult the wiki.

A short demonstration video is available here.

Running the Software

Running the tool will require installation of the necessary third-party packages including geopandas. The radian.py file must be run in the same directory as the parameters.json file. A detailed description of the running parameters and their expected values/limits is available here.

The once the parameters.json file is listed in the same directory as radian.py, the tool can be executed by simply runningthe radian.py file from the command line.

Sample Run

Let's say we wish to generate a synthetic dataset representing fast food restaurants in Glasgow. We will generate 400 points, each with a phone number, opening & closing time, area code, and finally the name of the restaurant. The parameters used for this generation can be viewed here.