The Global Open Simulator

A parallel agent-based platform for global social modelling.

Introduction

GOS is designed to enable high scale agent-based modeling in Python (and Jupyter). By using the multiprocessing library, GOS is able to run in parallel.

Installation

GOS is designed to run on Python 3.

The following are dependencies:

  • numpy
  • pandas

They can both be installed using pip:

pip3 install numpy pandas

Note: If you are running the code on a shared space, use the following:
    module load python/3.6.0       #enables usage of pip3
    pip3 install --user numpy pandas

Migration Example

Migration is the primary and most robust example of using the global open simulator (GOS).

GOS is designed to be run via commandline or more interactively Jupyter.

Command Line example

1. Clone into the git repository (https://github.com/paulrbrenner/GOS.git) using:

        git clone https://github.com/crcresearch/GOS

2. Install required packages (if you have not already done so). From within the GOS directory, use:

        pip3 install -r requirements.txt

3. Navigate to GOS/examples/migration

4. Execute migration.py with :

        python3 migration.py

For interactive Plotly Data Visualization in the command line

1. Install plotly with:

    pip3 install plotly

    If using a shared space use:
            pip3 install --user plotly

2. In VizDemo.py, comment out the basemap import. The imports should read:

    #import basemapviz as viz
    import plotlyviz as viz

3. Run VizDemo.py

Jupyter Example with Data Visualization

API