gama_tool

in test and development state

An agent-based simulation of GAMA platform deployed as a Galaxy tool on Siwaa platform

Begin

Galaxy tool

A Galaxy tool based on the GAMA platform has been created (as draft). Still in test and development state, this tool has been run into a local test environment but has not been deployed/installed into the .

GAMA platform

Siwaa platform

This tool allows to drive and observe the experiment of a GAMA agent-based model.

+ more

This tool is based on some GAMA softwares : gama-server, a "Predator Prey" agent-based model of the GAMA platform, the python wrapper and example code of gama-client-python (https://github.com/gitrdm/gama-client-python).

This tool can be accessed :

• through Siwaa Web interface : https://siwaa.toulouse.inrae.fr more exactly in test and development state : https://127.0.0.1:9090 (local test environment)

• through Siwaa API, from various clients (by HTTP requests).

+ more

Using the Galaxy tool by the Siwaa web interface

Home page of the Siwaa web interface

The user can select the gama tool appearing among the tools list

gama tool page

The user can choose the input parameter values and then command to run tool

gama tool input parameters

gama tool running

gama tool output datas appear into the history, they are not ready for the moment

Outputs datas into the history

End of gama tool running

gama tool output datas are ready now. The user can click on them to get more information.

Outputs datas into the history

The user can view the output datas, download them...

Using the Galaxy tool by Siwaa API

An example of using gama tool by Siwaa API has been developped, as a Python Client. The Python Client program use_gama_tool.py : prepares the gama tool inputs parameters, commands to run the gama tool, and then gets the gama tool outputs when ready (downloads them as local files and in memory).

use_gama_tool.py

+ more

Illustration :When running use_gama_tool.py program, containing as gama tool inputs parameters values :nb_preys_init = 40 , nb_predators_init = 4 , nb_more_steps = 10

The resulting downloaded gama tool outputs files :Galaxy1-[run_server_txt].txt , Galaxy2-[run_client_txt].txt , Galaxy3-[results_txt].txt+ Screen display RUN.txt

Overview

Siwaa USERS

Container - gama-server - gama-client

Siwaa web interface

Web Browser Client

gama tool

+ more

gama_tool.xml

Siwaa API

some embedded code

Example of a Python Client

MY_sequential_example.py

MY_predatorPrey.gaml

use_gama_tool.py

• gama tool inputs :
• nb_preys_init and nb_predators_init (integers) : will be sent to MY_predatorPrey.gaml by MY_sequential_example.py.

• nb_more_steps (integer) : directly used by MY_sequential_example.py.

• gama tool outputs :
• results_txt file : simulator production while running (cf MY_predatorPrey.gaml).
• run_server_txt and run_client_txt files (gama-server and gama-client screen display captures) : production defined at gama tool command level (cf <command> tag of gama_tool.xml)

• Documentation and code about the gama tool : https://forgemia.inra.fr/nathalie.rousse/use/-/blob/gama/GAMA/galaxy-tools/README.md

• The gama tool contains and runs a Python program that interacts with GAMA Headless server, as a GAMA Python client, concerning a "Predator Prey" agent-based simulation to :
• drive the experiment (initialization, parameters settings, running, pausing, running some more steps, stop).
• observe the experiment (getting information values such as : cycle, nb_preys/nb_preys_init...).
• Documentation and code about the gama tool : https://forgemia.inra.fr/nathalie.rousse/use/-/blob/gama/GAMA/galaxy-tools/README.md

• GAMA's website : https://gama-platform.org

• Siwaa (Workbench for Agroecosystems SImulations and Analytics) is a web site dedicated to host and share reproducible virtual experiments in the field of modelling and simulation.
• Siwaa is an instance of the Galaxy (https://galaxyproject.org) web framework for scientific workflow. The cluster MESO@LR (https://meso-lr.umontpellier.fr) is mobilized for access to computing resources.
• Siwaa is supported by the RECORD Team (https://record.toulouse.inrae.fr).
• Siwaa web site : https://siwaa.toulouse.inrae.fr.
• Siwaa documentation site : https://siwaa.pages.mia.inra.fr/com.