The GeoData Retriever Agent is a plugin integration of an autonomous GIS agent for geospatial data retrieval and QGIS. This plugin simplifies the process of retrieving and loading geospatial data into the QGIS environment through natural language commands. It leverages LLM-Find, an autonomous framework designed to select and fetch geospatial data by generating and executing programs with self-debugging capabilities. For more details on the framework, please refer to our preprint manuscript: Huan Ning, Zhenlong Li, Temitope Akinboyewa, and M. Naser Lessani. 2024. “An Autonomous GIS Agent Framework for Geospatial Data Retrieval.” arXiv. https://doi.org/10.48550/arXiv.2407.21024.
QGIS Plugin page: https://plugins.qgis.org/plugins/AutonomousGIS_GeodataRetrieverAgent/
- In QGIS,
select Plugins>Manage and Install Plugins... - Find
AutonomousGIS_GeoDataRetrieverAgentand clickInstall Plugin
Alternatively,
- Download the master repository of the plugin from github
- Launch QGIS software and navigate to
Plugin > Manage and install Plugins.. > Install from ZIP - Click on
...to select the directory of the downloaded zip file andInstall plugin
The User Manual is available here
Find some usage example on the Data Request Examples page
This framework is designed to be extendable to accommodate various data sources across different application scenarios. Therefore users are allowed to add new data source to the agent. Please find more details on the steps needed to add new data source in the Add a new data source page
After the installation of the plugin, you need to install the "nest_asyncio" manually. Follow the steps below:
-
Open the QGIS Python Console by navigating to
Plugins>Python Consoleor pressCtrl+Alt+P -
In the console, run these two lines of code:
import pip pip.main(['install', 'nest-asyncio'])
-
If you encounter any issue you can also try installing "nest_asyncio" via the terminal by pointing to QGIS python interpreter:
/Applications/QGIS3.38.1.app/Contents/MacOS/bin/python3 -m pip install nest_asyncio
Some video demonstrations have been created on our YouTubeChannel. Feel free to subscribe to the our YouTube channel dor regular updates.
