Welcome to AstronomicAL’s Documentation!

A plugin-based platform for interactive visualisation, inspection, labelling, and classification of scientific data.

AstronomicAL is a local, human-in-the-loop analysis platform for working with scientific datasets. It can be used to inspect records, combine contextual information, create reliable labels, and build model-assisted workflows such as active learning.

Although AstronomicAL has been heavily used for astronomy applications, the plugin-based system allows domain-specific tools to be created for other areas of scientific research. Astronomy remains an important part of the project, but astronomy-specific tools no longer need to be installed or used by every researcher.

Machine Learning	Hugging Face Importer
Spectroscopy and Photometry	Morphology

Statement of Need

Modern datasets are often too large to inspect manually, while their labels may be incomplete, noisy, or expensive to create. A model can only be as reliable as the data used to train and test it.

AstronomicAL brings the dataset, model, and expert into one workspace. Users can inspect difficult records alongside the information they need, record a decision, and immediately continue the workflow.

Note

Active learning remains a major use case, but it is now one workflow built from plugins rather than the fixed identity of the application.

What Has Changed?

Earlier versions of AstronomicAL used a fixed dashboard and a shared configuration object. The new application is assembled from plugins.

The platform provides shared infrastructure for:

datasets and semantic column mappings;
focused records and selection sets;
background jobs;
reusable artifacts and services;
events between independent components;
workspace layout and persistence;
plugin discovery and lifecycle management.

Plugins provide the visible panels, integrations, and research workflows.

Where Should I Start?

If you are new to AstronomicAL, start with Installation and then read Tour of the Interface.

For some quick-start example dataset and layout, start here: Example Datasets and Workspaces.

To understand the main ideas used throughout AstronomicAL, see Core Concepts.

For documentation on the plugins included with AstronomicAL, see Built-in Plugin Reference.

Plugin authors should begin with Building Your First AstronomicAL Plugin and then read Plugin API.

Contributors working on the shared application infrastructure should see Platform Architecture.

Documentation

Project Information