DISCO — Deprojection Image Software for Circumstellar Objects

Note

DISCO (v1.2.3) is an open-source Python package for the analysis and physical characterisation of protoplanetary disk observations from ALMA FITS data. This software is currently in active development.

DISCO integrates a convolutional neural network (DiscoNet) for rapid geometric parameter prediction with a hybrid numerical optimisation strategy, enabling robust deprojection and azimuthally-averaged radial profile extraction from continuum FITS images.

Two complementary operational modes are provided: a command-line interface (disco-start) designed for batch processing and reproducible automated pipelines, and a web-based GUI (disco-start gui) for interactive, exploratory analysis.

Two Modes of Operation

Feature	CLI (disco-start)	GUI (disco-start gui)
DiscoNet (CNN) geometry	✅	❌
Interactive visualisation	❌	✅
Batch processing	✅	❌
Multi-band support	✅	❌
Beam homogenisation	✅	❌
SIMBAD query	❌	✅
Session save / restore	❌	✅
Ease of use	Moderate	High

The GUI is recommended for exploratory analysis and first-time users. The CLI is designed for reproducible, automated pipelines.

Citation and Acknowledgements

If you use DISCO in published work, please cite this repository and acknowledge Jorge Luis Guzmán-Lazo, who developed the software within the YEMS Millennium Nucleus under the supervision of Sebastián Pérez and Camilo González-Ruilova.

Contact: jorge.guzman.l@usach.cl

If you use DISCO in your research, please cite the associated Zenodo record:

Guzmán-Lazo, J. L. (2026). DISCO: Deprojection Image Software for Circumstellar Objects (v1.2.3). Zenodo. https://doi.org/10.5281/zenodo.19999239

@software{guzman_lazo_2026_19999240,
  author    = {Guzmán-Lazo, Jorge Luis},
  title     = {DISCO: Deprojection Image Software for Circumstellar Objects},
  year      = {2026},
  version   = {v1.2.3},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.19999239},
  url       = {https://doi.org/10.5281/zenodo.19999239}
}

DISCO was developed by Jorge Luis Guzmán-Lazo within the YEMS Millennium Nucleus under the supervision of Sebastián Pérez and Camilo González-Ruilova.

---

Getting Started

• Installation
  • Requirements
  • Installation from PyPI
  • Keeping DISCO up to date
  • Included Static Assets
• Quick Start
  • Interactive GUI Mode
  • Automated CLI Mode

User Guide

• Architecture
  • Repository Structure
  • Module Relationships
  • Operational Modes
  • Frontend Build Pipeline
• Pipeline Description
  • Group Discovery
  • Phase 1 — FITS Ingestion
  • Phase 2 — Geometry Optimisation
  • Phase 3 — Uncertainty Estimation
  • Phase 4 — Profile Extraction and Beam Homogenisation
  • Phase 5 — Output Serialisation
• Command-Line Interface
  • Synopsis
  • Positional Arguments
  • Optional Arguments
  • Usage Examples
  • CNN Model Loading
• Graphical User Interface
  • Step-by-Step Guide
  • Application Layout
  • File Loading
  • Toolbar Reference
  • Geometry Controls (CONTROLS panel)
  • Auto-Tune Geometry
  • Analysis Dashboard (ANALYSIS panel)
  • Export Functions
  • SIMBAD Query

API Reference

• disco.core.fits_utils
  • Beam Utilities
  • Centroid and Parameter Detection
  • Profile Extraction
  • WCS and Coordinate Utilities
  • Gaia Proper Motion
• disco.core.cnn_inference
  • Architecture
  • Output Encoding
  • Inference Function
• disco.core.optimization
  • Geometric Loss Function
  • Hybrid Geometry Optimiser
  • Uncertainty Estimation
  • Centre Refinement
• disco.server — HTTP API Reference
  • Global State
  • Pydantic Request Models
  • Endpoints
  • Server Lifecycle
• disco.cli
  • discover_groups()
  • run_pipeline()
  • main()
• disco.main
  • run()

Supplementary

• DiscoNet Training Pipeline
• File Input and Output
• Changelog

Indices and Tables

• Index

• Module Index

• Search Page