DeepExtractor Documentation

Deep learning framework for reconstructing gravitational wave signals and glitches from LIGO detector data.

Built for LIGO’s O3 observing run (Hanford and Livingston detectors). Described in the paper: arXiv:2501.18423.

User Guide

• Overview
  • The key idea
  • Architecture
  • Pretrained models
  • Citation
• Installation
  • Requirements
  • Basic install
  • Install from source
  • Optional dependencies
  • Pretrained weights
• Quickstart
  • Pretrained model variants
  • Load a pretrained model and reconstruct
  • Notebooks
  • Bundled dataset
  • CLI tools
• Training
  • Data preparation
  • Training a model
  • Available models
  • Hyperparameter options

Notebooks

• Simulated example
• Reconstructing real LIGO O3 glitches
• Training from scratch
• DeepExtractor — Training Tutorial (Google Colab)
• Part II — Hackathon Tasks

API Reference

• API Reference

Indices and tables

• Index

• Module Index

• Search Page