space-dolphin 1.3.0

Automated pipeline for lens modeling based on lenstronomy

Welcome to dolphin, an AI-powered automated pipeline for strong gravitational lens modeling!

dolphin leverages lenstronomy as its core modeling engine, providing an accessible and scalable framework for studying galaxy-scale lenses.

What is Dolphin?

Strong gravitational lens modeling traditionally requires significant manual effort. dolphin changes this by providing an AI-driven approach to forward modeling, enabling researchers to process large samples of strong lenses with ease. Whether you want a fully hands-off automated pipeline or a semi-automated workflow where you can fine-tune the AI-generated configurations, dolphin gives you the flexibility and power you need.

Features

• 🤖 AI-Automated Modeling: Streamline forward modeling for large datasets of galaxy-scale lenses.

• 🎛️ Flexible Workflows: Choose between fully automated runs or semi-automated modes with manual overrides.

• 🌈 Multi-Band Support: Easily configure and model across multiple observing bands simultaneously.

• 🌌 Versatile Sources: Built-in support for both galaxy–galaxy and galaxy–quasar lens systems.

• 💻 HPC Ready: Seamlessly sync your setup between local machines and High-Performance Computing Clusters (HPCC).

• ✅ Tested & Reliable: Comprehensively tested with .

Installation

Installing dolphin is simple. You can install the latest stable release via pip:

pip install space-dolphin

Alternatively, to install the latest development version directly from GitHub:

git clone https://github.com/ajshajib/dolphin.git
cd dolphin
pip install .

For instructions on setting up your workspace and running your first model, please check out our Quickstart guide.

Citation

If you use dolphin in your research, please cite the main dolphin paper:

• Shajib et al. (2025)

Depending on the fitting recipe used, please additionally cite the following papers for the underlying modeling methodology:

• Galaxy-Quasar Recipe: Shajib et al. (2019)

• Galaxy-Galaxy Recipe: Shajib et al. (2021)