Herculens: Differentiable gravitational lensing

[owlice]

Herculens: Differentiable gravitational lensing

Abstract: Herculens models imaging data of strong gravitational lenses. The package supports various degrees of model complexity, ranging from standard smooth analytical profiles to pixelated models and machine learning approaches. In particular, it implements multiscale pixelated models regularized with sparsity constraints and wavelet decomposition, for modeling both the source light distribution and the lens potential. The code is fully differentiable - based on JAX (ascl:2111.002) - which enables fast convergence to the solution, access to the parameters covariance matrix, efficient exploration of the parameter space including the sampling of posterior distributions using variational inference or Hamiltonian Monte-Carlo methods.

Credit: Galan, Aymeric; Peel, Austin; Vernardos, Giorgos; Biggio, Luca; Courbin, Frédéric; Starck, Jean-Luc

Site: https://github.com/austinpeel/herculens
https://ui.adsabs.harvard.edu/abs/2022arXiv220705763G

Bibcode: 2022ascl.soft09002G

ID: ascl:2209.002

[owlice]

Original abstract: Herculens models current and future observations of strong gravitational lenses. It is based on the automatic differentiation and compilation features of JAX (ascl:2111.002), which enables faster convergence to the solution, more efficient exploration of the parameter space including the sampling of posterior distributions, and new ways to mitigate degeneracies that affect gravitational lensing. Herculens supports various degrees of model complexity, ranging from standard smooth analytical profiles to pixelated models combined with machine learning approaches. It also supports several of the most widely-used analytical profiles as well as multi-scale pixelated models regularized with wavelets.

Updated abstract: Herculens models imaging data of strong gravitational lenses. The package supports various degrees of model complexity, ranging from standard smooth analytical profiles to pixelated models and machine learning approaches. In particular, it implements multiscale pixelated models regularized with sparsity constraints and wavelet decomposition, for modeling both the source light distribution and the lens potential. The code is fully differentiable - based on JAX (ascl:2111.002) - which enables fast convergence to the solution, access to the parameters covariance matrix, efficient exploration of the parameter space including the sampling of posterior distributions using variational inference or Hamiltonian Monte-Carlo methods.