lenstronomy release 1.8.0

[Simon Birrer]

Dear lenstronomy community,

it’s my pleasure to announce lenstronomy version 1.8.0.

version 1.8.0 comes with many new features and speed-ups for certain tasks. All front-end APIs remain the same and the version should be fully backwards compatible.

The lenstronomy_extension notebooks are tested against this version (they did not need any changes).

Special thanks for this version goes to Ewoud Wempe, Daniel Gilman, Lyne Van de Vyvere, Jelle Alberts, Sebastian Carena-Wagner who all significantly contributed!

Especial shout-out to Ewoud Wempe for the numba implementations and speed-ups!

Find below the release notes. If you want to follow the lenstronomy development more closely (with this email, I intend to only announce PyPi releases) of pull requests, usage in the literature etc, we have a Slack channel (please let me know if you want to be added).

version 1.8.0 is named as a ‘major’ release. The reason is solely the configurable numba settings and the increased usage of the library.

A description of the configuration is given here: https://lenstronomy.readthedocs.io/en/latest/installation.html

Note: The default configurations are identical to previous versions and most likely your system is set-up to deal with them, except for HPC clusters.

Check also the requirements on the version of possible dependencies, as set here https://github.com/sibirrer/lenstronomy/blob/main/requirements.txt

Please let me know if you encounter problems or have questions!

Happy coding!

Simon

lenstronomy version 1.8.0 release notes

Top-level:

- numba global variable configurations

- installation documentation updated

- git branch master renamed to main and links updated

LensModel module:

- EPL numba version available (5x speed-up to EPL model)

- local curved arc models available (to be presented in an upcoming paper)

- improved latex documentation for some lens models

- vectorized interpolation model

- multi-plane cosmological settings allow for interpolated cosmology model from astropy

- speed up with numba for lens equation solver

- fast caustic calculation replacing matplotlib with skitlearn

- elliptical density slice lens model

LightModel:

- stability in Sersic profile improved

- vectorized interpolation model

Plots:

- multi-patch illustration class and plotting routines

FittingSequence:

- updated PSF iteration procedure with more settings