Weak-lensing shear measurement with machine learning: teaching artificial neural networks about feature noise

M. Tewes, T. Kuntzer, R. Nakajima, F. Courbin, H. Hildebrandt, and T. Schrabback
A&A 621, A36 (2019)

About the python implementation

The python code accompanying our paper is split into two github repositories and python packages:

• tenbilac is a simple artificial neural network library implementing the peculiar distinction between training cases and realizations, in python and numpy.
  This code is kept generic and ignorant of the application to galaxy shape measurement.
• momentsml is a toolbox for experimenting with shear and shape estimators, build around GalSim and Astropy.
  It includes a simple wrapper to process GREAT3 data, and an interface to tenbilac.

These packages provide a demonstration implementation of the algorithms described in the paper. They are oriented towards experimentation rather than being optimized for integration into a shear analysis pipeline.

Reproducing our results

The scripts and configuration files generating the results and figures from our paper are provided within the above repositories.
To use them, first get both tenbilac and momentsml (see their README files for instructions), and then follow the section Getting started in the MomentsML README.
In case of questions, don't hesitate to open issues on github, or to contact us.