Software package that implements the Quadratic Maximum Likelihood (QML) estimator and applies it
to simulated cosmic shear data and compares the results to a Pseudo-Cl implementation.
This software package is split into three distinct parts:
-
A "pre-processing" Python package. This script, located under
./python/PreProcessing.py
, computes and saves relevant data files for later processes, such as the fiduciary cosmic shear power spectrum used in the analysis, the sky mask, and computing an analytic version of the QML's covariance matrix. -
A C++ executable which forms the bulk of the QML implementation. This implements our conjugate-gradient approach for our quadratic estimator, and is parallelized for maximum performance. The code relies on the Eigen linear algebra package, and the HealPix spherical harmonic transform library.
-
A "post-processing" Python package. This script, located under
./python/PostProcessing.py
, analyses the results of the C++ code and compares the QML's estimates with those from the Pseudo-Cl estimator and produces an array of plots highlighting the results.
We provide an example Jupyter Notebook which runs you through the basic ideas and implementation of our method. The notebook starts out with generating a mask that's applicable for a space-based Stage-IV galaxy survey, computing an example power spectrum for cosmic shear, computing the Fisher matrix for this mask & power spectrum, and then finally estimating the power spectrum for a set of cosmic shear maps. Using this notebook, it should be easy to extend this to whatever way the user sees fit.
The algorithms and implementation is detailed in Maraio, Hall, and Taylor 2022 which can be found on the arXiv at https://arxiv.org/abs/2207.10412.
If you use all or part of our code/method then please consider citing the above paper as it demonstrates that this work is useful to the community, thank you!
If any issues are encountered with this software, feel free to either raise an issue on this GitHub repository or directly email the main author in the above paper.