Catalog information, model predictions, hierarchical inference chains for 'Lens Modeling Accuracy in Expected LSST Lensed AGN Sample'
收藏资源简介:
This repository contains catalog information, predictions and chains used for all figures in paper Lens Model Accuracy in the Expected LSST Lensed AGN Sample (Venkatraman P. et al 2025) produced within the Dark Energy Science Collaboration (DESC). For more details on how to use these files, please see notebooks in this github repository which uses these files as input, and shows how to re-create the figures mentioned above. - hierarchical_inference_chains.tgz contains MCMC chains for all lens mass population models. - model_predictions.zip contains all neural network predictions for all experiments (i.e. mass model posteriors) - test_train_data.tar.gz contains training data catalog and test data catalog - catalog_data.tar.gz contains the catalog information for the fiducial train and test set. It also contains information about the catalog we started from in order to construct the om10-cosmoDC2 catalog. - final_reweighted_posteriors_appendix.tar.gz contains samples from the reweighted NPE posteriors (reweighted with the inferred population model, dividing out the interim prior) for the fiducial test set



