Inference analysis

This module performs a likelihood analysis to derive Dark Energy constraints from (simulated) 3x2pt tomographic Pseudo-Cl (PCl) power spectra.

Following the creation of mock shear catalogues using run_cat_sim.sh in SWEPT/catalogue_sim, and the measurement of tomographic 3x2pt PCl spectra from the simulated mock data using SWEPT/pcl_measurement/run_3x2pt_tomo_measurement.sh, we can finally use the measured PCl data to place constraints on Dark Energy using this inference_analysis package.

Explicitly, run run_inference.sh in bash, which will perform a grid-based Gaussian likelihood analysis (presented in Upham+21, https://github.com/robinupham/gaussian_cl_likelihood) to derive constraints on the w0-wa parameters for a time-evolving Dark Energy equation of state. The parameters/set up of the inference analysis is controlled by the set_variables_inference.ini config file and the setup_inference_grid.py script. So, before executing run_inference.sh we will need to:

  • Set the path to the config file: in run_inference.sh, set PIPELINE_VARIABLES_PATH="/path-on-disk-to/SWEPT/inference_analysis/set_variables_inference.ini"

  • Specify inference analysis parameters in set_variables_inference.ini. Detailed description of parameters is provided in the header/comments in this config file

  • In setup_inference_grid.py, set the ranges/priors on the w0-wa grid to perform the likelihood analysis

By running run_inference.sh, the code will perform the Gaussian likelihood analysis, and save the posterior constraints on w0-wa for the given dataset/tomographic configuration in txt format on disk (location specified in the set_variables_inference.ini config file.