Simulation files for pair differencing evaluation

Scripts and parameter files used in the pair differencing project.

General

The simulation comprises only the SAT 90 GHz frequency band and spans one observing year.

We simulate the following data:

• noise: atmosphere, instrumental noise
• cmb

We only simulate the first calendar day of each month.

Software

The software packages used are TOAST 3, sotodlib and mappraiser. They are provided as submodules of this repository so that the exact setup can be reproduced easily.

The simulation worfklow, so_mappraiser.py, is a modified version of a script in sotodlib and can be obtained by running the following command:

patch sotodlib/sotodlib/toast/scripts/so_sim.py -o so_mappraiser.py < so_sim.patch

Files in this directory

Unless otherwise noted, all scripts should be run from the root of the repository.

Setup

• get_defaults.sh : Use so_mappraiser.py to generate a default parameter file for reference
• sat.toml : Master parameter file for the so_mappraiser.py workflow
• schedule.opti.txt : Schedule file with a single scan
• schedule.01.south.txt : Schedule file
• schedule.small.txt : Truncated schedule file for laptop tests
• ffp10_lensed_scl_100_nside0512.fits : Input map to be observed during simulation

Tests (laptop: truncated schedule, decimated focal plane)

• opti : Evaluate the optimality of pair-differencing compared to maximum-likelihood (single observation)
  • run.white.uniform.sh : all detector pairs have the same white noise level (but not detectors in a pair)
  • run.white.variable.sh : all detectors have different white noise levels
  • run.one_over_f.sh : all detectors have different 1/f noise parameters
• syst : Evaluate the impact of systematic effects on the pair-differencing approach
  • run.atm.cache.sh : simulate and cache the atmosphere simulation
  • run.baseline.sh : run the baseline configuration (ideal case)
  • run.gains.constant.sh : run with gain errors which are the same for all detector pairs

Execution (Jean-Zay: full schedule)

• slurm/run.atm.cache.slurm : Simulate and cache the atmosphere simulation
• slurm/run.baseline.slurm : Run the baseline configuration (ideal case)
• slurm/run.gains.constant.slurm : Run with gain errors which are the same for all detector pairs
• slurm/run.gains.random.slurm : Run with Gaussian distributed gain errors

Post-processing

• utils.py : Some utility routines
• plot_maps.py : Produce difference maps and histograms for a given run
• plot_maps_all.py : Plot difference maps and histograms for all runs under a given directory
• spectrum.py : Power spectrum routines
• get_mask_apo.py : Create and save a mask (requires NaMaster)
• compute_spectra.py : Compute and save power spectra for all runs
• slurm/run.spectra.slurm : Job script to compute power spectra