sim_covariance.py

Final step is to get errorbars on the measurement. For now we will use a gaussian approximation of the covariance matrix. In order to construct the covariance matrix, we need first to have a model for the power spectrum. We use the signal only spectra to provide such model. This is subpart, and we should have an actual analytic model at some point. Anyway to get the model we run:

salloc -N 1 -C haswell -q interactive -t 01:00:00
python sim_spectra_model.py global_sims_all_noise.dict

This should run in less than 5 minutes.

Then we want to compute all elements of the analytic covariance matrix, for this we will have to submit the code to the queue, since it is too long to be ran on interactive mode.

A submission script is provided in

https://github.com/simonsobs/PSpipe/blob/master/project/AnalyseSims/NERSC_run/COV_PREP.sb

submit it in the same directory as mcm and spectra

sbatch COV_PREP.sb