exposure

This is a simple library that's meant to compute the exposure of a network of GW detectors. In particular, this is decomposed into

• estimating the PSD within (join) lock segments as a function of time via massive parallelization

  • compute_psd (parallelized via condor-compute_psd)

• estimation of horizon for a particular detector given a particular PSD

  • compute_horizon (parallelized via condor-compute_horizon)

• estimation of network sensitivies given a set of detectors and their associated horizons

  • compute_network_sensitivity (parallelized via condor-compute_network_sensitivity)

• estimation of overall exposure by integrating network sensitivities over time

  • compute_exposure

The work flow will likely run in this order. We note that there is further parallelization available by writing a single DAG that combines compute_psd, compute_horizon, and compute_network_sensitivity together with appropriate parent/child relationships. We could add compute_exposure to that as well, but it will require all jobs to finish prior and therefore might as well just be a postscript or run by hand afterwards.

We also provide some basic functionality to estimate diurnal cycles.

• compute a stacked histogram (with some user-specified periodicity)

  • diurnal_hist

• compute the dft of lock segments and run a peak finding algorithm to identify periodic components

  • diuranl_dft

• compute a time-frequency representation of lock segments to look for variabilitoy of periodic elements

  • diurnal_timefreq

Dependencies

• skymap_statistics
• numpy
• matplotlib
• healpy
• lalsuite (lalframe.frread, gpstime.tconvert)
• corner.py

NOTE, the library is not Python3 compatible (it uses deprecated functionality only available in Python2)