Skip to content

Gravitational wave pipeline

Jump to bottom
Matteo Breschi edited this page Jul 18, 2022 · 5 revisions

bajes provides a pipeline targeted for GW parameter estimation.

Generate the pipeline

The submission file for the pipeline can be generated though a configuration file running the command

bajes_pipe.py config.ini

For the full options available for the configuration file see config_example.

In general, the pipeline includes specific functionalities for the GW events released with GWTC-1, such as power spectral densities and calibration envelopes. For these events, the user can used the specific flag, e.g. event=GW150914 in the [gw-data] section.

For the events that are not included in GWTC-1, the user has to specify the correct GPS time t-gps consistently with GWOSC. We recall that the nominal GPS time corresponds to the central value of the analyzed time axis. Moreover, for these events, the user has to provide the related data, e.g. power spectral densities. bajes includes a functionality, i.e. evaluate_psd, for the PSD estimation based Welch's method.

The available detecors are:

  • H1 for LIGO Hanford
  • L1 for LIGO Livingston
  • V1 for Virgo
  • K1 for KAGRA
  • G1 for GEO600
  • I1 for the fiducial location of LIGO India
  • ET1, ET2 and ET3 for the fiducial locations of Einstein Telescope
  • CE for the fiducial location of Cosmic Explorer
  • CE-North and CE-South for the updated fiducial locations of Cosmic Explorer. The corresponding fiducial design ASD can be selected using the flags: CE-North-40, CE-North-20, CE-North-lf, CE-North-pm, and analogously for CE-South.

Read the results

The final output of the bajes GW pipeline is a summary.hdf5 that can be converted to a dictionary using the function bajes.pipe.load_dict_from_hdf5.

The output summary contains the following keys:

  • parameters: parameter space information
    • names : list of parameters names
    • bounds : list of parameters bounds
  • constants : constant values
    • seglen : duration of the segment [seconds]
    • srate : sampling rate [Hertz]
    • t_gps : nominal GPS time
    • ...
  • data: analyzed data information
    • strain : information on the analyzed data
      • detector ifo, e.g. H1 [one entry for each detector]
        • time : time axis
        • series : data in time-domain
        • freq : frequency axis
        • spectrum : spectrum in frequency-domain
    • noise : information on the PSD
      • detector ifo, e.g. H1 [one entry for each detector]
        • freq : frequency axis
        • spectrum : amplitude spectral density
    • waveform : information on the reconstructed waveform
      • time : time axis
      • series : median reconstructed waveform
      • series_up : upper bound (90% c.l.) reconstructed waveform
      • series_lo : lower bound (90% c.l.) reconstructed waveform
      • freq : frequency axis
      • spectrum : spectrum reconstructed waveform
      • series_whiten : whitened median reconstructed waveform
      • series_whiten_up : whitened upper bound (90% c.l.) reconstructed waveform
      • series_whiten_lo : whitened lower bound (90% c.l.) reconstructed waveform
  • prior_samples : prior samples for each parameter
    • parameter name, e.g. mchirp [one entry for each parameter]
  • posterior_samples : posterior samples for each parameter
    • parameter name, e.g. mchirp [one entry for each parameter]
  • SNR : posterior on the SNR
    • indeces : corresponding indeces of the posterior file used to compute the SNR (for reproducibility)
    • network_mf : network matched-filtered SNR
    • network_opt : network optimal SNR
    • H1_mf : matched-filtered SNR for H1
    • H1_opt : optimal SNR for H1
    • (so on for the other detectors)
Clone this wiki locally