icecube.gulliver_bootstrap package

icecube.gulliver_bootstrap.I3BootstrappedFit(tray, name, seed_service, likelihood_service, pulses='OfflinePulses', If=<function <lambda>>, n_iterations=4, keep_intermediates=False, do_final_fit=False)

Run a series of bootstrapped gulliver fits (fits to resamplings of the existing pulses). The variation in direction of the fit results is used to compute an estimate of the angular error of the reconstruction of the original pulses (with the same likelihood). Optionally, run a reconstruction of the original pulses, using each of the resampled fit results (and the average of the resampled fit results) as a seed.

Parameters:
  • seed_service – The name of the service to use for the initial seed. If this is the empty string each resampled fit will be seeded with a linefit to the resampled pulses.

  • likelihood_service – The name of the likelihood service to use for all fitting

  • pulses – The I3RecoPulses to use for the resampling and fitting

  • If – A Python function, whose result determines whether to process each frame

  • n_iterations – The number of resampled fits to perform

  • keep_intermediates – Whether the resamped fit results and fit parameters should be left in the frame

  • do_final_fit – Whether to do a final fit of the original pulses, using all of the resampled fit results as seeds