gwnr.analysis — data-analysis tools

Matched-filter comparisons between waveform models, PSD utilities, GW transient catalog access, and template-bank construction. Source: gwnr/analysis/.

1. filter.py — matches, faithfulness and fitting factors
   1. calculate_faithfulness(m1, m2, s1x=0, ..., signal_approx='IMRPhenomD', signal_file='', signal_h=None, tmplt_approx='IMRPhenomC', tmplt_file='', tmplt_h=None, aligned_spin_tmplt_only=True, non_spin_tmplt_only=False, f_lower=15.0, sample_rate=4096, signal_duration=32, psd_string='aLIGOZeroDetHighPower', verbose=True, debug=False)
   2. calculate_fitting_factor(m1, m2, tmplt_approx, ..., vary_masses_only=True, vary_masses_and_aligned_spin_only=False, chirp_mass_window=0.2, effective_spin_window=0.75, f_lower=15.0, sample_rate=4096, signal_duration=16, psd_string='aLIGOZeroDetHighPower', ff_max=0.99999, pso_swarm_size=100, pso_omega=0.5, pso_phip=0.5, pso_phig=0.25, pso_minfunc=0.001, pso_n_processes=1, num_retries=5, verbose=True, debug=False)
   3. Other functions
2. psd.py — power spectral densities
   1. resample_and_extrapolate_psd(freq_vals, psd_vals, delta_f, f_max, precision=None, interpolation_func=scipy.interpolate.interp1d)
3. gw_transient_catalog.py — GW event catalogs
   1. class Merger(pycbc.catalog.Merger)
   2. class Catalog(pycbc.catalog.Catalog)
4. utils.py — bank/sim helpers
5. template_banks/ — template-bank construction
   1. template_banks/stochastic_bank/ — stochastic placement pipeline

---

filter.py — matches, faithfulness and fitting factors

calculate_faithfulness(m1, m2, s1x=0, ..., signal_approx='IMRPhenomD', signal_file='', signal_h=None, tmplt_approx='IMRPhenomC', tmplt_file='', tmplt_h=None, aligned_spin_tmplt_only=True, non_spin_tmplt_only=False, f_lower=15.0, sample_rate=4096, signal_duration=32, psd_string='aLIGOZeroDetHighPower', verbose=True, debug=False)

Computes the match (overlap maximized over time and phase) between a signal and a template with the same physical parameters, as modeled by two different approximants. Signals/templates may be specified by approximant name, by file, or as pre-generated waveforms (signal_h / tmplt_h). Template spin components transverse to the orbital angular momentum can be zeroed (aligned_spin_tmplt_only) or all spins dropped (non_spin_tmplt_only).

calculate_fitting_factor(m1, m2, tmplt_approx, ..., vary_masses_only=True, vary_masses_and_aligned_spin_only=False, chirp_mass_window=0.2, effective_spin_window=0.75, f_lower=15.0, sample_rate=4096, signal_duration=16, psd_string='aLIGOZeroDetHighPower', ff_max=0.99999, pso_swarm_size=100, pso_omega=0.5, pso_phip=0.5, pso_phig=0.25, pso_minfunc=0.001, pso_n_processes=1, num_retries=5, verbose=True, debug=False)

Computes the fitting factor: the match additionally maximized over the template’s physical parameters, using particle-swarm optimization (via pyswarm). The search space is either component masses only, or masses plus aligned spins, restricted to windows around the signal’s chirp mass and effective spin. PSO hyperparameters (swarm size, inertia omega, cognitive/social weights phip/phig) are tunable, and the optimization is retried up to num_retries times.

Other functions

• overlap_between_waveforms(wav1, wav2, psd, f_lower=15.0) — plain overlap between two already-generated waveforms.
• compute_snr_vs_time(wave, psd, time_step=0.01, f_lower=15.0) — accumulated SNR as a function of time.

---

psd.py — power spectral densities

resample_and_extrapolate_psd(freq_vals, psd_vals, delta_f, f_max, precision=None, interpolation_func=scipy.interpolate.interp1d)

Resamples measured PSD data onto a uniform frequency grid with spacing delta_f up to f_max, and extrapolates toward f = 0 when the measurement doesn’t extend below the physically measurable band. Returns a PSD usable with PyCBC filtering.

---

gw_transient_catalog.py — GW event catalogs

Extends pycbc.catalog with data-fetching conveniences.

class Merger(pycbc.catalog.Merger)

Information about a specific compact-binary merger, with methods to locate and download the public data around it:

• operating_ifos(ignore_ifos=['G1']), gpstime()
• frame_data_url(ifo, duration=32, sample_rate=4096) / frame_data_name(...)
• fetch_data(ifo, duration=32, sample_rate=4096, save_dir='') — download strain frames
• channel_name(ifo, sample_rate)
• psd_file_name(ifo) / fetch_psds(duration=32, sample_rate=4096, save_dir='') — download the event PSDs

class Catalog(pycbc.catalog.Catalog)

Catalog counterpart of the above. get_psd_url(source, name) resolves PSD download URLs per catalog release.

---

utils.py — bank/sim helpers

• get_uniform_mass_range(m_lower, m_upper, m_sep) — uniformly spaced mass values.
• outside_mchirp_window(bank, sim, w) — is a bank point outside a fractional chirp-mass window of an injection? (Used to prune match computations.)
• outside_tau0_window(bank, sim, window, f_lower) — same, in Newtonian chirp time τ₀.

---

template_banks/ — template-bank construction

Tools for building and manipulating template banks stored as LIGO_LW XML tables.

• ConvertTableType.py — convert between sngl_inspiral and sim_inspiral tables (invert_tabletype, new_row).
• MapTableToNRData.py — map template-bank rows to NR catalog simulations by matching physical parameters (does_this_map(p, c, param='eta', ...)).
• TurnCatalogIntoInjections.py — turn an NR catalog into an injection set, resolving each simulation’s waveform data location (CCE, extrapolated, or finite-radius files) from run metadata.

template_banks/stochastic_bank/ — stochastic placement pipeline

A self-contained pipeline for stochastic template-bank placement: propose random points in parameter space, reject those whose match with the existing bank exceeds the minimal-match threshold, and iterate to convergence. Key scripts:

Script	Role
script1.py … script4.py	Core proposal/rejection iterations (get_new_sample_point, reject_new_sample_point, waveform generation per point)
ChooseTestPoints.py	Draw random test points (masses, spins, angles) with acceptance regions for nonspinning / aligned-spin banks
banksim.py	Bank-simulation engine: pad waveforms to power-of-2 lengths, generate detector strain, compute matches
cut_bank.py, split_table.py	Restrict a bank to a parameter region; split tables for parallel jobs
push_eta_bank.py, push_chi_bank.py, move_eta_bank.py	Push bank boundaries in symmetric mass ratio / spin
take_uncovered_add_to_bank.py, match_combine.py, checkundone.py	Combine match results, add uncovered points, track job completion
make_dag.py	Generate the HTCondor DAG for the whole procedure
plotConvergance.py, plot_injection_match_mult.py, recovered_hist.py	Convergence and coverage diagnostics
EtasToQs.py, QsToEtas.py	Convert bank coordinates between η and q

These are batch scripts rather than an importable API; the installed command-line tools (gwnr_create_bank_workflow, choose_testpoints.py, …) drive them.