enterprise_extensions package

Subpackages

Submodules

enterprise_extensions.blocks module

enterprise_extensions.blocks.bwm_block(Tmin, Tmax, amp_prior='log-uniform', skyloc=None, logmin=- 18, logmax=- 11, name='bwm')[source]

Returns deterministic GW burst with memory model:: 1. Burst event parameterized by time, sky location, polarization angle, and amplitude

Parameters

Tmin – Min time to search, probably first TOA (MJD).
Tmax – Max time to search, probably last TOA (MJD).
amp_prior – Prior on log10_A. Default if “log-uniform”. Use “uniform” for upper limits.
skyloc – Fixed sky location of BWM signal search as [cos(theta), phi]. Search over sky location if None given.
logmin – log of minimum BWM amplitude for prior (log10)
logmax – log of maximum BWM amplitude for prior (log10)
name – Name of BWM signal.

enterprise_extensions.blocks.bwm_sglpsr_block(Tmin, Tmax, amp_prior='log-uniform', logmin=- 17, logmax=- 12, name='ramp', fixed_sign=None)[source]

enterprise_extensions.blocks.chromatic_noise_block(gp_kernel='nondiag', psd='powerlaw', nondiag_kernel='periodic', prior='log-uniform', dt=15, df=200, idx=4, include_quadratic=False, Tspan=None, name='chrom', components=30, coefficients=False)[source]

Returns GP chromatic noise model :

1. Chromatic modeled with user defined PSD with 30 sampling frequencies. Available PSDs are [‘powerlaw’, ‘turnover’ ‘spectrum’]

Parameters

gp_kernel – Whether to use a diagonal kernel for the GP. [‘diag’,’nondiag’]
nondiag_kernel – Which nondiagonal kernel to use for the GP. [‘periodic’,’sq_exp’,’periodic_rfband’,’sq_exp_rfband’]
psd – PSD to use for common red noise signal. Available options are [‘powerlaw’, ‘turnover’ ‘spectrum’]
prior – What type of prior to use for amplitudes. [‘log-uniform’,’uniform’]
dt – time-scale for linear interpolation basis (days)
df – frequency-scale for linear interpolation basis (MHz)
idx – Index of radio frequency dependence (i.e. DM is 2). Any float will work.
include_quadratic – Whether to include a quadratic fit.
name – Name of signal
Tspan – Tspan from which to calculate frequencies for PSD-based GPs.
components – Number of frequencies to use in ‘diag’ GPs.
coefficients – Whether to keep coefficients of the GP.

enterprise_extensions.blocks.common_red_noise_block(psd='powerlaw', prior='log-uniform', Tspan=None, components=30, combine=True, log10_A_val=None, gamma_val=None, delta_val=None, logmin=None, logmax=None, orf=None, orf_ifreq=0, leg_lmax=5, name='gw', coefficients=False, pshift=False, pseed=None)[source]

Returns common red noise model:

1. Red noise modeled with user defined PSD with 30 sampling frequencies. Available PSDs are [‘powerlaw’, ‘turnover’ ‘spectrum’]

Parameters

psd – PSD to use for common red noise signal. Available options are [‘powerlaw’, ‘turnover’ ‘spectrum’, ‘broken_powerlaw’]
prior – Prior on log10_A. Default if “log-uniform”. Use “uniform” for upper limits.
Tspan – Sets frequency sampling f_i = i / Tspan. Default will use overall time span for individual pulsar.
log10_A_val – Value of log10_A parameter for fixed amplitude analyses.
gamma_val – Value of spectral index for power-law and turnover models. By default spectral index is varied of range [0,7]
delta_val – Value of spectral index for high frequencies in broken power-law and turnover models. By default spectral index is varied in range [0,7]. :param logmin: Specify the lower bound of the prior on the amplitude for all psd but ‘spectrum’. If psd==’spectrum’, then this specifies the lower prior on log10_rho_gw
logmax – Specify the lower bound of the prior on the amplitude for all psd but ‘spectrum’. If psd==’spectrum’, then this specifies the lower prior on log10_rho_gw
orf – String representing which overlap reduction function to use. By default we do not use any spatial correlations. Permitted values are [‘hd’, ‘dipole’, ‘monopole’].
orf_ifreq – Frequency bin at which to start the Hellings & Downs function with numbering beginning at 0. Currently only works with freq_hd orf.
leg_lmax – Maximum multipole of a Legendre polynomial series representation of the overlap reduction function [default=5]
pshift – Option to use a random phase shift in design matrix. For testing the null hypothesis.
pseed – Option to provide a seed for the random phase shift.
name – Name of common red process

enterprise_extensions.blocks.dm_noise_block(gp_kernel='diag', psd='powerlaw', nondiag_kernel='periodic', prior='log-uniform', dt=15, df=200, Tspan=None, components=30, gamma_val=None, coefficients=False)[source]

Returns DM noise model:

DM noise modeled as a power-law with 30 sampling frequencies

Parameters

psd – PSD function [e.g. powerlaw (default), spectrum, tprocess]
prior – Prior on log10_A. Default if “log-uniform”. Use “uniform” for upper limits.
dt – time-scale for linear interpolation basis (days)
df – frequency-scale for linear interpolation basis (MHz)
Tspan – Sets frequency sampling f_i = i / Tspan. Default will use overall time span for indivicual pulsar.
components – Number of frequencies in sampling of DM-variations.
gamma_val – If given, this is the fixed slope of the power-law for powerlaw, turnover, or tprocess DM-variations

enterprise_extensions.blocks.red_noise_block(psd='powerlaw', prior='log-uniform', Tspan=None, components=30, gamma_val=None, coefficients=False, select=None, modes=None, wgts=None, combine=True, break_flat=False, break_flat_fq=None, logmin=None, logmax=None, dropout=False, k_threshold=0.5)[source]

Returns red noise model:: Red noise modeled as a power-law with 30 sampling frequencies

Parameters

psd – PSD function [e.g. powerlaw (default), turnover, spectrum, tprocess]
prior – Prior on log10_A. Default if “log-uniform”. Use “uniform” for upper limits.
Tspan – Sets frequency sampling f_i = i / Tspan. Default will use overall time span for indivicual pulsar.
components – Number of frequencies in sampling of red noise
gamma_val – If given, this is the fixed slope of the power-law for powerlaw, turnover, or tprocess red noise
coefficients – include latent coefficients in GP model?
dropout – Use a dropout analysis for intrinsic red noise models. Currently only supports power law option.
k_threshold – Threshold for dropout analysis.

enterprise_extensions.blocks.white_noise_block(vary=False, inc_ecorr=False, gp_ecorr=False, efac1=False, select='backend', tnequad=False, name=None)[source]

Returns the white noise block of the model:

EFAC per backend/receiver system

EQUAD per backend/receiver system

ECORR per backend/receiver system

Parameters

vary – If set to true we vary these parameters with uniform priors. Otherwise they are set to constants with values to be set later.
inc_ecorr – include ECORR, needed for NANOGrav channelized TOAs
gp_ecorr – whether to use the Gaussian process model for ECORR
efac1 – use a strong prior on EFAC = Normal(mu=1, stdev=0.1)
tnequad – Whether to use the TempoNest definition of EQUAD. Defaults to False to follow Tempo, Tempo2 and Pint definition.

enterprise_extensions.deterministic module

enterprise_extensions.deterministic.CWSignal(cw_wf, ecc=False, psrTerm=False, name='cw')[source]

enterprise_extensions.deterministic.bwm_delay(toas, pos, log10_h=- 14.0, cos_gwtheta=0.0, gwphi=0.0, gwpol=0.0, t0=55000, antenna_pattern_fn=None)

Function that calculates the earth-term gravitational-wave burst-with-memory signal, as described in: Seto et al, van haasteren and Levin, phsirkov et al, Cordes and Jenet. This version uses the F+/Fx polarization modes, as verified with the Continuous Wave and Anisotropy papers.

Parameters

toas – Time-of-arrival measurements [s]
pos – Unit vector from Earth to pulsar
log10_h – log10 of GW strain
cos_gwtheta – Cosine of GW polar angle
gwphi – GW azimuthal polar angle [rad]
gwpol – GW polarization angle
t0 – Burst central time [day]
antenna_pattern_fn – User defined function that takes pos, gwtheta, gwphi as arguments and returns (fplus, fcross)

Returns

the waveform as induced timing residuals (seconds)

enterprise_extensions.deterministic.bwm_sglpsr_delay(toas, sign, log10_A=- 15, t0=55000)

Function that calculates the earth-term gravitational-wave burst-with-memory signal for an optimally oriented source in a single pulsar

Parameters

toas – Time-of-arrival measurements [s]
log10_A – log10 of the amplitude of the ramp (delta_f/f)
t0 – Burst central time [day]

Returns

the waveform as induced timing residuals (seconds)

enterprise_extensions.deterministic.compute_eccentric_residuals(toas, theta, phi, cos_gwtheta, gwphi, log10_mc, log10_dist, log10_h, log10_F, cos_inc, psi, gamma0, e0, l0, q, nmax=400, pdist=1.0, pphase=None, pgam=None, psrTerm=False, tref=0, check=False)

Simulate GW from eccentric SMBHB. Waveform models from Taylor et al. (2015) and Barack and Cutler (2004). WARNING: This residual waveform is only accurate if the GW frequency is not significantly evolving over the observation time of the pulsar.

Parameters

toa – pulsar observation times
theta – polar coordinate of pulsar
phi – azimuthal coordinate of pulsar
gwtheta – Polar angle of GW source in celestial coords [radians]
gwphi – Azimuthal angle of GW source in celestial coords [radians]
log10_mc – Base-10 lof of chirp mass of SMBMB [solar masses]
log10_dist – Base-10 uminosity distance to SMBMB [Mpc]
log10_F – base-10 orbital frequency of SMBHB [Hz]
inc – Inclination of GW source [radians]
psi – Polarization of GW source [radians]
gamma0 – Initial angle of periastron [radians]
e0 – Initial eccentricity of SMBHB
l0 – Initial mean anomoly [radians]
q – Mass ratio of SMBHB
nmax – Number of harmonics to use in waveform decomposition
pdist – Pulsar distance [kpc]
pphase – Pulsar phase [rad]
pgam – Pulsar angle of periastron [rad]
psrTerm – Option to include pulsar term [boolean]
tref – Fidicuial time at which initial parameters are referenced [s]
check – Check if frequency evolves significantly over obs. time

Returns

Vector of induced residuals

enterprise_extensions.deterministic.cw_block_circ(amp_prior='log-uniform', dist_prior=None, skyloc=None, log10_fgw=None, psrTerm=False, tref=0, name='cw')[source]

Returns deterministic, cirular orbit continuous GW model:

Parameters

amp_prior – Prior on log10_h. Default is “log-uniform.” Use “uniform” for upper limits, or “None” to search over log10_dist instead.
dist_prior – Prior on log10_dist. Default is “None,” meaning that the search is over log10_h instead of log10_dist. Use “log-uniform” to search over log10_h with a log-uniform prior.
skyloc – Fixed sky location of CW signal search as [cos(theta), phi]. Search over sky location if None given.
log10_fgw – Fixed log10 GW frequency of CW signal search. Search over GW frequency if None given.
ecc – Fixed log10 distance to SMBHB search. Search over distance or strain if None given.
psrTerm – Boolean for whether to include the pulsar term. Default is False.
name – Name of CW signal.

enterprise_extensions.deterministic.cw_block_ecc(amp_prior='log-uniform', skyloc=None, log10_F=None, ecc=None, psrTerm=False, tref=0, name='cw')[source]

Returns deterministic, eccentric orbit continuous GW model:

Parameters

amp_prior – Prior on log10_h and log10_Mc/log10_dL. Default is “log-uniform” with log10_Mc and log10_dL searched over. Use “uniform” for upper limits, log10_h searched over.
skyloc – Fixed sky location of CW signal search as [cos(theta), phi]. Search over sky location if None given.
log10_F – Fixed log-10 orbital frequency of CW signal search. Search over orbital frequency if None given.
ecc – Fixed eccentricity of SMBHB search. Search over eccentricity if None given.
psrTerm – Boolean for whether to include the pulsar term. Default is False.
name – Name of CW signal.

enterprise_extensions.deterministic.cw_delay(toas, pos, pdist, cos_gwtheta=0, gwphi=0, cos_inc=0, log10_mc=9, log10_fgw=- 8, log10_dist=None, log10_h=None, phase0=0, psi=0, psrTerm=False, p_dist=1, p_phase=None, evolve=False, phase_approx=False, check=False, tref=0)

Function to create GW incuced residuals from a SMBMB as defined in Ellis et. al 2012,2013.

Parameters

toas – Pular toas in seconds
pos – Unit vector from the Earth to the pulsar
pdist – Pulsar distance (mean and uncertainty) [kpc]
cos_gwtheta – Cosine of Polar angle of GW source in celestial coords [radians]
gwphi – Azimuthal angle of GW source in celestial coords [radians]
cos_inc – cosine of Inclination of GW source [radians]
log10_mc – log10 of Chirp mass of SMBMB [solar masses]
log10_fgw – log10 of Frequency of GW (twice the orbital frequency) [Hz]
log10_dist – log10 of Luminosity distance to SMBMB [Mpc], used to compute strain, if not None
log10_h – log10 of GW strain, used to compute distance, if not None
phase0 – Initial GW phase of source [radians]
psi – Polarization angle of GW source [radians]
psrTerm – Option to include pulsar term [boolean]
p_dist – Pulsar distance parameter
p_phase – Use pulsar phase to determine distance [radian]
evolve – Option to include/exclude full evolution [boolean]
phase_approx – Option to include/exclude phase evolution across observation time [boolean]
check – Check if frequency evolves significantly over obs. time [boolean]
tref – Reference time for phase and frequency [s]

Returns

Vector of induced residuals

enterprise_extensions.deterministic.fdm_block(Tmin, Tmax, amp_prior='log-uniform', name='fdm', amp_lower=- 18, amp_upper=- 11, freq_lower=- 9, freq_upper=- 7, use_fixed_freq=False, fixed_freq=- 8)[source]

Returns deterministic fuzzy dark matter model:

FDM parameterized by frequency, phase,
and amplitude (mass and DM energy density).

Parameters

Tmin – Min time to search, probably first TOA (MJD).
Tmax – Max time to search, probably last TOA (MJD).
amp_prior – Prior on log10_A.
logmin – log of minimum FDM amplitude for prior (log10)
logmax – log of maximum FDM amplitude for prior (log10)
name – Name of FDM signal.
freq_lower (amp_upper, amp_lower, freq_upper,) – The log-space bounds on the amplitude and frequency priors.
use_fixed_freq – Whether to do a fixed-frequency run and not search over the frequency.
fixed_freq – The frequency value to do a fixed-frequency run with.

enterprise_extensions.deterministic.fdm_delay(toas, log10_A, log10_f, phase_e, phase_p)

Function that calculates the earth-term gravitational-wave fuzzy dark matter signal, as described in: Kato et al. (2020).

Parameters

toas – Time-of-arrival measurements [s]
log10_A – log10 of GW strain
log10_f – log10 of GW frequency
phase_e – The Earth-term phase of the GW
phase_p – The Pulsar-term phase of the GW

Returns

the waveform as induced timing residuals (seconds)

enterprise_extensions.deterministic.generalized_gwpol_psd(f, log10_A_tt=- 15, log10_A_st=- 15, log10_A_vl=- 15, log10_A_sl=- 15, kappa=3.3333333333333335, p_dist=1.0): PSD for a generalized mixture of scalar+vector dipole radiation and tensorial quadrupole radiation from SMBHBs.

enterprise_extensions.dropout module

enterprise_extensions.dropout.Dropout_PhysicalEphemerisSignal(frame_drift_rate=enterprise.signals.parameter.Uniform, d_jupiter_mass=enterprise.signals.parameter.Normal, d_saturn_mass=enterprise.signals.parameter.Normal, d_uranus_mass=enterprise.signals.parameter.Normal, d_neptune_mass=enterprise.signals.parameter.Normal, jup_orb_elements=enterprise.signals.parameter.Uniform, sat_orb_elements=enterprise.signals.parameter.Uniform, inc_jupiter_orb=True, inc_saturn_orb=False, use_epoch_toas=True, k_drop=enterprise.signals.parameter.Uniform, k_threshold=0.5, name='')[source]: Class factory for dropout physical ephemeris model signal.

enterprise_extensions.dropout.dropout_physical_ephem_delay(toas, planetssb, pos_t, frame_drift_rate=0, d_jupiter_mass=0, d_saturn_mass=0, d_uranus_mass=0, d_neptune_mass=0, jup_orb_elements=array([0.0, 0.0, 0.0, 0.0, 0.0, 0.0]), sat_orb_elements=array([0.0, 0.0, 0.0, 0.0, 0.0, 0.0]), inc_jupiter_orb=False, jup_orbelxyz=None, jup_mjd=None, inc_saturn_orb=False, sat_orbelxyz=None, sat_mjd=None, equatorial=True, k_drop=0.5, k_threshold=0.5): Dropout BayesEphem model. Switches BayesEphem on or off depending on whether k_drop exceeds k_threshold.

enterprise_extensions.dropout.dropout_powerlaw(f, name, log10_A=- 16, gamma=5, dropout_psr='B1855+09', k_drop=0.5, k_threshold=0.5)

Dropout powerlaw for a stochastic process. Switches a stochastic process on or off in a single pulsar depending on whether k_drop exceeds k_threshold.

Parameters: dropout_psr – Which pulsar to use a dropout switch on. The value ‘all’ will use the method on all pulsars.

enterprise_extensions.empirical_distr module

class enterprise_extensions.empirical_distr.EmpiricalDistribution1D(param_name, samples, bins)[source]

Bases: object

Class used to define a 1D empirical distribution based on posterior from another MCMC.

Parameters

samples – samples for hist
bins – edges to use for hist (left and right) make sure bins cover whole prior!

draw()[source]

logprob(params)[source]

prob(params)[source]

class enterprise_extensions.empirical_distr.EmpiricalDistribution1DKDE(param_name, samples, minval=None, maxval=None, bandwidth=0.1, nbins=40)[source]

Bases: object

Minvals and maxvals should specify priors for these. Should make these required.

draw()[source]

class enterprise_extensions.empirical_distr.EmpiricalDistribution2D(param_names, samples, bins)[source]

Bases: object

Class used to define a 1D empirical distribution based on posterior from another MCMC.

Parameters

samples – samples for hist
bins – edges to use for hist (left and right) make sure bins cover whole prior!

draw()[source]

logprob(params)[source]

prob(params)[source]

class enterprise_extensions.empirical_distr.EmpiricalDistribution2DKDE(param_names, samples, minvals=None, maxvals=None, bandwidth=0.1, nbins=40)[source]

Bases: object

Minvals and maxvals should specify priors for these. Should make these required.

Parameters

param_names – 2-element list of parameter names
samples – samples, with dimension (2 x Nsamples)

Return distr

list of empirical distributions

draw()[source]

logprob(params)[source]

prob(params)[source]

enterprise_extensions.empirical_distr.make_empirical_distributions(pta, paramlist, params, chain, burn=0, nbins=81, filename='distr.pkl', return_distribution=True, save_dists=True)[source]

Utility function to construct empirical distributions.

Parameters

pta – the pta object used to generate the posteriors
paramlist – a list of parameter names, either single parameters or pairs of parameters
chain – MCMC chain from a previous run
burn – desired number of initial samples to discard
nbins – number of bins to use for the empirical distributions

Return distr

list of empirical distributions

enterprise_extensions.empirical_distr.make_empirical_distributions_KDE(pta, paramlist, params, chain, burn=0, nbins=41, filename='distr.pkl', bandwidth=0.1, return_distribution=True, save_dists=True)[source]

Utility function to construct empirical distributions.

Parameters

paramlist – a list of parameter names, either single parameters or pairs of parameters
params – list of all parameter names for the MCMC chain
chain – MCMC chain from a previous run, has dimensions Nsamples x Nparams
burn – desired number of initial samples to discard
nbins – number of bins to use for the empirical distributions

Return distr

list of empirical distributions

enterprise_extensions.gp_kernels module

enterprise_extensions.gp_kernels.dmx_ridge_prior(avetoas, log10_sigma=- 7): DMX-like signal with Gaussian prior

enterprise_extensions.gp_kernels.get_tf_quantization_matrix(toas, freqs, dt=2592000, df=None, dm=False, dm_idx=2): Quantization matrix in time and radio frequency to cut down on the kernel size.

enterprise_extensions.gp_kernels.linear_interp_basis_dm(toas, freqs, dt=2592000)

enterprise_extensions.gp_kernels.linear_interp_basis_freq(freqs, df=64): Linear interpolation in radio frequency

enterprise_extensions.gp_kernels.periodic_kernel(avetoas, log10_sigma=- 7, log10_ell=2, log10_gam_p=0, log10_p=0): Quasi-periodic kernel for DM

enterprise_extensions.gp_kernels.se_dm_kernel(avetoas, log10_sigma=- 7, log10_ell=2): Squared-exponential kernel for DM

enterprise_extensions.gp_kernels.se_kernel(avefreqs, log10_sigma=- 7, log10_lam=3): Squared-exponential kernel for FD

enterprise_extensions.gp_kernels.sf_kernel(labels, log10_sigma=- 7, log10_ell=2, log10_ell2=4, log10_alpha_wgt=0): The product of a squared-exponential time kernel and a rational-quadratic frequency kernel.

enterprise_extensions.gp_kernels.tf_kernel(labels, log10_sigma=- 7, log10_ell=2, log10_gam_p=0, log10_p=0, log10_ell2=4, log10_alpha_wgt=0): The product of a quasi-periodic time kernel and a rational-quadratic frequency kernel.

enterprise_extensions.hypermodel module

enterprise_extensions.model_orfs module

enterprise_extensions.model_orfs.anis_orf(pos1, pos2, params, **kwargs): Anisotropic GWB spatial correlation function.

enterprise_extensions.model_orfs.bin_orf(pos1, pos2, params)

Agnostic binned spatial correlation function. Bin edges are placed at edges and across angular separation space. Changing bin edges will require manual intervention to create new function.

Param: params inter-pulsar correlation bin amplitudes.

Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.dipole_orf(pos1, pos2): Dipole spatial correlation function.

enterprise_extensions.model_orfs.freq_hd(pos1, pos2, params)

Frequency-dependent Hellings & Downs spatial correlation function. Implemented as a model that only enforces H&D inter-pulsar correlations after a certain number of frequencies in the spectrum. The first set of frequencies are uncorrelated.

Param: params params[0] is the number of components in the stochastic process. params[1] is the frequency at which to start the H&D inter-pulsar correlations (indexing from 0).

Reference: Taylor et al. (2017), https://arxiv.org/abs/1606.09180 Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.generalized_gwpol_psd(f, log10_A_tt=- 15, log10_A_st=- 15, alpha_tt=- 0.6666666666666666, alpha_alt=- 1, log10_A_vl=- 15, log10_A_sl=- 15, kappa=0, p_dist=1.0)

General powerlaw spectrum allowing for existence of all possible modes of gravity as predicted by a general metric spacetime theory and generated by a binary system. The SL and VL modes’ powerlaw relations are not normalized.

Param: f A list of considered frequencies
Param: log10_A_tt Amplitude of the tensor transverse mode
Param: log10_A_st Amplitude of the scalar transverse mode
Param: log10_A_vl Amplitude of the vector longitudinal mode
Param: log10_A_sl Amplitude of the scalar longitudinal mode
Param: kappa Relative amplitude of dipole radiation over quadrupolar radiation
Param: p_dist Pulsar distance in kpc
Param: alpha_tt spectral index of the TT mode.
Param: alpha_alt spectral index of the non-Einsteinian modes.

Reference: Cornish et al. (2017), https://arxiv.org/abs/1712.07132 Author: S. R. Taylor, N. Laal (2020)

enterprise_extensions.model_orfs.gt_orf(pos1, pos2, tau)

General Transverse (GT) Correlations. This ORF is used to detect the relative significance of all possible correlation patterns induced by the most general family of transverse gravitational waves.

Param: tau tau = 1 results in ST correlations while tau = -1 results in HD correlations.

Author: N. Laal (2020)

enterprise_extensions.model_orfs.gw_dipole_orf(pos1, pos2): GW-dipole Correlations. Author: N. Laal (2020)

enterprise_extensions.model_orfs.gw_monopole_orf(pos1, pos2): GW-monopole Correlations. This phenomenological correlation pattern can be used in Bayesian runs as the simplest type of correlations. Author: N. Laal (2020)

enterprise_extensions.model_orfs.hd_orf(pos1, pos2): Hellings & Downs spatial correlation function.

enterprise_extensions.model_orfs.legendre_orf(pos1, pos2, params)

Legendre polynomial spatial correlation function. Assumes process normalization such that autocorrelation signature is 1. A separate function is needed to use a “split likelihood” model with this Legendre process decoupled from the autocorrelation signature (“zero_diag_legendre_orf”).

Param: params Legendre polynomial amplitudes describing the Legendre series approximation to the inter-pulsar correlation signature. H&D coefficients are a_0=0, a_1=0, a_2=0.3125, a_3=0.0875, …

Reference: Gair et al. (2014), https://arxiv.org/abs/1406.4664 Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.monopole_orf(pos1, pos2): Monopole spatial correlation function.

enterprise_extensions.model_orfs.param_hd_orf(pos1, pos2, a=1.5, b=- 0.25, c=0.5)

Pre-factor parametrized Hellings & Downs spatial correlation function.

Param: a, b, c: coefficients of H&D-like curve [default=1.5,-0.25,0.5].

Reference: Taylor, Gair, Lentati (2013), https://arxiv.org/abs/1210.6014 Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.spline_orf(pos1, pos2, params)

Agnostic spline-interpolated spatial correlation function. Spline knots are placed at edges, zeros, and minimum of H&D curve. Changing locations will require manual intervention to create new function.

Param: params spline knot amplitudes.

Reference: Taylor, Gair, Lentati (2013), https://arxiv.org/abs/1210.6014 Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.st_orf(pos1, pos2): Scalar tensor correlations as induced by the breathing polarization mode of gravity. Author: N. Laal (2020)

enterprise_extensions.model_orfs.zero_diag_bin_orf(pos1, pos2, params)

Agnostic binned spatial correlation function. To be used in a “split likelihood” model with an additional common uncorrelated red process. The latter is necessary to regularize the overall Phi covariance matrix.

Param: params inter-pulsar correlation bin amplitudes.

Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.zero_diag_hd(pos1, pos2)

Off-diagonal Hellings & Downs spatial correlation function. To be used in a “split likelihood” model with an additional common uncorrelated red process. The latter is necessary to regularize the overall Phi covariance matrix.

Author: S. R. Taylor (2020)

enterprise_extensions.model_orfs.zero_diag_legendre_orf(pos1, pos2, params)

Legendre polynomial spatial correlation function. To be used in a “split likelihood” model with an additional common uncorrelated red process. The latter is necessary to regularize the overall Phi covariance matrix.

Param: params Legendre polynomial amplitudes describing the Legendre series approximation to the inter-pulsar correlation signature. H&D coefficients are a_0=0, a_1=0, a_2=0.3125, a_3=0.0875, …

Reference: Gair et al. (2014), https://arxiv.org/abs/1406.4664 Author: S. R. Taylor (2020)

enterprise_extensions.model_utils module

enterprise_extensions.models module

enterprise_extensions.sampler module

enterprise_extensions.sky_scrambles module

enterprise_extensions.sky_scrambles.compute_match(orf1, orf1_mag, orf2, orf2_mag)[source]: Computes the match between two different ORFs.

enterprise_extensions.sky_scrambles.compute_orf(ptheta, pphi)[source]

Computes the ORF coefficient. Takes different input than utils.hd_orf().

Parameters

ptheta – Array of values of pulsar positions theta
pphi – Array of values of pulsar positions phi

Returns

orf: ORF for the given positions orf_mag: Magnitude of the ORF

enterprise_extensions.sky_scrambles.get_scrambles(psrs, N=500, Nmax=10000, thresh=0.1, filename='sky_scrambles.npz', resume=False)[source]

Get sky scramble ORFs and matches.

Parameters

psrs – List of pulsar objects
N – Number of desired sky scrambles
Nmax – Maximum number of tries to get independent scrambles
thresh – Threshold value for match statistic.
filename – Name of the file where the sky scrambles should be saved. Sky scrambles should be saved in npz file.
resume – Whether to resume from an earlier run.

enterprise_extensions.sky_scrambles.make_true_orf(psrs)[source]: Computes the ORF by looping over pulsar pairs

enterprise_extensions.timing module

enterprise_extensions.timing.timing_block(tmparam_list=['RAJ', 'DECJ', 'F0', 'F1', 'PMRA', 'PMDEC', 'PX'])[source]: Returns the timing model block of the model :param tmparam_list: a list of parameters to vary in the model

enterprise_extensions.timing.tm_delay(residuals, t2pulsar, tmparams_orig, tmparams, which='all')

Compute difference in residuals due to perturbed timing model.

Parameters

residuals – original pulsar residuals from Pulsar object
t2pulsar – libstempo pulsar object
tmparams_orig – dictionary of TM parameter tuples, (val, err)
tmparams – new timing model parameters, rescaled to be in sigmas
which – option to have all or only named TM parameters varied

Returns

difference between new and old residuals in seconds