enterprise_extensions.frequentist package

Submodules

enterprise_extensions.frequentist.F_statistic module

class enterprise_extensions.frequentist.F_statistic.FpStat(psrs, params=None, psrTerm=True, bayesephem=True, pta=None, tnequad=False)

Bases: object

Class for the Fp-statistic.

Parameters

• psrs – List of enterprise Pulsar instances.

• noisedict – Dictionary of white noise parameter values. Default=None

• psrTerm – Include the pulsar term in the CW signal model. Default=True

• bayesephem – Include BayesEphem model. Default=True

compute_Fp(fgw)

Computes the Fp-statistic.

Parameters

fgw – GW frequency

Returns

fstat: value of the Fp-statistic at the given frequency

compute_fap(fgw)

Compute false alarm rate for Fp-Statistic. We calculate the log of the FAP and then exponentiate it in order to avoid numerical precision problems

Parameters

fgw – GW frequency

Returns

False alarm probability as defined in Eq (64) of Ellis, Seiemens, Creighton (2012)

get_Nmats()

Makes the Nmatrix used in the fstatistic

enterprise_extensions.frequentist.F_statistic.innerProduct_rr(x, y, Nmat, Tmat, Sigma, TNx=None, TNy=None)

Compute inner product using rank-reduced approximations for red noise/jitter Compute: x^T N^{-1} y - x^T N^{-1} T Sigma^{-1} T^T N^{-1} y

Parameters

• x – vector timeseries 1

• y – vector timeseries 2

• Nmat – white noise matrix

• Tmat – Modified design matrix including red noise/jitter

• Sigma – Sigma matrix (varphi^{-1} + T^T N^{-1} T)

• TNx – T^T N^{-1} x precomputed

• TNy – T^T N^{-1} y precomputed

Returns

inner product (x|y)

enterprise_extensions.frequentist.F_statistic.make_Nmat(phiinv, TNT, Nvec, T)

enterprise_extensions.frequentist.Fe_statistic module

class enterprise_extensions.frequentist.Fe_statistic.FeStat(psrs, params=None)

Bases: object

Class for the Fe-statistic.

Parameters

• psrs – List of enterprise Pulsar instances.

• params – Dictionary of noise parameters.

compute_Fe(f0, gw_skyloc, brave=False, maximized_parameters=False)

Computes the Fe-statistic (see Ellis, Siemens, Creighton 2012).

Parameters

• f0 – GW frequency

• gw_skyloc – 2x{number of sky locations} array containing [theta, phi] for each queried sky location, where theta=pi/2-DEC, phi=RA, for singlge sky location use gw_skyloc= np.array([[theta,],[phi,]])

• brave – Skip sanity checks in linalg for speedup if True.

• maximized_parameters – Calculate maximized extrinsic parameters if True.

Returns

fstat: value of the Fe-statistic

If maximized_parameters=True also returns

inc_max: Maximized value of inclination psi_max: Maximized value of polarization angle phase0_max: Maximized value of initial fhase h_max: Maximized value of amplitude

get_Nmats()

Makes the Nmatrix used in the fstatistic

enterprise_extensions.frequentist.Fe_statistic.innerProduct_rr(x, y, Nmat, Tmat, Sigma, TNx=None, TNy=None, brave=False)

Compute inner product using rank-reduced approximations for red noise/jitter Compute: x^T N^{-1} y - x^T N^{-1} T Sigma^{-1} T^T N^{-1} y

Parameters

• x – vector timeseries 1

• y – vector timeseries 2

• Nmat – white noise matrix

• Tmat – Modified design matrix including red noise/jitter

• Sigma – Sigma matrix (varphi^{-1} + T^T N^{-1} T)

• TNx – T^T N^{-1} x precomputed

• TNy – T^T N^{-1} y precomputed

Returns

inner product (x|y)

enterprise_extensions.frequentist.Fe_statistic.make_Nmat(phiinv, TNT, Nvec, T)

enterprise_extensions.frequentist.optimal_statistic module