feets.libs package

Submodules

feets.libs.ls_fap module

Utilities for computing periodogram statistics.

feets.libs.ls_fap.cdf_single(z, N, normalization, dH=1, dK=3)[source]

Cumulative distribution for the Lomb-Scargle periodogram

Compute the expected cumulative distribution of the periodogram for the null hypothesis - i.e. data consisting of Gaussian noise.

Parameters:

z : array-like

the periodogram value

N : int

the number of data points from which the periodogram was computed

normalization : string

The periodogram normalization. Must be one of [‘standard’, ‘model’, ‘log’, ‘psd’]

dH, dK : integers (optional)

The number of parameters in the null hypothesis and the model
Returns:

cdf : np.ndarray

The expected cumulative distribution function

Notes

For normalization=’psd’, the distribution can only be computed for periodograms constructed with errors specified. All expressions used here are adapted from Table 1 of Baluev 2008 [R7981].

References

[R7981](1, 2) Baluev, R.V. MNRAS 385, 1279 (2008)
feets.libs.ls_fap.false_alarm_probability(Z, fmax, t, y, dy, normalization, method='baluev', method_kwds=None)[source]

Approximate the False Alarm Probability

Parameters:TODO
Returns:TODO
feets.libs.ls_fap.fap_baluev(Z, fmax, t, y, dy, normalization='standard')[source]

Alias-free approximation to false alarm probability

(Eqn 6 of Baluev 2008)

feets.libs.ls_fap.fap_bootstrap(Z, fmax, t, y, dy, normalization='standard', n_bootstraps=1000, random_seed=None)[source]
feets.libs.ls_fap.fap_davies(Z, fmax, t, y, dy, normalization='standard')[source]

Davies upper-bound to the false alarm probability

(Eqn 5 of Baluev 2008)

feets.libs.ls_fap.fap_simple(Z, fmax, t, y, dy, normalization='standard')[source]

False Alarm Probability based on estimated number of indep frequencies

feets.libs.ls_fap.fap_single(z, N, normalization, dH=1, dK=3)[source]

Single-frequency false alarm probability for the Lomb-Scargle periodogram

This is equal to 1 - cdf, where cdf is the cumulative distribution. The single-frequency false alarm probability should not be confused with the false alarm probability for the largest peak.

Parameters:

z : array-like

the periodogram value

N : int

the number of data points from which the periodogram was computed

normalization : string

The periodogram normalization. Must be one of [‘standard’, ‘model’, ‘log’, ‘psd’]

dH, dK : integers (optional)

The number of parameters in the null hypothesis and the model
Returns:

fap : np.ndarray

The expected cumulative distribution function

Notes

For normalization=’psd’, the distribution can only be computed for periodograms constructed with errors specified. All expressions used here are adapted from Table 1 of Baluev 2008 [R8082].

References

[R8082](1, 2) Baluev, R.V. MNRAS 385, 1279 (2008)
feets.libs.ls_fap.pdf_single(z, N, normalization, dH=1, dK=3)[source]

Probability density function for Lomb-Scargle periodogram

Compute the expected probability density function of the periodogram for the null hypothesis - i.e. data consisting of Gaussian noise.

Parameters:

z : array-like

the periodogram value

N : int

the number of data points from which the periodogram was computed

normalization : string

The periodogram normalization. Must be one of [‘standard’, ‘model’, ‘log’, ‘psd’]

dH, dK : integers (optional)

The number of parameters in the null hypothesis and the model
Returns:

pdf : np.ndarray

The expected probability density function

Notes

For normalization=’psd’, the distribution can only be computed for periodograms constructed with errors specified. All expressions used here are adapted from Table 1 of Baluev 2008 [R8183].

References

[R8183](1, 2) Baluev, R.V. MNRAS 385, 1279 (2008)
feets.libs.ls_fap.tau_davies(Z, fmax, t, y, dy, normalization='standard', dH=1, dK=3)[source]

tau factor for estimating Davies bound (Baluev 2008, Table 1)

Module contents

External libs