search_joint_periodicity

ell1fit.spin_periodicity.search_joint_periodicity(t, ys, variances, min_period, max_period, oversample=5, n_shuffle=1000, fap_level=0.01, rng=None)

One period shared by several quantities measured at the same epochs.

Each quantity keeps its own mean, amplitude and phase; the joint power is the mean of their periodogram powers. The shuffles move all of an epoch’s values together, so a correlation between the quantities at each epoch – which makes their periodogram peaks coincide whether or not anything is periodic – is part of the null hypothesis, not mistaken for a signal.

Parameters:

tarray-like

Times, in days, shared by all quantities.

ys, varianceslist of array-like

Values and variances (None for equal weights) of each quantity.

Returns:

dict

As search_periodicity() (without the detectable amplitude), plus best_amplitudes and phases_of_maximum (cycles from t = 0) per quantity, phase_lags of each quantity’s maximum after the first one’s, in cycles within [-0.5, 0.5), and the arrays powers and best_coeffs per quantity.