Significance of the lines of a (de-chirped) EHA map against warped noise

The de-chirped map of analyse_chirp_eha warps the noise inside every window as well as the cycles, so wavy structure in it is not by itself evidence. This function runs the same analysis, with the same windows, gradients (or steer) and settings, on AR1 surrogates of the record and compares the observed line-to-continuum ratios with those of the surrogates cell by cell. Lines that are only artefacts of the warping appear in the surrogates as well and are not significant.

Usage

chirp_eha_significance(ce, data, n_sim = 20, rho = NULL, seed = 1)

Arguments

ce

Result of analyse_chirp_eha (list of class analyze.wavelet with its args element); its settings are reused for the surrogates. Required, no default.

data

The record ce was computed from: a matrix or data frame with depth in column 1 and the proxy value in column 2. Required, no default.

n_sim

Number of AR(1) surrogates (integer, > 0). More surrogates give more precise p-values and 95th percentiles but take longer (each one repeats the full analysis). Default=20.

rho

Lag-1 autocorrelation of the AR(1) surrogates (numeric, between -1 and 1); NULL estimates it from the record. Default=NULL.

seed

Seed of the random number generator for the surrogates (integer), so that the result can be reproduced. Default=1.

Value

ce with Power_sig (observed Power_norm divided by the 95th percentile of the surrogates: above 1 is significant at that level), Power_pval (fraction of surrogates at or above the observed value) and surrogate_q95.