Skip to contents

Windowed timeOpt with precession, short-eccentricity and obliquity interpretations

Source: R/win_timeOpt_multi.R
win_timeOpt_multi.Rd

Runs win_timeOpt with fit 1 (precession amplitude modulation), fit 2 (short-eccentricity amplitude modulation) and fit 3 (obliquity amplitude modulation) on the same record and compares their best sedimentation rates window by window.

Usage

win_timeOpt_multi(data = NULL, fits = c(1, 2, 3), ratio_tol = 0.15, ...)

Arguments

data

Record to analyse (matrix or data frame with two columns: depth in metres in column 1, proxy value in column 2), evenly spaced; passed to win_timeOpt. Must be given (the NULL default fails). Default=NULL.

fits

The timeOpt fits to run (numeric vector with any of 1, 2, 3): 1 precession amplitude modulation, 2 short-eccentricity amplitude modulation, 3 obliquity amplitude modulation (the fit argument of win_timeOpt). The agreement and ambiguity flags need the corresponding pairs. Default=c(1, 2, 3).

ratio_tol

Relative tolerance used to compare the best rates of two fits (numeric, > 0; 0.15 counts rates within a factor 1.15 as equal); used for the agree_* and ambiguous_* flags. Default=0.15.

...

Further arguments passed to win_timeOpt for every fit (e.g. window_size, sedmin, sedmax, numsed, n_null); fit is not allowed (use fits).

Value

A list of class win_timeOpt_multi.result with the win_timeOpt result of every fit (fit1, fit2, fit3) and summary, a data frame with per window the best rate and significance of each fit and the flags above.

Details

A record that holds only one cycle family can be read in two ways: obliquity cycles read as precession give a peak at about twice the true rate (fit 1), precession cycles read as obliquity give a peak at about half the true rate (fit 3). For every window the function flags

agree_12, agree_13, agree_23

both fits are significant and their best rates agree within ratio_tol: independent support for the same rate

ambiguous_13

both fit 1 and fit 3 are significant and the fit-1 rate is about twice the fit-3 rate: precession/obliquity 2:1 ambiguity, which needs other evidence (eccentricity cycles, an age tie point) to resolve

ambiguous_23

both fit 2 and fit 3 are significant and the fit-2 rate is about 0.41 times the fit-3 rate: obliquity (41 kyr with its 173-kyr modulation) read as short eccentricity (~100 kyr with the 405-kyr modulation)

Examples

# \donttest{
set.seed(1)
mag_multi <- win_timeOpt_multi(data = mag[1:250, ], fits = c(1, 2), window_size = 6,
    sedmin = 0.1, sedmax = 1, numsed = 5, n_null = 6, noise_model = "ar1", step = 25)
head(mag_multi$summary)
#>      x best_fit1 z_fit1 sig_fit1 best_fit2     z_fit2 sig_fit2 agree_12
#> 1 0.00        NA     NA    FALSE 0.3162278 -2.2376431    FALSE    FALSE
#> 2 1.25        NA     NA    FALSE 0.1778279 -1.3857398    FALSE    FALSE
#> 3 2.50        NA     NA    FALSE 0.5623413  0.3560039    FALSE    FALSE
#> 4 3.75        NA     NA    FALSE 0.5623413 -0.2735087    FALSE    FALSE
#> 5 5.00        NA     NA    FALSE 0.3162278 -0.3565558    FALSE    FALSE
#> 6 6.25        NA     NA    FALSE 0.5623413 -0.1424578    FALSE    FALSE
# }