LG_approx_extract: Extract data needed for Local Gaussian Spectral Densities.

Description Usage Arguments Value

View source: R/LG_approx_extract.R


The estimated Local Gaussian Approximations, created by LG_approx and LG_boot_approx, include a lot of information that we don't need in the computation of the Local Gaussian Spectral Densities. In order to keep the code later on as simple as possible, this function takes care of the extraction of the relevant values, i.e. the Local Gaussian Correlations.


LG_approx_extract(approx_data, LG_type = c("par_one", "par_five"),
  .method = c("all", "mean", "median"))



A list with Local Gaussian Approximations, which in general contains more information than needed for the computation of the Local Gaussian Spectral Densities.


One of c("par_five", "par_one"), i.e. should the Local Gaussian Autocorrelations be based on the approach using five parameters or the simplified approach using only one parameter. Note that even though approx_data contains information about two cases, the code later on will only work upon one case at the time (due to memory restrictions) and it's thus necessary to specify the part of interest.


One or more of the values in c("all", "mean", "median"), but the argument will only have an effect if the length of the content-dimension in the LG_type part of approx_data is larger than one. The idea is that this argument will be used with the values "mean" and "median" while investigating blocks, and the value "all" when doing a bootstrap-investigation.


A list with two components, a list info and an array pos_lag. The list info contains stuff related to reproducibility, whereas the array pos_lag contains the required Local Gaussian Autocorrelations.

LAJordanger/localgaussSpec documentation built on July 28, 2017, 12:15 a.m.