MNE: Modified natural estimates

In gajdosandrej/fdslrm: R package for modeling and prediction of time series using linear mixed models

Description

NE(X, F, V) calculates modified natural estimates (MNE) of variance parameters in general FDSLRM. MNE are an unbiased version of NE.

Usage

MNE(X, F, V)

Arguments

X	time series realization.
F	design matrix for fixed effects.
V	design matrix for random effects.

Value

Modified natural estimates of variance parameters σ^2, σ_1^2, ..., σ_l^2.

Note

Ver.: 21-Jan-2019 20:48:29.

Examples

## EXAMPLE 1
times <- c(1:40)
freq_fixed <- c(1/24)
freq_rand <- c(1/8, 1/6)
matF <- makeF(times, freq_fixed)
matV <- makeV(times, freq_rand)
X <- rnorm(40, 2, 1.25)
mne <- MNE(X, matF, matV)
print(mne)

gajdosandrej/fdslrm documentation built on April 28, 2020, 11:35 a.m.