slm (version 1.2.0)

cov_select: Covariances Selection

Description

Allows the user to select the lags of the autocovariance terms of the process to be kept.

Usage

cov_select(epsilon, model_selec, plot = FALSE)

Arguments

epsilon

numeric vector. An univariate process.

model_selec

a vector with the positive lags of the selected autocovariance terms. The variance (lag = 0) is automatically selected.

plot

logical. By default, plot = FALSE. If plot = TRUE the ACF of the process is plotted.

Value

This function returns the estimated autocovariance terms.

model_selec

the vector with the positive lag of the selected autocovariance terms.

cov_st

the vector of the selected autocovariances.

Details

In the framework of slm, this is a manual method for estimating the covariance matrix of the error process by only selecting some autocovariance terms from the residual autocovariances.

References

E. Caron, J. Dedecker and B. Michel (2019). Linear regression with stationary errors: the R package slm. arXiv preprint arXiv:1906.06583. https://arxiv.org/abs/1906.06583.

Examples

# NOT RUN {
x = arima.sim(list(ar=c(0.2,0.1,0.25)),1000)
cov_select(x, c(1,3,5))
# }