innovation: Innovation Algorithm

In adrian1econ/TimeSeries: What the Package Does (One Line, Title Case)

Description

Description Details

Usage

innovation(ts, lag.max = NA)

Arguments

ts	A numeric vector containing a time series or an object of class "arma".
lag.ma	Number of recursions to determined Prediction.
lag.na	Number of recursions to determined prediction.

Value

Numeric vector containing the Prediction determined by the Innovation algorithm. The innovation algorithm is algorithm, which determinate thetas and mean squared errors.They are used to predict the next values in the function innovation_prediction. General the innovation algorithm is good for Ma-processes

Details The innovation algorithm is algorithm, which determinate from previous elements and their predictors the predictor of the (n-th) next element. Therefore the algorithm need coefficients (Thetas), which are determinate in this algorithm. They are determinate with help from the previous Thetas, the ACF-function and mean squared errors (between predictor and the elements). The mean square errors are also calculated in this algorithm.

A list with the Matrix with the calculated Thetas and a vector with a mean squared error

Examples

innovation(arma_sim(theta = c(0.8,-0.3),n = 1000,burnin = 1000))

adrian1econ/TimeSeries documentation built on Aug. 25, 2020, 5:18 p.m.