# StatDiffTSPred: Method to predict according to the stational difference time... In elisa-esteban/TSPred: Point and std prediction of time series

## Description

This method implements the predicted value and their standard deviation according to the stational difference time series model (1-B)^s (1-B)y<sub>t</sub>=a<sub>t</sub>.

## Usage

 ```1 2 3 4 5 6 7 8 9``` ```StatDiffTSPred(x, StatDiff = 12L, forward = 2L, VarNames = NULL) ## S4 method for signature 'vector' StatDiffTSPred(x, StatDiff = 12L, forward = 2L, VarNames = NULL) ## S4 method for signature 'StQList' StatDiffTSPred(x, StatDiff = 12L, forward = 2L, VarNames = NULL) ```

## Arguments

 `x` object upon which the prediction will be made. `StatDiff` stational differences of the time series; by default it is 12L. `forward` integer indicating the number of periods ahead when the prediction will be made; by default it is 2L. `VarNames` character vector with the variable names for which the prediction will be made; by default it is NULL

## Value

It returns a list with components Pred and STD, containing the point prediction and the estimated standard deviations, respectively. Depending on the class of the input parameter x, it returns:

• For input class vector, it returns numeric vectors.

• For input class matrix, it returns matrices.

• For input class StQList, it returns list whose components are data.tables.

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20``` ```# Predicting one and two months ahead in time data(Example1.TS) StatDiffTSPred(Example1.TS, forward = 1L) StatDiffTSPred(Example1.TS, forward = 2L) # Predicting upon a times series with many NA values data(Example2.TS) StatDiffTSPred(Example2.TS, forward = 1L) # On a matrix Mat <- rbind(Example1.TS, Example2.TS) StatDiffTSPred(Mat, forward = 1L) ## Not run: # With an object of class StQList data(StQListExample) VarNames <- c('ActivEcono_35._6._2.1.4._0', 'GeoLoc_35._6._2.1._1.2.5.') StatDiffTSPred(StQListExample, VarNames = VarNames) ## End(Not run) ```

elisa-esteban/TSPred documentation built on Dec. 8, 2018, 9:25 p.m.