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

Description Usage Arguments Value Examples

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>.

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)

`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

It returns a data.table with components Pred and STD, containing the point prediction and the estimated standard deviations, respectively, for each variable.

# 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)