StatDiffTSPred: Method to predict according to the stational difference time...

Description Usage Arguments Value Examples

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:

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 July 20, 2018, 5:11 a.m.