combined ets-arima-theta forecasts

combined ets, arima, and theta (eat) forecasting

eatf(
  y,
  h = 5,
  level = c(80, 95),
  method = c("E", "A", "T", "EAT"),
  weights = rep(1/3, 3),
  ci = c("E", "A", "T"),
  ...
)

Arguments

y

a univariate time series

h

number of periods for forecasting

level

confidence levels for prediction intervals

method

forecasting method: "E": forecast::ets; "A": forecast::auto.arima; "T": forecast::thetaf; or "EAT" for the combination of the three methods (with weights)

weights

weights of each method, in method 'EAT'

...

additional parameters to be passed to forecast::ets, forecast::auto.arima, forecast::thetaf and forecast::forecast

Value

a list with point forecasts and prediction intervals; returns from forecast::ets, forecast::auto.arima, forecast::theta, only point forecasts for eat

mean point forecasts as a time series

Details

ensemble forecasts obtained from forecast::ets, forecast::auto.arima and forecast::theta

Examples

require(forecast)
#> Loading required package: forecast
#> 
#> Attaching package: ‘forecast’
#> The following object is masked from ‘package:after’:
#> 
#>     is.constant

print(eatf(WWWusage, method = "EAT",
weights = c(0.5, 0, 0.5)))
#>     Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
#> 101       219.3380 213.8856 222.8471 211.5136 225.2190
#> 102       218.8270 207.7611 226.3090 202.8518 231.2183
#> 103       218.4393 201.4969 230.4034 193.8458 238.0544
#> 104       218.1519 195.2867 234.8452 184.8162 245.3157
#> 105       217.9463 189.2333 239.4575 175.9398 252.7510

print(eatf(WWWusage, method = "EAT"))
#>     Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
#> 101       219.1855 213.8856 222.8471 211.5136 225.2190
#> 102       218.6022 207.7611 226.3090 202.8518 231.2183
#> 103       218.1858 201.4969 230.4034 193.8458 238.0544
#> 104       217.8916 195.2867 234.8452 184.8162 245.3157
#> 105       217.6877 189.2333 239.4575 175.9398 252.7510

print(eatf(WWWusage, method = "T"))
#>     Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
#> 101       220.3096 212.8385 227.7808 208.8835 231.7357
#> 102       220.6190 210.0538 231.1843 204.4609 236.7772
#> 103       220.9284 207.9889 233.8679 201.1392 240.7177
#> 104       221.2378 206.2967 236.1789 198.3873 244.0883
#> 105       221.5472 204.8426 238.2518 195.9997 247.0947


obj <- eatf(WWWusage, method = "EAT",
weights = c(0, 0.5, 0.5),
ci = "T")
plot(obj)



obj <- eatf(WWWusage, method = "EAT",
weights = c(0, 0.5, 0.5))
plot(obj)



par(mfrow=c(3, 2))
plot(eatf(USAccDeaths, h=10, level=c(80, 90, 95, 99)))
plot(eatf(AirPassengers, h=10, level=c(80, 90, 95, 99)))
plot(eatf(lynx, h=10, level=c(80, 90, 95, 99)))
plot(eatf(WWWusage, h=10, level=c(80, 90, 95, 99)))
plot(eatf(Nile, h=10, level=c(80, 90, 95, 99)))
plot(eatf(fdeaths, h=10, level=c(80, 90, 95, 99)))