Daniel James :
How do I check what the RMSE alone is in ARIMA(p,d,q). If I simulate a time series data with arima.sim like this.
wn <- rnorm(10, mean = 0, sd = 1)
ar <- wn[1:2]
for (i in 3:10){
ar<-arima.sim(n=10,model=list(ar=-0.7048,order=c(1,1,0)),start.innov=4.1,n.start=1,innov=wn)
}
ar <-ar[-1]
I have tried the r code bellow:
mis <- auto.arima(ar)
summary(auto.arima(ar))
mod1 <- auto.arima(ar)
refit <- Arima(ar, model=mod1)
acu<-accuracy(refit)
acu$
I want a function like rmse<-function(mis,...) that will print out just the value of RMSE
user63230 :
All you need is acu[1, 2] but if you want a function:
library(forecast)
set.seed(100)
wn <- rnorm(10, mean = 0, sd = 1)
ar <- wn[1:2]
for (i in 3:10){
ar<-arima.sim(n=10,model=list(ar=-0.7048,order=c(1,1,0)),start.innov=4.1,n.start=1,innov=wn)
}
ar <-ar[-1]
ar
mis <- auto.arima(ar)
acu <- accuracy(mis)
acu
# ME RMSE MAE MPE MAPE MASE ACF1
# Training set -0.06866332 0.5832581 0.510061 -15.2432 52.34 0.2901498 0.5778458
acu[1, 2]
# [1] 0.5832581
rmse_mis_fun <- function(x) {
a <- accuracy(x)
a[1, 2]
}
rmse_mis_fun(mis)
# [1] 0.5832581
being able to pass ar might be a more useful function:
rmse <- function(x) {
m <- auto.arima(x)
acu <- accuracy(m)
acu[1, 2]
}
rmse(ar)
# [1] 0.5832581
Guess you like
Origin http://10.200.1.11:23101/article/api/json?id=391082&siteId=1