r - Fitted Arima Model with NA values -
hey have short question fitting arima model r. have large dataset multiple na values can not remove.
so need figure way model data arima missing values. of aic value figured best model arima 2,0,2 if try fit model result: null
does has idea how can work that?
my code fit model
fitted(final.arima) i tried know work missing values
thats code use can not give dataset use ... daten1<-test
final.aic <- inf final.order <- c(0,0,0) (p in 0:2) (q in 0:2) for(z in 0:1) { current.aic <- aic(arima(daten1, order=c(p, z, q))) if (current.aic < final.aic) { final.aic <- current.aic final.order <- c(p,z, q) final.arima <- arima(daten1, order=final.order) } } final.aic final.order final.arima acf(resid(final.arima)) box.test(resid(final.arima), lag=20, type="ljung-box") ############rmse und r² testen für arima model############### ffa<-fitted(final.arima) daten1.mean<-mean(daten1) r1<-sum((ffa-daten1.mean)^2) r2<-sum((daten1-ffa)^2) rsquared<-r1/(r1+r2) rsquared
Comments
Post a Comment