Hodrick Prescott Decomposition and ARIMA Forecast

1) Hodrick Prescott Filter for Ethiopian GDP

2) ARIMA - FORECASTING

The pictures below show the different ARIMA models in order to choose the best fit that could support our argument.

fit1 <- Arima(GDP_4, order=c(0,1,3), seasonal=c(0,0,0))
fit1 %>% forecast(h=12) %>% autoplot()

fit2 <- Arima(GDP_4, order=c(0,1,2), seasonal=c(0,0,0))
  #checkresiduals(fit2)
    fit2 %>% forecast(h=12) %>% autoplot()

fit3 <- Arima(GDP_4, order=c(1,1,0), seasonal=c(0,0,0))
    fit3 %>% forecast(h=12) %>% autoplot()   

fit4 <- Arima(GDP_4, order=c(1,1,2), seasonal=c(0,0,0))
    fit4 %>% forecast(h=12) %>% autoplot() 

fit5 <- Arima(GDP_4, order=c(2,1,2), seasonal=c(0,0,0))
    fit5 %>% forecast(h=12) %>% autoplot() 

fit6 <- Arima(GDP_4, order=c(1,1,1), seasonal=c(0,0,0))
    fit6 %>% forecast(h=12) %>% autoplot()