Skip to content
Numerical Method , home
Menu
Search

Forum Numerical Method Statistics

ARIMAForecast failure

  1. demonwavParticipant 2013-02-27 at 10:03 pm

    ARIMAForecast function fails on this data:

    import com.numericalmethod.suanshu.stats.timeseries.datastructure.univariate.realtime.inttime.IntTimeTimeSeries;
import com.numericalmethod.suanshu.stats.timeseries.datastructure.univariate.realtime.inttime.SimpleTimeSeries;
import com.numericalmethod.suanshu.stats.timeseries.linear.univariate.arima.*;

    public class Arimasho {

    /** * @param args */ public static void main(String[] args) { // TODO Auto-generated method stub

    IntTimeTimeSeries xt = new SimpleTimeSeries(new double[]{2.0, 3349.0, 4321.0});
ARIMAForecast ar = new ARIMAForecast(xt,1,1,1,0);
double forecast=ar.next().xHat();

    System.out.println(forecast);

    }

    }

  2. RyuParticipant 2013-02-28 at 1:13 pm

    You need at least 4 points to estimate an ARIMA(1,1,1) model.

  3. demonwavParticipant 2013-03-12 at 3:34 am

    Thank you. Really appreciate your kind reply.

  4. demonwavParticipant 2013-03-12 at 3:37 am

    However, i wish it should be an easy to understand error message, maybe through exception handling. Again many thanks.