> eq_gclock<-arima(gclock,order=c(2,1,0))
> eq_gclock

Call:
arima(x = gclock, order = c(2, 1, 0))

Coefficients:
         ar1     ar2
      0.3428  0.2975
s.e.  0.1066  0.1126

sigma^2 estimated as 1.602e-22:  log likelihood = 1396.53,  aic = -2787.06
> tsdiag(eq_gclock)
> resid=residuals(eq_gclock)
> Box.test(resid,lag=10,type="Ljung-Box",fitdf=2)
        Box-Ljung test

data:  resid
X-squared = 0.0022396, df = 8, p-value = 1


> predict(eq_gclock,n.ahead=5)
$pred
Time Series:
Start = 61
End = 65
Frequency = 1
[1] 4.664359e-05 4.664359e-05 4.664359e-05 4.664359e-05 4.664359e-05

$se
Time Series:
Start = 61
End = 65
Frequency = 1
[1] 1.265646e-11 2.119008e-11 3.072383e-11 3.982425e-11 4.865927e-11


         请问专家，我使用R语言进行预测的这个程序，哪方面存在错误吗？为何我预测的5个值或者更多的值，结果都是相同的，这样预测还有效果吗。我应该怎么改正啊，比较着急啊，万分感谢大家！