如何制作每个序列的循环？

Question

我尝试使用循环for()来预测我使用ARIMA的所有数据，但是我得到了一个错误。

请参见以下脚本：

library(forecast)

Exval<-c("CLI", "SPY", "USDEUR", "CPI", "CONSCONF", "FEDFUNDS", "HOUST", "UNRATE", "INDPRO", "VIX")
fcst<-c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)

for(i in Exval)
  {for (y in fcst){
    fit_arima<-auto.arima(DATA$i,d=1,D=1,stepwise = FALSE,approximation = FALSE, trace = TRUE)
    y<-forecast(fit_arima,h=36)
    print(summary(y))}}

Error in ts(x) : 'ts' object must have one or more observations

在我看来，错误来自于i。循环立即执行next i，而许多观察必须在相同的I上进行。例如，当我不使用循环时，结果是：

> fit_arima<-auto.arima(DATA$VIX,d=1,D=1,stepwise = FALSE,approximation = FALSE, trace = TRUE)

 ARIMA(0,1,0)                    : 2017.494
 ARIMA(0,1,0) with drift         : 2019.493
 ARIMA(0,1,1)                    : 2014.853
 ARIMA(0,1,1) with drift         : 2016.861
 ARIMA(0,1,2)                    : 1997.6
 ARIMA(0,1,2) with drift         : 1999.6
 ARIMA(0,1,3)                    : 1991.777
 ARIMA(0,1,3) with drift         : 1993.766
 ARIMA(0,1,4)                    : 1992.797
 ARIMA(0,1,4) with drift         : 1994.786
 ARIMA(0,1,5)                    : 1994.604
 ARIMA(0,1,5) with drift         : 1996.598
 ARIMA(1,1,0)                    : 2016.692
 ARIMA(1,1,0) with drift         : 2018.701
 ARIMA(1,1,1)                    : 1990.448
 ARIMA(1,1,1) with drift         : 1992.359
 ARIMA(1,1,2)                    : 1989.871
 ARIMA(1,1,2) with drift         : 1991.819
 ARIMA(1,1,3)                    : 1991.932
 ARIMA(1,1,3) with drift         : 1993.892
 ARIMA(1,1,4)                    : 1993.856
 ARIMA(1,1,4) with drift         : 1995.819
 ARIMA(2,1,0)                    : 2007.087
 ARIMA(2,1,0) with drift         : 2009.103
 ARIMA(2,1,1)                    : 1989.959
 ARIMA(2,1,1) with drift         : 1991.912
 ARIMA(2,1,2)                    : 1991.932
 ARIMA(2,1,2) with drift         : 1993.894
 ARIMA(2,1,3)                    : Inf
 ARIMA(2,1,3) with drift         : Inf
 ARIMA(3,1,0)                    : 2002.725
 ARIMA(3,1,0) with drift         : 2004.75
 ARIMA(3,1,1)                    : 1991.889
 ARIMA(3,1,1) with drift         : 1993.845
 ARIMA(3,1,2)                    : Inf
 ARIMA(3,1,2) with drift         : Inf
 ARIMA(4,1,0)                    : 2002.643
 ARIMA(4,1,0) with drift         : 2004.677
 ARIMA(4,1,1)                    : 1993.877
 ARIMA(4,1,1) with drift         : 1995.839
 ARIMA(5,1,0)                    : 2002.835
 ARIMA(5,1,0) with drift         : 2004.879



 Best model: ARIMA(1,1,2)

然后，我想问你们，我怎样才能做一个关于观察序列的循环。

提前谢谢你。

亚历克斯

Answer 1

不能使用DATA$i引用数据帧的列，其中i是字符串。相反，使用DATA[[i]]作为向量提取列。