ARIMA double seasonality with dummy in R error xreg

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

I'm playing with hourly univariate data and try to fit an arima model with more than one seasonality (daily, weekly) using a dummy for the weekly seasonality. I found a very good post explaining a similar problem but after many trials I'm running out of ideas with the following error:

1. Error in optim(init[mask], armaCSS, method = optim.method, hessian = FALSE,  :   non-finite value supplied by optim

复制代码

My code is the following:

1. library(lubridate)
   
2. start=dmy_hms("25/02/2011 00:00:00")
   
3. index=start + c(0:1000) * hours(1)
   
4. week = wday(index)
   
5. xreg_w = model.matrix(~0+as.factor(week))
   
6. colnames(xreg_w) = c("Mon","Tue","Wed","Thu","Fri","Sat","Sun")
   
7. 
   
8. freq=24
   
9. set.seed(1234)
   
10. y=ts(log(35+10*rnorm(1001)),f=freq)
    
11. 
    
12. library(forecast)
    
13. model = Arima(y, order=c(0,0,0), seasonal=list(order=c(1,0,0), period=24),
    
14.               xreg=xreg_w)

复制代码

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：seasonal double Dummy Eason Error double error

Your xreg_w matrix contains dummy variables for all seven days, but you only need six days due to the inclusion of a constant in your ARIMA model. Thus, the design matrix is not of full rank. Either leave the constant out, or use only six days.

This will work:

1. model <- Arima(y, order=c(0,0,0), seasonal=list(order=c(1,0,0), period=24),
   
2.               xreg=xreg_w, include.mean=FALSE)
   
3. Or this (using only six days):
   
4. 
   
5. xreg_w <- seasonaldummy(ts(y,f=7))
   
6. model <- Arima(y, order=c(0,0,0), seasonal=list(order=c(1,0,0), period=24),
   
7.               xreg=xreg_w)

复制代码