Question

I have a time series data:

output of x data frame:

Date  vol
1990  12
1991  13
1992  15
1994  18
1995  20
1996  35

I am trying to plot this data and predict 4 year ahead as below:

plot(x$Date, x$vol, col="blue")
x.lm<-lm(x$Vol ~ x$Date)
x.pre<-predict(x.lm, n.ahead=4)
abline(x.pre, col="red")

I get this error:

Error in int_abline(a = a, b = b, h = h, v = v, untf = untf, ...) : 
  invalid a=, b= specification

can anybody tell me what I am doing wrong?

Answer 1

You capitalized “Vol” in x.lm<-lm(x$Vol ~ x$Date).

Try

x.lm<-lm(x$vol ~ x$Date)

Also, you’re not going to be able to pass the value of predict() into abline like that without some extra modification. Since you’re not doing any sophisticated prediction, but really just wanting to plot the linear fit, you could plot the line using

abline(x.lm, col="red")

If you separate your variables and then use the newdata parameter in predict, you should be able to get the actual predictions. That would probably be the cleanest way to do it. For instance:

y <- x$vol
x <- x$Date
x.lm <- lm(y~x)
predict(x.lm, data.frame(x=1997:2000))

This will return the predictions for 1997 – 2000. I believe the “x” in data.frame(x=1997:2000) must match the variable you put into lm() as your x variable. In your case, that has a dollar sign accessor, which makes the whole thing a bit more complex. I’d just take the approach above and rename the x component fed into your lm() function as a valid variable name which can be referenced later.