2
$\begingroup$

I'm new to stats. I want to get clarity about Ljung Box test. Its a omnibus test.

I'm using R to test autocorrelations of a time series. Using below command

Box.test(LogReturns, lag = 10, type = c("Ljung-Box"))

To test autocorrelations upto lag 10, can I infer with above single command or I need to run it in loop from lag = 1 to 10?

R documentation is not clear to me.

Any suggestion?

Thanks

$\endgroup$
2
$\begingroup$

If you want to test the null of no autocorrelation using 10 lags, it is enough to run the command as stated in the question. You should not run it once for every lag.

Edit: To be clear, including 10 lags tests the null hypothesis of no autocorrelation against the alternative that at least one of the included lags exhibits non-zero autocorrelation, which to me sounds like what you want.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.