[R] Confidence intervals for ccf()

Gustaf Rydevik gustaf.rydevik at gmail.com
Mon Aug 27 13:44:43 CEST 2007


Hello,

This is not a purely R-question, but perhaps someone can help me anyway.

I am trying to estimate the correlation between two time series (which
are both basically different types of  measurements of the same
phenomena), using both cor.test() (with pearson as method) and ccf().

Now, cor.test gives a confidence interval for the pearson correlation,
while ccf does not. I've tried to use bootstrap methods to get
confidence interval for the ccf function, but no luck. It is a bit
tricky, since the time series are non-stationary, and so I'm not sure
how to go about to generate the bootstrap-sample.

Does anyone have any ideas on how to do this, i.e get a confidence
interval for the ccf at different time lags?

Many thanks in advance,

Gustaf

-- 
Gustaf Rydevik, M.Sci.
tel: +46(0)703 051 451
address:Essingetorget 40,112 66 Stockholm, SE
skype:gustaf_rydevik



More information about the R-help mailing list