time series - cross-correlation using ccf in R -
I want to use CCF in R to calculate cross-correlation on two sets of time-series data. The question is how can I know that the dash falls out of the blue lines without having to manually see any correlation coefficient in the plot? Since I have thousands of sets of time-series data, to deal with it thank you in advanced!
Here's how to calculate confidence interval:
res & Lt; - ccf (mdeaths, fdeaths, ylab = "cross-correlation") upper ci & lt; - qnorm ((1 + 0.95) / 2) / sqrt (res $ n.used) lower CI & lt; However, help (plot.acf) warning: (1 + 0.95) / 2) / sqrt (res $ n. Used) The confidence interval plotted in plot.acf is based on an unorganized chain and should be treated with due care. Using Ci.type = "ma" may be misleading.
type "ma" to see getAnywhere (plot.acf) .
Comments
Post a Comment