0
$\begingroup$

What does it mean to compute the autocorrelation between two different time istances in a time series? I don't well understand the concept, which is the sense of measuring between two time istances of the same series? Thanks!

$\endgroup$
0
$\begingroup$

auto-correlation is computed on the n pairs of values k periods apart . For example if we have 60 observations we can compute the auto-correlation of lag 1 using 59 pairs. likewise we can use 48 pairs to compute the relationship between values 12 period apart.

autocorrelation is an ensemble measure based upon a number of relationships for a particular lag (temporal distance)

$\endgroup$
  • $\begingroup$ I don't catch well the idea. For example, in a sample autocorrelation plot, the presence of autocorellation at lag 3, what does tell to us? That there is a correlation between what? Thanks $\endgroup$ – White Noise Nov 26 '18 at 22:43
  • $\begingroup$ That would indicate a correlation between observations that are 3 time units apart. $\endgroup$ – Joey F. Nov 26 '18 at 23:28

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.