TOPICS
Search

Sample Autocorrelation


The sample autocorrelation at lag h of observations x_1, ..., x_n from a time series is

 r_h=(sum_(t=h+1)^(n)(x_t-x^_)(x_(t-h)-x^_))/(sum_(t=1)^(n)(x_t-x^_)^2),

where x^_ is the sample mean. It estimates the autocorrelation between observations separated by h time steps.


See also

Autocorrelation, Box-Pierce Test, Portmanteau Test, Time Series Analysis

Explore with Wolfram|Alpha

References

Hamilton, J. D. Time Series Analysis. Princeton, NJ: Princeton University Press, 1994.

Cite this as:

Weisstein, Eric W. "Sample Autocorrelation." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/SampleAutocorrelation.html

Subject classifications