Autocorrelation Calculator

About this calculator

Autocorrelation measures how correlated a time series is with a delayed copy of itself, at a chosen lag k — computed as the covariance between the series and its k-step-shifted version, normalized by the series' own variance. A high autocorrelation at lag 1 means consecutive values tend to move together (momentum), while a strong autocorrelation at a longer lag can reveal seasonality or a repeating cycle in the data.

This calculator takes a numeric series and a lag value, then computes the sample autocorrelation at that lag. Data scientists and econometricians identifying ARIMA model orders through the Box-Jenkins method, signal processing engineers checking for periodicity, and analysts screening time series for hidden patterns use it as a fundamental diagnostic step before building any forecasting model, since it reveals the dependence structure the model needs to capture.