Comparing Time Series: A Univariate ApproachMany studies require a method for deciding if two time series come from the same stochastic process. Our approach here is to fit two time series to separate autoregressive (AR) models, and then test if the parameters of the AR model are equal. For Gaussian distributions, the test can be decomposed into two independent tests, one for testing equality of noise variance, and another for testing equality of AR coefficients. On this page, we provide links to our paper and numerical codes of these new methods. If you use these algorithms and find bugs or improvements, then please let us know. Main references
Software
|