“A stochastic process (a collection of random variables ordered in time) is said to be (weakly) stationary if its mean and variance are constant over time. By contrast, a nonstationary time series will have a time-varying mean or a time-varying variance or both.”