Introduction
Variance describes one variable. Covariance describes how two variables move together: do they rise and fall in sync, in opposition, or independently? Collected for many variables, covariances form the covariance matrix — the central object of PCA, Gaussian models, and the Mahalanobis distance.