canova

Testing dependence/correlation of two variables is one of the fundamental tasks in statistics. We focus on testing nonlinear dependence/correlation of two continuous variables (X and Y). We address this problem by a framework named CANOVA (continuous analysis of variance). In the CANOVA framework, we first define a neighborhood of each data point according to its X value, and then calculate the variance of the Y value within the neighborhood, finally perform a permutation test for the significance of the observed within neighborhood variance. CANOVA is efficient in testing nonlinear correlation and has its own advantages in real data applications.