statsmodels.distributions.copula.api.GaussianCopula.tau¶ GaussianCopula.tau(corr=None)¶ Bivariate kendall’s tau based on correlation coefficient. Parameters: corrNone or floatPearson correlation. If corr is None, then the correlation will be taken from the copula attribute. Returns: Kendall’s tau that corresponds to pearson correlation in the elliptical copula.