Central and \(L^p\)-concentration of 1-Lipschitz maps into \(R\)-trees (Q1030019): Difference between revisions

Let \(\{(X_n, d_{X_n}, \mu_{X_n})\}_{n=1}^\infty\) be a sequence of metric measure spaces and \(\{(Y_n, d_{Y_n})\}_{n=1}^\infty\) be a sequence of metric spaces. We say that \(1\)-Lipschitz maps from \(\{X_n\}_n\) to \(\{Y_n\}_n\) have a concentration property if, for an arbitrary sequence \(\{f_n\}_{n=1}^\infty\) of \(1\)-Lipschitz maps \(f_n:X_n\to Y_n\), there exist points \(m_{f_n}\in Y_n\) such that \[ \mu_{X_n}(\{x\in X_n: d_{Y_n}(f_n(x),m_{f_n})\geq\varepsilon\})\to 0\text{ as }n\to \infty \] for any \(\varepsilon > 0\). The main result of the paper is that the concentration property of \(1\)-Lipschitz maps from \(\{X_n\}\) to \(\{Y_n\}\) with each \(Y_n\) being the real line is equivalent to the concentration property of \(1\)-Lipschitz maps from \(\{X_n\}\) into \(\{Y_n\}\), with each \(Y_n\) being an \(\mathbb{R}\)-tree. This result is proved in terms of the observable diameter. Similar results are proved for the observable central radius and the observable \(L_p\)-variation.