arXiv · 2611.15672
Nagata dimension, separation modulus and stochastic embeddings into trees
Abstract
The separation modulus of a metric space, introduced by Bartal in 1996, is related to different notions of dimensionality. Most generally, Naor and Silberman gave in 2011 an upper bound on the separation modulus in terms of Nagata dimension. We prove that this bound cannot be reversed; that is, there are metric spaces with arbitrarily high Nagata dimension and stochastically embeddable into trees with uniform distortion, which therefore have uniformly bounded separation modulus. Similarly, we construct spaces with arbitrarily high Nagata dimension whose spaces of probability measures endowed with the Earth Mover's Distance (equivalently, their Lipschitz-free spaces) are embeddable into $L_1$ with uniformly bounded distortion. By gluing these spaces together, we obtain infinite Nagata dimensional metric spaces whose separation modulus is finite and whose Lipschitz-free space is isomorphic to $\ell_1$, respectively. Finally, we provide relations between separation modulus and stochastic embeddability into trees, also known as random approximability by tree metrics, in particular for the case of snowflake metrics.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Ramón J. Aliaga, Rubén Medina. 2026-11-13. Nagata dimension, separation modulus and stochastic embeddings into trees. https://searcharxiv.com/abs/2611.15672
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.