Abstract:SIAM Journal on Computing, Ahead of Print. The random geometric graph model [math] is a distribution over graphs in which the edges capture a latent geometry. To sample [math], we identify each of our [math] vertices with an independently and uniformly sampled vector from the [math]-dimensional unit sphere [math], and we connect pairs of vertices whose vectors are "sufficiently close," such that the marginal probability of an edge is [math]. Because of the underlying geometry, this model is natural for applications in data science and beyond. We investigate the problem of testing for this latent geometry, or, in other words, distinguishing an Erdős–Rényi graph [math] from a random geometric graph [math]. It is not too difficult to show that if [math] while [math] is held fixed, the two distributions become indistinguishable; we wish to understand how fast [math] must grow as a function of [math] for indistinguishability to occur. When [math] for constant [math], we prove that if [math], the total variation distance between the two distributions is close to 0; this improves upon the best previous bound of Brennan, Bresler, and Nagaraj (2020), which required [math], and further our result is nearly tight, resolving a conjecture of Bubeck, Ding, Eldan, and Rácz (2016) up to logarithmic factors. We also obtain improved upper bounds on the statistical indistinguishability thresholds in [math] for the full range of [math] satisfying [math], improving upon the previous bounds by polynomial factors. Our analysis uses the belief propagation algorithm to characterize the distributions of (subsets of) the random vectors conditioned on producing a particular graph. In this sense, our analysis is connected to the "cavity method" from statistical physics. To analyze this process, we rely on novel sharp estimates for the area of the intersection of a random sphere cap with an arbitrary subset of [math], which we prove using optimal transport maps and entropy-transport inequalities on the unit sphere. We believe these techniques may be of independent interest.

Reconstructing a Point Set from a Random Subset of Its Pairwise Distances

Reconstructing almost all of a point set in $\mathbb{R}^d$ from randomly revealed pairwise distances

Revisiting Random Points: Combinatorial Complexity and Algorithms

Random zero sets with local growth guarantees

Transforming the Challenge of Constructing Low-Discrepancy Point Sets into a Permutation Selection Problem

Planar point sets with forbidden $4$-point patterns and few distinct distances

Porous Sets for Mutually Nearest Points in Banach Spaces

Maximum interpoint distance of high-dimensional random vectors

Testing Thresholds for High-Dimensional Sparse Random Geometric Graphs

Missing digits points near manifolds

On integer distance sets

Partial isometries, duality, and determinantal point processes

Reconstructing point patterns from incomplete nonmetric representations

Sample-Efficient Geometry Reconstruction from Euclidean Distances using Non-Convex Optimization

Improved Algorithms for Distance Selection and Related Problems

Distinct Distances Between a Circle and a Generic Set

Sharp density discrepancy for cut and project sets: An approach via lattice point counting

Stabbing Pairwise Intersecting Disks by Four Points

Reconstructing random graphs from distance queries

Reconstructing the Geometry of Random Geometric Graphs

A Note on Interference in Random Point Sets