“…More recently, in applied and computational topology, Vietoris-Rips and Čech complexes have been used to recover the "shape" of a dataset. Indeed, there are theoretical guarantees that if X is a sufficiently nice sample from an unknown underlying space M , then one can recover the homotopy types, homology groups, or approximate persistent homology of M from X [14,15]. In data analysis contexts, instead of letting r be arbitrarily small (as for (co)homology theories), and instead of letting r be sufficiently large (as in geometric group theory), we instead are interested in an intermediate range of scale parameters r. Indeed, if r is smaller than the distance between any two data points in X, then VR(X; r) = X is a disjoint union of points.…”