Daniel Eisenstein & Piet Hut
The HOP algorithm is described in our paper to appear in the Astrophysical Journal (May 1, 1998 issue). It's available here as Postscript (319k), gzipped-Postscript (78k), and PDF (218k) files, as well as from astro-ph.
Abstract: We describe a new method (HOP) for identifying groups of particles in N-body simulations. Having assigned to every particle an estimate of its local density, we associate each particle with the densest of the Nh particles nearest to it. Repeating this process allows us to trace a path, within the particle set itself, from each particle in the direction of increasing density. The path ends when it reaches a particle that is its own densest neighbor; all particles reaching the same such particle are identified as a group. Combined with an adaptive smoothing kernel for finding the densities, this method is spatially adaptive, coordinate-free, and numerically straight-forward. One can proceed to process the output by truncating groups at a particular density contour and combining groups that share a (possibly different) density contour. While the resulting algorithm has several user-chosen parameters, we show that the results are insensitive to most of these, the exception being the outer density cutoff of the groups.
Our implementation of HOP is currently at version 1.1 (released 4/2/2002). This version fixes one bug in the code regroup.c that could crash the program when reading large density files. No results could be corrupted, as the program would simply crash.
HOP Source Code -- tar file, with gzip (48k) or without (185k).
HOP On-line Documentation.
Please contact me with any questions or problems.