Version 1.3.2 (Dec 10, 2010):
1) Added vozisol, a version of voz* designed to work on an isolated
bunch of points not in a periodic box.
2) Some changes in zobovpostproc.py to work with current versions of
python packages (matplotlib, numpy)
3) Added a void-finding example illustrating vozisol and
zobovpostproc.py to /examples, poisson_vloidsters.bash
4) Incorporated the new qhull version 2010.1
Version 1.3.1 (Apr 19, 2008):
1) jozov.c edited to work for both voids and clusters.
Version 1.3 (Feb 29, 2008):
Version 1.2: (Jan 16, 2008) VOBOZ, ZOBOV, and Qhull are now in a single package. Most of the changes were implemented by Rick Wagner.
Version 1.1.1 (Jan 5, 2008) A bug in the not-entirely-essential vozinit program fixed.
Version 1.1: (Jan 16, 2007) (not well-tested) Many
little changes, including decreased memory waste, and reporting of the
most-bound particles. boz.c also tries unbinding with both the
velocity centroid of the core zone of a halo, and the entire halo. If you want to know more about 1.1, please email me.
Version 1.03: (Jan 16, 2007) In boz.c, fixed
another small bug in the kinetic energy estimation. The bug would slightly
affect spread-out but unmassive small haloes on the edge of a periodic
box.
Version 1.02: (Dec 31, 2004) In
boz.c, fixed a bug in the kinetic energy calculation when
using the grid (i.e. for the largest haloes). The kinetic energy had
been somewhat underestimated, in a nonisotropic way.
Version 1.01: (Feb 22, 2004) Fixed a
couple of minor bugs, made Omega_matter a command-line parameter
instead of hard-coded. Qhull was not included with VOBOZ before v.1.2. You can find the two latest versions of it at qhull2002.1.tgz and qhull-2003.1-src.tgz. Alternatively, you can pick them up at http://www.qhull.org.
By Rick Wagner:
1) voz1b1 now has an MPI version, voz1b1_mpi. The arguments for the parallel version are
2) Three Python scripts are added in the bin directory:
cull_subhalos--remove subhalos (not the particles, just the groups)
writevtkpoints--write a VTK PolyData file of all particle positions
writevtkhalos--write VTK PolyData files of all halo particle positions
(The last two require the Python VTK bindings.)
3) To reduce their memory footprint, vozinit and voz1b1 now read in the position files in chunks. The size is defined in voz.h via N_CHUNK.
4) Made voz1b1 more flexibly handle period boundary conditions, by
duplicating buffer particles when needed. The maximum buffer size
is now (1 - 1/ndiv). This is useful for very small (32^3 or less)
particle counts.
5) Made the simulation grid size a command line argument to boz.
6) Corrected print statement in voztie.c
7) Corrected order of argument to fread and fwrite in several places.
8) Added memory deallocation to voz1b1.c to help parallel version.
9) Changed boz scale factor in posread to boxsize.
By Mark Neyrinck:
1) Reduced the number of command-line arguments in jozov.c. Now it assumes that the output of voz1b1, etc. is blabla.adj and blabla.vol; jozov simply takes blabla as an arg.
2) Changed output format for ZOBOV's jozov.c. Now all three output
files are in in ASCII format:
.txt: list of voids along with data about them
.zone: particles in each zone
.void: zones in each void, in order of addition to the void. Each
zone-adding event is also now marked with the density contrast
at which it happens
3) Included post-processing Python routines (in bin/zobovpostproc.py)
that implement two strategies for determining edges of large
voids, described in Sect. 2.5 of arXiv:0712.3049. These require
the new .void format.
His changes:
1) VOBOZ and ZOBOV can be either single- or double-precision; they accept the precision settings of Qhull.
2) Corrected particle indexing in boz.c.*
3) Moved qhull_a.h into voz.h.
4) Moved common defines into voz.h.
5) Fixed first array error in readfiles.c:posread*
6) Makefile changes, including clean and install targets.
7) Included some example particle sets.
Mine:
8) voztie.c no longer dies if the volumes measured for a particle differ. Tiny differences now ignored.
9) VOBOZ, ZOBOV combined into a single package
* - these appear to be bugs in v.1.1 or v.1.1.1, but probably not v.1.03.
Also in boz.c, made the scale factor a into a command-line parameter.