mirror/MUSIC
1
0
Fork 0
mirror of https://github.com/cosmo-sims/MUSIC.git synced 2024-09-19 17:03:46 +02:00
Code Releases Activity
162 commits 1 branch 1 tag 2.4 MiB
Commit graph

162 commits

This branch
This branch
All branches
Author SHA1 Message Date
Oliver Hahn
2236b9f29b fixed some code causing compiler warnings 2012-11-20 16:02:32 +01:00
Oliver Hahn
cd902e787b added CIC deconvolution step to hybrid Poisson solver. This is necessary when glass or resampling is employed via CIC particle interpolation 2012-11-20 16:01:45 +01:00
Oliver Hahn
4e1f4224fa more realistic calculation of multigrid error, no more 'no convergence' when using single precision 2012-11-20 16:00:53 +01:00
Oliver Hahn
0807dd6861 fixed a bug that required either ref_extent or ref_dims to be present even when setting up a unigrid simulation 2012-10-08 15:26:59 +02:00
Oliver Hahn
d9fbe83bbc added new parameter [setup]/ref_dims=NX,NY,NZ to explicitly specify the resoluti 
on of the high resolution region.
added output of comoving size of high-res region to screen
added output of densest point also when using the generic output plugin
 2012-09-28 15:17:25 +02:00
Oliver Hahn
449dbfc431 fixed another bug with the hash-key random numbers, when coarsening of noise is needed 2012-09-10 16:26:33 +02:00
Oliver Hahn
6e8234c49e fixed a bug that caused the new hash-key based random number scheme to not work with noise read from file 2012-09-10 15:02:16 +02:00
Oliver Hahn
5589695ef6 updated code version to 1.1b, first version with significantly reduced memory usage for very deep zooms. 2012-09-07 16:53:43 +02:00
Oliver Hahn
aff77c90dc added hash-key based generation of random number subcubes, based on need 
fixed a few minor bugs that could lead to crashes for small refinement regions close to the boundary
 2012-09-07 16:39:53 +02:00
Oliver Hahn
6db9e358fc fixed a bug that white noise was not read from file if levelmin_TF or levelmin was lower than the level with the white noise from file 2012-09-06 15:32:38 +02:00
Oliver Hahn
22ab605e1e can now read white noise file with 64bit Fortran unformatted block size indicators 2012-09-06 15:24:28 +02:00
Oliver Hahn
a51296c172 added new parameter [random]/grafic_sign = yes|no, (default=no) 
to set the sign of the input white noise file consistent with grafic2
 2012-09-04 15:19:16 +02:00
Oliver Hahn
9275aa4529 added output of density extrema to ENZO plugin 2012-08-24 12:11:17 -07:00
Oliver Hahn
8040f40104 Gadget-2 plugin: added new parameters for length and mass units, allowed gadget coarse particle type to be specified in parameter file. Fixed bug that H0 instead of h was written to Gadget header. 2012-08-19 13:38:49 -07:00
Oliver Hahn
e52e6ffd0b updated sample parameter file to features of current version of MUSIC 2012-08-18 16:27:57 -07:00
Oliver Hahn
bc7d5de346 added baryon support to tipsy output plugin 2012-07-23 12:17:24 -07:00
Oliver Hahn
057e35eb6d Added tag UCSC_comparison_project for changeset 72c7e4e35665 2012-07-11 17:19:48 -07:00
Oliver Hahn
3802e03da2 merge 2012-07-11 17:18:20 -07:00
Oliver Hahn
52c7d83f88 Added tag UCSC_comparison_project for changeset ff41c20c58ba 2012-07-11 17:15:59 -07:00
Oliver Hahn
f10f8f9a6c version freeze for Santa Cruz Code Comparison Project 2012-07-11 17:15:45 -07:00
Oliver Hahn
0754897c60 added possibility to restart using previously generated white noise fields 
via option [random]/restart = yes/no
if set to yes, it will try to load the data, but all files have to exist,
and the refinement hierarchy cannot be changed.
 2012-03-28 14:55:16 -07:00
Oliver Hahn
a7535f3d89 fixed a bug that caused the refinement region to be always larger than specified, even 
if it would fit exactly. THIS SHOULD BE TESTED MORE THOROUGHLY.
 2012-03-28 13:56:34 -07:00
Oliver Hahn
19cc12a603 moved all omp function calls into #ifndef SINGLETHREAD_FFTW so that the code also compiles on non-openmp compilers 2012-03-28 12:38:31 -07:00
Oliver Hahn
0d6d79e6bc merge 2011-09-19 18:53:48 -07:00
Oliver Hahn
924092d96c Fixed a missing velocity unit factor when using "linger++" transfer functions, which 
have different velocity units than the other transfer functions, and 2LPT.
 2011-09-19 18:49:43 -07:00
Oliver Hahn
db54d01b2a Re-enabled the possibility to apply a deconvolution filter when using CIC interpolation to particles positions 
(e.g. for a glass) by setting [output]/glass_cicdeconvolve=yes/no, default is off
 2011-09-16 17:19:23 -07:00
Oliver Hahn
3c12876d5e Added a temporary workaround for the 'music' transfer function plugin that 
catches cases where the transfer functions become negative at very large k.
 2011-09-07 22:28:39 -07:00
Oliver Hahn
06d1f88aa2 Gadget plugin: Fixed an overflow bug when using more than 2048**3 particles and baryons. 2011-09-06 17:08:36 -07:00
Oliver Hahn
cfc16f509d Added new option cosmology/z0norm to 'linger++' plugin. This allows to choose 
the normalization at z=0 or the linearly scaled from z_ini.
 2011-09-06 16:39:41 -07:00
Oliver Hahn
e7f8c2e1ec Added a first DM only version of a tipsy output plugin. 
Credits go to Kyle Stewart and Shea Garrison-Kimmel.
 2011-09-02 15:21:55 -07:00
Oliver Hahn
eb6dbfd317 Code now writes out input power spectra "input_powerspec_{cdm,baryon,total}" rather than 
the "transfer_k.dat" file which was confusing as it did NOT contain the k-space transfer
function.
 2011-08-29 18:38:38 -07:00
Oliver Hahn
826071634a updated version to 1.01 2011-08-29 16:12:40 -07:00
Oliver Hahn
e6137dcf07 Fixed a bug that prevented the 'MUSIC' transfer function plugin to be used when 
also the 'LINGER++' one was present (they used the same class name, so which one
was actually used depended on the compile order rather than the user choice).
 2011-08-29 16:12:14 -07:00
Oliver Hahn
29822726ec Removed option "shift_back" for Gadget2 output as it can cause problems with zoom sims overlapping the periodic boundaries. 2011-08-29 14:00:47 -07:00
Oliver Hahn
8ee0772b81 Removed defaut value of Omega_radiation of 8.3e-5 and replaced with zero. 
Nonzero Omega_r leads to small differences in transfer function normalisation
when used with zero Omega_r in the simulation code, dependent on the
starting redshift.
 2011-08-29 12:25:23 -07:00
Oliver Hahn
e5f663ab0a Added a missing pragma that leads to errors when compiling for single-threaded FFTW3 2011-07-21 10:43:00 -07:00
Oliver Hahn
ebd300f2fa Some more bugfixes for Gadget-2 output with very large grids. 
Verified to work with 2048**3.
 2011-06-09 12:08:24 -07:00
Oliver Hahn
6bc18e1aaf More 64 bit compliance fixes for output plugins. 
Gadget-2 output plugin now can write multiple output files.
Saved a bit of memory in main loop
 2011-06-08 20:58:42 -07:00
Oliver Hahn
2ab09dd484 Cleaned up duplicate code for FFTW2/FFTW3 using some macros 2011-06-03 19:02:56 -07:00
Oliver Hahn
03fb47cc3d Added support for very large arrays to enzo output plugin. 
Enzo plugin now writes in slabs rather than in one monoloithic block (saving memory).
 2011-06-03 18:41:39 -07:00
Oliver Hahn
00fd30f333 Added tag release_candidate for changeset 2d8c8adfe858 2011-06-03 07:20:06 -07:00
Oliver Hahn
3d02bca2b9 fixed an overflow in the FFT poisson solver normalization when using very large unigrids 2011-06-03 07:19:52 -07:00
Oliver Hahn
f397ff096e Added tag release_candidate for changeset 6072084560e1 2011-06-02 15:24:48 -07:00
Oliver Hahn
0e180443d1 fixed a forgotten test flag in the code from the commit 3 minutes ago 2011-06-02 15:24:41 -07:00
Oliver Hahn
fcda6360e9 Added tag release_candidate for changeset 58af2cbe51f3 2011-06-02 15:23:17 -07:00
Oliver Hahn
1044b448de branch merge with development branch brings in 
- improved 64bit compliance
- support for FFTW3 single and double precision
- support for new transfer function formation from linger++
 2011-06-02 15:20:13 -07:00
Oliver Hahn
0f012ee9f9 a 64bit compliance fix in convolution_kernel 
fixed a bug introduced in this branch in random.cc
 2011-06-02 15:10:05 -07:00
Oliver Hahn
e470be5da7 added support for new linger++ transfer function file format 
cleaned up a bit of code
 2011-06-02 11:12:44 -07:00
Oliver Hahn
5181ca7b60 removed meshvar_bnd_double 2011-06-02 10:50:00 -07:00
Oliver Hahn
d6da7b16fc More 64 bit compliance changes. 
New convergence criterium for multi-grid (No longer reduction of initial residual by X, but rel. error of X).
 2011-06-01 18:05:12 -07:00