Commits

Show all
Author Commit Message Labels Comments Date
Michael Kuhlen
Added support for OmegaRadiationNow > 0, which might be important for early (z>100) starting redshifts. This requires a numerical integration of the Friedman equation, which is too slow for on the fly computation. Hence I've introduced a lookup table for t(a), a(t), and da/dt(a) in CosmologyComputeExpansionFactor(), CosmologyComputeTimeFromRedshift(). This lookup table can be turned on using 'gmake use-cosmotable-yes'.
Branches
cosmotable
Michael Kuhlen
(1) Added support for the 1998 version of the Eisenstein & Hu transfer function (1998ApJ...496..605E), which includes baryonic oscillations. (PowerSpectrumType=15). (2) Output the z=0 power spectrum to a file, in addition to the InitialRedshift one. For CMBFAST TF's the two can differ by more than D(a)^2.
Michael Kuhlen
Added support for Pleiades' HDF5 path (requires module load hdf5/1.6.5) instead of John Wise's local version in Make.mach.nasa-pleiades. Also added -xSSE4.1 to MACH_OPT_AGGRESSIVE.
Michael Kuhlen
Added a few '#ifdef TRANSFER' statements, so that the compilation completes even with CONFIG_PHOTON = no.
John Wise
Adding config file for Pleiades.
John Wise
Merging to tip.
Elizabeth Harper-Clark
Removed an empty #define in m_and_p
Elizabeth Harper-Clark
Merged with Tip
Matthew Turk
Moving "ROUNDOFF" to macros_and_parameters and changing the name to [BP]FLOAT_EPSILON. Thanks to Ji-hoon for helping me track down all of these.
Ji-hoon Kim
Working version of RadiativeTransferTraceSpectrum method; currently the test is done only with the MBH particle, and only with the sample table of a monochromatic spectrum. Implemented in a non-invasive way, so it shouldn't affect anybody. More to come as advanced tests are carried out.
Ji-hoon Kim
We now trace the "spectrum" of the photon package, not the energy bin(s) of monochromatic rays. The fraction of photons absorbed in the cell, and the mean energy of the spectrum at the cell is a function of ColumnDensity and dColumnDensity (column density added in this cell). These two values are read in from the pre-made table (spectrum_table.dat) that contains the fraction of photons left in rays, and the mean energy of the spectrum at a given column density. Natur…
Ji-hoon Kim
initial kick-start for tracing spectrum (instead of energy bins) in RT; certainly more machinery to come
Ji-hoon Kim
merged to tip
Matthew Turk
Merging in to tip
Matthew Turk
Merging in many of the checkpointing changes along with other bugfixes.
John Wise
Cut-n-paste, not copy-paste!
Matthew Turk
This made it into public trunk but not here
John Wise
Four changes. (1) CPU time outputs won't happen if already written on this cycle. (2) Fixed a couple more round-off errors in the optimized CTP when using a non power-of-2 topgrid resolution. (3) Added a Multispecies-dependent minimum temperature for cluster_maker. (4) When checking for star deaths, the order of the linked list wasn't preserved. This caused a mismatch between the stars and AddedFeedback[] array.
John Wise
Merged.
John Wise
Updating autoconf and automake files. No more -lsz in libraries. I should learn configure better and have it check for this...
John Wise
Fixed a couple of things in nested grid sims. When reconstructing the hierarchy at the end of initialization, round-off errors in runs with non-powers of 2 top grid resolutions could cause orphaned child grids! This no longer happens. We now use the grid centers instead of the edges to determine if a child is contained within a parent grid. Second, inserted a missing nint() in the calculation of grid edges in the "CONTAINED_WITHIN_PARENT" define. Quietened things …
John Wise
I prematurely pushed the last commit before adding all of the changes to the single precision version. Sorry!
John Wise
These includes aren't needed and cause problems.
John Wise
degraf now works in single precision (--enable-single in configure) and fixed a bug where degraf didn't shift the coarse grids by 1/2 cell width.
John Wise
Forgot to change file precision for new grid I/O.
John Wise
Fixed newly introduced bug where ParticleType used the PINT temporary array instead of tempint. Need to initialize the dimensions when writing the boundary.
John Wise
Forgot to change some MPI calls to use the particle ID int type (PINT) in the halo finder.
John Wise
Can now add external boundaries for the out-of-code boundary conditions.
John Wise
Forgot to include the stellar ejection mass fraction in the metal feedback of the "star cluster" particles.
Elizabeth Harper-Clark
added cloud type 7 - only turb k=1-2
  1. Prev
  2. Next