Permalink
Browse files

Added OS files and data files to .gitignore

  • Loading branch information...
csullivan committed Jun 24, 2013
1 parent 678f735 commit d90e05242afc2e9cfa483d5c457f0e0495ddb900
Showing with 132,090 additions and 0 deletions.
  1. +6 −0 Makefile
  2. +305 −0 README
  3. +17 −0 make.inc
  4. +62 −0 src/Makefile
  5. +8 −0 src/Makefile_test
  6. +3 −0 src/Makefile~
  7. +394 −0 src/absorption_crosssections.F90
  8. +24 −0 src/constants.inc
  9. +207 −0 src/electron_positron_annihilation.F90
  10. +403 −0 src/emissivities.F90
  11. +401 −0 src/emissivities.F90~
  12. +403 −0 src/emissivities_commit.F90
  13. +219 −0 src/fermi.F90
  14. +106 −0 src/gauss_laguerre_helpers.F90
  15. +104 −0 src/gauss_legendre_helpers.F90
  16. +112 −0 src/helpers.F90
  17. +442 −0 src/inelastic_electron_scattering.F90
  18. +113 −0 src/linterp_many_mod.F90
  19. +681 −0 src/make_table_example.F90
  20. +448 −0 src/make_table_example.F90~
  21. +50 −0 src/microphysical_ec_routine_temp.F90
  22. +30 −0 src/nuc_eos/Makefile
  23. +30 −0 src/nuc_eos/Makefile.orig
  24. +163 −0 src/nuc_eos/bisection.F90
  25. +163 −0 src/nuc_eos/bisection.F90.orig
  26. +74 −0 src/nuc_eos/driver.F90
  27. +74 −0 src/nuc_eos/driver.F90.orig
  28. +60 −0 src/nuc_eos/eosmodule.F90
  29. +60 −0 src/nuc_eos/eosmodule.F90.orig
  30. +112 −0 src/nuc_eos/findrho.F90
  31. +112 −0 src/nuc_eos/findrho.F90.orig
  32. +207 −0 src/nuc_eos/findtemp.F90
  33. +207 −0 src/nuc_eos/findtemp.F90.orig
  34. +132 −0 src/nuc_eos/linterp.f
  35. +132 −0 src/nuc_eos/linterp.f.orig
  36. +125 −0 src/nuc_eos/linterp_many.F90
  37. +125 −0 src/nuc_eos/linterp_many.F90.orig
  38. +358 −0 src/nuc_eos/nuc_eos.F90
  39. +358 −0 src/nuc_eos/nuc_eos.F90.orig
  40. +235 −0 src/nuc_eos/readtable.F90
  41. +235 −0 src/nuc_eos/readtable.F90.orig
  42. +122 −0 src/nuclei_distribution_helpers.F90
  43. +715 −0 src/nulib.F90
  44. +217 −0 src/nulibtable.F90
  45. +95 −0 src/nulibtable_driver.F90
  46. +203 −0 src/nulibtable_reader.F90
  47. +206 −0 src/point_example.F90
  48. +60,500 −0 src/rates-ext.out
  49. +60,500 −0 src/rates-ext.out~
  50. +68 −0 src/requested_interactions.inc
  51. +60 −0 src/requested_interactions.inc~
  52. +764 −0 src/scattering.F90
  53. +63 −0 src/test.F90
  54. +14 −0 src/test.F90~
  55. +202 −0 src/weak_magnetism_correction.F90
  56. +524 −0 src/weak_rates.F90
  57. +337 −0 src/weak_rates.F90~
View
@@ -0,0 +1,6 @@
all:
$(MAKE) -C src
clean:
$(MAKE) -C src clean
View
305 README
@@ -0,0 +1,305 @@
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!! Welcome to NuLib !!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
DISCLAIMER:
-----------
Please note that the routines provided here come with absolutely no
warranty and we are unable to guarantee that we will be able to
provide support or help if you run into problems integrating them with
your simulation code. If you decide to use the provided routines in
published work, it is YOUR responsibility to check their physical
correctness and consistency.
If you have any questions or have discovered a bug in our routines,
please e-mail us at evanoc@tapir.caltech.edu or cott@tapir.caltech.edu
COPYRIGHT:
----------
While NuLib is open source, its copyright is held by Evan O’Connor and
Christian Ott. In the absence of suitable open scientific software
licenses, we release this version of NuLib to the community under the
Creative Commons attribution-noncommercial-share alike license:
http://creativecommons.org/licenses/by-nc-sa/3.0/us
Essentially, you may use NuLib, but must make reference to our work,
must not use NuLib for commercial purposes, and any code including or
using our routines or part of them may be made publically available,
and if so, only under the same license.
Introduction:
-------------
<<<<<<< HEAD
The goal of NuLib is to provide a basic standard set of neutrino
=======
The goal of NULIB is to provide a basic standard set of neutrino
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
matter interaction routines that can be readily incorporated in
radiation-hydrodynamics codes for physics benchmarking.
NuLib v1.0 includes the basic neutrino emissivities and absorption
<<<<<<< HEAD
opacities (including pair processes) as well as neutrino-nucleon,
eutrino-nucleus elastic scattering processes and neutrino-electron
inelastic scattering. Other inelastic processes will also be
including in future versions.
=======
opacities (including pair processes) as well as neutrino-nucleon and
neutrino-nucleus scattering processes. Inelastic scattering (with
energy redistribution) is neglected, but will be included in future
revisions. Other inelastic processes will also be including in future
versions.
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
If anyone would like to contribute to the development of NuLib please
let me know, I'm hosting this on GitHub with this in mind. I am
currently developing a neutrino transport code that will make use of
all these interactions, that being said it is not finished and since
I've never done such a task, it may well be the case that methodlogies
of coding these interactions will change slightly (and have not been
fully tested) If you have advice on this front please fill me in, I am
more than happy to make NuLib as accessible as it needs to be to allow
others to benefit from it.
NuLib, in its current form, is used by me to make tables of neutrino
<<<<<<< HEAD
emissivities, opacities (scattering and absorption), and scattering
kernels. It is not yet optimized for on-the-fly calculations of
quantities. In fact, the routines are coded in such a way as to be as
clear and accurate as possble, with little or no regard for
computational speed. For example, the weak magnitism correction for
scattering processes is small but many terms long, I do the full
calculation. In the future I hope NuLib will have routines capable of
on the fly calculations, I expect this is necessary if one whats fully
differential cross sections (in energy and angle, as a function of
energy).
=======
emissivities and opacities (scattering and absorption). It is not yet
optimized for on-the-fly calculations of quantities. In fact, the
routines are coded in such a way as to be as clear and accurate as
possble, with little or no regard for computational speed. For
example, the weak magnitism correction for scattering processes is
small but many terms long, I do the full calculation. In the future I
hope NuLib will have routines capable of on the fly calculations, I
expect this is necessary if one whats fully differential cross
sections (in energy and angle, as a function of energy).
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
NuLib v1.0 Neutrino Interactions.
---------------------------------
Emissivities:
-------------
1. electron-positron annhilation to \nu - \bar{\nu}, this
follows Bruenn 1985, only the isotropic case is currently coded but
the extension to higher legendre mode is trivial as the structure of
the calculation is laid out.
2. Nucleon-Nucleon Bremsstrahlung, this is an approximation used in
Burrows Reddy, and Thompson (2006) [BRT06]. A full calculation of
this reaction would be great!
Scattering Opacities:
---------------------
For the scattering opacities, I list only the base interactions, the
neutrino type plays a role in the calculations, see the code for a
full description of the interaction. The cross sections all come from
BRT06 with appropiate corrections (e.g. weak magnetism [has a logical
flag to turn off if desired])
1. neutrino scattering on neutrons
2. neutrino scattering on protons
3. neutrino scattering on heavy nuclei (this includes lots of
corrections, see BRT06 and the code for details)
4. neutrino scattering on electrons (elastic, Thomspon, T. PhD)
5. neutrino scattering on alphas
Absorption Opacities:
---------------------
1. \nu_e absorption neutrons: This currently includes a stimulated
absorption term as described in BRT06, final
state electron blocking and also final state proton blocking. Weak
magnitism, phase space and recoil corrections [optional via flag] are
applied via Horowitz (2002).
2. \bar{\nu}_e absorption protons: This currently includes a
stimulated absorption term as described in Burrows, Reddy, and
Thompson, final state positorn blocking and also final state neutron
blocking. Weak magnitism, phase space and recoil corrections
[optional via flag] are applied via Horowitz (2002).
3. \nu_e absorption on heavy nuclei: This follows the simple treatment
of Bruenn 85 (among others), placing cuts on the cross section based
on the average A and average Z of the nucleus. Much better treatment
is desired and someday will be implemented.
<<<<<<< HEAD
Neutrino-Electron Inelastic Scattering Kernels:
-----------------------------------------------
For a temperature and electron chemical potential, NuLib calculates
the first two terms in a Legendre expansion of the scattering kernels
for neutrinos on electrons. We essentially follow Bruenn 1985 and
references there in.
=======
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
Sample Executables:
-------------------
make_table_example: by default this makes a horribly under resolved
<<<<<<< HEAD
10x10x10x24 (rho,temp,ye,energy) + (10x10x24*24)
(temp,eta,energy_in,energy_out) NuLib table in h5 format. The
calculations takes abut 1 minute to generate. The table boundaries,
and number of data points are changable in the make_table_example.F90
file. To get enough accuracy in the interpolation I expect at least
10 points per decade in rho, 20 in temperature and 1 for every 0.01 or
0.02 in ye. This makes a table ~1GB in size with 24 energy bins. The
energy spacing is changable in nulib.F90, right now it is a 4MeV bin,
then a logarithmic spacing starting at 1MeV going to ~300MeV, this may
not be the best choice, if you have a better suggestion, let me know,
or code up a routine to generate good energy spacing and send a pull
request (?).
=======
10x10x10x24 (rho,temp,ye,energy) NuLib table in h5 format, the
1000*24*3*3 calculations takes abut 1 minute to generate. The table
boundaries, and number of data points are changable in the
make_table_example.F90 file. To get enough accuracy in the
interpolation I expect at least 10 points per decade in rho, 20 in
temperature and 1 for every 0.01 or 0.02 in ye. This makes a table
~1GB in size with 24 energy bins. The energy spacing is changable in
nulib.F90, right now it is a 4MeV bin, then a logarithmic spacing
starting at 1MeV going to ~300MeV, this may not be the best choice, if
you have a better suggestion, let me know, or code up a routine to
generate good energy spacing and send a pull request (?).
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
You must specify an equation of state, NuLib is set up to read in the
EOS tables on stellarcollapse.org, the filename is set in
make_table_example.F90. For each EOS you must set the reference mass,
this is used to convert the density into a number density for the
scattering and absorption cross sections.
The main routine that make_table_example.F90 calls is
single_point_return_all. This routine takes as input all of the
equation of state variables and returns the emissivity, absorption
opacity and scattering opacity for all neutrino species and energies.
You also must specify the neutrino scheme, this is what sets the
number of species. Here the comments regarding the different neutrino
scheme currently available in NuLib, again if you have a request let
me know, I want to make this as useful as possible.
! many people use different number of species, this is the possible
! summing scheme NuLib can currently do
!
! mytable_neutrino_scheme = 1 (three output species)
! species #1: electron neutrino #2 electron antineutrino
! #3: muon+tau neutrino+antineutrino
!
! neutrino_scheme = 2 (four output species)
! species #1: electron neutrino #2 electron antineutrino
! #3: muon+tau neutrino #4 mu and tau antineutrino
!
! neutrino_scheme = 3 (six output species)
! species #1: electron neutrino #2 electron antineutrino
! #3: muon neutrino #4 mu antineutrino
! #5: tau neutrino #6 tau antineutrino
single_point_return_all appies Kirchoff's law to the emissivities and
absorption cross sections. This adds an contribution to the
emissivity from the absorption cross section (and vice-versa). This
is explained in BRT06 and explicitly showed in the
<<<<<<< HEAD
single_point_return_all routine. There is a similar routine for the
inelastic scattering kernels, single_Ipoint_return_all. This routine
only calculates half of the terms, we use symmetry laws to calculate
the other half.
=======
single_point_return_all routine.
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
point_example: this program shows examples of how to call the NuLib
routines for a single point. Again, the energy spacing is changable in
nulib.F90, right now it is a 4MeV bin, then a logarithmic spacing
starting at 1MeV going to ~300MeV, this may not be the best choice, if
you have a better suggestion, let me know, or code up a routine to
generate good energy spacing and send a pull request (?). You must
specify an equation of state, NuLib is set up to read in the EOS
tables on stellarcollapse.org, the filename is set in
point_example.F90. For each EOS you must set the reference mass, this
is used to convert the density into a number density for the
scattering and absorption cross sections.
Unlike single_point_return_all, the individual calls to emissivity
(e.g. return_emissivity_spectra_given_neutrino_scheme) or the cross
section routines
(e.g. return_absorption_opacity_spectra_given_neutrino_scheme) do not
apply Kirchoff's law.
nulibtable_driver: This routine is a driver routine for reading in a
NuLib table and using a trilinear interpolation (log rho, log temp,
ye) routine to interpolate the emissivities and cross sections to any
rho,temp,ye. This is extermely useful for transport simulations and
prevents on the fly calculations of the neutrino interaction terms.
It does not currently interpolate in energy, this would be a useful
feature to add, it would require slightly adjusting the units of the
emissivities, in addition to writing a 4th order interpolator. There
are several routines available in nulibtable.F90 for accessing the
table. The large number of variables can lead to long times spent in
interpolating. I've tried to optimize this but more could be done I'm
sure.
Installation.
-------------
If you are reading this then you are halfway there. You must set the
F90 and F90FLAGS compiler variables in the make.inc file in this
directory to point to your Fortran compiler. Also, you must have HDF5
compiled with the _same_ compiler. This usually means downloading the
source from http://www.hdfgroup.org/HDF5/release/obtain5.html,
configuring with your version of:
./configure --enable-fortran FC=ifort --prefix=/Users/evanoc/opt/hdf5-current-ifort12
and then
make
make install
the HDF5DIR variable in make.inc would then be set to /Users/evanoc/opt/hdf5-current-ifort12
After this, a simple make should create three executables in the main
directory, a brief explanation of these is in the section
`executables' above.
<<<<<<< HEAD
Extras
------
There is support in NuLib for using Matthias Hempel's NSE mass
distributions available from
http://phys-merger.physik.unibas.ch/~hempel/eos.html. To enable these
use must download his code and tables from his website, place them in
the directory src/extra_code_and_tables/ and enable the preprocessor
flag NUCLEI_HEMPEL. We use the SFHo table as an example in the code,
you can change this by editting nuclei_distribution_helpers.F90
directly. Please see this file for more details.
=======
>>>>>>> 3f541161a90c79c7f07bf6d07b37717da38bc659
View
@@ -0,0 +1,17 @@
F90=gfortran
F90FLAGS= -O3 -ffree-line-length-none -fopenmp
#if you want to take advantage of openmp use this flag
#F90FLAGS= -O3 -openmp
MODINC="-I ./"
OMP_NUM_THREADS=12
#You must have a HDF5 version installed with
#the _same_ compiler, this often means compiling
#it by yourself.
#See README file for compiling HDF5
HDF5DIR=/projects/ceclub/gr1dnulib/mesasdk
HDF5INCS=-I$(HDF5DIR)/include
HDF5LIBS=-L$(HDF5DIR)/lib -lhdf5 -lhdf5_fortran -lhdf5 -lz
View
@@ -0,0 +1,62 @@
include ../make.inc
SOURCES=nulib.F90 \
fermi.F90 \
helpers.F90 \
gauss_laguerre_helpers.F90 \
gauss_legendre_helpers.F90 \
electron_positron_annihilation.F90 \
absorption_crosssections.F90 \
emissivities.F90 \
scattering.F90 \
weak_magnetism_correction.F90 \
weak_rates.F90 \
inelastic_electron_scattering.F90 \
NT_SOURCES=nulibtable.F90 \
nulibtable_reader.F90 \
linterp_many_mod.F90
EXTRADEPS = requested_interactions.inc constants.inc
OBJECTS=$(SOURCES:.F90=.o )
NT_OBJECTS=$(NT_SOURCES:.F90=.o )
#take care of EOS dependences etc
DEFS = -DHAVE_NUC_EOS
EXTRAINCS = $(HDF5INCS) -I./nuc_eos
EXTRADEPS += nuc_eos/nuc_eos.a
EXTRAOBJECTS = nuc_eos/nuc_eos.a $(HDF5LIBS)
all: nulibtable_driver point_example make_table_example test
nulibtable_driver: $(EXTRADEPS) $(NT_OBJECTS)
$(F90) $(F90FLAGS) $(MODINC) $(EXTRAINC) -o ../nulibtable_driver nulibtable_driver.F90 $(NT_OBJECTS) $(EXTRAOBJECTS)
point_example: $(EXTRADEPS) $(OBJECTS) point_example.F90
$(F90) $(F90FLAGS) $(DEFS) $(MODINC) $(EXTRAINCS) -o ../point_example point_example.F90 $(OBJECTS) $(EXTRAOBJECTS)
make_table_example: $(EXTRADEPS) $(OBJECTS) make_table_example.F90
$(F90) $(F90FLAGS) $(DEFS) $(MODINC) $(EXTRAINCS) -o ../make_table_example make_table_example.F90 $(OBJECTS) $(EXTRAOBJECTS)
test: $(EXTRADEPS) $(OBJECTS) test.F90
$(F90) $(F90FLAGS) $(DEFS) $(MODINC) $(EXTRAINCS) -o test test.F90 $(OBJECTS)
$(OBJECTS): %.o: %.F90 $(EXTRADEPS)
$(F90) $(F90FLAGS) $(DEFS) $(MODINC) $(EXTRAINCS) -c $< -o $@
$(NT_OBJECTS): %.o: %.F90 $(EXTRADEPS)
$(F90) $(F90FLAGS) $(DEFS) $(MODINC) $(EXTRAINCS) -c $< -o $@
nuc_eos/nuc_eos.a: nuc_eos/*.F90 nuc_eos/*.f
$(MAKE) -C nuc_eos
clean:
rm -rf ../make_table_example
rm -rf ../point_example
rm -rf ../nulibtable_driver
rm -rf test
rm -rf *.o
rm -rf *.mod
rm -rf *.a
$(MAKE) -C nuc_eos clean
Oops, something went wrong.

0 comments on commit d90e052

Please sign in to comment.