It is currently Wed Dec 01, 2021 12:21 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 25 posts ]  Go to page Previous  1, 2
Author Message
 Post subject:
PostPosted: Fri Jun 29, 2007 6:09 am 
Offline

Joined: Sun Feb 11, 2007 4:13 pm
Posts: 9
Location: Minneapolis, MN
I just want to make sure I did this right...

I was plotting B/C data with a GALPROP predicted B/C. I wanted to show how differing diffusion indices changed the B/C prediction. I started with the 599278 galdef file and changed D_g_2 only to 0.4,0.45,0.55. As shown in this B/C

plot:http://webusers.physics.umn.edu/~childers/BCRatioB.jpg

It seems to show that the HEAO-3 data corresponds to an index of 0.4 instead of 0.6 as shown in your 1998 ApJ Paper.

I tried looking at the 2006 Ptuskin et al. (COSPAR) and 2004 Strong (Astrophys. and Sp. Sci.) papers, but didn't see any references to B/C data runs with parameters listed. I did find a 2002 ApJ paper of yours that does have such parameters. I'm running them now.

Taylor


Top
 Profile  
 
 Post subject: B/C
PostPosted: Sat Jun 30, 2007 2:17 am 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
Igor, can you run his galdef file, when I ran it everything was OK (see above).

_________________
Andy Strong, MPE


Top
 Profile  
 
 Post subject: other nuclei
PostPosted: Tue Jul 03, 2007 12:38 am 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
from your other nuclei plots above, even He is wrong, much too low and with strange breaks. So we can concentrate on just p,He runs for debugging, it's obviously the same effect.

a conjecture: I wonder if it's something with the FITS output: it looks a bit like a numerical scaling problem, that typically causes such jumps. There is a debug option to print the spectra before outputting to FITS, will try to find it.

_________________
Andy Strong, MPE


Top
 Profile  
 
 Post subject: debug printout
PostPosted: Tue Jul 03, 2007 5:38 am 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
OK let's try just p,He:
max_Z = 2

you can get printout by setting e.g.
timestep_print =20 number of timesteps between printings

in the printout this produces (for my successful run) - after the setup phase:

alpha1_r:0
alpha1_r:0.135476
alpha2_r:0.541955
alpha3_r:0.406427
alpha1_z:27.0952
alpha2_z:54.1904
alpha3_z:27.0952
alpha1_p:9.68781
alpha2_p:16.7869
alpha3_p:7.17452
galdef_ID 51_600003 start timestep dt=3.15576e+14 Helium_4
12345678910111213141516171819
particle.cr_density after iteration 20 with time step 1e+07 yr
Distribution.print: n_spatial_dimensions=2
Distribution with 2 spatial dimensions:
ip=0
0 9.63046e+11 4.31368e+12 5.41142e+12 5.89545e+12 6.05776e+12 6.02605e+12 5.87237e+12 5.6418e+12 5.36414e+12 5.05975e+12 4.74282e+12 4.42325e+12 4.10799e+12 3.80183e+12 3.50798e+12 3.22849e+12 2.96446e+12 2.71498e+12 2.46474e+12 2.05323e+12 0
1 9.40895e+17 9.83399e+17 1.01208e+18 1.02242e+18 1.01748e+18 1.00609e+18 9.94332e+17 9.8224e+17 9.67616e+17 9.47908e+17 9.19654e+17 8.82731e+17 8.39353e+17 7.93887e+17 7.4671e+17 6.92909e+17 6.25676e+17 5.3622e+17 4.12623e+17 2.41756e+17 0
2 1.41675e+18 1.48164e+18 1.52531e+18 1.54101e+18 1.53349e+18 1.51624e+18 1.49848e+18 1.48025e+18 1.4582e+18 1.4285e+18 1.38588e+18 1.33016e+18 1.2647e+18 1.19615e+18 1.1251e+18 1.04413e+18 9.42964e+17 8.08382e+17 6.22358e+17 3.64868e+17 0

etcetc

Compare this with what you get. It will show if something is wrong at the start. I will send more output by email if necessary.

_________________
Andy Strong, MPE


Top
 Profile  
 
PostPosted: Tue Jul 03, 2007 5:55 am 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
childers wrote:
I've also been trying to reproduce your diffusion coefficient work from your 1998 ApJ paper.

If I take 599278 and change its parameters to match those of your run with the index 01000.I don't get the same results

Here is my galdef file

http://webusers.physics.umn.edu/~childe ... 50p_600025

here is a plot of its output:

http://webusers.physics.umn.edu/~childers/BCRatioA.jpg

As you can see, it doesn't reproduce your fitting results of the HEAO-3 data.

Thanks,
Taylor



This one has convection (convection=1) which will make a big difference.

_________________
Andy Strong, MPE


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 03, 2007 6:12 am 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
childers wrote:
I just want to make sure I did this right...

I was plotting B/C data with a GALPROP predicted B/C. I wanted to show how differing diffusion indices changed the B/C prediction. I started with the 599278 galdef file and changed D_g_2 only to 0.4,0.45,0.55. As shown in this B/C

plot: http://webusers.physics.umn.edu/~childers/BCRatioB.jpg

It seems to show that the HEAO-3 data corresponds to an index of 0.4 instead of 0.6 as shown in your 1998 ApJ Paper.

I tried looking at the 2006 Ptuskin et al. (COSPAR) and 2004 Strong (Astrophys. and Sp. Sci.) papers, but didn't see any references to B/C data runs with parameters listed. I did find a 2002 ApJ paper of yours that does have such parameters. I'm running them now.

Taylor


If you change D_g_2 you may also have to adjust the normalization of D to optimize the fit.
Also, how do you handle the solar modulation in these plots ?

_________________
Andy Strong, MPE


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jul 12, 2007 10:19 am 
Offline

Joined: Sun Feb 11, 2007 4:13 pm
Posts: 9
Location: Minneapolis, MN
I just got back from the ICRC so I'll try and get back to you on your last few posts. Meanwhile, I decided to try and install GALPROP on another computer in order to verify it wasn't just my computer.

I've been trying to compile it, but am having linking issues because of the fortran/C interactions.

Everything compiled fine, but the linking produces this:

g++ *.o -o galprop -L/home/creamdata/GALPROP/cfitsio -lcfitsio -Wl,-rpath,/home/creamdata/GALPROP/cfitsio -lm -lg2c
antiproton.o(.text+0x2d8): In function `antiproton_':
: undefined reference to `nucleon_cs__'
antiproton.o(.text+0x359): In function `antiproton_':
: undefined reference to `nucleon_cs__'
cfactor.o(.text+0x1d): In function `emiss_':
: undefined reference to `isrf_energy_density__'
fort_interface1.o(.text+0x14): In function `set_sigma_cc()':
: undefined reference to `set_sigma_'
fort_interface1.o(.text+0x4f): In function `bremss_spec_cc(double, double, int, int)':
: undefined reference to `bremss_spec_'
fort_interface1.o(.text+0x3f0): In function `e_loss_compton_cc(double, double)':
: undefined reference to `e_loss_compton_'
fort_interface1.o(.text+0x55b): In function `pp_meson_cc(double, double, int, int, int)':
: undefined reference to `pp_meson_'
collect2: ld returned 1 exit status
make: *** [galprop] Error 1


Any ideas. I seem to remember having this problem before, but I don't remember how I fixed it. Perhaps it had to do with the underscores in the function names. I'm using g++ 3.3.3

Thanks,
Taylor


Top
 Profile  
 
 Post subject: compiling/linking
PostPosted: Thu Jul 12, 2007 1:28 pm 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
it looks like it has to do with this routine, which has to be modified
as stated depending on the way the compiler/linker handles fortran

/**.****|****.****|****.****|****.****|****.****|****.****|****.****|****.****|
// * fort_interface1.cc * galprop package * 2001/05/11
//**"****!****"****!****"****!****"****!****"****!****"****!****"****!****"****|

// C++ wrapper for fortran routines 2001/05/11
// These routines are to be renamed: "_(" should be replaced with "__("
// and vise versa when changing platform (e.g. SUN UNIX, Linux).

_________________
Andy Strong, MPE


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jul 13, 2007 3:44 pm 
Offline

Joined: Sun Feb 11, 2007 4:13 pm
Posts: 9
Location: Minneapolis, MN
I figured out that my problem is my FITS file extraction.

I wrote my own software to extract the FITS data, but after downloading Igor's plotting routine and seeing that the output looked OK I realized it was my software.

I'm sorry for all the confusion. The software seemed to work on some level, so I didn't question it, but I should have realized it.

Is the energy in the FITS file in units of MeV/nucleon (I know its Log(E))? I'm assuming so, since the manual says the flux in (MeV/nuc)^2/(cm^2 sr s MeV/nuc).

I'm going to try to fix the problem and remake my B/C plots with the diffusion indices.

Thanks,
Taylor


Top
 Profile  
 
 Post subject: B/C
PostPosted: Fri Jul 13, 2007 10:41 pm 
Offline

Joined: Wed May 17, 2006 7:02 am
Posts: 285
Location: MPE Garching
good that you found the problem and that galprop is OK.
I also suspected the FITS interface, as you see above.

Yes the energy units are MeV/nucleon and the flux is multiplied by the energy squared
to reduce the dynamic range.

_________________
Andy Strong, MPE


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 25 posts ]  Go to page Previous  1, 2

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group