Several points:
0) The files are at: ki-jmck.slac.stanford.edu
1
a) Old data
for MB09 model is in /data2/jmckinne/runlocaldipole3dfiducial/
Read description file:
less /data2/jmckinne/runlocaldipole3dfiducial/datadesc.txt
1b) I've copied over the "thickdisk7" (272x128x256) model
"fieldline????.bin" files. The format is given in item #6. More
models are being copied and I'll report here when they are ready.
Change directly to it since you won't be able to read the intermediate
directories:
dir: /data2/jmckinne/thickdisk7/dumps/
Rest, directly view or copy, can't list directory:
b) coordparms.dat file(shows parameters
for coord.c file):
/data2/jmckinne/thickdisk7/coordparms.dat
c) Basic description file (this file):
/data2/jmckinne/thickdisk7/desc.txt
d) List of all models (more copied as time goes on) in latex format.
/data2/jmckinne/table_models.share.tex
2) Header:
The dump, gdump, and fieldline files have
a new header format. Still,
all files are composed of
a 1 text text header with \n at the end.
Rest of file beyond \n is data that is binary doubles for gdump and dump and floats for fieldline.
C code for header for all files is:
fprintf(headerptr, "%21.15g %d %d %d %21.15g %21.15g %21.15g %21.15g
%21.15g %21.15g %ld %21.15g %21.15g %21.15g %21.15g %21.15g %21.15g
%21.15g %d %21.15g %21.15g %d %d %d %d %d %d %d %d %d\n",
tsteppartf,
realtotalsize[1], realtotalsize[2], realtotalsize[3], realstartx[1],
realstartx[2], realstartx[3],
dx[1],
dx[2],
dx[3],
localrealnstep,
gam,
a,
R0,
Rin,
Rout,
hslope,localdt,
defcoord,
MBH,
QBH,is,ie,js,je,ks,ke,whichdump,whichdumpversion,
numcolumns);
4) For 272x128x256 run, I changed the frequency of
dumps at some point
since it was costing too much computational time. The fieldline
dumpswent from every 2
M to every 4
M at some point (see time in header that
is first line in file as text header).
5) Recall that column data
index runs as fastest
index in files, then
6) I also realized you probably need
a dump0000.bin file so you can
extract the
gdet=sqrt(-g) value at least. Or did you use
a gdump.bin
file before since you needed metric quantities?
FYI, gdump.bin is 8.4
G for 272x128x256 run. dump0000.bin is 4.9
G.
Each fieldline file is 375
M and there are 3647 of them (again, dump
period went from 2
M to 4
M at some point).
7) Here's the format for the fieldline, dump, and gdump files as
read-in by supermongo:
fieldline data format (floats as binary data):
{rho u -u_x0 -T^x1_x0/(rho u^x1) u^
t v^x1 v^x2 v^x3 B^x1 B^x2 B^x3}
B^x? are 3-field in lab-
frame (
i.e. B^
i = *F^{it} or whatever
sign HARM paper has.)
dump data format: (doubles as binary data):
{
ti tj tk x1 x2 x3 r h ph rho u v1 v2 v3 \
p cs2 Sden \
divb \
uu0
uu1 uu2 uu3 ud0 ud1 ud2 ud3 \
bu0 bu1 bu2 bu3 bd0 bd1 bd2 bd3 \
v1m v1p v2m v2p v3m v3p gdet \
ju0 ju1 ju2 ju3 \
jd0 jd1 jd2 jd3 \
fu0 fu1 fu2 fu3 fu4 fu5 \
fd0 fd1 fd2 fd3 fd4 fd5 }
gdump data format: (doubles as binary data):
{
ti tj tk x1 x2 x3 V1 V2 V3 \
c000 c001 c002 c003 \
c010 c011 c012 c013 \
c020 c021 c022 c023 \
c030 c031 c032 c033 \
c100 c101 c102 c103 \
c110 c111 c112 c113 \
c120 c121 c122 c123 \
c130 c131 c132 c133 \
c200 c201 c202 c203 \
c210 c211 c212 c213 \
c220 c221 c222 c223 \
c230 c231 c232 c233 \
c300 c301 c302 c303 \
c310 c311 c312 c313 \
c320 c321 c322 c323 \
c330 c331 c332 c333 \
gn300 gn301 gn302 gn303 \
gn310 gn311 gn312 gn313 \
gn320 gn321 gn322 gn323 \
gn330 gn331 gn332 gn333 \
gv300 gv301 gv302 gv303 \
gv310 gv311 gv312 gv313 \
gv320 gv321 gv322 gv323 \
gv330 gv331 gv332 gv333 \
gdet \
ck0 ck1 ck2 ck3 \
dxdxp00 dxdxp01 dxdxp02 dxdxp03 \
dxdxp10 dxdxp11 dxdxp12 dxdxp13 \
dxdxp20 dxdxp21 dxdxp22 dxdxp23 \
dxdxp30 dxdxp31 dxdxp32 dxdxp33}
1.8.3.1
