Hi,
The first problem was solved as I reported, by disabling the QuasiLocalMeasures. Some MPI problem. Btw, which information am I missing without QuasiLocalMeasures? (the output file was the wrong one, of a successful simulation without WeyScal4)
About the .par file. I found out (please correct me if necessary) that I need to change the name of the run by the simfactory submission (here static_tov): ./simfactory/bin/sim submit *static_tov* --parfile=par/static_tov.par --procs=32 --walltime=8:0:0
Otherwise the run works with the original .par (first used with this run) , and not to the newly specified .par file. So, it worked by doing ./simfactory/bin/sim submit *static_tov01* --parfile=par/*static_tov_new.par * --procs=32 --walltime=8:0:0
Best, Jose
On Wed, Feb 8, 2012 at 10:05 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 22:36, Jose Fiestas Iquira wrote:
Hi,
In order to use the wave generator, I am probably not submitting the job properly.
I had modified the .par file located at: /home/ettest24/Cactus/par/qc0-mclachlan.par since I run the example like this: ./simfactory/bin/sim submit qc0-mclachlan --parfile=par/qc0-mclachlan.par --procs=32 --walltime=8:0:0
Here the run stopped with errors, as previously reported.
Hi Jose,
Could you update me on your status? Did you solve this problem? In the standard output file that you sent to the ET mailing list on 16-Jan-2012, I do not see any problem. The run reached 100 M of evolution time. The only problem is a non-fatal warning message which is caused by some problem with the way that Carpet keeps track of points which should or should not be included in norm operations. You can ignore this warning.
Could be possible that the proper .par is located somewhere else? I found it also here:
/home/ettest24/Cactus/simfactory/etc/par/qc0-mclachlan.par /home/ettest24/Cactus/arrangements/McLachlan/par/qc0-mclachlan.par
Btw, in the last one WeylScal4 is activated, in a similar way as you told me.
SimFactory takes the parameter file that you specify on the command line - it does not look anywhere else. Can you be more specific about the problem you are having?
PS: if you ask these questions on the mailing list, other people can help and benefit from the discussion.
Best, Jose
On Mon, Jan 23, 2012 at 1:11 PM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 14:57, Jose Fiestas Iquira wrote:
Hello,
Regarding the Binary Black Hole coalescence example from the ETK.
In the BH diagnostics files are columns with quadrupole terms, which
are not constant. Do you know of some tool in ETK to calculate the GW
emission from these terms?
No, not that I am aware of. These are the quadrupole moments of the
individual apparent horizons, rather than the quadrupole moment of the
system as a whole. The gravitational wave emission can be computed by using
the WeylScal4 and Multipole thorns. These give you the spherical harmonic
multipoles of Psi4 (e.g. l = 2, m = 2 is the dominant one), which can then
be converted to strain by "two integrations in time" [this is actually quite
subtle and hard to get right - see Reisswig and Pollney, arXiv:1006.1632,
for an overview and solution).
Thanks,
Jose
On Tue, Jan 17, 2012 at 5:18 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 17 Jan 2012, at 13:37, Jose Fiestas Iquira wrote:
The Einstein Analysis guide is very useful. I found here
http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech142.html#x15...
the units used in 'Hydrobase'.
If I assume these are the units for all thorns, I obtain from data in
BH_diagnostics.ah1.gp a unit of distance of about 1 Km. To estimate
the time I am using a mass of ~0.5 M_sun for both BHs (m_irreducible
?) , and I read a final time (cctk_time) ~ 1microsec. Are these
quantities in the usual range?
The length unit is G M / c^2 and the time unit is G M / c^3. For M = M_sun,
I get the length unit as 1.5 km and the time unit as 4.9 us. The final time
in the simulation is 100 M (i.e. 100 G M / c^3), which would be 0.5 ms.
Jose
On Tue, Jan 17, 2012 at 4:04 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
I sent the wrong file, here is the right one.
$ hg tip
gives me:
changeset: 3421:dff4df2b3687
tag: tip
user: Erik Schnetter schnetter@cct.lsu.edu
date: Thu Jan 12 14:11:27 2012 -0500
summary: CarpetLib: Allocate space for dstbox before
restricting/prolongating
[40514 refs]
I copied this version from:
./GetComponents -a
http://svn.einsteintoolkit.org/manifest/trunk/einsteintoolkit.th
Thanks,
Jose
On Tue, Jan 17, 2012 at 1:13 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 16 Jan 2012, at 23:21, Jose Fiestas Iquira wrote:
Thanks, I am attaching the output file.
The output file looks fine - I can't see any problem with the simulation.
According to the output file, it reached t = 100 M, and in the parameter
file that is listed as cctk_final_time, so all looks well!
I did copy the development version. As I can see, it was already
fixed, so I should not get this warning, since I copied the
development version.
Yes, if you are using the current development version, then I don't know why
you get the warning. Can you check which version of Carpet you have by
doing
cd repos/carpet
hg tip
to be sure you are using the development version?
Is the puncture tracker not a better way to plot BH trajectories.
It's cheaper to use the puncture tracker, because computing the horizon
centroid is more expensive, but it has to be done every iteration, whereas
the horizon can be found less frequently. For comparison with
point-particle results, I don't think it makes much difference; the results
are gauge dependent anyway so differences on the order of the horizon size
cannot be interpreted physically. Since the parameter file already has
horizon-finding built in, it is easier for you to use that.
If you want to figure out how to add the puncture tracker to the simulation,
you can start with these parameters:
ActiveThorns = "PunctureTracker"
PunctureTracker::track [0] = yes
PunctureTracker::initial_x [0] = $xp
PunctureTracker::which_surface_to_store_info[0] = 0
PunctureTracker::track [1] = yes
PunctureTracker::initial_x [1] = $xm
PunctureTracker::which_surface_to_store_info[1] = 1
IOASCII::out0D_vars = "PunctureTracker::pt_loc"
Replace $xp and $xm with the initial x coordinates of the +x and -x
punctures (this will be \pm TwoPunctures::par_b). You then need to choose
two spherical surfaces
(https://docs.einsteintoolkit.org/et-docs/Einstein_Toolkit_standards) which
are not being used currently to store the puncture locations in. If there
is already 0D output, just add the puncturetracker variable separated with a
space.
I
did some plots, what are really the combinations xx, xy, xz, yy, zz ?
Are they related to the trajectories of one BH (projections)? What
about the second BH?
You should have BH_diagnostics.ah1.gp and BH_diagnostics.ah2.gp. These
correspond to the first and second BHs. If you look at the top of these
files, you will see an explanation of the columns. More information about
each column is available in the AHFinderDirect documentation
( http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14... ).
Specifically, information about the BH_diagnostics output files is
at http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14...
To plot the trajectory of one of the BHs, you want to plot column 3 and
column 4 in a parametric plot.
Cheers,
Jose
On Mon, Jan 16, 2012 at 4:39 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 14 Jan 2012, at 00:41, Jose Fiestas Iquira wrote:
Hi Ian,
I did the changes but it started and finished early. I am attaching
the qc0-mclachlan.err file. At the end there is a warning in one of
the Carpet files (mask_test.c)
The warning you see,
WARNING level 1 in thorn CarpetReduce processor 0 host qb517
(line 120 of
/home/ettest24/Cactus/arrangements/Carpet/CarpetReduce/src/mask_test.c):
-> Simulation domain volume and reduction weight sum differ
is a known issue (https://trac.einsteintoolkit.org/ticket/434) which was
fixed after the last release and the fix is only available in the
development version. I believe that the warning is not fatal, and it should
not have terminated the simulation. Can you also post the qc0-mclachlan.out
file?
Where can I find the puncture tracker?
The PunctureTracker thorn is part of the Einstein Toolkit thornlist, so you
should have it already. However, I think the best thing for you to do
initially would be to look at the BH_diagnostics files for the apparent
horizon centroid locations for tracking the locations of the BHs.
Thanks,
Jose
On Fri, Jan 13, 2012 at 4:38 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 13 Jan 2012, at 12:32, Jose Fiestas Iquira wrote:
Yes, trajectories and wave forms would be enough for now. Which are
the names of this files? Many of the .asc files have data like kxx,
kxy, kxz, etc. Is there any description of them?
Not that I am aware of. Files with extension .x.asc are 1D ascii output of
grid functions from CarpetIOASCII. kxy is the variable name for the x,y
component of the extrinsic curvature tensor K. The waveform file should be
in mp_psi4_l2_m2_r*.asc where * corresponds to the radius of the extraction
sphere. However, looking at the qc0-mclachlan.par file, I see that there is
no wave extraction switched on. You should be able to enable wave
extraction by adding something like the following to the parameter file:
ActiveThorns = "WeylScal4 CarpetInterp AEILocalInterp Multipole"
Multipole::nradii = 3
Multipole::radius[0] = 30
Multipole::radius[1] = 40
Multipole::radius[2] = 50
Multipole::variables = "WeylScal4::Psi4r{sw=-2 cmplx=’WeylScal4::Psi4i’}"
Multipole::l_max = 4
WeylScal4::fdOrder = 4
(I haven't tested this, so let me know if it doesn't work)
I've just found that the usual PunctureTracker thorn is not enabled in this
example parameter file. So to look at trajectories you will have to use the
apparent horizon centroids. These are in the files BH_diagnostics.ah*.gp.
The meanings of the columns are at the top of the files. You want the
centroid x and y columns to plot in GnuPlot.
Thanks,
Jose
On Fri, Jan 13, 2012 at 12:41 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 12 Jan 2012, at 23:46, Jose Fiestas Iquira wrote:
Dear all,
I run the tutorial and found other interesting examples like the
McLachlan tool for Binary Black Holes.
Regarding the McLachlan tool,
Could somebody help me finding a description of the output? Is all
contained in the .asc files? I would like to visualize it.
Hi Jose,
What sort of output do you want to visualise? Cactus thorns typically
output "lightweight" information such as BH trajectories and waveforms as
ASCII, and more heavyweight data -- like gridfunctions -- as HDF5. For the
trajectories (from thorn Puncture Tracker) and waveforms (from Multipole)
you can use gnuplot or anything that can read standard ASCII output to
visualise them. For gridfunctions in HDF5 format, I think the easiest way
to get started is to use VisIt, for which there is a Carpet plugin (though
installing VisIt is sometimes not so easy!).
Does anyone know if the Carpet plugin is distributed with VisIt yet? I
believe that was the intention.
Cheers,
Jose
On Wed, Jan 11, 2012 at 7:06 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Thanks to Erik for advise. ETK example simulation run successfully.
On Wed, Jan 11, 2012 at 11:49 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Right, it looks good now:
./simfactory/bin/sim list-simulations
static_tov [ACTIVE (RUNNING), restart 0000, job id 595490]
Let see how it looks when it finishes. The example was set to 8 hours.
Thanks,
Jose
On Wed, Jan 11, 2012 at 11:45 AM, Erik Schnetter schnetter@cct.lsu.edu
wrote:
On Wed, Jan 11, 2012 at 2:42 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
The only thing is that now
$ ./simfactory/bin/sim list-simulations
returns
Error: unknown command list-simulation
Aborting Simfactory.
Did you really type "list-simulations", and the error message says
"list-simulation" (without the final "s")?
-erik
--
Erik Schnetter schnetter@cct.lsu.edu http://www.cct.lsu.edu/~eschnett/
Users mailing list
Users@einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.err>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.out>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
On 8 Feb 2012, at 21:05, Jose Fiestas Iquira wrote:
Hi,
The first problem was solved as I reported, by disabling the QuasiLocalMeasures. Some MPI problem. Btw, which information am I missing without QuasiLocalMeasures? (the output file was the wrong one, of a successful simulation without WeyScal4)
About the .par file. I found out (please correct me if necessary) that I need to change the name of the run by the simfactory submission (here static_tov): ./simfactory/bin/sim submit static_tov --parfile=par/static_tov.par --procs=32 --walltime=8:0:0
Otherwise the run works with the original .par (first used with this run) , and not to the newly specified .par file. So, it worked by doing ./simfactory/bin/sim submit static_tov01 --parfile=par/static_tov_new.par --procs=32 --walltime=8:0:0
Simulation output goes to a directory of the same name as the simulation, so if you want a new simulation, you need to give it a different name. SimFactory interprets a second "submit" command on a simulation as an attempt to "resubmit" the simulation, which means to start a second Cactus run which continues the simulation from checkpoint files written by the first one. So if you have forgotten to change the name or to delete the first simulation directory, simfactory doesn't do what you wanted it to. I don't usually use the submit command for starting simulations, because of this problem. There is a command called "create-submit" which gives an error message if the simulation already exists. Maybe we should change simfactory to only accept explicit "submit" and "resubmit" commands to help this.
Best, Jose
On Wed, Feb 8, 2012 at 10:05 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 22:36, Jose Fiestas Iquira wrote:
Hi,
In order to use the wave generator, I am probably not submitting the job properly.
I had modified the .par file located at: /home/ettest24/Cactus/par/qc0-mclachlan.par since I run the example like this: ./simfactory/bin/sim submit qc0-mclachlan --parfile=par/qc0-mclachlan.par --procs=32 --walltime=8:0:0
Here the run stopped with errors, as previously reported.
Hi Jose,
Could you update me on your status? Did you solve this problem? In the standard output file that you sent to the ET mailing list on 16-Jan-2012, I do not see any problem. The run reached 100 M of evolution time. The only problem is a non-fatal warning message which is caused by some problem with the way that Carpet keeps track of points which should or should not be included in norm operations. You can ignore this warning.
Could be possible that the proper .par is located somewhere else? I found it also here:
/home/ettest24/Cactus/simfactory/etc/par/qc0-mclachlan.par /home/ettest24/Cactus/arrangements/McLachlan/par/qc0-mclachlan.par
Btw, in the last one WeylScal4 is activated, in a similar way as you told me.
SimFactory takes the parameter file that you specify on the command line - it does not look anywhere else. Can you be more specific about the problem you are having?
PS: if you ask these questions on the mailing list, other people can help and benefit from the discussion.
Best, Jose
On Mon, Jan 23, 2012 at 1:11 PM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 14:57, Jose Fiestas Iquira wrote:
Hello,
Regarding the Binary Black Hole coalescence example from the ETK.
In the BH diagnostics files are columns with quadrupole terms, which are not constant. Do you know of some tool in ETK to calculate the GW emission from these terms?
No, not that I am aware of. These are the quadrupole moments of the individual apparent horizons, rather than the quadrupole moment of the system as a whole. The gravitational wave emission can be computed by using the WeylScal4 and Multipole thorns. These give you the spherical harmonic multipoles of Psi4 (e.g. l = 2, m = 2 is the dominant one), which can then be converted to strain by "two integrations in time" [this is actually quite subtle and hard to get right - see Reisswig and Pollney, arXiv:1006.1632, for an overview and solution).
Thanks, Jose
On Tue, Jan 17, 2012 at 5:18 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 17 Jan 2012, at 13:37, Jose Fiestas Iquira wrote:
The Einstein Analysis guide is very useful. I found here
http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech142.html#x15...
the units used in 'Hydrobase'.
If I assume these are the units for all thorns, I obtain from data in
BH_diagnostics.ah1.gp a unit of distance of about 1 Km. To estimate
the time I am using a mass of ~0.5 M_sun for both BHs (m_irreducible
?) , and I read a final time (cctk_time) ~ 1microsec. Are these
quantities in the usual range?
The length unit is G M / c^2 and the time unit is G M / c^3. For M = M_sun,
I get the length unit as 1.5 km and the time unit as 4.9 us. The final time
in the simulation is 100 M (i.e. 100 G M / c^3), which would be 0.5 ms.
Jose
On Tue, Jan 17, 2012 at 4:04 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
I sent the wrong file, here is the right one.
$ hg tip
gives me:
changeset: 3421:dff4df2b3687
tag: tip
user: Erik Schnetter schnetter@cct.lsu.edu
date: Thu Jan 12 14:11:27 2012 -0500
summary: CarpetLib: Allocate space for dstbox before
restricting/prolongating
[40514 refs]
I copied this version from:
./GetComponents -a
http://svn.einsteintoolkit.org/manifest/trunk/einsteintoolkit.th
Thanks,
Jose
On Tue, Jan 17, 2012 at 1:13 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 16 Jan 2012, at 23:21, Jose Fiestas Iquira wrote:
Thanks, I am attaching the output file.
The output file looks fine - I can't see any problem with the simulation.
According to the output file, it reached t = 100 M, and in the parameter
file that is listed as cctk_final_time, so all looks well!
I did copy the development version. As I can see, it was already
fixed, so I should not get this warning, since I copied the
development version.
Yes, if you are using the current development version, then I don't know why
you get the warning. Can you check which version of Carpet you have by
doing
cd repos/carpet
hg tip
to be sure you are using the development version?
Is the puncture tracker not a better way to plot BH trajectories.
It's cheaper to use the puncture tracker, because computing the horizon
centroid is more expensive, but it has to be done every iteration, whereas
the horizon can be found less frequently. For comparison with
point-particle results, I don't think it makes much difference; the results
are gauge dependent anyway so differences on the order of the horizon size
cannot be interpreted physically. Since the parameter file already has
horizon-finding built in, it is easier for you to use that.
If you want to figure out how to add the puncture tracker to the simulation,
you can start with these parameters:
ActiveThorns = "PunctureTracker"
PunctureTracker::track [0] = yes
PunctureTracker::initial_x [0] = $xp
PunctureTracker::which_surface_to_store_info[0] = 0
PunctureTracker::track [1] = yes
PunctureTracker::initial_x [1] = $xm
PunctureTracker::which_surface_to_store_info[1] = 1
IOASCII::out0D_vars = "PunctureTracker::pt_loc"
Replace $xp and $xm with the initial x coordinates of the +x and -x
punctures (this will be \pm TwoPunctures::par_b). You then need to choose
two spherical surfaces
(https://docs.einsteintoolkit.org/et-docs/Einstein_Toolkit_standards) which
are not being used currently to store the puncture locations in. If there
is already 0D output, just add the puncturetracker variable separated with a
space.
I
did some plots, what are really the combinations xx, xy, xz, yy, zz ?
Are they related to the trajectories of one BH (projections)? What
about the second BH?
You should have BH_diagnostics.ah1.gp and BH_diagnostics.ah2.gp. These
correspond to the first and second BHs. If you look at the top of these
files, you will see an explanation of the columns. More information about
each column is available in the AHFinderDirect documentation
(http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14...).
Specifically, information about the BH_diagnostics output files is
at http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14...
To plot the trajectory of one of the BHs, you want to plot column 3 and
column 4 in a parametric plot.
Cheers,
Jose
On Mon, Jan 16, 2012 at 4:39 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 14 Jan 2012, at 00:41, Jose Fiestas Iquira wrote:
Hi Ian,
I did the changes but it started and finished early. I am attaching
the qc0-mclachlan.err file. At the end there is a warning in one of
the Carpet files (mask_test.c)
The warning you see,
WARNING level 1 in thorn CarpetReduce processor 0 host qb517
(line 120 of
/home/ettest24/Cactus/arrangements/Carpet/CarpetReduce/src/mask_test.c):
-> Simulation domain volume and reduction weight sum differ
is a known issue (https://trac.einsteintoolkit.org/ticket/434) which was
fixed after the last release and the fix is only available in the
development version. I believe that the warning is not fatal, and it should
not have terminated the simulation. Can you also post the qc0-mclachlan.out
file?
Where can I find the puncture tracker?
The PunctureTracker thorn is part of the Einstein Toolkit thornlist, so you
should have it already. However, I think the best thing for you to do
initially would be to look at the BH_diagnostics files for the apparent
horizon centroid locations for tracking the locations of the BHs.
Thanks,
Jose
On Fri, Jan 13, 2012 at 4:38 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 13 Jan 2012, at 12:32, Jose Fiestas Iquira wrote:
Yes, trajectories and wave forms would be enough for now. Which are
the names of this files? Many of the .asc files have data like kxx,
kxy, kxz, etc. Is there any description of them?
Not that I am aware of. Files with extension .x.asc are 1D ascii output of
grid functions from CarpetIOASCII. kxy is the variable name for the x,y
component of the extrinsic curvature tensor K. The waveform file should be
in mp_psi4_l2_m2_r*.asc where * corresponds to the radius of the extraction
sphere. However, looking at the qc0-mclachlan.par file, I see that there is
no wave extraction switched on. You should be able to enable wave
extraction by adding something like the following to the parameter file:
ActiveThorns = "WeylScal4 CarpetInterp AEILocalInterp Multipole"
Multipole::nradii = 3
Multipole::radius[0] = 30
Multipole::radius[1] = 40
Multipole::radius[2] = 50
Multipole::variables = "WeylScal4::Psi4r{sw=-2 cmplx=’WeylScal4::Psi4i’}"
Multipole::l_max = 4
WeylScal4::fdOrder = 4
(I haven't tested this, so let me know if it doesn't work)
I've just found that the usual PunctureTracker thorn is not enabled in this
example parameter file. So to look at trajectories you will have to use the
apparent horizon centroids. These are in the files BH_diagnostics.ah*.gp.
The meanings of the columns are at the top of the files. You want the
centroid x and y columns to plot in GnuPlot.
Thanks,
Jose
On Fri, Jan 13, 2012 at 12:41 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 12 Jan 2012, at 23:46, Jose Fiestas Iquira wrote:
Dear all,
I run the tutorial and found other interesting examples like the
McLachlan tool for Binary Black Holes.
Regarding the McLachlan tool,
Could somebody help me finding a description of the output? Is all
contained in the .asc files? I would like to visualize it.
Hi Jose,
What sort of output do you want to visualise? Cactus thorns typically
output "lightweight" information such as BH trajectories and waveforms as
ASCII, and more heavyweight data -- like gridfunctions -- as HDF5. For the
trajectories (from thorn Puncture Tracker) and waveforms (from Multipole)
you can use gnuplot or anything that can read standard ASCII output to
visualise them. For gridfunctions in HDF5 format, I think the easiest way
to get started is to use VisIt, for which there is a Carpet plugin (though
installing VisIt is sometimes not so easy!).
Does anyone know if the Carpet plugin is distributed with VisIt yet? I
believe that was the intention.
Cheers,
Jose
On Wed, Jan 11, 2012 at 7:06 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Thanks to Erik for advise. ETK example simulation run successfully.
On Wed, Jan 11, 2012 at 11:49 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Right, it looks good now:
./simfactory/bin/sim list-simulations
static_tov [ACTIVE (RUNNING), restart 0000, job id 595490]
Let see how it looks when it finishes. The example was set to 8 hours.
Thanks,
Jose
On Wed, Jan 11, 2012 at 11:45 AM, Erik Schnetter schnetter@cct.lsu.edu
wrote:
On Wed, Jan 11, 2012 at 2:42 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
The only thing is that now
$ ./simfactory/bin/sim list-simulations
returns
Error: unknown command list-simulation
Aborting Simfactory.
Did you really type "list-simulations", and the error message says
"list-simulation" (without the final "s")?
-erik
--
Erik Schnetter schnetter@cct.lsu.edu http://www.cct.lsu.edu/~eschnett/
Users mailing list
Users@einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.err>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.out>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
On Wed, Feb 8, 2012 at 3:37 PM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 8 Feb 2012, at 21:05, Jose Fiestas Iquira wrote:
Hi,
The first problem was solved as I reported, by disabling the QuasiLocalMeasures. Some MPI problem. Btw, which information am I missing without QuasiLocalMeasures? (the output file was the wrong one, of a successful simulation without WeyScal4)
About the .par file. I found out (please correct me if necessary) that I need to change the name of the run by the simfactory submission (here static_tov): ./simfactory/bin/sim submit static_tov --parfile=par/static_tov.par --procs=32 --walltime=8:0:0
Otherwise the run works with the original .par (first used with this run) , and not to the newly specified .par file. So, it worked by doing ./simfactory/bin/sim submit static_tov01 --parfile=par/static_tov_new.par --procs=32 --walltime=8:0:0
Simulation output goes to a directory of the same name as the simulation, so if you want a new simulation, you need to give it a different name. SimFactory interprets a second "submit" command on a simulation as an attempt to "resubmit" the simulation, which means to start a second Cactus run which continues the simulation from checkpoint files written by the first one. So if you have forgotten to change the name or to delete the first simulation directory, simfactory doesn't do what you wanted it to. I don't usually use the submit command for starting simulations, because of this problem. There is a command called "create-submit" which gives an error message if the simulation already exists. Maybe we should change simfactory to only accept explicit "submit" and "resubmit" commands to help this.
Yes, we should. Do you want to add this to the Trac?
We also should complain if a user specifies a parameter file option that is ignored.
-erik
Best, Jose
On Wed, Feb 8, 2012 at 10:05 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 22:36, Jose Fiestas Iquira wrote:
Hi,
In order to use the wave generator, I am probably not submitting the job properly.
I had modified the .par file located at: /home/ettest24/Cactus/par/qc0-mclachlan.par since I run the example like this: ./simfactory/bin/sim submit qc0-mclachlan --parfile=par/qc0-mclachlan.par --procs=32 --walltime=8:0:0
Here the run stopped with errors, as previously reported.
Hi Jose,
Could you update me on your status? Did you solve this problem? In the standard output file that you sent to the ET mailing list on 16-Jan-2012, I do not see any problem. The run reached 100 M of evolution time. The only problem is a non-fatal warning message which is caused by some problem with the way that Carpet keeps track of points which should or should not be included in norm operations. You can ignore this warning.
Could be possible that the proper .par is located somewhere else? I found it also here:
/home/ettest24/Cactus/simfactory/etc/par/qc0-mclachlan.par /home/ettest24/Cactus/arrangements/McLachlan/par/qc0-mclachlan.par
Btw, in the last one WeylScal4 is activated, in a similar way as you told me.
SimFactory takes the parameter file that you specify on the command line - it does not look anywhere else. Can you be more specific about the problem you are having?
PS: if you ask these questions on the mailing list, other people can help and benefit from the discussion.
Best, Jose
On Mon, Jan 23, 2012 at 1:11 PM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 23 Jan 2012, at 14:57, Jose Fiestas Iquira wrote:
Hello,
Regarding the Binary Black Hole coalescence example from the ETK.
In the BH diagnostics files are columns with quadrupole terms, which are not constant. Do you know of some tool in ETK to calculate the GW emission from these terms?
No, not that I am aware of. These are the quadrupole moments of the individual apparent horizons, rather than the quadrupole moment of the system as a whole. The gravitational wave emission can be computed by using the WeylScal4 and Multipole thorns. These give you the spherical harmonic multipoles of Psi4 (e.g. l = 2, m = 2 is the dominant one), which can then be converted to strain by "two integrations in time" [this is actually quite subtle and hard to get right - see Reisswig and Pollney, arXiv:1006.1632, for an overview and solution).
Thanks, Jose
On Tue, Jan 17, 2012 at 5:18 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 17 Jan 2012, at 13:37, Jose Fiestas Iquira wrote:
The Einstein Analysis guide is very useful. I found here
http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech142.html#x15...
the units used in 'Hydrobase'.
If I assume these are the units for all thorns, I obtain from data in
BH_diagnostics.ah1.gp a unit of distance of about 1 Km. To estimate
the time I am using a mass of ~0.5 M_sun for both BHs (m_irreducible
?) , and I read a final time (cctk_time) ~ 1microsec. Are these
quantities in the usual range?
The length unit is G M / c^2 and the time unit is G M / c^3. For M = M_sun,
I get the length unit as 1.5 km and the time unit as 4.9 us. The final time
in the simulation is 100 M (i.e. 100 G M / c^3), which would be 0.5 ms.
Jose
On Tue, Jan 17, 2012 at 4:04 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
I sent the wrong file, here is the right one.
$ hg tip
gives me:
changeset: 3421:dff4df2b3687
tag: tip
user: Erik Schnetter schnetter@cct.lsu.edu
date: Thu Jan 12 14:11:27 2012 -0500
summary: CarpetLib: Allocate space for dstbox before
restricting/prolongating
[40514 refs]
I copied this version from:
./GetComponents -a
http://svn.einsteintoolkit.org/manifest/trunk/einsteintoolkit.th
Thanks,
Jose
On Tue, Jan 17, 2012 at 1:13 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 16 Jan 2012, at 23:21, Jose Fiestas Iquira wrote:
Thanks, I am attaching the output file.
The output file looks fine - I can't see any problem with the simulation.
According to the output file, it reached t = 100 M, and in the parameter
file that is listed as cctk_final_time, so all looks well!
I did copy the development version. As I can see, it was already
fixed, so I should not get this warning, since I copied the
development version.
Yes, if you are using the current development version, then I don't know why
you get the warning. Can you check which version of Carpet you have by
doing
cd repos/carpet
hg tip
to be sure you are using the development version?
Is the puncture tracker not a better way to plot BH trajectories.
It's cheaper to use the puncture tracker, because computing the horizon
centroid is more expensive, but it has to be done every iteration, whereas
the horizon can be found less frequently. For comparison with
point-particle results, I don't think it makes much difference; the results
are gauge dependent anyway so differences on the order of the horizon size
cannot be interpreted physically. Since the parameter file already has
horizon-finding built in, it is easier for you to use that.
If you want to figure out how to add the puncture tracker to the simulation,
you can start with these parameters:
ActiveThorns = "PunctureTracker"
PunctureTracker::track [0] = yes
PunctureTracker::initial_x [0] = $xp
PunctureTracker::which_surface_to_store_info[0] = 0
PunctureTracker::track [1] = yes
PunctureTracker::initial_x [1] = $xm
PunctureTracker::which_surface_to_store_info[1] = 1
IOASCII::out0D_vars = "PunctureTracker::pt_loc"
Replace $xp and $xm with the initial x coordinates of the +x and -x
punctures (this will be \pm TwoPunctures::par_b). You then need to choose
two spherical surfaces
(https://docs.einsteintoolkit.org/et-docs/Einstein_Toolkit_standards) which
are not being used currently to store the puncture locations in. If there
is already 0D output, just add the puncturetracker variable separated with a
space.
I
did some plots, what are really the combinations xx, xy, xz, yy, zz ?
Are they related to the trajectories of one BH (projections)? What
about the second BH?
You should have BH_diagnostics.ah1.gp and BH_diagnostics.ah2.gp. These
correspond to the first and second BHs. If you look at the top of these
files, you will see an explanation of the columns. More information about
each column is available in the AHFinderDirect documentation
(http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14...).
Specifically, information about the BH_diagnostics output files is
at http://einsteintoolkit.org/documentation/ThornGuide/ThornGuidech128.html#x14...
To plot the trajectory of one of the BHs, you want to plot column 3 and
column 4 in a parametric plot.
Cheers,
Jose
On Mon, Jan 16, 2012 at 4:39 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 14 Jan 2012, at 00:41, Jose Fiestas Iquira wrote:
Hi Ian,
I did the changes but it started and finished early. I am attaching
the qc0-mclachlan.err file. At the end there is a warning in one of
the Carpet files (mask_test.c)
The warning you see,
WARNING level 1 in thorn CarpetReduce processor 0 host qb517
(line 120 of
/home/ettest24/Cactus/arrangements/Carpet/CarpetReduce/src/mask_test.c):
-> Simulation domain volume and reduction weight sum differ
is a known issue (https://trac.einsteintoolkit.org/ticket/434) which was
fixed after the last release and the fix is only available in the
development version. I believe that the warning is not fatal, and it should
not have terminated the simulation. Can you also post the qc0-mclachlan.out
file?
Where can I find the puncture tracker?
The PunctureTracker thorn is part of the Einstein Toolkit thornlist, so you
should have it already. However, I think the best thing for you to do
initially would be to look at the BH_diagnostics files for the apparent
horizon centroid locations for tracking the locations of the BHs.
Thanks,
Jose
On Fri, Jan 13, 2012 at 4:38 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 13 Jan 2012, at 12:32, Jose Fiestas Iquira wrote:
Yes, trajectories and wave forms would be enough for now. Which are
the names of this files? Many of the .asc files have data like kxx,
kxy, kxz, etc. Is there any description of them?
Not that I am aware of. Files with extension .x.asc are 1D ascii output of
grid functions from CarpetIOASCII. kxy is the variable name for the x,y
component of the extrinsic curvature tensor K. The waveform file should be
in mp_psi4_l2_m2_r*.asc where * corresponds to the radius of the extraction
sphere. However, looking at the qc0-mclachlan.par file, I see that there is
no wave extraction switched on. You should be able to enable wave
extraction by adding something like the following to the parameter file:
ActiveThorns = "WeylScal4 CarpetInterp AEILocalInterp Multipole"
Multipole::nradii = 3
Multipole::radius[0] = 30
Multipole::radius[1] = 40
Multipole::radius[2] = 50
Multipole::variables = "WeylScal4::Psi4r{sw=-2 cmplx=’WeylScal4::Psi4i’}"
Multipole::l_max = 4
WeylScal4::fdOrder = 4
(I haven't tested this, so let me know if it doesn't work)
I've just found that the usual PunctureTracker thorn is not enabled in this
example parameter file. So to look at trajectories you will have to use the
apparent horizon centroids. These are in the files BH_diagnostics.ah*.gp.
The meanings of the columns are at the top of the files. You want the
centroid x and y columns to plot in GnuPlot.
Thanks,
Jose
On Fri, Jan 13, 2012 at 12:41 AM, Ian Hinder ian.hinder@aei.mpg.de wrote:
On 12 Jan 2012, at 23:46, Jose Fiestas Iquira wrote:
Dear all,
I run the tutorial and found other interesting examples like the
McLachlan tool for Binary Black Holes.
Regarding the McLachlan tool,
Could somebody help me finding a description of the output? Is all
contained in the .asc files? I would like to visualize it.
Hi Jose,
What sort of output do you want to visualise? Cactus thorns typically
output "lightweight" information such as BH trajectories and waveforms as
ASCII, and more heavyweight data -- like gridfunctions -- as HDF5. For the
trajectories (from thorn Puncture Tracker) and waveforms (from Multipole)
you can use gnuplot or anything that can read standard ASCII output to
visualise them. For gridfunctions in HDF5 format, I think the easiest way
to get started is to use VisIt, for which there is a Carpet plugin (though
installing VisIt is sometimes not so easy!).
Does anyone know if the Carpet plugin is distributed with VisIt yet? I
believe that was the intention.
Cheers,
Jose
On Wed, Jan 11, 2012 at 7:06 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Thanks to Erik for advise. ETK example simulation run successfully.
On Wed, Jan 11, 2012 at 11:49 AM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
Right, it looks good now:
./simfactory/bin/sim list-simulations
static_tov [ACTIVE (RUNNING), restart 0000, job id 595490]
Let see how it looks when it finishes. The example was set to 8 hours.
Thanks,
Jose
On Wed, Jan 11, 2012 at 11:45 AM, Erik Schnetter schnetter@cct.lsu.edu
wrote:
On Wed, Jan 11, 2012 at 2:42 PM, Jose Fiestas Iquira jafiestas@lbl.gov
wrote:
The only thing is that now
$ ./simfactory/bin/sim list-simulations
returns
Error: unknown command list-simulation
Aborting Simfactory.
Did you really type "list-simulations", and the error message says
"list-simulation" (without the final "s")?
-erik
--
Erik Schnetter schnetter@cct.lsu.edu http://www.cct.lsu.edu/~eschnett/
Users mailing list
Users@einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.err>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
<qc0-mclachlan.out>
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
--
Ian Hinder
http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
-- Ian Hinder http://numrel.aei.mpg.de/people/hinder
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Ian Hinder -
Jose Fiestas Iquira