Hi Roland,
I use VisIt too for the HDF5 output and sure, it does pick up the files from all processes (var.file_*.h5) to open them as as a single (virtual) database.
Your second point does answer my question anyway. Thanks. I'll have to figure out some work-around, e.g. processing the output from each individual checkpoint directory.
Regards, Dumsani
On 30/09/2016 22:54, Roland Haas wrote:
Hello Dumsani,
Now, for purposes of visualizing the data, for instance, one would be interested in collecting all these data chunks into a single file. The question is how to bundle these into a single file. I wanted to know if any of the available Carpet tools can help with this.
This depends on the visualization software you would like to use. VisIt will automatically use all rho files if you open the rho.file_0.h5 file so you typically do not have to do anything. It will however not combine files from the different output-???? directories. I do not know of any way to have VisIt read them all and present them as a single database.
If you want to combine really combine the files, you can use the hdf5_slicer command that is part of CarpetIOHDF5 and that you can find in exe/sim/hdf5_slicer (it can be used to combine files if you give it the out3d-cube option). For your files, if you want a single file (keep in mind Ian's comment it will just copy the datasets, no actual merging of the datasets happens) you can use:
hdf5_slicer --out3d-cube output-????/<parfile>/{rho,eps}.file_*.h5 rho.h5
Yours, Roland
On Sat, Oct 01, 2016 at 01:38:25PM +0200, dumsani wrote:
I use VisIt too for the HDF5 output and sure, it does pick up the files from all processes (var.file_*.h5) to open them as as a single (virtual) database.
Your second point does answer my question anyway. Thanks. I'll have to figure out some work-around, e.g. processing the output from each individual checkpoint directory.
Take a look at the following:
http://lists.einsteintoolkit.org/pipermail/users/2016-May/004802.html
And especially this script:
http://lists.einsteintoolkit.org/pipermail/users/attachments/20160517/3d6634...
It might not do exactly what you want, but it should be easy to adapt.
Frank
Hi Frank, Thanks for the links. I'll have a look.
On 04/10/2016 16:24, Frank Loeffler wrote:
On Sat, Oct 01, 2016 at 01:38:25PM +0200, dumsani wrote:
I use VisIt too for the HDF5 output and sure, it does pick up the files from all processes (var.file_*.h5) to open them as as a single (virtual) database.
Your second point does answer my question anyway. Thanks. I'll have to figure out some work-around, e.g. processing the output from each individual checkpoint directory.
Take a look at the following:
http://lists.einsteintoolkit.org/pipermail/users/2016-May/004802.html
And especially this script:
http://lists.einsteintoolkit.org/pipermail/users/attachments/20160517/3d6634...
It might not do exactly what you want, but it should be easy to adapt.
Frank
Hi Frank, Everyone,
I have looked at the links you sent through. Exactly the same topic I'm interested in. I'm sorry not to have seen and read that thread first.
To begin with, I am trying the option where I must compile the updated CarpetHDF5 reader plugin into my VisIt installation. I traced that version down to this site https://bitbucket.org/rhaas80/carpethdf5/overview (thanks to the communication thread). At the moment, I need to resolve a few issues in the build (it fails due to a missing library):
*** No rule to make target `/usr/local/visit/2.10.3/linux-x86_64/lib/libdl.so', needed by `/home/dumsani/.visit/2.10.3/linux-x86_64/plugins/databases/libECarpetHDF5Database_par.so'. Stop. make[1]: *** [CMakeFiles/ECarpetHDF5Database_par.dir/all] Error 2
But I think the VisIt people would be the ones to help with this build issue. I'm just mentioning it here in case someone else happens to have come across and solved this before.
Thanks for the python script as well.
On 04/10/2016 17:05, dumsani wrote:
Hi Frank, Thanks for the links. I'll have a look.
On 04/10/2016 16:24, Frank Loeffler wrote:
On Sat, Oct 01, 2016 at 01:38:25PM +0200, dumsani wrote:
I use VisIt too for the HDF5 output and sure, it does pick up the files from all processes (var.file_*.h5) to open them as as a single (virtual) database.
Your second point does answer my question anyway. Thanks. I'll have to figure out some work-around, e.g. processing the output from each individual checkpoint directory.
Take a look at the following:
http://lists.einsteintoolkit.org/pipermail/users/2016-May/004802.html
And especially this script:
http://lists.einsteintoolkit.org/pipermail/users/attachments/20160517/3d6634...
It might not do exactly what you want, but it should be easy to adapt.
Frank
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
Regards,
users@lists.einsteintoolkit.org
-
dumsani -
Frank Loeffler