Hi Bruno.

I saw that VisiT wanted to open all the files as a single "database" by default, but I explicitly told it *not* to do this. Perhaps it disobeyed me and did it anyway? The impression I had was that the database assumed the different files were a time series rather than a collection of 3D patches of a single time level.

Beany

On 21 June 2014 04:11, Bruno Giacomazzo bruno.giacomazzo@unitn.it wrote:

Hi Beany,

On Jun 20, 2014, at 10:20 PM, Bernard Kelly physicsbeany@gmail.com wrote:

A summer intern here has been doing some visualizations with CarpetIOHDF5 2D data in VisiT, and it all seems to be going well. However, we're having trouble using the 3D output.

• using the default "chunked" output, and looking at individual files

*seems* to load OK, but the file contents are suspiciously similar across all cores (e.g., all show the same domain size, rather than specific chunks of it as I'd expect)

this is the normal behavior. If you open *file_0.h5 (or other files) the plugin will open automatically all the other files and show you the full domain.

• switching to the "unchunked" output (which takes a long time for Carpet

to create), loading goes OK, but any attempt to visualize hangs, with an error message along the lines of "engine abnormally exited"

no experience on this.

Cheers, Bruno

We're using fairly up-to-date VisiT (v. 2.7.2 / 2.7.3).

Any insight welcomed. Thanks,

Bernard _______________________________________________ Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users

Dr. Bruno Giacomazzo Department of Physics University of Trento via Sommarive 14 38123 Trento Italy

Tel. : +39 0461281631 email : bruno.giacomazzo@unitn.it web: http://www.brunogiacomazzo.org

---

There are only 10 types of people in the world: Those who understand binary, and those who don't

---

Thanks, Roland. So resistance is futile, then?

We'll have to take a closer look at the resulting data, which appeared to be trivial at first glance.

This means that taking the time to output unchunked 3D HDF5 data from Carpet is doubly stupid ...

B

On 21 June 2014 11:26, Roland Haas roland.haas@physics.gatech.edu wrote:

Hello Bernard,

...

I saw that VisiT wanted to open all the files as a single "database" by default, but I explicitly told it *not* to do this. Perhaps it disobeyed

me

...

and did it anyway? The impression I had was that the database assumed the different files were a time series rather than a collection of 3D patches of a single time level.

The CarpetHDF5 VisIt plugin contains explicit code to ignore VisIt's attempts to treat file_0.h5 file_1.h5 etc as a time series and will always behave (at least in the current version) as if only file_0 was given. It will then behave as Bruno described and on its own seek out file_1 file_2 etc.

I am not even sure if you can open just file_1 or if it will automatically revert to opening all files. You'd have to physically move file_0 etc out of the way to test how it behaves when opening file_1 alone I think.

Yours, Roland

-- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://keys.gnupg.net.

---

Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users