open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I just tried it once more on the lock-nonhydro test case and it fails (v6p3r2). As mentionned in the first posts of this thread, this seems to be a python-specific problem in 6.3. I am on Win7 with gfortran and python.
What is strange is that yugi managed to have it working on 6.3 with python but I never did. I also had feedback by personal message from other user that also have the same problem.

And I never faced compilation/installation problems so far...

Thanks in advance for your help...!

Regards
PL

And the problem is the same on vent, see the attached screenshot from the console.





Regards,
PL

Good afternoon all,

By any chance has anyone beeing able to overcome this issue? I´m have the same problem.

Thank you very much.
Marcus

Hi Marcus,

Just a quick note. I had some problems with Postel3d as well at the beginning. I am not sure if it is same as you have. I tried to use Postel3d in parallel and it is not possible, I am afraid. I use v6p3, on Linux with Python. So if you are trying to run in parallel then it is not designed for it yet, but it is working in scalar. So I run the case in parallel and afterwards extract cross sections with Postel3d in scalar. For that you need to have to modes in your configuration file. But maybe your problem is different than mine. Good luck! Let me know if you need my config file as an example if you are using Linux and Python.

Kind Regards!

Violet

Hi Marcus & Violet,

I guess no (I am on win7/python/gfortran). Maybe the python bug can be fixed in a new release (v6p3r3...)?
In the meanwhile, you need to run postel3d with v6p2...

Good luck,
PL

Dear Violet,

Thank you for your response and will to help, but my problem is similar to PL´s. I´m using Python in scalar mode and POSTEL crashes just as PL posted some time ago.

Dear PL,

Thank you too, I have the same issue as yours. I will solve it by using V6p2 until we have a better understanding of the issue.

Best regards to both,
Marcus

Good Luck Marcus!:)

Violet

Hi all!

I have been working on a complex T3D model using version 6.3.2. As I don't have access to the trunk (in which it seems the the python-related bug has been corrected according to a post by JMH), I thought of extracting my cross sections by running postel3d with 6.2.

BUT this did not work, it seems that there is a compatibility problem when running postel3d in 6.2 using a result file created with 6.3. The programs says that all of my cross section coordinates don't intersect the domain (which is not relevant), and the program stops after a call to LIT which points out a READ ERROR, with problem of type R4 on LU 4.

This problem is solved by running postel3d on a T3D result file created with a restart procedure with 6.2. BUT, as my model is complex, the restart procedure brings errors compared with 6.3...

So, my question is: is it 100% confirmed that the postel3d version available on the trunk fully works with python? If yes, how is it possible to get it?

A last question: is it possible to force the reading of PRIVE variables in a restart procedure?

Thanks very much in advance!

Regards
PL

Hi!

Does someone have any tips regarding the previous post?

Thanks in advance!

Regards
PL

Dear colegues,

Has anyone been able to use POSTE3D in v6p3? Or we still have to use it on v692?

Thank you all
Marcus