Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update Ocean2D example in FabUQCampaign.py to use the refactored EasyVVUQ #100

Closed
djgroen opened this issue May 17, 2019 · 9 comments
Closed

Comments

@djgroen
Copy link
Owner

@djgroen djgroen commented May 17, 2019

It's this function:
def ocean2D(**args):
in this file:
http://www.github.com/wedeling/FabUQCampaign.py

SOFT DEADLINE: 23rd of May
HARD DEADLINE: 29th of May (but then you'll also have to revise the tutorial document in FabUQCampaign! ;))

@wedeling
Copy link
Collaborator

@wedeling wedeling commented May 17, 2019

I'll have a look at this early next week

Loading

@wedeling
Copy link
Collaborator

@wedeling wedeling commented May 20, 2019

Hi Derek, quick question. How do I run your ocean2D subroutine? My old script doesn't work any more now that run_UQ_sample is removed.

Loading

@djgroen
Copy link
Owner Author

@djgroen djgroen commented May 20, 2019

Loading

@djgroen
Copy link
Owner Author

@djgroen djgroen commented May 21, 2019

Okay, I committed the fix now. You can run the script through FabSim3 using

cd config_files/ocean2d
fab localhost ocean2d:ocean2d:label=1

I hope to be able to remove the first line after the refactoring with EasyVVUQ is done.

Loading

@wedeling
Copy link
Collaborator

@wedeling wedeling commented May 27, 2019

Hi Derek, I have Stochastic Collocation working with the refactored EasyVVUQ, although I haven't pushed the new version yet. Anyway, when I run your command above I get a Command(s) not found:ocean2d error, and the same when I try to run uq_ensemble from within my EasyVVUQ script. It seems I cannot access any subroutine in FabUQCampaign.py. Is there something which I might be messing up?

Loading

@wedeling
Copy link
Collaborator

@wedeling wedeling commented May 28, 2019

I've pushed both the refactored advection diffusion and ocean model examples. I also updated both tutorials (the readme and the separate ocean tutorial). I'll fine tune them once we work out the issue above.

Loading

@wedeling
Copy link
Collaborator

@wedeling wedeling commented Jun 4, 2019

Hi @djgroen

I have managed to fix the issues above with the help of Hamid. Turns out that the import statements above the ocean_2d task in FabUQCampaign,py were creating a problem. If I comment them out, FabSim recognizes the other subroutines in the same file again.

I had another issues as well, which I fixed and pushed:

  1. If I keep sim_ID fixed (i.e. the sim_ID needed for campaign2ensemble), and rerun a campaign with a different polynomial order, the old samples get executed again, along side the new ones. This is because the old samples do not get deleted from <fab_home>/config_files/<sim_ID>/SWEEP. I fixed this by assigning a dynamic sim_ID (the same name used for the EasyVVUQ campaign), such that it is different every time a campaign is launched.

  2. The previous version had quite a lot of hard coded variables. I now made a more generic run_FabUQ_ensemble subroutine to run the ensemble. It is located in the same file as the example scripts. Only the fab home directory needs to be hard coded now.

I've updated the advection diffusion tutorial (https://github.com/wedeling/FabUQCampaign/blob/master/README.md) and the ocean tutorial (https://github.com/wedeling/FabUQCampaign/blob/master/Tutorial_ocean.md)

Loading

@djgroen
Copy link
Owner Author

@djgroen djgroen commented Jun 10, 2019

@wedeling I think we have this tutorial fully working, correct? If so, I think we can close this now :).

Loading

@wedeling
Copy link
Collaborator

@wedeling wedeling commented Jun 10, 2019

I agree :)

Loading

@wedeling wedeling closed this Jun 10, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Linked pull requests

Successfully merging a pull request may close this issue.

None yet
4 participants