New tutorial: Partitioned Burgers eq. 1D by vidulejs · Pull Request #670 · precice/tutorials

vidulejs · 2025-09-29T15:15:41Z

New tutorial case: Partitioned Burgers' Equation in 1D solved with SciPy Finite Volume implementation and a Neural Network surrogate model !

Fix README.md to comply with the standard
How to run surrogate,
Improve README.md further
Clean up the scripts in base directory
What to do with the initial condition generation? Steps are too complex currently
How to reproduce the offline model, i.e., training scripts
Either remove old .solver-nutils-dgalerkin or fix
Read through again https://precice.org/community-contribute-to-precice.html
Allow skipping venv setup

Checklist:

I added a summary of any user-facing changes (compared to the last release) in the changelog-entries/<PRnumber>.md.
I will remember to squash-and-merge, providing a useful summary of the changes of this PR.

… in visualize_partitioned_domain.py. Wrap boundary conditions in new class BoundaryWrapper because they are dependent on the solver iterations. Implement implicit Euler (and Jacobian and residual for the root finding iteration. Initialize participants. Zero gradient condition on the outside boundaries.

…-surrogate.sh, generate-training-data.sh. Fix floating point issue in t_index. Add --savefile argument to solver-scipy-fvolumes.py to save data for training.

…n scripts. Fix requirements.txt for scipy and surrogate solvers. Add example images for README

…ipt to check for initial conditions after installing environment.

uekerman

Thanks for the very nice addition already! I only had brief look in this first iteration and did not yet try to run things. Please see comments below.

Let's indeed remove the Nutils solver. Could always be added later again.

partitioned-burgers-1d/README.md

partitioned-burgers-1d/solver-scipy/solver.py

partitioned-burgers-1d/precice-config.xml

partitioned-burgers-1d/neumann-surrogate/run.sh

uekerman · 2025-10-21T11:10:06Z

partitioned-burgers-1d/neumann-surrogate/clean.sh

+rm -rf precice-profiling
+rm -f neumann-surrogate.log precice-Neumann-convergence.log precice-Neumann-iterations.log
+rm -f surrogate.npz


Compare how other tutorials do this.

Renamed clean.sh to clean-tutorial.sh. But should it be symlinked to clean-tutorial-base.sh? I need these additional commands.

Use tooling in ../tools/cleaning-tools.sh and add a function there if necessary.

partitioned-burgers-1d/clean.sh

…arate from images directory.

… versions in requirements. Remove commented out code from solvers.

partitioned-burgers-1d/README.md

#680

uekerman

I was able to run the cases and got meaningful results. We are probably close to merging this. Mainly cosmetics now.

I think a few scripts of utils could go into solver-scipy. Only keep those in utils that are used by neumann-surrogate as well.

uekerman · 2026-02-06T15:02:18Z

changelog-entries/670.md

@@ -0,0 +1 @@
+New tutorial case: Partitioned Burgers' Equation in 1D solved with Finite Volumes and a Neural Network surrogate model !


Suggested change

New tutorial case: Partitioned Burgers' Equation in 1D solved with Finite Volumes and a Neural Network surrogate model !

- Added new case: 1D partitioned Burgers' equation with one finite volume and one NN surrogate participant.

uekerman · 2026-02-06T15:10:02Z

partitioned-burgers-1d/README.md

+
+---


Suggested change

---

Not used in the docs, but should also not be necessary here.

uekerman · 2026-02-06T15:18:44Z

partitioned-burgers-1d/README.md

+The conservative formulation of the Burgers' equation `solver-scipy` is implemented in a first-order finite volume code using Lax-Friedrichs fluxes and implicit Euler time stepping.
+
+This tutorial includes two versions for the Neumann participant:
+
+- A standard finite volume solver (`neumann-scipy`).
+- A pre-trained neural network surrogate that approximates the solver (`neumann-surrogate`).


Suggested change

The conservative formulation of the Burgers' equation `solver-scipy` is implemented in a first-order finite volume code using Lax-Friedrichs fluxes and implicit Euler time stepping.

This tutorial includes two versions for the Neumann participant:

- A standard finite volume solver (`neumann-scipy`).

- A pre-trained neural network surrogate that approximates the solver (`neumann-surrogate`).

Currently, the SciPy solver can be used for both side, the NN surrogate solver only for the Neumann side.

- SciPy. Simple finite volume solver using Lax-Friedrichs fluxes and implicit Euler time stepping.

- Surrogate. Pre-trained neural network surrogate.

To be consistent with other tutorials.
TODO: It would be nice to write a few more words on the surrogate here.

uekerman · 2026-02-06T15:20:20Z

partitioned-burgers-1d/README.md

+<p align="center">
+  <img src="images/tutorials-partitioned-burgers-1d-precice-config.png" alt="preCICE configuration visualization" width="600"/>
+</p>


I don't think we need html for the images in the README, we normally don't use it here.

Suggested change

<p align="center">

<img src="images/tutorials-partitioned-burgers-1d-precice-config.png" alt="preCICE configuration visualization" width="600"/>

</p>

![preCICE configuration visualization](images/tutorials-partitioned-burgers-1d-precice-config.png)

Similarly everywhere here.

uekerman · 2026-02-06T15:20:31Z

partitioned-burgers-1d/README.md

+  <img src="images/tutorials-partitioned-burgers-1d-precice-config.png" alt="preCICE configuration visualization" width="600"/>
+</p>
+
+## Available Solvers


Suggested change

## Available Solvers

## Available solvers

uekerman · 2026-02-06T15:52:42Z

partitioned-burgers-1d/README.md

+  <img src="images/tutorials-partitioned-burgers-1d-full-domain-timestep-slice.png" alt="Full Domain Timestep Slice" width="400"/>
+</p>
+
+- `gradient-timestep-slice.png`: Gradient $du/dx$ at a selected timestep


Suggested change

- `gradient-timestep-slice.png`: Gradient $du/dx$ at a selected timestep

- `gradient-timestep-slice.png`: Gradient $du/dx$ at the selected timestep

uekerman · 2026-02-06T15:55:41Z

partitioned-burgers-1d/README.md

+## Visualization
+
+After both participants (and/or monolithic simulation) have finished, you can run the visualization script.
+`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot:


Suggested change

`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot:

`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot. Call from the root of the tutorial:

uekerman · 2026-02-06T15:55:51Z

partitioned-burgers-1d/README.md

+`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot:
+
+```bash
+python3 visualize_partitioned_domain.py --neumann neumann-surrogate/surrogate.npz [timestep]


Suggested change

python3 visualize_partitioned_domain.py --neumann neumann-surrogate/surrogate.npz [timestep]

python3 utils/visualize_partitioned_domain.py --neumann neumann-surrogate/surrogate.npz [timestep]

uekerman · 2026-02-06T16:01:18Z

partitioned-burgers-1d/neumann-surrogate/requirements.txt

+pyprecice==3.3.1
+scipy==1.16.3
+matplotlib==3.10.8
+torch==2.9.1


Is there a variant to get PyTorch without all GPU builds? Just to make it smaller here.

uekerman · 2026-02-06T16:05:02Z

partitioned-burgers-1d/README.md

+## Visualization
+
+After both participants (and/or monolithic simulation) have finished, you can run the visualization script.
+`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot:


How can I compare partitioned and monolithic solutions? If I just run it after one setup, I only get that one setup.

Partitioned Burgers eq. 1D initial commit. Dirichlet Neumann coupling.

4149918

vidulejs requested a review from uekerman September 29, 2025 15:15

vidulejs changed the title ~~Partitioned Burgers eq. 1D~~ New tutorial: Partitioned Burgers eq. 1D Oct 1, 2025

vidulejs added 7 commits October 14, 2025 15:57

Fully working Neumann surrogate participant. Scripts: run-partitioned…

6c8660c

…-surrogate.sh, generate-training-data.sh. Fix floating point issue in t_index. Add --savefile argument to solver-scipy-fvolumes.py to save data for training.

Rename surrogate burgers to neumann-surrogate

82accc2

activate-env.sh script

2c425e4

Add readme and requirements.txt. Move DG old files

5ab08f6

Rename solver-scipy-fvolumes to solver-scipy. Add proper run and clea…

c5844c1

…n scripts. Fix requirements.txt for scipy and surrogate solvers. Add example images for README

Update README.md. Conform to tutorial images standard. Update run scr…

8b78407

…ipt to check for initial conditions after installing environment.

add equation to readme

2357008

vidulejs marked this pull request as ready for review October 15, 2025 13:04

uekerman reviewed Oct 21, 2025

View reviewed changes

vidulejs added 8 commits December 15, 2025 11:41

Remove DG solver

5d66f9f

Create utils directory and move scripts. Create output directory, sep…

8213367

…arate from images directory.

Update run scripts to conform to other preCICE tutorials. Fix package…

6546901

… versions in requirements. Remove commented out code from solvers.

clean.sh to clean-tutorial.sh

20cab48

remove comment from precice-config.xml

fd63b1d

Update README and add domain diagram image

9beb26b

Specify IC script in readme

1e8b1ea

Changelog entry

11b91fc

MakisH added the new-tutorial label Dec 21, 2025

MakisH reviewed Feb 3, 2026

View reviewed changes

partitioned-burgers-1d/README.md Outdated Show resolved Hide resolved

vidulejs added 6 commits February 3, 2026 12:59

Updated format: #704

854ad44

Pre-commit formatting

292a672

Pre-commit formatting fix 2

b193d05

Allow skipping venv setup

6c99c35

#680

Update readme ## Visualization

3c66728

Pre-commit formatting fix

bb434d9

uekerman requested changes Feb 6, 2026

View reviewed changes

		@@ -0,0 +1 @@
		New tutorial case: Partitioned Burgers' Equation in 1D solved with Finite Volumes and a Neural Network surrogate model !

	New tutorial case: Partitioned Burgers' Equation in 1D solved with Finite Volumes and a Neural Network surrogate model !
	- Added new case: 1D partitioned Burgers' equation with one finite volume and one NN surrogate participant.

	- `gradient-timestep-slice.png`: Gradient $du/dx$ at a selected timestep
	- `gradient-timestep-slice.png`: Gradient $du/dx$ at the selected timestep

	`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot:
	`visualize_partitioned_domain.py` generates plots comparing the partitioned and monolithic solutions. You can specify which timestep to plot. Call from the root of the tutorial:

	python3 visualize_partitioned_domain.py --neumann neumann-surrogate/surrogate.npz [timestep]
	python3 utils/visualize_partitioned_domain.py --neumann neumann-surrogate/surrogate.npz [timestep]

Uh oh!

Conversation

vidulejs commented Sep 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

uekerman left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

uekerman left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

vidulejs commented Sep 29, 2025 •

edited

Loading