Update tutorials 1-12 to current version (no Plotter) #463

MatteB03 · 2025-02-26T15:30:04Z

Changed imports to match current folders
Changed notation where version differed from master
Deleted unnecessary inputs and modified outdated ones
Minor changes to solve tutorial-specific errors
Accordingly changed tutorial.py files
Added tmp_poisson_inverse folder in tutorial7 to store epochs log

dario-coscia

Hi! Thank you for updating the tutorials, it is really an incredible work :) I have made few comments on the PR, in particular the missing plots and some others. For plotting the solution you can use matplotlib, while for the loss the tensorboard extension should be fine.

dario-coscia · 2025-02-26T15:41:27Z

pina/model/block/convolution_2d.py

        """
        # initialize to all zeros
-        tmp = torch.zeros_like(X)
+        tmp = torch.zeros_like(X).as_subclass(torch.Tensor)


remove this, and use use_lt=False

So, this block should work only without LabelTensor? I did this to make compatible both with and without LabelTensor

all blocks should work with both, if it is needed ok!

pina/label_tensor.py

dario-coscia · 2025-02-26T15:42:57Z

tutorials/tutorial1/tutorial.py


-pl = Plotter()
-pl.plot_samples(problem=problem)
+#pl = Plotter()


we should be able to plot the solution, please do the plot with matplotlib

dario-coscia · 2025-02-26T15:45:32Z

tutorials/tutorial14/tutorial.py

+get_ipython().run_line_magic('matplotlib', 'inline')
+
+import matplotlib.pyplot as plt
+plt.style.use('tableau-colorblind10')


Maybe we should use the same palette in every tutorial for consistency. We can think of having a PINA palette but maybe we can do it in a second iteration.

dario-coscia · 2025-02-26T15:48:47Z

tutorials/tutorial14/tutorial.py

+# In[2]:
+
+
+get_ipython().system('pip install git+https://github.com/mathLab/Smithers.git #if required --break-system-packages')


Let's remove this, maybe we can create optional dependencies for pina tutorials (see here)

dario-coscia · 2025-02-26T15:53:27Z

Also @MatteB03 and @FilippoOlivo there are conflicts to be solver inside pina/label_tensor.py. I think they are basically the same code, do you agree?

dario-coscia · 2025-03-04T10:34:01Z

@MatteB03 how are you doing with this? Should we convert it as draft?

MatteB03 · 2025-03-04T11:06:47Z

@MatteB03 how are you doing with this? Should we convert it as draft?

Aside from tutorials 2 and 13 everything else works, if I revert tutorial 2 and keep working on those on my local could we still update everything else?

MatteB03 · 2025-03-04T11:08:48Z

@MatteB03 how are you doing with this? Should we convert it as draft?

Also, I think the issue in tutorial2 is the same as in tutorial13... the phys_loss and train_loss are very high while the bound_losses are low, similar to what we marked as bug in tutorial13. Did you get what the bug was?

FilippoOlivo · 2025-03-05T14:37:14Z

tutorials/tutorial4/data/MNIST/raw/t10k-images-idx3-ubyte

Please @MatteB03, remove folder tutorial/tutorial4/data. @dario-coscia, should we update .gitignore to prevent this behavior in the future?

Tutorial10 also has data, although the folder is created in the initial imports. @FilippoOlivo want me to delete that as well?

No, they must be there

Does tutorial 4 requires data?

Yes, MNIST but it is available in torchvision

Yes, it's the one that takes MNIST from Torchvision. The data is saved while being downloaded, I can try to look up a way to prevent this or to erase the folder automatically if u want

Just deleting the folder is fine for me. Saving data locally is the correct approach, but we don’t want to upload it to PINA

MatteB03 · 2025-03-05T15:01:43Z

Also, I just saw a strange thing... On every epoch of training the train function prints
Validation: | | 0/? [00:00<?, ?it/s] and at the end of the training there's an empty output for each epoch... Yesterday when I committed they weren't visible, I'm not sure of what happened but if you open tutorial1/3/4 you can see very well how bad it is... Do you know why does it print an empty dataloader? It didn't do this before yesterday...

MatteB03 · 2025-03-06T14:35:53Z

Sorry guys, the strange behaviour with the validation print keeps happening even after rebasing everything up to the latest PR. I'll attach a screenshot I took earlier to show you what happens, now it does so at every training in the 5 tutorials I've gone through this afternoon so far... do you know what can I do not to have all these useless prints? This issue started at the beginning of this week, maybe something needs to be fixed in the validation part of the training @dario-coscia @FilippoOlivo

GiovanniCanali · 2025-03-07T11:03:55Z

Hi @MatteB03! Please be mindful not to upload raw files, logs, or other large object files to PINA. They take up a lot of space and can be tricky to remove once uploaded.

MatteB03 · 2025-03-07T11:25:53Z

Hi @MatteB03! Please be mindful not to upload raw files, logs, or other large object files to PINA. They take up a lot of space and can be tricky to remove once uploaded.

Yep, this happened in the tutorials I just mentioned on Slack as I really didn't notice the empty outputs, once solved the big issue of this week I'll make sure to run everything again to delete those space-consuming abominations

MatteB03 · 2025-03-07T15:23:25Z

Ok now that I removed all hidden outputs and extra files the dimensions of the PR are acceptable, but I did NOT solve the validation problem that I exposed this morning on pina... I simply deleted those outputs, and I don't think it would be useful to put train=1.0 everywhere like @FilippoOlivo told me to do in tutorial 13. I need help solving this problem, I can't figure out what is causing it among the changes of this past week...

dario-coscia · 2025-03-10T17:15:23Z

@MatteB03 can I start reviewing?

MatteB03 · 2025-03-10T17:34:46Z

@MatteB03 can I start reviewing?

I'd say so, aside from that issue with tensorboard everything works.

--------- Co-authored-by: Dario Coscia <[email protected]>

dario-coscia requested changes Feb 26, 2025

View reviewed changes

dario-coscia mentioned this pull request Feb 26, 2025

Add tutorial14 on Lid-Cavity dataset snapshot prediction #464

Closed

dario-coscia assigned dario-coscia and MatteB03 Feb 26, 2025

dario-coscia added pr-to-fix Label for PR that needs modification v0.2 labels Feb 26, 2025

ndem0 force-pushed the 0.2 branch from bd5e4a6 to 1bbba45 Compare March 3, 2025 14:42

MatteB03 force-pushed the matteo0.2 branch 2 times, most recently from 409f56d to 782194e Compare March 4, 2025 18:05

FilippoOlivo reviewed Mar 5, 2025

View reviewed changes

MatteB03 force-pushed the matteo0.2 branch from 782194e to c414248 Compare March 6, 2025 15:04

dario-coscia force-pushed the matteo0.2 branch from 2e1fafc to faa2773 Compare March 6, 2025 16:30

GiovanniCanali self-requested a review March 7, 2025 11:04

dario-coscia mentioned this pull request Mar 7, 2025

Condition notation updates in tutorials #476

Closed

MatteB03 force-pushed the matteo0.2 branch from a4ae164 to b91381e Compare March 7, 2025 13:27

dario-coscia force-pushed the matteo0.2 branch 2 times, most recently from 670781f to 69a37a7 Compare March 7, 2025 17:29

MatteB03 force-pushed the matteo0.2 branch 3 times, most recently from 4ea6df6 to c9880eb Compare March 8, 2025 18:40

Changes to Tensor labels handling

4e7ef41

MatteB03 and others added 15 commits March 10, 2025 15:41

Update tutorials 1 through 12 to current version 0.2

2cfbcfc

New tutorial14 on Lid-Cavity snapshot prediction

0398315

Add plot in tutorials 1,3,4,9

b119fb6

Update tutorials 1 through 12 to current version 0.2

1bac728

Add plot in tutorials 1,3,4,9

bc67ef3

Update tutorial 1,3,4,9 plots

b23a458

Tutorials WIP

32d2ee9

rm unnecessary files

c816ee2

Slim plots in tutorial3

633b433

Update Condition notation, delete error outputs and output folders

5970d30

Working Tutorial13

ae2dd6f

update tutorial3

5811c00

Update tutorials 1 through 7

07fa92c

Update tutorials 8 through 14

1e20f89

Updates to tutorial and run post codacy changes

5c7aaf3

MatteB03 force-pushed the matteo0.2 branch from e834619 to 5c7aaf3 Compare March 10, 2025 16:19

Update Tensorboard use

48724f4

ndem0 merged commit 1a1f0d8 into mathLab:0.2 Mar 12, 2025
17 checks passed

ndem0 mentioned this pull request Mar 12, 2025

Revert "Update tutorials 1-12 to current version (no Plotter)" #485

Closed

FilippoOlivo pushed a commit that referenced this pull request Mar 13, 2025

Update tutorials (#463)

bffb5ba

--------- Co-authored-by: Dario Coscia <[email protected]>

FilippoOlivo pushed a commit to FilippoOlivo/PINA that referenced this pull request Mar 21, 2025

Update tutorials (mathLab#463)

a06ac27

FilippoOlivo pushed a commit to FilippoOlivo/PINA that referenced this pull request Apr 17, 2025

Update tutorials (mathLab#463)

bd9b495

--------- Co-authored-by: Dario Coscia <[email protected]>

		# In[2]:


		get_ipython().system('pip install git+https://github.com/mathLab/Smithers.git #if required --break-system-packages')

Update tutorials 1-12 to current version (no Plotter) #463

Update tutorials 1-12 to current version (no Plotter) #463

Uh oh!

Conversation

MatteB03 commented Feb 26, 2025

Uh oh!

dario-coscia left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

FilippoOlivo Feb 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

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!

dario-coscia commented Feb 26, 2025 • edited by FilippoOlivo Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dario-coscia commented Mar 4, 2025

Uh oh!

MatteB03 commented Mar 4, 2025

Uh oh!

MatteB03 commented Mar 4, 2025

Uh oh!

FilippoOlivo Mar 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

MatteB03 Mar 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

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!

FilippoOlivo Mar 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

FilippoOlivo Mar 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

MatteB03 commented Mar 5, 2025

Uh oh!

MatteB03 commented Mar 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

GiovanniCanali commented Mar 7, 2025

Uh oh!

MatteB03 commented Mar 7, 2025

Uh oh!

MatteB03 commented Mar 7, 2025

Uh oh!

dario-coscia commented Mar 10, 2025

Uh oh!

MatteB03 commented Mar 10, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

FilippoOlivo Feb 26, 2025 •

edited

Loading

dario-coscia commented Feb 26, 2025 •

edited by FilippoOlivo

Loading

FilippoOlivo Mar 5, 2025 •

edited

Loading

MatteB03 Mar 5, 2025 •

edited

Loading

FilippoOlivo Mar 5, 2025 •

edited

Loading

FilippoOlivo Mar 5, 2025 •

edited

Loading

MatteB03 commented Mar 6, 2025 •

edited

Loading