Diagnostics#10
Open
vtommasini wants to merge 1 commit into
Open
Conversation
There was a problem hiding this comment.
Pull request overview
Note
Copilot was unable to run its full agentic suite in this review.
Introduces a GPR diagnostics/checkpointing API and replaces the previous single GPR unit test with broader tests covering diagnostics, core training/prediction behavior, and checkpoint save/load.
Changes:
- Added diagnostics utilities (LOO predictions/cross-validation + residual plotting) and corresponding unit tests.
- Added checkpoint save/load functions for trained GPR models (including normalization stats and training data) with unit tests.
- Reorganized exports via a new
SimulationSupport.gprpackage interface and removed the oldtest_gpr.py.
Reviewed changes
Copilot reviewed 6 out of 6 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
| tests/test_gpr.py | Removed legacy GPR unit test file in favor of more granular coverage. |
| tests/test_diagnostics.py | Adds unit tests for new diagnostics helpers (LOO + residuals). |
| tests/test_core.py | Adds unit tests for core GPR behavior and checkpoint save/load round-trips. |
| src/SimulationSupport/gpr/diagnostics.py | New diagnostics module implementing LOO prediction/CV and residual plotting. |
| src/SimulationSupport/gpr/init.py | New package export surface for core + diagnostics. |
| src/SimulationSupport/gpr.py | Updates documentation/formatting and adds checkpoint save/load functions. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+473
to
+516
| def setUp(self): | ||
| self.df = make_df() | ||
|
|
||
| def tearDown(self): | ||
| plt.close("all") | ||
|
|
||
| def test_returns_four_outputs(self): | ||
| # Test that the function returns (X, Y, predictions_LOO, uncertainties_LOO) | ||
| result = loo_predictions( | ||
| self.df, | ||
| inputVar="initial_separation", | ||
| outputVar="initial_orbital_frequency", | ||
| ) | ||
| self.assertEqual(len(result), 4) | ||
|
|
||
| def test_output_types_are_arrays(self): | ||
| # Test that all four outputs are numpy arrays (and not lists or tensors) | ||
| X, Y, preds, uncertainties = loo_predictions( | ||
| self.df, | ||
| inputVar="initial_separation", | ||
| outputVar="initial_orbital_frequency", | ||
| ) | ||
| for arr in (X, Y, preds, uncertainties): | ||
| self.assertIsInstance(arr, np.ndarray) | ||
|
|
||
| def test_output_shapes_match_input(self): | ||
| # Test that each returned array has one entry per row in the DataFrame | ||
| X, Y, preds, uncertainties = loo_predictions( | ||
| self.df, | ||
| inputVar="initial_separation", | ||
| outputVar="initial_orbital_frequency", | ||
| ) | ||
| n = len(self.df) | ||
| for arr in (X, Y, preds, uncertainties): | ||
| self.assertEqual(arr.shape, (n,)) | ||
|
|
||
| def test_X_matches_input_column(self): | ||
| # Returned X should be exactly the values of the inputVar column | ||
| X, _, _, _ = loo_predictions( | ||
| self.df, | ||
| inputVar="initial_separation", | ||
| outputVar="initial_orbital_frequency", | ||
| ) | ||
| np.testing.assert_array_equal(X, self.df["initial_separation"].values) |
Comment on lines
+617
to
+647
| def setUp(self): | ||
| """ | ||
| Train a model and save a checkpoint once for all tests in this class. | ||
| Individual tests inspect different aspects of the saved file. | ||
| """ | ||
| # Temporary directory | ||
| self.tmp_dir = tempfile.TemporaryDirectory() | ||
| self.ckpt_path = str(Path(self.tmp_dir.name) / "test_model.pt") | ||
|
|
||
| self.model, self.likelihood, self.features, self.X, self.Y = ( | ||
| _train_test_model() | ||
| ) | ||
| self.run_col = "initial_orbital_frequency" | ||
| self.pn_col = "spec_pn_guess_omega" | ||
|
|
||
| # Save once - all tests below read this file | ||
| save_gpr_checkpoint( | ||
| self.model, | ||
| self.likelihood, | ||
| self.features, | ||
| "omega", | ||
| self.run_col, | ||
| self.pn_col, | ||
| self.X, | ||
| self.Y, | ||
| self.ckpt_path, | ||
| ) | ||
|
|
||
| def tearDown(self): | ||
| # Temporary directory | ||
| self.tmp_dir.cleanup() |
| self.assertFalse(model.training) | ||
|
|
||
| def test_likelihood_in_eval_mode(self): | ||
| # Test lihelihood.training is False |
vtommasini
force-pushed
the
diagnostics
branch
3 times, most recently
from
795ff5e to
99a78b8
Compare
nilsvu
requested changes
Jun 6, 2026
vtommasini
force-pushed
the
diagnostics
branch
2 times, most recently
from
bf62555 to
5d719fd
Compare
nilsvu
requested changes
Jun 24, 2026
nilsvu
requested changes
Jul 1, 2026
vtommasini
force-pushed
the
diagnostics
branch
2 times, most recently
from
04268d8 to
f4ff5a5
Compare
Simplified parallelization and tests, and decreased number of test_data points to make tests run faster Rebased, shortened tests
nilsvu
approved these changes
Jul 2, 2026
Member
|
@vtommasini for some reason the CI tests take forever and time out at 6 hours. Any idea what's happening? Doe the tests run quick on your machine? |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Functions to carry out leave-one-out cross validation for GPR model