Custom pedestal api #1752
Open
Custom pedestal api #1752
Conversation
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
This commit implements a Custom Pedestal Model API that allows users to define custom pedestal scaling laws without modifying TORAX source code. Fixes google-deepmind#1711 ## Changes - Add CustomPedestalModel class supporting user-defined callable functions - Add CustomPedestal Pydantic configuration - Update PedestalConfig union to include CustomPedestal - Add comprehensive unit tests (7 test cases) - Add example configuration with EPED-like scaling - Add complete API documentation ## Features Users can now provide Python functions to compute: - Ion temperature at pedestal (T_i_ped) - Electron temperature at pedestal (T_e_ped) - Electron density at pedestal (n_e_ped) - Optional dynamic pedestal location (rho_norm_ped_top) Functions receive full access to runtime parameters, geometry, and core profiles, enabling machine-specific scaling laws (e.g., STEP pedestal models with Europed data fits). ## API Design Follows the transport model pattern with: - JAX Model Layer: CustomPedestalModel (frozen dataclass) - Pydantic Config Layer: CustomPedestal (validation) - Runtime Parameters: time-varying support Fully backwards compatible - no changes to existing models.
- Add public pedestal API (torax/pedestal.py) following transport model pattern - Remove PR documentation files (PR_SUMMARY.md, PR_TEXT.md, CUSTOM_PEDESTAL_API.md) - Add RST documentation (docs/custom_pedestal_models.rst) - concise and properly formatted - Simplify example to single model using public API and StaticDataclass - Fix test imports (removed 'as pm' alias) - Revert PedestalConfig formatting to single line Addresses all review comments from @tamaranorman
- Use public API throughout (Geometry, CoreProfiles, JAX_STATIC from torax) - Remove StaticDataclass from inheritance (parent already has it) - Simplify example config (minimal pedestal-only config) - Add test for custom_pedestal_example in examples_test.py - Clarify in RST docs that config example is minimal Addresses all review comments from @tamaranorman
| register_model.register_pedestal_model(TestPedestal) | ||
| | ||
| # Verify it can be instantiated | ||
| config = TestPedestal(test_value=99.0) |
Member
There was a problem hiding this comment.
What we need to test is that it can be instantiated through the ModelConfig API not directly which works even without the test
| register_model.register_pedestal_model(Config1) | ||
| register_model.register_pedestal_model(Config2) | ||
| | ||
| # Verify both can be instantiated |
Member
There was a problem hiding this comment.
Add a reference to this to init.py
Member
| Can you also fix the broken tests |
…le to __init__.py - Update register_model_test.py to test model instantiation through ToraxConfig.from_dict() instead of direct class instantiation - Add pedestal module import to torax/__init__.py for better API accessibility - Tests now properly verify that registered models work through the pydantic discriminator system Addresses maintainer feedback on PR.
Aaryan-549 force-pushed the custom-pedestal-api branch from 8768c31 to fc93e9f Compare | | ||
| # Verify it can be instantiated through the ModelConfig API | ||
| # Create a minimal ToraxConfig using from_dict to test the registration | ||
| minimal_config_dict = { |
Member
There was a problem hiding this comment.
Just use the config from test_utils.default_configs here and below
Member
| Please also fix the tests |
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
2 participants
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.
No description provided.