Add qfm surface functionality #1264
Open
Conversation
This file contains 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
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #1264 +/- ##
==========================================
- Coverage 95.44% 95.44% -0.01%
==========================================
Files 95 95
Lines 23423 23437 +14
==========================================
+ Hits 22357 22369 +12
- Misses 1066 1068 +2
|
| benchmark_name | dt(%) | dt(s) | t_new(s) | t_old(s) |
| -------------------------------------- | ---------------------- | ---------------------- | ---------------------- | ---------------------- |
test_build_transform_fft_lowres | -1.86 +/- 4.86 | -9.98e-03 +/- 2.61e-02 | 5.26e-01 +/- 1.0e-02 | 5.36e-01 +/- 2.4e-02 |
test_equilibrium_init_medres | +0.08 +/- 2.61 | +3.55e-03 +/- 1.10e-01 | 4.21e+00 +/- 1.1e-01 | 4.21e+00 +/- 2.3e-02 |
test_equilibrium_init_highres | -0.03 +/- 4.05 | -1.63e-03 +/- 2.25e-01 | 5.57e+00 +/- 2.0e-01 | 5.57e+00 +/- 1.0e-01 |
test_objective_compile_dshape_current | -0.56 +/- 6.06 | -2.21e-02 +/- 2.38e-01 | 3.90e+00 +/- 2.4e-01 | 3.92e+00 +/- 2.4e-02 |
test_objective_compute_dshape_current | +1.40 +/- 1.56 | +5.04e-05 +/- 5.62e-05 | 3.66e-03 +/- 4.9e-05 | 3.61e-03 +/- 2.7e-05 |
test_objective_jac_dshape_current | +3.53 +/- 6.88 | +1.44e-03 +/- 2.82e-03 | 4.24e-02 +/- 2.4e-03 | 4.09e-02 +/- 1.5e-03 |
test_perturb_2 | +2.49 +/- 2.98 | +4.38e-01 +/- 5.22e-01 | 1.80e+01 +/- 4.3e-01 | 1.76e+01 +/- 3.0e-01 |
test_proximal_freeb_jac | +0.35 +/- 1.57 | +2.66e-02 +/- 1.18e-01 | 7.56e+00 +/- 4.8e-02 | 7.54e+00 +/- 1.1e-01 |
test_solve_fixed_iter | +0.07 +/- 60.15 | +3.37e-03 +/- 3.04e+00 | 5.05e+00 +/- 2.1e+00 | 5.05e+00 +/- 2.2e+00 |
test_build_transform_fft_midres | -6.04 +/- 2.90 | -4.00e-02 +/- 1.92e-02 | 6.23e-01 +/- 1.6e-02 | 6.63e-01 +/- 1.0e-02 |
test_build_transform_fft_highres | -3.85 +/- 3.23 | -4.11e-02 +/- 3.44e-02 | 1.03e+00 +/- 2.7e-02 | 1.07e+00 +/- 2.1e-02 |
test_equilibrium_init_lowres | -4.94 +/- 2.92 | -2.14e-01 +/- 1.27e-01 | 4.13e+00 +/- 1.2e-01 | 4.34e+00 +/- 5.2e-02 |
test_objective_compile_atf | -1.72 +/- 3.54 | -1.45e-01 +/- 2.99e-01 | 8.28e+00 +/- 1.3e-01 | 8.43e+00 +/- 2.7e-01 |
test_objective_compute_atf | +1.74 +/- 3.38 | +1.93e-04 +/- 3.76e-04 | 1.13e-02 +/- 1.9e-04 | 1.11e-02 +/- 3.2e-04 |
test_objective_jac_atf | -0.50 +/- 3.47 | -1.01e-02 +/- 6.99e-02 | 2.01e+00 +/- 5.2e-02 | 2.02e+00 +/- 4.7e-02 |
test_perturb_1 | +0.21 +/- 4.02 | +2.86e-02 +/- 5.48e-01 | 1.37e+01 +/- 1.9e-01 | 1.36e+01 +/- 5.1e-01 |
test_proximal_jac_atf | +0.24 +/- 0.83 | +2.00e-02 +/- 6.94e-02 | 8.36e+00 +/- 4.7e-02 | 8.34e+00 +/- 5.1e-02 |
test_proximal_freeb_compute | +1.49 +/- 1.00 | +2.78e-03 +/- 1.87e-03 | 1.90e-01 +/- 1.2e-03 | 1.87e-01 +/- 1.5e-03 | |
f0uriest
reviewed
Sep 25, 2024
| Equilibrium upon whose surface the normal field error will be minimized. | ||
| The equilibrium is kept fixed during the optimization with this objective. | ||
| eq : Equilibrium or FourierRZToroidalSurface | ||
| Equilibrium (or QFM surface) upon whose surface the normal field error |
There was a problem hiding this comment.
should maybe add a note to use BoundaryError or VacuumBoundaryError for free boundary / single stage optimization
f0uriest
reviewed
Sep 25, 2024
| self._vacuum = vacuum or qfm_surface | ||
| self._qfm_surface = qfm_surface | ||
| errorif( | ||
| qfm_surface and hasattr(eq, "L_lmn"), |
There was a problem hiding this comment.
would prefer an actual instance check here instead of relying on attributes to be/not be there
f0uriest
reviewed
Sep 25, 2024
| @@ -1182,8 +1198,23 @@ def __init__( | |||
| self._eq = eq | |||
| self._field = field | |||
| self._field_grid = field_grid | |||
| self._vacuum = vacuum | |||
| things = [field] | |||
| self._vacuum = vacuum or qfm_surface | |||
There was a problem hiding this comment.
might be better to throw an error here if they're incompatible?
There was a problem hiding this comment.
The thing is, by default vacuum=False, so an error would be thrown if qfm_surface=True is passed in.
I guess that is fine if we want users to have to be verbose about asking for qfm surface and specifying they are for a vacuum field
f0uriest
reviewed
Sep 25, 2024
| @@ -1282,14 +1323,25 @@ def compute(self, field_params, constants=None): | |||
| """ | |||
| if constants is None: | |||
| constants = self.constants | |||
|
|
|||
| field_params = params_1 if not self._qfm_surface else params_2 | |||
There was a problem hiding this comment.
don't you also need to check if the field is fixed? in which case the only DOFs will be the surface?
f0uriest
reviewed
Sep 25, 2024
| x = jnp.array([eval_data["R"], eval_data["phi"], eval_data["Z"]]).T | ||
|
|
||
| # B_ext is not pre-computed because field is not fixed | ||
| if not self._qfm_surface: |
There was a problem hiding this comment.
minor point but I think its clearer if this is re-ordered to be
if qfm_surface:
...
else:
...
f0uriest
reviewed
Sep 25, 2024
| field_params = params_1 if not self._qfm_surface else params_2 | ||
| surf_params = params_2 if not self._qfm_surface else params_1 | ||
|
|
||
| if not self._qfm_surface: |
|
make a new objective for this specifically |
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.
Resolve #1162
ToroidalFluxto acceptFourierRZToroidalSurfaceso that we can specify the toroidal flux through the QFM (we can do the vector potential method on the surface just fine so this should be easy, and there is no plasma current to consider either)