-
Notifications
You must be signed in to change notification settings - Fork 26
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Vectorize omnigenity objective over multiple surfaces #1225
Conversation
- [x] Allows the calculation of `B_mn` on tensor product grids with multiple flux surfaces. - [x] Allows `QuasisymmetryBoozer` objective to target multiple surfaces at once ~- [ ] Allows `Omnigenity` objective to target multiple surfaces at once~ moved to #1225 - [x] Simplify boozer/QS plotting functions Resolves #754
| benchmark_name | dt(%) | dt(s) | t_new(s) | t_old(s) |
| -------------------------------------- | ---------------------- | ---------------------- | ---------------------- | ---------------------- |
test_build_transform_fft_lowres | -8.52 +/- 6.33 | -4.99e-02 +/- 3.71e-02 | 5.36e-01 +/- 3.2e-02 | 5.86e-01 +/- 1.9e-02 |
test_equilibrium_init_medres | -7.54 +/- 7.44 | -3.44e-01 +/- 3.40e-01 | 4.22e+00 +/- 1.7e-01 | 4.57e+00 +/- 2.9e-01 |
test_equilibrium_init_highres | -0.15 +/- 2.34 | -8.03e-03 +/- 1.27e-01 | 5.43e+00 +/- 5.7e-02 | 5.43e+00 +/- 1.1e-01 |
test_objective_compile_dshape_current | +0.85 +/- 2.78 | +3.37e-02 +/- 1.10e-01 | 4.00e+00 +/- 9.9e-02 | 3.96e+00 +/- 4.8e-02 |
test_objective_compute_dshape_current | +1.53 +/- 2.00 | +7.92e-05 +/- 1.03e-04 | 5.25e-03 +/- 7.6e-05 | 5.17e-03 +/- 6.9e-05 |
test_objective_jac_dshape_current | +0.05 +/- 5.95 | +2.34e-05 +/- 2.58e-03 | 4.34e-02 +/- 2.0e-03 | 4.34e-02 +/- 1.6e-03 |
test_perturb_2 | +2.95 +/- 7.88 | +6.09e-01 +/- 1.63e+00 | 2.13e+01 +/- 1.5e+00 | 2.07e+01 +/- 5.9e-01 |
test_proximal_freeb_jac | -0.66 +/- 1.51 | -4.94e-02 +/- 1.13e-01 | 7.44e+00 +/- 8.2e-02 | 7.49e+00 +/- 7.8e-02 |
test_solve_fixed_iter | +2.60 +/- 2.52 | +8.67e-01 +/- 8.41e-01 | 3.42e+01 +/- 6.4e-01 | 3.33e+01 +/- 5.5e-01 |
test_LinearConstraintProjection_build | +4.44 +/- 3.97 | +4.57e-01 +/- 4.09e-01 | 1.08e+01 +/- 2.8e-01 | 1.03e+01 +/- 3.0e-01 |
test_build_transform_fft_midres | +5.23 +/- 7.98 | +3.21e-02 +/- 4.89e-02 | 6.45e-01 +/- 4.7e-02 | 6.13e-01 +/- 1.4e-02 |
test_build_transform_fft_highres | +1.05 +/- 3.29 | +1.03e-02 +/- 3.22e-02 | 9.88e-01 +/- 2.0e-02 | 9.78e-01 +/- 2.5e-02 |
test_equilibrium_init_lowres | -0.66 +/- 3.93 | -2.58e-02 +/- 1.54e-01 | 3.88e+00 +/- 7.1e-02 | 3.91e+00 +/- 1.4e-01 |
test_objective_compile_atf | -0.01 +/- 4.53 | -5.95e-04 +/- 3.67e-01 | 8.10e+00 +/- 2.5e-01 | 8.10e+00 +/- 2.7e-01 |
test_objective_compute_atf | +1.63 +/- 2.28 | +2.57e-04 +/- 3.60e-04 | 1.60e-02 +/- 2.7e-04 | 1.57e-02 +/- 2.4e-04 |
test_objective_jac_atf | -0.46 +/- 3.23 | -8.93e-03 +/- 6.30e-02 | 1.94e+00 +/- 4.0e-02 | 1.95e+00 +/- 4.8e-02 |
test_perturb_1 | -2.25 +/- 1.15 | -3.40e-01 +/- 1.73e-01 | 1.47e+01 +/- 1.2e-01 | 1.51e+01 +/- 1.3e-01 |
test_proximal_jac_atf | +0.34 +/- 0.86 | +2.84e-02 +/- 7.12e-02 | 8.28e+00 +/- 4.0e-02 | 8.25e+00 +/- 5.9e-02 |
test_proximal_freeb_compute | -0.65 +/- 1.40 | -1.32e-03 +/- 2.82e-03 | 2.00e-01 +/- 2.2e-03 | 2.02e-01 +/- 1.7e-03 |
test_solve_fixed_iter_compiled | +0.05 +/- 0.61 | +9.97e-03 +/- 1.34e-01 | 2.20e+01 +/- 1.2e-01 | 2.20e+01 +/- 5.4e-02 | |
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #1225 +/- ##
==========================================
+ Coverage 95.56% 95.58% +0.02%
==========================================
Files 98 98
Lines 25127 25156 +29
==========================================
+ Hits 24012 24045 +33
+ Misses 1115 1111 -4
|
@dpanici change test tols for QFM here |
why would the qfm test be failing here? nothing that changed should have any effect on that |
No idea, I will see and possibly adjust the test today |
I don't know why it was failing on this PR but I guess it was due to some numerical thing with CI, and not an actual problem. The test passes now. |
No description provided.