Effective ripple ε

[unalmis]

Resolves #519 . Will be replaced by #1290 soon, so users are advised to wait for that.

[github-actions]

|             benchmark_name             |         dt(%)          |         dt(s)          |        t_new(s)        |        t_old(s)        | 
| -------------------------------------- | ---------------------- | ---------------------- | ---------------------- | ---------------------- |
 test_build_transform_fft_midres         |     +0.41 +/- 4.91     | +2.66e-03 +/- 3.20e-02 |  6.55e-01 +/- 3.1e-02  |  6.53e-01 +/- 9.6e-03  |
 test_build_transform_fft_highres        |     -0.38 +/- 2.33     | -3.88e-03 +/- 2.36e-02 |  1.01e+00 +/- 1.4e-02  |  1.01e+00 +/- 1.9e-02  |
 test_equilibrium_init_lowres            |     +0.83 +/- 1.97     | +3.60e-02 +/- 8.56e-02 |  4.39e+00 +/- 6.2e-02  |  4.35e+00 +/- 5.9e-02  |
 test_objective_compile_atf              |     -0.09 +/- 4.68     | -7.98e-03 +/- 3.99e-01 |  8.53e+00 +/- 3.0e-01  |  8.53e+00 +/- 2.6e-01  |
 test_objective_compute_atf              |     -1.00 +/- 7.03     | -1.17e-04 +/- 8.18e-04 |  1.15e-02 +/- 7.3e-04  |  1.16e-02 +/- 3.7e-04  |
 test_objective_jac_atf                  |     +5.46 +/- 4.21     | +1.07e-01 +/- 8.25e-02 |  2.06e+00 +/- 5.6e-02  |  1.96e+00 +/- 6.1e-02  |
 test_perturb_1                          |     +1.04 +/- 1.71     | +1.63e-01 +/- 2.67e-01 |  1.58e+01 +/- 2.5e-01  |  1.56e+01 +/- 1.0e-01  |
 test_proximal_jac_atf                   |     +0.66 +/- 1.13     | +5.52e-02 +/- 9.43e-02 |  8.42e+00 +/- 7.6e-02  |  8.36e+00 +/- 5.6e-02  |
 test_proximal_freeb_compute             |     +0.29 +/- 1.71     | +5.68e-04 +/- 3.41e-03 |  2.00e-01 +/- 2.0e-03  |  1.99e-01 +/- 2.8e-03  |
 test_solve_fixed_iter_compiled          |     +0.58 +/- 1.40     | +1.03e-01 +/- 2.48e-01 |  1.78e+01 +/- 1.1e-01  |  1.77e+01 +/- 2.2e-01  |
 test_build_transform_fft_lowres         |     -0.17 +/- 5.07     | -9.21e-04 +/- 2.75e-02 |  5.41e-01 +/- 2.1e-02  |  5.42e-01 +/- 1.8e-02  |
 test_equilibrium_init_medres            |     -0.12 +/- 4.07     | -5.04e-03 +/- 1.75e-01 |  4.30e+00 +/- 1.1e-01  |  4.30e+00 +/- 1.3e-01  |
 test_equilibrium_init_highres           |     +0.26 +/- 2.42     | +1.43e-02 +/- 1.35e-01 |  5.60e+00 +/- 1.1e-01  |  5.59e+00 +/- 7.3e-02  |
 test_objective_compile_dshape_current   |     -0.62 +/- 2.02     | -2.45e-02 +/- 7.98e-02 |  3.92e+00 +/- 3.7e-02  |  3.95e+00 +/- 7.1e-02  |
 test_objective_compute_dshape_current   |     +1.12 +/- 2.59     | +4.04e-05 +/- 9.37e-05 |  3.65e-03 +/- 7.5e-05  |  3.61e-03 +/- 5.6e-05  |
 test_objective_jac_dshape_current       |     -2.17 +/- 5.37     | -9.15e-04 +/- 2.27e-03 |  4.13e-02 +/- 9.1e-04  |  4.23e-02 +/- 2.1e-03  |
 test_perturb_2                          |     -0.23 +/- 2.09     | -4.58e-02 +/- 4.13e-01 |  1.97e+01 +/- 2.6e-01  |  1.98e+01 +/- 3.2e-01  |
 test_proximal_freeb_jac                 |     +0.58 +/- 1.98     | +4.33e-02 +/- 1.49e-01 |  7.55e+00 +/- 1.3e-01  |  7.51e+00 +/- 7.5e-02  |
 test_solve_fixed_iter                   |     +1.28 +/- 1.41     | +3.67e-01 +/- 4.03e-01 |  2.90e+01 +/- 2.6e-01  |  2.86e+01 +/- 3.1e-01  |
 test_LinearConstraintProjection_build   |     -0.67 +/- 1.37     | -1.53e-01 +/- 3.13e-01 |  2.27e+01 +/- 2.8e-01  |  2.29e+01 +/- 1.4e-01  |

[gretahibbard]

Resolves #1288

[unalmis]

#1119 and upward is now broken due to git corruption. So let's merge this instead.

[f0uriest]

Can we try to make the objective function api forward compatible with the 2d version?

[unalmis]

Ok I fixed the issue on the other branch.

[unalmis]

Can we try to make the objective function api forward compatible with the 2d version?

I have pushed an update to make it as compatible as possible. The remaining difference is that this objective wants rho and alpha while the 2d methods want an fft compatible (rho, theta, zeta) grid. This difference cannot be made forward compatible.

[rahulgaur104]

I have reviewed this once before and used the API enough to be satisfied with this PR.
The discrepancy with WISTELL's ripple calculator is due to significant differences in the numerical methods and I am pretty sure ours are better.
Merge at will!

outdated

[unalmis]

You all have reviewed this multiple times already, and the content of the PR has been split apart and merged via other PRs. At this point this PR is an inoffensive skeleton, and will be replaced by #1290, so I am merging with one approval. No objectives have been added to public api.

[YigitElma]

Why the tests of this fail on master? https://github.com/PlasmaControl/DESC/actions/runs/12193541254

EDIT: Oh maybe related to #1434