Umbilic fourier bounce #1349
Draft
Umbilic fourier bounce #1349
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
Due to using |B|^2 instead of jnp.sqrt(|B|^2)**2 in compute functions
the GitHUB CI fails on test_compute everything with he following errors.
So we increase the tolerance.
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J^rho.
Mismatched elements: 5 / 660 (0.758%)
Max absolute difference: 2.06564437e-08
Max relative difference: 1.97650683e-09
x: array([-4.121012e+02, -1.125881e+03, -1.537982e+03, -1.537982e+03,
-1.125881e+03, -4.121012e+02, -6.978519e+02, -1.423246e+03,
-1.440419e+03, -2.124563e+03, -2.423728e+03, -1.091549e+03,...
y: array([-4.121012e+02, -1.125881e+03, -1.537982e+03, -1.537982e+03,
-1.125881e+03, -4.121012e+02, -6.978519e+02, -1.423246e+03,
-1.440419e+03, -2.124563e+03, -2.423728e+03, -1.091549e+03,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: F_helical.
Mismatched elements: 5 / 660 (0.758%)
Max absolute difference: 3.14466888e-08
Max relative difference: 1.97650684e-09
x: array([-1.033077e-09, 3.099232e-09, -4.648847e-09, 1.549616e-09,
1.033077e-09, -4.648847e-09, -1.095852e+02, -2.269565e+02,
-2.359033e+02, -3.581634e+02, -4.179317e+02, -1.904907e+02,...
y: array([-1.033077e-09, 3.099232e-09, -4.648847e-09, 1.549616e-09,
1.033077e-09, -4.648847e-09, -1.095852e+02, -2.269565e+02,
-2.359033e+02, -3.581634e+02, -4.179317e+02, -1.904907e+02,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: F_theta.
Mismatched elements: 3 / 660 (0.455%)
Max absolute difference: 1.67492544e-08
Max relative difference: 1.97650688e-09
x: array([-4.687658e-10, 1.406297e-09, -2.109446e-09, 7.031486e-10,
4.687658e-10, -2.109446e-09, -4.906491e+01, -1.020250e+02,
-1.067778e+02, -1.633828e+02, -1.919069e+02, -8.779719e+01,...
y: array([-4.687658e-10, 1.406297e-09, -2.109446e-09, 7.031486e-10,
4.687658e-10, -2.109446e-09, -4.906491e+01, -1.020250e+02,
-1.067778e+02, -1.633828e+02, -1.919069e+02, -8.779719e+01,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J^zeta.
Mismatched elements: 1 / 660 (0.152%)
Max absolute difference: 1.62435754e-09
Max relative difference: 2.39065063e-10
x: array([ 6.169842e+01, 6.169842e+01, 6.169842e+01, 6.169842e+01,
6.169842e+01, 6.169842e+01, 1.094439e+04, 8.052767e+03,
3.164126e+03, -2.186919e+03, -6.393220e+03, -8.582584e+03,...
y: array([ 6.169842e+01, 6.169842e+01, 6.169842e+01, 6.169842e+01,
6.169842e+01, 6.169842e+01, 1.094439e+04, 8.052767e+03,
3.164126e+03, -2.186919e+03, -6.393220e+03, -8.582584e+03,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J.
Mismatched elements: 1 / 1980 (0.0505%)
Max absolute difference: 2.29847501e-08
Max relative difference: 2.3906495e-10
x: array([[-1.179022e-12, 3.689596e+02, 1.645854e+03],
[ 7.763537e-10, 3.689596e+02, 1.645854e+03],
[-8.259202e-10, 3.689596e+02, 1.645854e+03],...
y: array([[-1.179022e-12, 3.689596e+02, 1.645854e+03],
[ 7.763537e-10, 3.689596e+02, 1.645854e+03],
[-8.259202e-10, 3.689596e+02, 1.645854e+03],...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: F_rho.
Mismatched elements: 6 / 660 (0.909%)
Max absolute difference: 2.30247679e-08
Max relative difference: 3.13879473e-10
x: array([ 1.044221e+03, 7.644227e+02, 2.797981e+02, -2.797981e+02,
-7.644227e+02, -1.044221e+03, 8.748079e+02, 6.315174e+02,
8.434725e+02, 9.728103e+02, 8.031394e+01, -9.269602e+02,...
y: array([ 1.044221e+03, 7.644227e+02, 2.797981e+02, -2.797981e+02,
-7.644227e+02, -1.044221e+03, 8.748079e+02, 6.315174e+02,
8.434725e+02, 9.728103e+02, 8.031394e+01, -9.269602e+02,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J_phi.
Mismatched elements: 1 / 660 (0.152%)
Max absolute difference: 9.02218744e-09
Max relative difference: 2.3906495e-10
x: array([ 3.689596e+02, 3.689596e+02, 3.689596e+02, 3.689596e+02,
3.689596e+02, 3.689596e+02, 6.574153e+04, 4.829832e+04,
1.893106e+04, -1.305157e+04, -3.808409e+04, -5.107903e+04,...
y: array([ 3.689596e+02, 3.689596e+02, 3.689596e+02, 3.689596e+02,
3.689596e+02, 3.689596e+02, 6.574153e+04, 4.829832e+04,
1.893106e+04, -1.305157e+04, -3.808409e+04, -5.107903e+04,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J*sqrt(g).
Mismatched elements: 1 / 1980 (0.0505%)
Max absolute difference: 3.49973561e-08
Max relative difference: 2.39065196e-10
x: array([[-2.577098e-10, 0.000000e+00, 2.590525e-10],
[ 5.659700e-10, 5.283297e-10, -1.975447e-09],
[-3.107391e-10, -1.188742e-09, 4.018067e-09],...
y: array([[-2.577098e-10, 0.000000e+00, 2.590525e-10],
[ 5.659700e-10, 5.283297e-10, -1.975447e-09],
[-3.107391e-10, -1.188742e-09, 4.018067e-09],...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: J_rho.
Mismatched elements: 1 / 660 (0.152%)
Max absolute difference: 1.72271939e-08
Max relative difference: 1.49087068e-10
x: array([-4.127112e+02, -1.127548e+03, -1.540259e+03, -1.540259e+03,
-1.127548e+03, -4.127112e+02, 2.171914e+04, 4.405792e+04,
2.387998e+04, -1.988408e+04, -4.441021e+04, -2.299714e+04,...
y: array([-4.127112e+02, -1.127548e+03, -1.540259e+03, -1.540259e+03,
-1.127548e+03, -4.127112e+02, 2.171914e+04, 4.405792e+04,
2.387998e+04, -1.988408e+04, -4.441021e+04, -2.299714e+04,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: F_zeta.
Mismatched elements: 3 / 660 (0.455%)
Max absolute difference: 1.86773832e-08
Max relative difference: 1.97650681e-09
x: array([-4.897653e-10, 1.469296e-09, -2.203944e-09, 7.346479e-10,
4.897653e-10, -2.203944e-09, -4.639982e+01, -9.736332e+01,
-1.050441e+02, -1.705320e+02, -2.151820e+02, -1.034384e+02,...
y: array([-4.897653e-10, 1.469296e-09, -2.203944e-09, 7.346479e-10,
4.897653e-10, -2.203944e-09, -4.639982e+01, -9.736332e+01,
-1.050441e+02, -1.705320e+02, -2.151820e+02, -1.034384e+02,...
Not equal to tolerance rtol=1e-10, atol=1e-10
Parameterization: desc.equilibrium.equilibrium.Equilibrium. Name: F.
Mismatched elements: 14 / 1980 (0.707%)
Max absolute difference: 7.244671e-08
Max relative difference: 3.84854726e-09
x: array([[ 4.185952e+03, -8.299412e-11, 3.296918e-12],
[ 4.185952e+03, 8.262415e-10, -2.076149e-09],
[ 4.185952e+03, -9.693418e-10, 2.147146e-09],...
y: array([[ 4.185952e+03, -8.299412e-11, 3.296918e-12],
[ 4.185952e+03, 8.262415e-10, -2.076149e-09],
[ 4.185952e+03, -9.693418e-10, 2.147146e-09],
For unoptimized |B|, typically higher N is needed to capture all oscillations. θ will likely always be fine with low N since it's essentially just proportional to iota at fixed alpha, so Chebyshev converges fast.
Rather than doing coordinate mapping on M * N tensor grid each large, do it on M * (N_{\theta} where N θ is small. Then use the resulting, accurate, θ(α, ζ) to compute the interpolation points in (α, ζ) so that |B| (α, ζ) can be constructed with large Chebyshev resolution. This reduces the number of points to perform root finding on each time we want to compute a neoclassical objective, decreases interpolation error, increases speed etc.
… should pass all tests
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
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.
This is a dummy PR that combines fourier bounce with rg/NFP_fac so we can find umbilic equilibria with good omnigenity.
The purpose right now is to see if I can do such an optimization. The optimization throws a cryptic error when I combine combine umbilic with ripple (spline field.)