We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
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
Hi, when I switch to fmm assembly for operators defined on segments, I get very different results.
fmm
segments
I have attached a very simple script that reproduces the behavior.
import bempp.api import numpy as np import scipy from matplotlib import pyplot as plt # Define parameters: Even with a coarse resolution, # where the FMM is expected to be dense, the results vary significantly across the multishape. k0 = 1 precision = 1 #h = 2 * np.pi/(precision*k0) h = 1 grid = bempp.api.shapes.multitrace_cube(h=h) # Set the segments and swapped_normals corresponding to \Omega_1, the first half-cube in the multitrace_cube. segments = [[1, 2, 3, 4, 5, 6]] swapped_normals = [[]] # The same half cube without more triangles. grid1=bempp.api.shapes.cuboid(length = (1,1,0.5), origin=(0,0,0), h=h) # Define the function spaces. rwg = [bempp.api.function_space(grid, "RWG", 0, segments=seg, swapped_normals=normals, include_boundary_dofs=True) for seg, normals in zip(segments, swapped_normals)] snc = [bempp.api.function_space(grid, "SNC", 0, segments=seg, swapped_normals=normals,include_boundary_dofs=True) for seg, normals in zip(segments, swapped_normals)] rwg1 = bempp.api.function_space(grid1, "RWG", 0,include_boundary_dofs=True) snc1 = bempp.api.function_space(grid1, "SNC", 0,include_boundary_dofs=True) # Assembly A_1 in dense mode and fmm mode for both grids efiem_dense = bempp.api.operators.boundary.maxwell.electric_field(rwg[0], rwg[0], snc[0], k0).weak_form() efiem_fmm = bempp.api.operators.boundary.maxwell.electric_field(rwg[0], rwg[0], snc[0], k0, assembler='fmm').weak_form() efie1m_dense = bempp.api.operators.boundary.maxwell.electric_field(rwg1, rwg1, snc1, k0).weak_form() efie1m_fmm = bempp.api.operators.boundary.maxwell.electric_field(rwg1, rwg1, snc1, k0, assembler='fmm').weak_form() N = efiem_dense.shape[0] k = min(100,N-2) # Plot the k first eigenvalues eigvalsm_dense, eivects = scipy.sparse.linalg.eigs(efiem_dense, k = k) eigvalsm_fmm, eivects = scipy.sparse.linalg.eigs(efiem_fmm, k = k) eigvals1m_dense, eivects = scipy.sparse.linalg.eigs(efie1m_dense, k = k) eigvals1m_fmm, eivects = scipy.sparse.linalg.eigs(efie1m_fmm, k = k) fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, sharex=True, sharey=True) ax1.scatter(np.real(eigvalsm_dense), np.imag(eigvalsm_dense), label='efiem_dense') ax2.scatter(np.real(eigvalsm_fmm), np.imag(eigvalsm_fmm), label='efiem_fmm') ax3.scatter(np.real(eigvals1m_dense), np.imag(eigvals1m_dense), label='efiem_dense') ax4.scatter(np.real(eigvals1m_fmm), np.imag(eigvals1m_fmm), label='efiem_fmm')
which results in:
For the multishape with FMM (top-right), there is a clear problem.
The text was updated successfully, but these errors were encountered:
map_to_points
No branches or pull requests
Hi, when I switch to
fmm
assembly for operators defined onsegments
, I get very different results.I have attached a very simple script that reproduces the behavior.
which results in:
For the multishape with FMM (top-right), there is a clear problem.
The text was updated successfully, but these errors were encountered: