print("level.nw_tensor \n", level.nw_tensor)
level.pad()
print("level.arr after padding\n", level.arr)
print("north_east\n", level.ne)
print("north_west\n", level.nw)
print("south_east\n", level.se)
print("south_west\n", level.sw)
print("evaluable view of stencil\n", level.evaluable_view(laplace_stencil))
is_on_border = level.border_function_generator(laplace_stencil)
border_truth = [[is_on_border((i, j)) for j in range(level.arr.shape[1])] for i in range(level.arr.shape[0])]
print("border_truth\n", np.asarray(border_truth, dtype=np.int))
level.rhs[:] = 0.0
laplace_stencil.modify_rhs(level)
print("level.rhs after modification\n", level.rhs)
print("==== DirectSolver ====")
direct_solver = DirectSolverSmoother(laplace_stencil, level)
laplace_stencil.modify_rhs(level)
direct_solver.relax()
print("level.arr after direct solution\n", level.mid)
print("test of the solution Ax=b by convolve\n ", laplace_stencil.eval_convolve(level.mid, "same"))
rhs_test = np.zeros(level.rhs.shape)
laplace_stencil.eval_sparse(level.mid, rhs_test)
print("test of the solution Ax=b by sparse matrix application \n", rhs_test)
print("==== SplitSmoother ====")
omega = 2.0/3.0
top_jacobi_smoother = SplitSmoother(l_plus / top_level.h**2,
l_minus / top_level.h**2,
top_level)
mid_jacobi_smoother = SplitSmoother(l_plus / mid_level.h**2,
l_minus / mid_level.h**2,
mid_level)
low_jacobi_smoother = SplitSmoother(l_plus / low_level.h**2,
l_minus / low_level.h**2,
low_level)
low_direct_smoother = DirectSolverSmoother(laplace_stencil, low_level)
# time to test the relaxation methods
print("===== DirectSolverSmoother Test =====")
low_level.rhs[:] = 0.0
low_level.pad()
print("arr:", low_level.arr)
laplace_stencil.modify_rhs(low_level)
print("rhs:", low_level.rhs)
low_direct_smoother.relax()
print(low_level.arr)
low_level.pad()
# Lets test the SplitSmoother by using the jacobi smoother
# but for this case we need an initial guess
print("===== JacobiSmoother Test =====")
# define the 3 different JacobiSmoother Implementations
jacobi_loop = WeightedJacobiSmoother(laplace_stencil,
low_level, 0.5, "loop")
jacobi_matrix = WeightedJacobiSmoother(laplace_stencil,
low_level, 0.5, "matrix")
jacobi_convolve = WeightedJacobiSmoother(laplace_stencil,
low_level, 0.5, "convolve")
ipl_mid_to_top = InterpolationByStencilForLevelsClassical(mid_level, top_level, ipl_stencil_list_standard,
pre_assign=iadd)
ipl_low_to_mid = InterpolationByStencilForLevelsClassical(low_level, mid_level, ipl_stencil_list_standard,
pre_assign=iadd)
n_pre = 1
n_post = 1
print("\t \t --Nothing done--\n")
print_all(top_level)
top_level.mid[:] = 1
top_level.pad()
mg_problem.fill_rhs(top_level)
top_stencil.modify_rhs(top_level)
print("\t \t --Initialisation of top level --\n")
print_all(top_level)
top_jacobi_smoother.relax(n_pre)
print("\t\t --After %d Jacobi iterations --" % n_pre)
print_all(top_level)
top_level.compute_residual(top_stencil)
print("\t\t --After the computation of the residual--")
print_all(top_level)
rst_top_to_mid.restrict()
