Last active
August 29, 2015 13:58
-
-
Save MiroK/10117588 to your computer and use it in GitHub Desktop.
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
from dolfin import * | |
from numpy.random import random, choice | |
class RandomF(Expression): | |
def eval(self, values, x): | |
values[0] = random() | |
def make_random(n_cells, n_random): | |
mesh = UnitIntervalMesh(n_cells) | |
V = FunctionSpace(mesh, 'CG', 1) | |
f = Function(V) | |
indices = choice(V.dim(), n_random) | |
values = random(n_random) | |
f.vector()[indices] = values | |
return f | |
def get_hm1_norm(f, n_cells): | |
mesh = UnitIntervalMesh(n_cells) | |
V = FunctionSpace(mesh, 'CG', 1) | |
u = TrialFunction(V) | |
v = TestFunction(V) | |
# Form for H^1_0 norm | |
a = inner(grad(u), grad(v))*dx | |
# Form for L^2 inner product | |
m = inner(u, v)*dx | |
L = inner(Constant(0.), v)*dx | |
bc = DirichletBC(V, Constant(0.), DomainBoundary()) | |
A, _ = assemble_system(a, L, bc) | |
M, _ = assemble_system(m, L, bc) | |
f_V = interpolate(f, V) | |
F = f_V.vector() | |
bc.apply(F) | |
plot(f_V, interactive=True) | |
g = Function(V) | |
G = g.vector() | |
A_dot_G = Function(V) | |
A_DOT_G = A_dot_G.vector() | |
M_dot_G = Function(V) | |
M_DOT_G = M_dot_G.vector() | |
functionals = [] | |
norms = [] | |
winner = (0, 0) | |
# Compute the H^{-1 norm} | |
hm1_norm = -1 | |
for i in range(1, V.dim()-1): | |
# Get basis function of V | |
G.zero() | |
G[i] = 1 | |
A_DOT_G = A*G | |
M_DOT_G = M*G | |
g_norm = sqrt(A_DOT_G.inner(G)) | |
functional = abs(M_DOT_G.inner(F)) | |
value = functional/g_norm | |
functionals.append(functional) | |
norms.append(g_norm) | |
if value > hm1_norm: | |
hm1_norm = value | |
winner = (functional, g_norm) | |
print '(f, v) min max', min(functionals), max(functionals) | |
print '(v, v)_1 min max ', min(norms), max(norms) | |
print '(f_v), (v, v)_1 for norm', winner | |
# Compute the L^{2} norm | |
M_DOT_G = M*F | |
l2_norm = sqrt(M_DOT_G.inner(F)) | |
# Compute the H^{1} norm | |
A_DOT_G = A*F | |
h1_norm = sqrt(A_DOT_G.inner(F)) | |
return h1_norm, l2_norm, hm1_norm | |
#------------------------------------------------------------------------------ | |
# Notes | |
print '\033[1;37;32m%s\033[0m' % 'sin(k*pi*x)' | |
u = Expression('sin(k*pi*x[0])', k=1) | |
print '\033[1;37;34m\t%s %s\t%s\t%s\033[0m' %\ | |
('k', 'H^{1} norm', 'L^{2} norm', 'H^{-1} norm') | |
for k in [1, 100, 100]: | |
u.k = k | |
h1_norm, l2_norm, hm1_norm = get_hm1_norm(u, n_cells=1000) | |
print '\033[1;37;34m%10d %10E %10E %10E\033[0m'\ | |
% (k, h1_norm, l2_norm, hm1_norm) | |
# Random exercise | |
print '\033[1;37;32m%s\033[0m' % 'random' | |
u = RandomF() | |
print '\033[1;37;34m\t%s %s\t%s\t%s\033[0m' %\ | |
('k', 'H^{1} norm', 'L^{2} norm', 'H^{-1} norm') | |
for k in [10, 100, 1000]: | |
h1_norm, l2_norm, hm1_norm = get_hm1_norm(u, n_cells=k) | |
print '\033[1;37;34m%10d %10E %10E %10E\033[0m'\ | |
% (k, h1_norm, l2_norm, hm1_norm) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment