def plotSolution(path, error):
"""
Definition space of the solution is assumed to be [0, 1)
Length of loaded array is assumed to be N = N_sim+2, where 2 comes
from adding the boundaries after calculations.
h = 1/(N_sim+1) = 1/(N-1)
Plotting only solutions with high steplength to avoid crash of gnuplot
"""
dataSolution = np.fromfile(path, dtype=np.float64)
N = len(dataSolution)
h = 1.0 / (N - 1)
x = np.linspace(0, 1, N)
solution = analytic_solution(x)
if N < 10000:
figure()
plot(x, dataSolution, 'o')
hold('on')
plot(x, solution)
title('Solution to -u\'\'=f(x), h=%6.2g' % h)
xlabel('x')
ylabel('u(x)')
legend('Approximation', 'Analytic solution')
error_array = solution - dataSolution
error_array = np.abs(error_array / solution)
error_array[0] = 0
error_array[len(error_array) - 1] = 0
error.append([h, np.max(error_array)])
def plotSolution(path, error):
"""
Definition space of the solution is assumed to be [0, 1)
Length of loaded array is assumed to be N = N_sim+2, where 2 comes
from adding the boundaries after calculations.
h = 1/(N_sim+1) = 1/(N-1)
Plotting only solutions with high steplength to avoid crash of gnuplot
"""
dataSolution = np.fromfile(path, dtype=np.float64);
N = len(dataSolution);
h = 1.0/(N-1);
x = np.linspace(0, 1, N);
solution = analytic_solution(x);
if N<10000:
figure()
plot(x, dataSolution, 'o');
hold('on');
plot(x, solution);
title('Solution to -u\'\'=f(x), h=%6.2g' %h)
xlabel('x')
ylabel('u(x)')
legend('Approximation', 'Analytic solution')
error_array = solution-dataSolution
error_array = np.abs(error_array/solution)
error_array[0] = 0;
error_array[len(error_array)-1] =0;
error.append([h, np.max(error_array)])
def plot_fitted_velocities(self, destination = False):
if self.P == 0:
self.fit_velocities();
data = self.model.modelFunction([self.v_mean, self.v_r, self.P, self.t_0], self.t);
figure();
plot(self.t, self.velocities, '--r');
hold('on');
plot(self.t, data, '-b')
title("Least square modeled velocities");
xlabel('t [s]');
ylabel('v [m/s]');
legend('v_mean=%g \n v_r=%g \n P=%g \n t_0=%g' % (self.v_mean, self.v_r, self.P, self.t_0));
if destination:
hardcopy(destination);
def plot_fitted_velocities(self, destination=False):
if self.P == 0:
self.fit_velocities()
data = self.model.modelFunction(
[self.v_mean, self.v_r, self.P, self.t_0], self.t)
figure()
plot(self.t, self.velocities, '--r')
hold('on')
plot(self.t, data, '-b')
title("Least square modeled velocities")
xlabel('t [s]')
ylabel('v [m/s]')
legend('v_mean=%g \n v_r=%g \n P=%g \n t_0=%g' %
(self.v_mean, self.v_r, self.P, self.t_0))
if destination:
hardcopy(destination)
def plotSolution(path, name, rho, eigenvalues):
dataSolution = np.fromfile(path + name, dtype=np.float64)
N = len(dataSolution)
x = rho
dataSolution = dataSolution / np.sqrt(trapz(dataSolution ** 2 / x ** 2, x))
state_number = round(float((name.strip(".bin").strip("state"))))
integral = trapz(dataSolution ** 2 / x ** 2, x)
print state_number
if N < 10000:
figure()
plot(x, dataSolution ** 2 / x ** 2)
hold("on")
xlabel(r"$\rho$")
ylabel(r"$R^2(\rho)$")
title(name.strip(".bin") + ", " + "$\lambda =$ %g" % eigenvalues[state_number])
legend("integral = %g" % integral)
else:
print "Stopped plot, N>10000"
def plotSolution(path, name, rho, eigenvalues):
dataSolution = np.fromfile(path + name, dtype=np.float64)
N = len(dataSolution)
x = rho
dataSolution = dataSolution / np.sqrt(trapz(dataSolution**2 / x**2, x))
state_number = round(float((name.strip(".bin").strip("state"))))
integral = trapz(dataSolution**2 / x**2, x)
print state_number
if N < 10000:
figure()
plot(x, dataSolution**2 / x**2)
hold('on')
xlabel(r'$\rho$')
ylabel(r'$R^2(\rho)$')
title(
name.strip('.bin') + ", " +
'$\lambda =$ %g' % eigenvalues[state_number])
legend("integral = %g" % integral)
else:
print "Stopped plot, N>10000"
prec_gauleg = np.asarray(prec_gauleg)
prec_is_mc = np.asarray(prec_is_mc)
prec_bf_mc = np.asarray(prec_bf_mc)
sort_data(prec_gaulag)
sort_data(prec_gauleg)
sort_data(prec_is_mc)
sort_data(prec_bf_mc)
figure()
loglog(prec_gaulag[:,0], prec_gaulag[:,1])
hold("all")
loglog(prec_gauleg[:,0], prec_gauleg[:,1])
loglog(prec_bf_mc[:,0], prec_bf_mc[:,1])
loglog(prec_is_mc[:,0], prec_is_mc[:,1])
legend("Gauss Laguerre", "Gauss Legendre", "Brute force Monte Carlo", "Importance sampling Monte Carlo")
xlabel("Number of function evaluations")
ylabel("Relative error")
title("Relative error vs number of calculations for different methods")
hold("off")
# Plotting function evaluations and cpu_time
prec_gaulag = []
prec_gauleg = []
prec_is_mc = []
prec_bf_mc = []
for sim in simulations:
if sim.ID == "gaulag":
prec_gaulag.append([sim.N, sim.cpu_time])
if sim.ID == "gauleg":
