-
Notifications
You must be signed in to change notification settings - Fork 1
/
script_noise_test1.m
182 lines (168 loc) · 5.39 KB
/
script_noise_test1.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
%% Script to investigate accuracy
%% Setting up the script
clc, clear
% free parameters
dim = 2; % dimension (1,2,3)
points = 'equid'; % points (equid, uniform, Halton)
noise_level = 10^(-6);
CC = 50; % number of tests
% fixed parameters
domain = 'cube'; % domain (cube, ball)
volume = 2^dim;
weightFun = '1'; % weight function - 1, C2k, sqrt(r)
omega = generate_weightFun( weightFun, dim);
if dim == 1
f = @(x) 1./(1+x.^2); % test function
I = 2*atan(1); % exact integral
n = 20;
n_max = 400;
elseif dim == 2
f = @(x,y) (1./(1+x.^2)).*(1./(1+y.^2)); % test function
I = (2*atan(1))^2; % exact integral
n = 4;
n_max = 40;
else
f = @(x,y,z) (1./(1+x.^2)).*(1./(1+y.^2)).*(1./(1+z.^2)); % test function
I = (2*atan(1))^3; % exact integral
n = 4;
n_max = 16;
end
NN_Leg = []; NN_MC = []; NN_LS = []; NN_l1 = []; % number of data points
err_Leg = []; err_MC = []; err_LS = []; err_l1 = []; % errors
while n <= n_max
% Legendre rule
example = matfile(['CFs/CF_Leg_dim=',num2str(dim),'_',domain,'_n=',num2str(n),'.mat']);
C = example.CF_Leg;
[ N, aux] = size(C);
NN_Leg = [NN_Leg; N];
X = C(:,1:dim); % data points
w = C(:,dim+1); % weights
% function values
f_values = zeros(N,1);
for m = 1:N
if dim == 1
f_values(m) = f( X(m,1) ).*omega( X(m,1) );
elseif dim == 2
f_values(m) = f( X(m,1), X(m,2) ).*omega( X(m,1), X(m,2) );
elseif dim == 3
f_values(m) = f( X(m,1), X(m,2) , X(m,3) ).*omega( X(m,1), X(m,2) , X(m,3) );
else
error('Desired dimension not yet implemented!')
end
end
% generate and add uniform noise
error = 0;
for c=1:CC
noise = noise_level*(2*rand(N,1)-1);
f_values = f_values + noise;
error = error + abs( I - dot(w,f_values) );
end
err_Leg = [err_Leg; error/CC]; % absolute error
% LS rule
example = matfile(['CFs/CF_LS_dim=',num2str(dim),'_',domain,'_',weightFun,'_',points,'_n=',num2str(n),'.mat']);
C = example.CF_LS;
[ N, aux] = size(C);
NN_LS = [NN_LS; N];
X = C(:,1:dim); % data points
w = C(:,dim+1); % weights
% function values
f_values = zeros(N,1);
for m = 1:N
if dim == 1
f_values(m) = f( X(m,1) );
elseif dim == 2
f_values(m) = f( X(m,1), X(m,2) );
elseif dim == 3
f_values(m) = f( X(m,1), X(m,2) , X(m,3) );
else
error('Desired dimension not yet implemented!')
end
end
% generate and add uniform noise
error = 0;
for c=1:CC
noise = noise_level*(2*rand(N,1)-1);
f_values = f_values + noise;
error = error + abs( I - dot(w,f_values) );
end
err_LS = [err_LS; error/CC]; % absolute error
% l1 rule
example = matfile(['CFs/CF_l1_dim=',num2str(dim),'_',domain,'_',weightFun,'_',points,'_n=',num2str(n),'.mat']);
C = example.CF_l1;
[ N, aux] = size(C);
NN_l1 = [NN_l1; N];
X = C(:,1:dim); % data points
w = C(:,dim+1); % weights
% function values
f_values = zeros(N,1);
for m = 1:N
if dim == 1
f_values(m) = f( X(m,1) );
elseif dim == 2
f_values(m) = f( X(m,1), X(m,2) );
elseif dim == 3
f_values(m) = f( X(m,1), X(m,2) , X(m,3) );
else
error('Desired dimension not yet implemented!')
end
end
% generate and add uniform noise
error = 0;
for c=1:CC
noise = noise_level*(2*rand(N,1)-1);
f_values = f_values + noise;
error = error + abs( I - dot(w,f_values) );
end
err_l1 = [err_l1; error/CC]; % absolute error
% MC integration
NN_MC = [NN_MC; N];
% MC weights
for m = 1:N
if dim == 1
w(m) = volume*omega( X(m,1) )/N;
elseif dim == 2
w(m) = volume*omega( X(m,1), X(m,2) )/N;
elseif dim == 3
w(m) = volume*omega( X(m,1), X(m,2) , X(m,3) )/N;
else
error('Desired dimension not yet implemented!')
end
end
% generate and add uniform noise
error = 0;
for c=1:CC
noise = noise_level*(2*rand(N,1)-1);
f_values = f_values + noise;
error = error + abs( I - dot(w,f_values) );
end
err_MC = [err_MC; error/CC]; % absolute error
% increase n
if dim == 1
n = n + 20;
elseif dim == 2
n = n + 2;
else
n = n + 1;
end
end
% Plot the results
figure(1)
p = plot( NN_MC,err_MC,'ms', NN_LS,err_LS,'r+', NN_l1,err_l1,'b^', NN_Leg,err_Leg,'ko');
set(p, 'LineWidth',1.5)
set(p, 'markersize',8)
set(gca, 'FontSize', 20) % Increasing ticks fontsize
xlim([ max([NN_Leg(1);NN_LS(1)]), min([NN_Leg(end);NN_LS(end)]) ])
ylim([ min([err_MC;err_LS;err_l1;err_Leg])/10, max([err_MC;err_LS;err_l1;err_Leg])*10 ])
xlabel('$N$','Interpreter','latex')
ylabel('$|C[f] - I[f]|$','Interpreter','latex')
set(gca, 'XScale', 'log')
set(gca, 'YScale', 'log')
if strcmp( points, 'Halton')
id = legend('QMC','LS','$\ell^1$','Legendre','Interpreter','latex','Location','southwest');
else
id = legend('MC','LS','$\ell^1$','Legendre','Interpreter','latex','Location','southwest');
end
set(id, 'Interpreter','latex', 'FontSize',26)
grid on
str = sprintf( ['plots_noise/accuracy_test1_noisy_dim=',num2str(dim),'_',points,'.fig'] );
%savefig(str);