My MATLAB code for fft and ifft below has a problem with the inverse Fourier signal y not matching the in put signal x. Is there any solution to resolve this?
N = 1000;
t0 = 1e-13;
tau = 2*1e-14;
n = [0:t0/40:2*1e-13-t0/40];
f0 = 3*1e8/(150*1e-9);
x = cos(2*pi*f0*n);
x = x.*exp((-(n-t0).^2)./(tau^2));
X = abs(fft(x,N));
F = [-N/2 : N/2 - 1]/N;
X = fftshift(X);
y=ifft(X,80);
figure(3)
plot(n,y)
I see a number of issues here:
N = 1000;
t0 = 1e-13;
tau = 2*1e-14;
n = [0:t0/40:2*1e-13-t0/40];
f0 = 3*1e8/(150*1e-9);
x = cos(2*pi*f0*n);
x = x.*exp((-(n-t0).^2)./(tau^2));
% X = abs(fft(x,N)); <-- Not seen this technique before, and why N=1000?
% try something more like:
X = fft(x);
F = [-N/2 : N/2 - 1]/N;
% this is fine to shift and plot the function
Xshifted = fftshift(X);
plot( abs( Xshifted ) )
% now you're taking the inverse of the shifted function, not what you want
% y=ifft(X,80); also not sure about the 80
y = ifft(X);
figure(3)
plot(n,y)
figure(4)
plot( n, x ); hold on; plot( n, y, 'o' )
That's all I see at first. HTH!
If you take the absolute value of the fft, you destroy the phase information needed to reconstruct the original signal, i.e. the moment you compute
X = abs(fft(x,N));
You cannot go back via ifft, because now you only have the magnitude.
Also, the inverse transformation only works if you use the same number of FFT bins with NFFT>=length(x).
y=ifft(fft(x));
should be exactly the same as x.
Related
This is my code for finding the centered coefficients for lagrange polynomial interpolation:
% INPUT
% f f scalar - valued function
% interval interpolation interval [a, b]
% n interpolation order
%
% OUTPUT
% coeff centered coefficients of Lagrange interpolant
function coeff = lagrangeInterp (f, interval , n)
a = interval(1);
b = interval(2);
x = linspace(a,b,n+1);
y = f(x);
coeff(1,:) = polyfit(x,y,n);
end
Which is called in the following script
%Plot lagrangeInterp and sin(x) together
hold on
x = 0:0.1*pi:2*pi;
for n = 1:1:4
coeff = lagrangeInterp(#(x)sin(x),[0,2*pi],n);
plot(x,polyval(coeff,x,'-'));
end
y = sin(x);
plot(x,y);
legend('1st order','2nd order','3rd order','4th order','sin(x)');
To check for stability I would like to perturb the function (eg g(x) = f(x) + epsilon). How would I go about this?
Well, a little trick for you.
You know randn([m,n]) in matlab generate a m*n random matrix. The point is to generate a random vector, and interp1 to a function of x. Like this:
x = linspace(a,b,n+1); % Your range of input
g = #(ep,xx)f(xx)+interp1(x,ep*randn([length(x),1]),xx);
I want to understand the discrete fourier transform by implementing it by myself.
While the result returned by my DFT is not correct the in matlab included version returns the correct frequencies of the original signal.
So the question is, where went I wrong. Is it a math or a implementation problem?
%% Initialisation
samples=2000;
nfft = 1024;
K = nfft / 2 + 1;
c = 264;
e = 330;
t = -1:1/samples:1-1/samples;
[~, N] = size(t);
f = (sin(2*c*pi*t)+cos(2*e*pi*t)).*exp(-pi*(2*t-1).^2);
X = zeros(nfft, 1);
%% Discrete Fourier Transform
if true
for k=1:nfft
for n=1:nfft
X(k) = X(k) + f(n)*exp(-j*2*pi*(k-1)*(n-1)/N);
end
end
else
X=fft(f, nfft);
end
R = abs(X(1:K));
[V,I] = sort(R,'descend');
F1 = samples*(I(1)-1)/nfft;
F2 = samples*(I(2)-1)/nfft;
disp(F1)
disp(F2)
plot(1:K, R, 1:K, real(X(1:K)), 1:K, imag(X(1:K)))
The issue lies in the number of samples for which the transform is done.
Xall = fft(f);
plot(abs(Xall(1:500)),'b');
hold on
plot(abs(X(1:500)),'r');
What you compute matches the result from the FFT done on all samples (i.e. with 4000 real samples in and 4000 complex values out).
Now, if you read the documentation of FFT with doc fft you will see that the signal is truncated if the output size is smaller than the input size. If you try:
Y = zeros(nfft, 1);
for k=1:nfft
for n=1:nfft
Y(k) = Y(k) + f(n)*exp(-1j*2*pi*(k-1)*(n-1)/nfft);
end
end
Y2 = fft(f(:),nfft); %make it a column
abs(sum(Y-Y2)) %6.0380e-12 , result within precision of the double float format
I am coding a Gaussian Process regression algorithm. Here is the code:
% Data generating function
fh = #(x)(2*cos(2*pi*x/10).*x);
% range
x = -5:0.01:5;
N = length(x);
% Sampled data points from the generating function
M = 50;
selection = boolean(zeros(N,1));
j = randsample(N, M);
% mark them
selection(j) = 1;
Xa = x(j);
% compute the function and extract mean
f = fh(Xa) - mean(fh(Xa));
sigma2 = 1;
% computing the interpolation using all x's
% It is expected that for points used to build the GP cov. matrix, the
% uncertainty is reduced...
K = squareform(pdist(x'));
K = exp(-(0.5*K.^2)/sigma2);
% upper left corner of K
Kaa = K(selection,selection);
% lower right corner of K
Kbb = K(~selection,~selection);
% upper right corner of K
Kab = K(selection,~selection);
% mean of posterior
m = Kab'*inv(Kaa+0.001*eye(M))*f';
% cov. matrix of posterior
D = Kbb - Kab'*inv(Kaa + 0.001*eye(M))*Kab;
% sampling M functions from from GP
[A,B,C] = svd(Kaa);
F0 = A*sqrt(B)*randn(M,M);
% mean from GP using sampled points
F0m = mean(F0,2);
F0d = std(F0,0,2);
%%
% put together data and estimation
F = zeros(N,1);
S = zeros(N,1);
F(selection) = f' + F0m;
S(selection) = F0d;
% sampling M function from posterior
[A,B,C] = svd(D);
a = A*sqrt(B)*randn(N-M,M);
% mean from posterior GPs
Fm = m + mean(a,2);
Fmd = std(a,0,2);
F(~selection) = Fm;
S(~selection) = Fmd;
%%
figure;
% show what we got...
plot(x, F, ':r', x, F-2*S, ':b', x, F+2*S, ':b'), grid on;
hold on;
% show points we got
plot(Xa, f, 'Ok');
% show the whole curve
plot(x, fh(x)-mean(fh(x)), 'k');
grid on;
I expect to get some nice figure where the uncertainty of unknown data points would be big and around sampled data points small. I got an odd figure and even odder is that the uncertainty around sampled data points is bigger than on the rest. Can someone explain to me what I am doing wrong? Thanks!!
There are a few things wrong with your code. Here are the most important points:
The major mistake that makes everything go wrong is the indexing of f. You are defining Xa = x(j), but you should actually do Xa = x(selection), so that the indexing is consistent with the indexing you use on the kernel matrix K.
Subtracting the sample mean f = fh(Xa) - mean(fh(Xa)) does not serve any purpose, and makes the circles in your plot be off from the actual function. (If you choose to subtract something, it should be a fixed number or function, and not depend on the randomly sampled observations.)
You should compute the posterior mean and variance directly from m and D; no need to sample from the posterior and then obtain sample estimates for those.
Here is a modified version of the script with the above points fixed.
%% Init
% Data generating function
fh = #(x)(2*cos(2*pi*x/10).*x);
% range
x = -5:0.01:5;
N = length(x);
% Sampled data points from the generating function
M = 5;
selection = boolean(zeros(N,1));
j = randsample(N, M);
% mark them
selection(j) = 1;
Xa = x(selection);
%% GP computations
% compute the function and extract mean
f = fh(Xa);
sigma2 = 2;
sigma_noise = 0.01;
var_kernel = 10;
% computing the interpolation using all x's
% It is expected that for points used to build the GP cov. matrix, the
% uncertainty is reduced...
K = squareform(pdist(x'));
K = var_kernel*exp(-(0.5*K.^2)/sigma2);
% upper left corner of K
Kaa = K(selection,selection);
% lower right corner of K
Kbb = K(~selection,~selection);
% upper right corner of K
Kab = K(selection,~selection);
% mean of posterior
m = Kab'/(Kaa + sigma_noise*eye(M))*f';
% cov. matrix of posterior
D = Kbb - Kab'/(Kaa + sigma_noise*eye(M))*Kab;
%% Plot
figure;
grid on;
hold on;
% GP estimates
plot(x(~selection), m);
plot(x(~selection), m + 2*sqrt(diag(D)), 'g-');
plot(x(~selection), m - 2*sqrt(diag(D)), 'g-');
% Observations
plot(Xa, f, 'Ok');
% True function
plot(x, fh(x), 'k');
A resulting plot from this with 5 randomly chosen observations, where the true function is shown in black, the posterior mean in blue, and confidence intervals in green.
I've got a vector and I want to calculate the moving average of it (using a window of width 5).
For instance, if the vector in question is [1,2,3,4,5,6,7,8], then
the first entry of the resulting vector should be the sum of all entries in [1,2,3,4,5] (i.e. 15);
the second entry of the resulting vector should be the sum of all entries in [2,3,4,5,6] (i.e. 20);
etc.
In the end, the resulting vector should be [15,20,25,30]. How can I do that?
The conv function is right up your alley:
>> x = 1:8;
>> y = conv(x, ones(1,5), 'valid')
y =
15 20 25 30
Benchmark
Three answers, three different methods... Here is a quick benchmark (different input sizes, fixed window width of 5) using timeit; feel free to poke holes in it (in the comments) if you think it needs to be refined.
conv emerges as the fastest approach; it's about twice as fast as coin's approach (using filter), and about four times as fast as Luis Mendo's approach (using cumsum).
Here is another benchmark (fixed input size of 1e4, different window widths). Here, Luis Mendo's cumsum approach emerges as the clear winner, because its complexity is primarily governed by the length of the input and is insensitive to the width of the window.
Conclusion
To summarize, you should
use the conv approach if your window is relatively small,
use the cumsum approach if your window is relatively large.
Code (for benchmarks)
function benchmark
clear all
w = 5; % moving average window width
u = ones(1, w);
n = logspace(2,6,60); % vector of input sizes for benchmark
t1 = zeros(size(n)); % preallocation of time vectors before the loop
t2 = t1;
th = t1;
for k = 1 : numel(n)
x = rand(1, round(n(k))); % generate random row vector
% Luis Mendo's approach (cumsum)
f = #() luisMendo(w, x);
tf(k) = timeit(f);
% coin's approach (filter)
g = #() coin(w, u, x);
tg(k) = timeit(g);
% Jubobs's approach (conv)
h = #() jubobs(u, x);
th(k) = timeit(h);
end
figure
hold on
plot(n, tf, 'bo')
plot(n, tg, 'ro')
plot(n, th, 'mo')
hold off
xlabel('input size')
ylabel('time (s)')
legend('cumsum', 'filter', 'conv')
end
function y = luisMendo(w,x)
cs = cumsum(x);
y(1,numel(x)-w+1) = 0; %// hackish way to preallocate result
y(1) = cs(w);
y(2:end) = cs(w+1:end) - cs(1:end-w);
end
function y = coin(w,u,x)
y = filter(u, 1, x);
y = y(w:end);
end
function jubobs(u,x)
y = conv(x, u, 'valid');
end
function benchmark2
clear all
w = round(logspace(1,3,31)); % moving average window width
n = 1e4; % vector of input sizes for benchmark
t1 = zeros(size(n)); % preallocation of time vectors before the loop
t2 = t1;
th = t1;
for k = 1 : numel(w)
u = ones(1, w(k));
x = rand(1, n); % generate random row vector
% Luis Mendo's approach (cumsum)
f = #() luisMendo(w(k), x);
tf(k) = timeit(f);
% coin's approach (filter)
g = #() coin(w(k), u, x);
tg(k) = timeit(g);
% Jubobs's approach (conv)
h = #() jubobs(u, x);
th(k) = timeit(h);
end
figure
hold on
plot(w, tf, 'bo')
plot(w, tg, 'ro')
plot(w, th, 'mo')
hold off
xlabel('window size')
ylabel('time (s)')
legend('cumsum', 'filter', 'conv')
end
function y = luisMendo(w,x)
cs = cumsum(x);
y(1,numel(x)-w+1) = 0; %// hackish way to preallocate result
y(1) = cs(w);
y(2:end) = cs(w+1:end) - cs(1:end-w);
end
function y = coin(w,u,x)
y = filter(u, 1, x);
y = y(w:end);
end
function jubobs(u,x)
y = conv(x, u, 'valid');
end
Another possibility is to use cumsum. This approach probably requires fewer operations than conv does:
x = 1:8
n = 5;
cs = cumsum(x);
result = cs(n:end) - [0 cs(1:end-n)];
To save a little time, you can replace the last line by
%// clear result
result(1,numel(x)-n+1) = 0; %// hackish way to preallocate result
result(1) = cs(n);
result(2:end) = cs(n+1:end) - cs(1:end-n);
If you want to preserve the size of your input vector, I suggest using filter
>> x = 1:8;
>> y = filter(ones(1,5), 1, x)
y =
1 3 6 10 15 20 25 30
>> y = (5:end)
y =
15 20 25 30
This is my code:
T
F = 1/T
y
L = length(y)
Y=fft(y);
f=(0:F:(L-1)*F)
plot(f, Y)
T is the sampling time (with its value), F is the frequency and y is the discrete signal.
Is it the correct way to compute DFT using Matlab? I haven't passed F or T to the function so I'm not sure if the results Y correspond to their respective multiple frequencies of F stored in f.
F = 1/T
L = length(y)
t = linspace(0,L/F,L);
NFFT = 2^nextpow2(L);
f = F/2*linspace(0,1,NFFT/2+1);
fft_y = fft(y, NFFT)/L;
magnitude_y = = 2*abs(fft_y(1:NFFT/2+1));
plot(f, magnitude_y);
This is the correct way.