How to create nonlinear spaced vector in Matlab? - matlab

I'm trying to create a contour plot with focus around a particular finite range from 1 to 1.05. At the same time, I need very high resolution closer to 1. I thought I could use something like the following but the spacing still looks linear
out=exp(linspace(log(1),log(1.05),100))
plot(diff(out))
What is the best way to enhance the nonlinearity of the spacing when the bounds are so tight? Again, I need to maintain high density near 1 with the resolution tapering off in a nonlinear way. I have a few ideas but I thought someone might have a quick 2 liner or something of the sort.

instead of applying the function f(x) = ex, to get a 'steeper' non-linearity, rather apply f(x) = eax
n = 20;
a = 100;
lower = 1;
upper = 1.05;
temp = exp(linspace(log(1)*a,log(1.05)*a,n))
% re-scale to be between 0 and 1
temp_01 = temp/max(temp) - min(temp)/max(temp)
% re-scale to be between your limits (i.e. 1 and 1.05)
out = temp_01*(upper-lower) + lower
now plot(diff(out),diff(out),'o') produces
Note that you can use the exact same scaling scheme above with logspace so just use
temp = logspace(...)
and then the rest is the same

You can generate a logarithmic distribution between, for example, 1 and 1000 and then scale it back to [1, 1.05]:
out = logspace(0, 3, 100);
out = ( (out-min(out(:)))*(1.05-1) ) / ( max(out(:))-min(out(:)) ) + 1;
Result:
plot(diff(out));

Related

Metropolis-Hastings in matlab

I am trying to use the Metropolis Hastings algorithm with a random walk sampler to simulate samples from a function $$ in matlab, but something is wrong with my code. The proposal density is the uniform PDF on the ellipse 2s^2 + 3t^2 ≤ 1/4. Can I use the acceptance rejection method to sample from the proposal density?
N=5000;
alpha = #(x1,x2,y1,y2) (min(1,f(y1,y2)/f(x1,x2)));
X = zeros(2,N);
accept = false;
n = 0;
while n < 5000
accept = false;
while ~accept
s = 1-rand*(2);
t = 1-rand*(2);
val = 2*s^2 + 3*t^2;
% check acceptance
accept = val <= 1/4;
end
% and then draw uniformly distributed points checking that u< alpha?
u = rand();
c = u < alpha(X(1,i-1),X(2,i-1),X(1,i-1)+s,X(2,i-1)+t);
X(1,i) = c*s + X(1,i-1);
X(2,i) = c*t + X(2,i-1);
n = n+1;
end
figure;
plot(X(1,:), X(2,:), 'r+');
You may just want to use the native implementation of matlab mhsample.
Regarding your code, there are a few things missing:
- function alpha,
- loop variable i (it might be just n but it is not suited for indexing since it starts at zero).
And you should always allocate memory in matlab if you want to fill it dynamically, i.e. X in your case.
To expand on the suggestions by #max, the code appears to work if you change the i indices to n and replace
n = 0;
with
n = 2;
X(:,1) = [.1,.1];
It would probably be better to assign X(:,1) to random values within your accept region (using the same code you use later), and/or include a burn-in period.
Depending upon what you are going to do with this, it may also make things cleaner to evaluate the argument to sin in the f function to keep it within 0 to 2 pi (likely by shifting the value by 2 pi if it exceeds those bounds)

Approximation of cosh and sinh functions that give large values in MATLAB

My calculation involves cosh(x) and sinh(x) when x is around 700 - 1000 which reaches MATLAB's limit and the result is NaN. The problem in the code is elastic_restor_coeff rises when radius is small (below 5e-9 in the code). My goal is to do another integral over a radius distribution from 1e-9 to 100e-9 which is still a work in progress because I get stuck at this problem.
My work around solution right now is to approximate the real part of chi_para with a step function when threshold2 hits a value of about 300. The number 300 is obtained from using the lowest possible value of radius and look at the cut-off value from the plot. I think this approach is not good enough for actual calculation since this value changes with radius so I am looking for a better approximation method. Also, the imaginary part of chi_para is difficult to approximate since it looks like a pulse instead of a step.
Here is my code without an integration over a radius distribution.
k_B = 1.38e-23;
T = 296;
radius = [5e-9,10e-9, 20e-9, 30e-9,100e-9];
fric_coeff = 8*pi*1e-3.*radius.^3;
elastic_restor_coeff = 8*pi*1.*radius.^3;
time_const = fric_coeff/elastic_restor_coeff;
omega_ar = logspace(-6,6,60);
chi_para = zeros(1,length(omega_ar));
chi_perpen = zeros(1,length(omega_ar));
threshold = zeros(1,length(omega_ar));
threshold2 = zeros(1,length(omega_ar));
for i = 1:length(radius)
for k = 1:length(omega_ar)
omega = omega_ar(k);
fric_coeff = 8*pi*1e-3.*radius(i).^3;
elastic_restor_coeff = 8*pi*1.*radius(i).^3;
time_const = fric_coeff/elastic_restor_coeff;
G_para_func = #(t) ((cosh(2*k_B*T./elastic_restor_coeff.*exp(-t./time_const))-1).*exp(1i.*omega.*t))./(cosh(2*k_B*T./elastic_restor_coeff)-1);
G_perpen_func = #(t) ((sinh(2*k_B*T./elastic_restor_coeff.*exp(-t./time_const))).*exp(1i.*omega.*t))./(sinh(2*k_B*T./elastic_restor_coeff));
chi_para(k) = (1 + 1i*omega*integral(G_para_func, 0, inf));
chi_perpen(k) = (1 + 1i*omega*integral(G_perpen_func, 0, inf));
threshold(k) = 2*k_B*T./elastic_restor_coeff*omega;
threshold2(k) = 2*k_B*T./elastic_restor_coeff*(omega*time_const - 1);
end
figure(1);
semilogx(omega_ar,real(chi_para),omega_ar,imag(chi_para));
hold on;
figure(2);
semilogx(omega_ar,real(chi_perpen),omega_ar,imag(chi_perpen));
hold on;
end
Here is the simplified function that I would like to approximate:
where x is iterated in a loop and the maximum value of x is about 700.

Reverse-calculating original data from a known moving average

I'm trying to estimate the (unknown) original datapoints that went into calculating a (known) moving average. However, I do know some of the original datapoints, and I'm not sure how to use that information.
I am using the method given in the answers here: https://stats.stackexchange.com/questions/67907/extract-data-points-from-moving-average, but in MATLAB (my code below). This method works quite well for large numbers of data points (>1000), but less well with fewer data points, as you'd expect.
window = 3;
datapoints = 150;
data = 3*rand(1,datapoints)+50;
moving_averages = [];
for i = window:size(data,2)
moving_averages(i) = mean(data(i+1-window:i));
end
length = size(moving_averages,2)+(window-1);
a = (tril(ones(length,length),window-1) - tril(ones(length,length),-1))/window;
a = a(1:length-(window-1),:);
ai = pinv(a);
daily = mtimes(ai,moving_averages');
x = 1:size(data,2);
figure(1)
hold on
plot(x,data,'Color','b');
plot(x(window:end),moving_averages(window:end),'Linewidth',2,'Color','r');
plot(x,daily(window:end),'Color','g');
hold off
axis([0 size(x,2) min(daily(window:end))-1 max(daily(window:end))+1])
legend('original data','moving average','back-calculated')
Now, say I know a smattering of the original data points. I'm having trouble figuring how might I use that information to more accurately calculate the rest. Thank you for any assistance.
You should be able to calculate the original data exactly if you at any time can exactly determine one window's worth of data, i.e. in this case n-1 samples in a window of length n. (In your case) if you know A,B and (A+B+C)/3, you can solve now and know C. Now when you have (B+C+D)/3 (your moving average) you can exactly solve for D. Rinse and repeat. This logic works going backwards too.
Here is an example with the same idea:
% the actual vector of values
a = cumsum(rand(150,1) - 0.5);
% compute moving average
win = 3; % sliding window length
idx = hankel(1:win, win:numel(a));
m = mean(a(idx));
% coefficient matrix: m(i) = sum(a(i:i+win-1))/win
A = repmat([ones(1,win) zeros(1,numel(a)-win)], numel(a)-win+1, 1);
for i=2:size(A,1)
A(i,:) = circshift(A(i-1,:), [0 1]);
end
A = A / win;
% solve linear system
%x = A \ m(:);
x = pinv(A) * m(:);
% plot and compare
subplot(211), plot(1:numel(a),a, 1:numel(m),m)
legend({'original','moving average'})
title(sprintf('length = %d, window = %d',numel(a),win))
subplot(212), plot(1:numel(a),a, 1:numel(a),x)
legend({'original','reconstructed'})
title(sprintf('error = %f',norm(x(:)-a(:))))
You can see the reconstruction error is very small, even using the data sizes in your example (150 samples with a 3-samples moving average).

sum of absolute differences of two images

I want to find how similar a picture is to some samples that I have (black and white).
I use the sum of absolute difference code, but because I'm new to MATLAB I didn't find out how to use it. How does this algorithm work? Does it give a measure of how similar the pics are?
I= imread('img1.jpg');
image2= imread('img2.jpg');
% J = uint8(filter2(fspecial('gaussian'), I));
K = imabsdiff(I,image2);
figure, imshow(K,[])
Well I think you pretty much answered your question yourself. It is the sum of the absolute difference. So let say you have img1 and img2 which are the same size and type.
To find the difference, do subtraction
img1-img2
To find the absolute difference, use the absolute value function abs
abs(img1-img2)
To find the sum, use the sum function. Note that you will need to do this for each "dimension" your image has. If you are not sure, type size(img1) and see if there are 2 or 3 numbers that show up, this corresponds to how many sum(...) you need to use.
For a color image (3 dimensions):
sum(sum(sum(abs(img1-img2))))
^^ Is the sum of the absolute differences. Whichever has the smallest sum can be considered the closest image.
If you have different sized images, you need to use the normxcorr2 function. This function will return a matrix of the same size with how well the template (small) image fits into the big image at each different point. Find the maximum value of that matrix and that is how well that image fits.
For instance:
correlation = normxcorr2(smallImg, bigImg);
compareMe = max(correlation(:))
It is best practice to use MATLAB's build-in function imabsdiff. In contrast to the other suggested answers, it takes care of the range boundaries if your image is formatted as uint8. Consider:
img1 = uint8(10);
img2 = uint8(20);
sum(abs(img1(:)-img2(:)))
gives you 0, whereas
imabsdiff(img1(:),img2(:))
correctly gives 10.
You should use the command im2col in MATLAB you will be able to do so in Vectorized manner.
Just arrange each neighborhood in columns (For each frame).
Put them in 3D Matrix and apply you operation on the 3rd dimension.
Code Snippet
I used Wikipedia's definition of "Sum of Absolute Differences".
The demo script:
```
% Sum of Absolute Differences Demo
numRows = 10;
numCols = 10;
refBlockRadius = 1;
refBlockLength = (2 * refBlockRadius) + 1;
mImgSrc = randi([0, 255], [numRows, numCols]);
mRefBlock = randi([0, 255], [refBlockLength, refBlockLength]);
mSumAbsDiff = SumAbsoluteDifferences(mImgSrc, mRefBlock);
```
The Function SumAbsoluteDifferences:
```
function [ mSumAbsDiff ] = SumAbsoluteDifferences( mInputImage, mRefBlock )
%UNTITLED2 Summary of this function goes here
% Detailed explanation goes here
numRows = size(mInputImage, 1);
numCols = size(mInputImage, 2);
blockLength = size(mRefBlock, 1);
blockRadius = (blockLength - 1) / 2;
mInputImagePadded = padarray(mInputImage, [blockRadius, blockRadius], 'replicate', 'both');
mBlockCol = im2col(mInputImagePadded, [blockLength, blockLength], 'sliding');
mSumAbsDiff = sum(abs(bsxfun(#minus, mBlockCol, mRefBlock(:))));
mSumAbsDiff = col2im(mSumAbsDiff, [blockLength, blockLength], [(numRows + blockLength - 1), (numCols + blockLength - 1)]);
end
```
Enjoy...

Finding the difference between two signals

I have two signals, let's call them 'a' and 'b'. They are both nearly identical signals (recorded from the same input and contain the same information) however, because I recorded them at two different 'b' is time shifted by an unknown amount. Obviously, there is random noise in each.
Currently, I am using cross correlation to compute the time shift, however, I am still getting improper results.
Here is the code I am using to calculate the time shift:
function [ diff ] = FindDiff( signal1, signal2 )
%FINDDIFF Finds the difference between two signals of equal frequency
%after an appropritate time shift is applied
% Calculates the time shift between two signals of equal frequency
% using cross correlation, shifts the second signal and subtracts the
% shifted signal from the first signal. This difference is returned.
length = size(signal1);
if (length ~= size(signal2))
error('Vectors must be equal size');
end
t = 1:length;
tx = (-length+1):length;
x = xcorr(signal1,signal2);
[mx,ix] = max(x);
lag = abs(tx(ix));
shifted_signal2 = timeshift(signal2,lag);
diff = signal1 - shifted_signal2;
end
function [ shifted ] = timeshift( input_signal, shift_amount )
input_size = size(input_signal);
shifted = (1:input_size)';
for i = 1:input_size
if i <= shift_amount
shifted(i) = 0;
else
shifted(i) = input_signal(i-shift_amount);
end
end
end
plot(FindDiff(a,b));
However the result from the function is a period wave, rather than random noise, so the lag must still be off. I would post an image of the plot, but imgur is currently not cooperating.
Is there a more accurate way to calculate lag other than cross correlation, or is there a way to improve the results from cross correlation?
Cross-correlation is usually the simplest way to determine the time lag between two signals. The position of peak value indicates the time offset at which the two signals are the most similar.
%// Normalize signals to zero mean and unit variance
s1 = (signal1 - mean(signal1)) / std(signal1);
s2 = (signal2 - mean(signal2)) / std(signal2);
%// Compute time lag between signals
c = xcorr(s1, s2); %// Cross correlation
lag = mod(find(c == max(c)), length(s2)) %// Find the position of the peak
Note that the two signals have to be normalized first to the same energy level, so that the results are not biased.
By the way, don't use diff as a name for a variable. There's already a built-in function in MATLAB with the same name.
Now there are two functions in Matlab:
one called finddelay
and another called alignsignals that can do what you want, I believe.
corr finds a dot product between vectors (v1, v2). If it works bad with your signal, I'd try to minimize a sum of squares of differences (i.e. abs(v1 - v2)).
signal = sin(1:100);
signal1 = [zeros(1, 10) signal];
signal2 = [signal zeros(1, 10)];
for i = 1:length(signal1)
signal1shifted = [signal1 zeros(1, i)];
signal2shifted = [zeros(1, i) signal2];
d2(i) = sum((signal1shifted - signal2shifted).^2);
end
[fval lag2] = min(d2);
lag2
It is computationally worse than cross-calculation which can be speeded up by using FFT. As far as I know you can't do this with euclidean distance.
UPD. Deleted wrong idea about cross-correlation with periodic signals
You can try matched filtering in frequency domain
function [corr_output] = pc_corr_processor (target_signal, ref_signal)
L = length(ref_signal);
N = length(target_signal);
matched_filter = flipud(ref_signal')';
matched_filter_Res = fft(matched_filter,N);
corr_fft = matched_filter_Res.*fft(target_signal);
corr_out = abs(ifft(corr_fft));
The peak of the matched filter maximum-index of corr_out above should give you the lag amount.