How to find a minimum amount of equal adjacent values in an array?

Hi all,

I have a column vector X of 3000 data points long (15 min averaged data). To check whether a certain system freezed for at least a certain period of time, I want to check this array for a minimum of 100 adjacent data points with the exact same value.

I started like this:

diff_X = diff(X); 
  
movingsum_X = movsum(diff_X,100); 

flatlines = find(movingsum_X == 0);

Despite this giving results that are partly satisfying, it doesn't do the trick all the way unfortunately.

Anyone that can help me out extending this to a fully working script? Other ways to reach the goal are welcome too, thanks.

Lennard

[Sourabh Kondapaka]

Hi Lennard Pol,

Could you upload the dataset that you are working with ?

[Bruno Luong]

Your code is flawed and fails if the signal is sum of pure oscillations (signal without DC component). Change

diff_X = diff(X)

to

diff_X = abs(diff(X))

will fix the flaw.

[Lennard Pol]

This improved the performance, thanks!

Answer 1

[Setsuna Yuuki.]

You can test this code with different A matrix:

x = 2;
% X x 3000 matrix of example
% A = ones(x,1000); %there is always error
% A = [3*ones(2,25) 4*ones(2,48) 2*ones(2,2901)]; % there is error after 147 cycles
% A = rand(x,3000); % there is never error
B = reshape(A,1,x*length(A));
cont = 0; contM = 0;
for n = 1:length(B)-99 % n indicates the number that is repeated 100 times
   for m = n:(n+99)
      if(B(n) == B(m)) %Compare wit future values
         cont = cont+1; 
      end
   end
   if (cont == 100) %If cont == 100 the system has stopped
        fprintf("Error in system \n");
        break; %Only if you want stop analysis
   else
        fprintf("Good luck \n"); %Good luck
   end
    cont = 0;
end

[Lennard Pol]

Valuable contribution, thanks Bastian!