Why does my code produce a different graph every time I run it?

I am currently trying to produce a graph of acceleration. I initially had 66000 samples, and cropped it to 20001:66000 as this is where the interesting data is. There are also 3 other sensors with the data in a matrix, acceleration is the last column in said matrix. To perform an analysis I wished to average every 460 points into 1, to reduce the white noise on an fft. The code for this section was provided in a separate question and I am unsure whether it is correct. After the averaging I am trying to plot this graph (altering number of samples to a time domain). However, every time I run my code the graph produced keeps changing and I cannot figure out why. Any help would be greatly appreciated. Code is as below.

LongTestArray = table2array(LongTest);

A = LongTestArray(20001:66000, 1:4);

M = mean(A);

A(:,3) = A(:,3) - M(:,3);

A(:,4) = A(:,4) - M(:,4);

B = A;

B = randi(9, 46000, 4); %Create Matrix

Ar = reshape(B, [], 4, 4600); % Reshape

% A_RMS = sqrt(mean(Br.^2, 3)); % Calculate RMS

X=Ar(:,4);

f = 43.5 ; % reduced from sample rate of 20000 due to averaged results

samples = 0:size(X, 1)-1 ;

t = samples/f ;

%plot column 1

plot(t, X(:, 1) ) grid on ;

hold on ;

title('Low Frequency Accelerometer Test 1')

xlabel('Time (s)') % x-axis label

ylabel('Voltage (V)')

Answer 1

[KALYAN ACHARJYA]

% In your code having a statement to create a random matrix, each execution it generates the different random matrix.

B = randi(9, 46000, 4)

[Lara Severne]

So to change this would I need to simply reshape the original matrix? Wasn't made clear in original answer that this was for that purpose. Very new to coding so did not realise, sorry

[KALYAN ACHARJYA]

If B matrix is fixed in each execution, then only you will get same results.Do you have any fixed matrix value for B?

B = randi(maximum value element, no of rows, no of colm

Try this one

B = randi(1, 46000, 4); %It generates Unit Matrix

But I am not sure whether OK with your result or not.

[Lara Severne]

okay perfect thanks so much for your help!