peak average of consecutive values in a matrix array

Question

Jorge Luis Paredes Estacio 2023-7-19

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1997748-peak-average-of-consecutive-values-in-a-matrix-array

评论： Mathieu NOE 2023-7-19

Hello, I would like to calculate the average of the peaks (positive and negative values) of consecutive values from a predifined number of elements from a Nx1 matrix, where N is the number of rows. For example, let's say I have a matrix with the following form (24x1):

A[1;2;3;4;5;6;7;8;9;10;11;12;13;14;15;16;17;18;19;20;21;22;23;24]

I want to obtain the peak and the average for every 5 elements,

B=[mean(peak(1,2,3,4,5)); mean(peak((2,3,4,5,6)); mean(peak((3,4,5,6,7));...........mean(peak((20,21,22,23,24))]

I am chosing every 5 elements to simplify the example. But, this number number of elements can be higher than 100 and the points vary dramatically from positive to negative as they come from acceleration earthquake signals from N higher than 245000.

I would appreciate the help.

6 个评论
显示 4更早的评论隐藏 4更早的评论

Dyuman Joshi 2023-7-19

在 MATLAB Online 中打开

Suppose this is the data in hand -

x = randi([-20 20],1,15)
x = 1×15
    -5   -16    14   -18    19    -3    18     1   -10    -6   -12    16     5     8     0

Assume the moving window is 4, what should be the output in thise case? and what is the logic behind it?

Jorge Luis Paredes Estacio 2023-7-19

编辑：Jorge Luis Paredes Estacio 2023-7-19

Thank you for your help.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Mathieu NOE 2023-7-19

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1997748-peak-average-of-consecutive-values-in-a-matrix-array#answer_1275518

在 MATLAB Online 中打开

hello

maybe this ?

I am using peakseek , this fex submission is available here

PeakSeek - File Exchange - MATLAB Central (mathworks.com)

, but you can use the standard (slower) findpeaks function offered by TMW

the code provided below will buffer your data with length = 51 samples , then split the signal for each cahnnel into a positive and negative signal and find the positive and negative peaks , then take the mean of them (separately not mixing pos and neg peaks !)

example for X channel :

data = readmatrix('Signal.txt'); % acceleration earthquake signal in cm/sec2 ,X, Y and Z
[samples, channels] = size(data);
dt = 1; % sample rate (s)
t = dt*(0:samples-1);  %  time vector 
%% home made solution (you choose the amount of overlap)
buffer_size = 51; % how many samples 
overlap = 0; % overlap expressed in samples
%%%% main loop %%%%
[new_time,data_out_pos,data_out_neg] = my_peakmean(t,data,buffer_size,overlap);
figure(1),
plot(t,data(:,1),new_time,data_out_pos(:,1),'*-g',new_time,data_out_neg(:,1),'*-r');
title('X');
legend('raw data','pos peaks mean','neg peaks mean');
xlabel('Time(s)');
ylabel('cm/sec2');
figure(2),
plot(t,data(:,2),new_time,data_out_pos(:,2),'*-g',new_time,data_out_neg(:,2),'*-r');
title('Y');
legend('raw data','pos peaks mean','neg peaks mean');
xlabel('Time(s)');
ylabel('cm/sec2');
figure(3),
plot(t,data(:,3),new_time,data_out_pos(:,3),'*-g',new_time,data_out_neg(:,3),'*-r');
title('Z');
legend('raw data','pos peaks mean','neg peaks mean');
xlabel('Time(s)');
ylabel('cm/sec2');
%%%%%%%%%% my functions %%%%%%%%%%%%%%
function [new_time,data_out_pos,data_out_neg] = my_peakmean(t,data_in,buffer_size,overlap)
% NB : buffer size and overlap are integer numbers (samples)
% data (in , out) are 1D arrays (vectors)
shift = buffer_size-overlap;    % nb of samples between 2 contiguous buffers  
[samples,channels] = size(data_in);
nb_of_loops = fix((samples-buffer_size)/shift +1);
    for k=1:nb_of_loops
        start_index = 1+(k-1)*shift;
        stop_index = min(start_index+ buffer_size-1,samples);
        x_index(k) = round((start_index+stop_index)/2);  
        % find peaks 
        for n = 1:channels
            tmp = data_in(start_index:stop_index,n);
            tmp_pos = tmp;
            tmp_pos(tmp_pos<0) = 0;
            tmp_neg = tmp;
            tmp_neg(tmp_neg>0) = 0;
            
            [locsp, pksp]= peakseek(tmp_pos,1,max(tmp_pos)/10); % pos peaks above 10% of max pos amplitude
            [locsn, pksn]= peakseek(-tmp_neg,1,max(-tmp_neg)/10); % neg peaks above 10% of max neg amplitude
            data_out_pos(k,n) = mean(pksp); % mean of pos peaks 
            data_out_neg(k,n) = mean(-pksn); % mean of neg peaks 
       
        end
        
    end
new_time = t(x_index); % time values are computed at the center of the buffer
end