How can I get the mean data from under the graphs?

Hi,

I am struggeling with getting the mean values from my data.

When plotted my data looks like this (this is only a small sample of the whole thing). I want to get the mean values of the signal from everything above 0. But not the noise you see around the 0 yline, the data is already filtered at the frequency it needs to be. I have tried it with manually looking up the range of the data which I needed and then:

data = [all_data(39:111) all_data(150:171)]

meandata = mean(data(data>0))

But this would take me years to manually look up every data point I need. I thought maybe a loop could help by telling MATLAB if the value is >0 for a period of 10 or more sample then sum/mean but I can't figure out how to do this?

Hope anyone can help me, thanks in advance!

Answer 1

[Star Strider]

To isolate the regions around the peaks, do something like this —

x = linspace(0, 150, 500);                                      % Create Data (Row)
y = @(v) sinc((x+v)-mean((x+v))/5);                             % Create Data (Rows)
v = (0:10:1500).';

x = x(:);                                                       % Convert To Column
y = sum(y(v)).';                                                % Convert To Column
y = y + rand(size(y))*0.2;                                      % Add Noise

figure
plot(x, y)
grid
xlabel('X')
ylabel('Y')
title('Original Data')

[pks,plocs] = findpeaks(y, 'MinPeakProminence',1);              % Desired Peak & Index

for k = 1:numel(pks)
    [vys,vlocs] = findpeaks(-y, 'MinPeakProminence',0.1);       % Valleys & Indices
    [vlc,ixv] = mink(abs(vlocs-plocs(k)),2);                    % Choose Two Closest Valleys To Desired Peak
    vlcs(:,k) = plocs(k) + [-1; 1].*vlc;                        % Calculate Indices  
end
% vlcs

figure
plot(x, y)
hold on
for k = 1:size(vlcs,2)
    idxr = vlcs(1,k) : vlcs(2,k);                               % Initial Index Range
    xv = x(idxr);
    yv = y(idxr);
    Apos(k,:) = trapz(x(yv>=0),yv(yv>=0));                      % Calculate Positive Peak Areas
    Mpos(k,:) = mean(yv(yv>=0));                                % Calculate Mean Value
    plot(x(idxr), y(idxr), '-r')                                % Plot Desired Peak Region
end
hold off
grid
xlabel('X')
ylabel('Y')
title('Original Data With Selected Regions Higlighted')

PeakAreas = table(x(plocs),Apos,Mpos, 'VariableNames',{'Peak X Location','Peak Area','Peak Mean'}) 

PeakAreas = 11×3 table
    Peak X Location    Peak Area    Peak Mean
    _______________    _________    _________

        12.625          1.7361       0.58128 
        25.251          1.9327        0.6023 
        37.575          1.7997       0.60411 
          50.2          1.7286       0.58274 
        62.826          1.6858       0.57318 
         75.15          1.7205       0.65916 
        87.475          1.7189       0.58363 
         100.1          1.6778       0.63264 
        112.42           1.753       0.53994 
        125.05          1.7449       0.59364 
        137.37          1.9823       0.45056 

The regions in red in the second plot are the segments used for the ‘Apos’ area and ‘Mpos’ mean calculations.

.

[Carmen Brouwers]

Thank you so much! This works :)

[Star Strider]

As always, my pleasure!

Answer 2

[Jonas]

what about a general threshold like 10 here?

mean(all_data(all_data>10))

[Carmen Brouwers]

yeah thought about that aswell, but I still need the points between 0-10 under the graphs if a want an accurate mean value right?

[Jonas]

but you wrote you want to ignore the noise around the 0 value?