How to plot outliers from mean

I have a code that looks like this:

figure
for ii = 1:12
    y =Data.StrideTimeIntervals_15minTrial.PD(:,ii);
    subplot (3,4,ii),plot (y, 'r')
    title (sprintf('PD %d',ii))
    xlabel ('ISI #')
    ylabel ('ISI (s)')
    axis ([0,500,0.8,1.2])
    disp (sprintf('PD %02d Done',ii))
end

How would I plot the outliers in red circles on all 12 graphs on the subplots. Here is what the subplots look like.

Answer 1

[the cyclist]

There are three steps:

1. Create an algorithm to define the outliers
2. Identify the points that meet that definition
3. Plot those points

Here is an example. You'll need to adapt to your case, for example, by doing this calculation inside of each for loop, and using your own definition of what constitutes an outlier.

Also, I would recommend against using red circles to identify outliers, since your plot is a red line.

% Make up some data
N = 1000;
y = randn(N,1);
y_mean = mean(y);
y_std = std(y);

% Define "outlier" as greater than 2 standard deviations away from the
% mean.
% Find the index to those points
outlierIndex = find(abs(y - y_mean) > 2*y_std);

figure
hold on

% Plot the data
plot(y,'.')

% Plot a red circle around the outliers
plot(outlierIndex,y(outlierIndex),'ro')

[timetry2]

I ended with a code that looks like this:

figure

for ii = 1:12

y =Data.StrideTimeIntervals_15minTrial.PD(:,ii);

subplot (3,4,ii),plot (y, 'r')

y_mean = mean(y);

y_std = std(y);

title (sprintf('PD %d',ii))

xlabel ('ISI #')

ylabel ('ISI (s)')

axis ([0,500,0.8,1.2])

disp (sprintf('PD %02d Done',ii))

outlierIndex = find(abs(y - y_mean) > 2*y_std);

hold on

plot(y,'.')

plot(outlierIndex,y(outlierIndex),'ro')

end

From here how would I clean the data using a moving average window filter to create a window size of 5 as a new figure? Thank you so much for the help!

[the cyclist]

I don't understand this new question. Since your original question seems to be answered, can I suggest you accept my answer, and start a new one, with a more complete description of what you are trying to do?

[timetry2]

Yes, thank you for your help