Make a Truth Table

OK so I would like to clarify my question about Using mat2cell in a proper way.

To be honest what I want (in general) is to make something like a truth table for my final year project. I will take the data from MATLAB database as an example, since I cannot upload the data to this community due to the file size. The approachment is similar to what I have done in my project

So far what I have done is following (I divide the code in several parts so that all of you can understand what will be produced in each part):

tt = edfread('example.edf');
info = edfinfo("example.edf");
fs = info.NumSamples/seconds(info.DataRecordDuration);

for k=1:width(tt)
    for l=2:4
        recnum(l) = l;
        signum(k) = k;
        %t{k,l} = (0:info.NumSamples(signum(k))-1)/fs(signum(k));
        y{l,k} = tt.(signum(k)){recnum(l)};
    end
end
y = cell2mat(y);

waveletFunction = 'sym8';

for k = 1:width(tt)
    [C{k},L{k}] = wavedec(y(:,k),6,waveletFunction);
    s{k} = detcoef(C{k},L{k},6);
end

tc_d = linspace(0,20,74);
plot(tc_d,abs(s{1})); grid on; xticks(0:20); title(info.SignalLabels(1))

plot(tc_d,abs(s{2})); grid on; xticks(0:20); title(info.SignalLabels(2))

Then I need to create a truth table to indicate whether the high voltage occurs (in what seconds). For example, please refer to my drawing:

If you refer to the graph, the voltage that falls above the half will be considered as "high" (and the others considered as "low").

Actually, I can do it manually by looking at the graph and make the tables in Excel, but I hope there is an inbuilt tool in MATLAB to do the same things more efficiently, and in the other hand I can avoid trivial mistake like mistakenly read the graph.

Answer 1

[Steven Lord]

See the islocalmax function.

Answer 2

[chicken vector]

From the graph I see that the peaks marked are, as you say, over the half, but of the limits of the y axis, not of the maximum value. For this reason allow me to do something that I honestly don't like, that is to find the peaks by looking for valeus over 0.6 in the first graph, and over 0.06 in the second.

% Peaks thresholds:
threshold1 = 0.6;
threshold2 = 0.06;

% Determine respective timestamps:
times1 = unique(floor(tc_d(abs(s{1}) >= threshold1)));
times2 = unique(floor(tc_d(abs(s{2}) >= threshold2)));

% Initialise results cell:
truthCell = cell(20,3);

% Loop over time intervals:
for time = 1 : 20
    
    % Build time interval name:
    truthCell{time,1} = [num2str(time-1) '-' num2str(time) ' min'];
    
    % Get result of ECG1 for current time interval:
    if ismember(time,times1)
        truthCell{time,2} = 'True';
    else
        truthCell{time,2} = 'False';
    end
    
    % Get result of ECG2 for current time interval:
    if ismember(time,times2)
        truthCell{time,3} = 'True';
    else
        truthCell{time,3} = 'False';
    end
end

% Convert cell to table:
truthTable = cell2table(truthCell, ...
                        'VariableNames', ['Time', 'ECG1', 'ECG2']);

% Optional if you want to write data fo .xlsx:
writetable(truthTable,'TruthTable.xlsx');

[Edoardo]

@chicken vector if the peaks are the maximum value, is it possible, if yes, how?

[chicken vector]

Not sure i understand what you are asking.

If you want to find the peaks and the corresponding time instants, you can use islocalmax as suggested by @Steven Lord.

% Base data:
threshold = 0.6;
ECG1 = abs(s{1});

% Find idx that are local max and over threshold:
localMaxIdx = islocalmax(ECG1) & tc_d(abs(s{1}) >= threshold));

% Find time instants corresponding to peaks over threshold:
peakTimes = tc_d(localMaxIdx);