How to calculate FWHM and rise time?

Hello, I have a lot of signals (.txt) with the values of EDA in the six columm of the text files. I need to create a script that opens all the text files and then calculate the amplitude, time of rise and FWHM, like in the photo. The amplitude is ok, but the time of rise and FWHM is wrong.

I hope you can help me, thank you very much!!!

clear
close all;
%%
pasta = '/Users/sofiasantos/Desktop/sinais'; %directory
dados=dir(fullfile(pasta,'*.txt'));
for k=1:numel(dados)
  signals=dados(k).name
  data_org = importdata(signals,'\t',7); %text file with 7 columms
EDA=data_org.data; 
EDA= EDA(:,6);
EDAuS=((EDA/2^10)*3)/0.12; %convert to uS
%%
N= length(EDAuS); 
fs = 1000; 
T=N/fs;
t = (0:N-1)/fs; 
f = (0:N-1)/T;

%% FILTER
[b,a] = butter(6,0.04,'low');
EDAuS_filer = filtfilt(b,a,EDAuS);  %filtering signal EDA
 
%% AMPLITUDE (good)
minimo= min(EDAuS_filer);
maximo= max(EDAuS_filer);
tempo=find(EDAuS==EDAuS);
amp=[];
amp=abs(maximo-minimo);
disp('amplitude=');disp(amp); %amplitude da SCR

%% RISE TIME (wrong)
index_a=find(t==minimo);
disp(index_a);
index_b=find(t==maximo);
rise_time=abs(index_b-index_a);
disp('rise time='); disp(rise_time);

%% LARGURA A MEIA ALTURA (FWHM) (wrong)
halfMax = (minimo + maximo) / 2; 
disp(meia_altura);
index1 = find(EDAuS_filer >= halfMax, 1, 'first');
disp(index1);
index2 = find(EDAuS_filer >= halfMax, 1, 'last');
disp(index2);
FWHM = index2-index1 + 1;
disp('FWHM='); disp(FWHM);
end

[Star Strider]

The image you posted does not at all look like the waveforms in your file.

I managed to figure out how to read the file, however I have no idea what to do with it.

[Sofia Santos]

Thank you for trying, I changed the txt file for another that looks like the imagem I posted.

[Star Strider]

The new file does not look at all like the image you posted, any more than the previous one did. It looks like some sort of digital pulse train or digital communications signal rather than the transient in the image you posted. (Fortunately, it has the same strange format.) The pulses also do not have the same phases and on-off lengths, so they appear apparently randomly.

Answer 1

[Star Strider]

The Signal Processing Toolbox has everything you need to analyse your signals.

After reading and reshaping your file so it is possible to work with it:

filename1 = 'opensignals_201505292090_2018-02-25_18-09-25.txt';
filename2 = 'lixa3.1.txt';
txt = fileread(filename1);
strt = strfind(txt,'# EndOfHeader') + numel('# EndOfHeader');
Dc = textscan(txt(strt:end), '%f%f%f%f%f%f%f', 'CollectOutput',1);
D = cell2mat(Dc); 
[~,ia,ix] = unique(D(:,1));
fullend = find(D(:,1) == 15,1,'last');
Dr = reshape(D(1:fullend,[1 6 7])', 3, 16, []);
Dr = permute(Dr,[2,1,3]);

you can get all the informaiton you want from these functions (pulsewidth, risetime, and others):

for k = 1:2
    RT{k,:} = risetime(Dr(:,2,k), Dr(:,1,k));
    [W,INITCROSS,FINALCROSS,MIDLEV] = pulsewidth(Dr(:,2,k), Dr(:,1,k));
    PW{k,:} = {W,INITCROSS,FINALCROSS,MIDLEV};
end

Add other function calls as necessary to get the information you want. Note how I saved the pulsewidth output to be certain that all of them are saved, even those with empty fields. This allows you to make appropriate assessments of each pulse. (The permute call is not absolutely required. It just makes the matrices a bit more intuitive to work with.)

I tested this code with both files, and it works with both of them.

[Sofia Santos]

Thank you very much for your answer!!

It was such a big help! :)

[Star Strider]

As always, my pleasure!