hello, I have a project and I'm extracting the features from emg signal, accelerometer signal and goniometer signal.

I extracting feature from 7 signals in a row for each signal, my question is how can I save the feature extracted for every signal in for loop because when I want to see the features, its showing the features of the 7th signal only

I will share accelerometer code and goniometer code.

accelerometer code:

clear all; close all;

for i = 1:7

eq = load ((['TBGACCPATIENT',num2str(i),'.txt']))';

Fs=400; % frequence d'echantillonage

len=length(eq);

mini=min(eq);

maxi=max(eq);

moy= mean(eq);

vari=var(eq);

stda= std(eq);

h = spectrum.welch; % Create a Welch spectral estimator.

Hpsd = psd(h,eq,'Fs',Fs); % Calculate the PSD

dd=Hpsd.Data;

Freq=Hpsd.Frequencies;

f=Freq;

Df=f(2)-f(1);

s=dd;

len1=length(dd);

M0=sum(s); %power of the signal

M1=sum(f.*s); % 2MPF

mpf=M1/M0; % mean frequency

cd1= sum((f-mpf).^3.*s);

cd2=sum((f-mpf).^2.*s);

Cd=cd1/sqrt(cd2^3);

% median of frequency Fmed

fm=0;

milieu=sum(dd)/2;

i=1;

fm=0;

while (( fm< milieu) & (i< length(eq)-1));

fm=fm+s(i);

i=i+1;

end

fmed=i*Df;

% kurtosis

ca1= sum((f-mpf).^4.*s);

ca2= sum((f-mpf).^2.*s);

Ca=ca1/(ca2^2);

% peak of frequency

pf = find(dd==max(dd));

pf=pf*Df;

% calcul of spectral entropy H

H=0;

ll=length(eq);

ss=length(s);

for i=1:length(s)

HH=length(H);

H=H+(s(i)*log(s(i)));

HHH=length(H);

end

%percentiles or fractiles fk

ss=cumsum(s);

frp=[];

for i=1:9

f1=find(ss>=(i/10)*sum(s));

frp=[frp;f1(1)*Df];

end

% % % relative energy by frequency band

Len=length(s) ;

Lseg=Len/10 ;

for i=0:9

w(i+1)=sum(s(Lseg*i+1:Lseg*(i+1)));

end

w1=w/M0;

[c,l]=wavedec(eq,5,'sym1');

ap5=appcoef(c,l,'sym1',5);

ap4=appcoef(c,l,'sym1',4);

ap3=appcoef(c,l,'sym1',3);

ap2=appcoef(c,l,'sym1',2);

ap1=appcoef(c,l,'sym1',1);

d5=detcoef(c,l,5);

d4=detcoef(c,l,4);

d3=detcoef(c,l,3);

d2=detcoef(c,l,2);

d1=detcoef(c,l,1);

format long g

V5=var(d5);

V4=var(d4);

V3=var(d3);

V2=var(d2);

V1=var(d1);

ftu=[mini;maxi;moy;vari;Cd;pf; stda ;mpf; fmed; Ca; H; V1; V2; V3; V4; V5;frp;w1'];

end

goniometer code:

clear all;close all;

for i = 1:7

x = load ((['TBGGONIOPATIENT',num2str(i),'.txt']))';

plot (x);

Fs = 4*1000; %Hz -- put in real data

t = (0 : length(x) - 1) / Fs;

%cftool (t,x)

[xData, yData] = prepareCurveData( t, x );

% Set up fittype and options.

ft = fittype( 'sin8' );

opts = fitoptions( 'Method', 'NonlinearLeastSquares' );

opts.Display = 'Off';

opts.Lower = [-Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0 -Inf];

opts.StartPoint = [260.175794786051 1.25676273770969 0.0629107881533813 113.29478585637 2.51352547541938 1.59529193400343 22.4164334568717 5.02705095083875 1.45853359383986 12.9543596116753 20.108203803355 2.49920460557347 8.23752641768903 7.54057642625813 1.2050686662606 7.53966278924057 17.5946783279356 0.399135169386131 5.17482134679428 12.5676273770969 1.54711062715199 5.17211566529762 15.0811528525163 1.05486183464122];

% Fit model to data.

[fitresult, gof] = fit( xData, yData, ft, opts );

% Plot fit with data.

figure( 'Name', 'untitled fit i' );

h = plot( fitresult, xData, yData );

legend( h, 'x vs. t', 'untitled fit i', 'Location', 'NorthEast', 'Interpreter', 'none' );

% Label axes

xlabel( 't', 'Interpreter', 'none' );

ylabel( 'x', 'Interpreter', 'none' );

%pause;

end

[dpb]

You've not saved space for multiple passes through the loop -- when you write a variable on the LHS w/o any subscripting, MATLAB replaces the variable in its entirety with the content of the RHS. That's why you only have the last values of all variables.

Depending upon the variable, the output size will be either a vector of 7 results or a 2D array if the result is a vector or a 3D array by plane if the variable is 2D array. Or use cell arrays to contain the results of whatever the RHS values are...

[mohammad mehio]

what I meant is in the for loop when it finishing from every signal I want it to save the features for each signal

[dpb]

Yes, I just described the changes you need to make to do so...you have to create arrays of some sort inside the loop for the computed results of each step you need to save.

These are the most basic of concepts, if they're unfamiliar to you, I suggest going to <matlab-onramp> tutorial and spend the time to go through it.

There are rudiments of examples in the doc for the for...end statement and in all the examples of cell arrays and so on, as well.