How can I fill an empty matrix?

% Hi everyone I am trying to classfify a matrix M=21x2560 from an EEG. My aim is to divide the 21 channels into four different empty matrix. How can I fill temporal, occipital, parietal and frontal?. Then, I want to do the mean of every matrix created before. If I use "mean" only one value comes up but I want one value for each column of the matrix. Pleasee helpp meee! Thankyouuu!

load AD1.mat

M=X; [n,m]=size(M);

Temporal=[]; Frontal=[]; Parietal=[]; Occipital=[];

for i=1:n

    c=M(i:2520);  
    signal=c; 
    Fs=128; 
    band=[1 60]; 
    frequency_band=[1 4; 4 8; 8 13; 13 30; 30 40];  
    [ RelativePower ] = Compute_RP( signal,Fs,band,frequency_band);
    RP_c=RelativePower;     

    if i==8 || i==13 || i==3 || i==4 || i==9 ||i==14
        Temporal=[RP_c];

    end

    if i==1 || i==20 || i==2 || i==5 || i==6 ||i==7
       Frontal=[RP_c];
    end 

    if i==10 || i==11 || i==12 || i==15 || i==16 ||i==17
        Parietal=[RP_c];
    end 

    if i==18 || i==21 || i==19
        Occipital=[RP_c];
    end

end

RP_Temporal=Temporal; RP_Frontal=Frontal; RP_Parietal=Parietal; RP_Occipital=Occipital;

figure; bar(RP_Temporal); title('Relative Power Temporal lobe')

figure; bar(RP_Frontal); title ('Relative Power Frontal lobe')

figure; bar(RP_Parietal); title('Relative Power Parietal lobe')

figure; bar(RP_Occipital); title ('Relative Power Occipital lobe')

Answer 1

[Harish Ramachandran]

There are two points to note:

1. The piece of code below will overwrite Temporal's existing value at different values of i (with value of RP_c). Instead from your question, I gather that your requirement is to concatenate the value of Temporal with the i'th column of RP_c

    if i==8 || i==13 || i==3 || i==4 || i==9 ||i==14
        Temporal=[RP_c];
    end

In order to concatenate the values, do take a look at this link . Your code would look something like this:

    if i==8 || i==13 || i==3 || i==4 || i==9 ||i==14
        Temporal =[ Temporal    RP_c(:,i) ];
    end

2. With respect to your statement:

If I use "mean" only one value comes up but I want one value for each column of the matrix.

Once the values are filled up and you get a matrix with N columns, you can calculate the mean of each column using the mean function.

>> x = rand(50,5);
>> Y = mean(x)

Y =
    0.5413    0.5499    0.4918    0.4321    0.5170

Hope this helps.