As per my understanding you want your above code to run in a loop with "trialNum" ranging from 1 to the desired number of iterations. You can do the following modifications to your code:
modes = 2:4;
i = 3;
j = 7;
numIterations = 3; % Number of iterations
All_Data = [];
All_Labels = [];
for trialNum = 1:numIterations
data_i = [save_sx1_hat{i, trialNum}(modes, 1:500)];
data_j = [save_sx1_hat{j, trialNum}(modes, 1:500)];
Alldata_i = data_i';
Alldata_j = data_j';
All_Data = [All_Data; Alldata_i; Alldata_j];
All_Labels = [All_Labels; zeros(size(Alldata_i, 1), 1) + 1; zeros(size(Alldata_j, 1), 1) + 2];
end
Here, the "numIterations" variable takes the number of trials you want to run and then encloses your concatenation logic in a for loop for the same number of trials. I am not sure there is any other way to transpose the vector more efficiently than how it is done here.
