Creating row-concatenated matrix in Matlab

Question

How do I create a concatenated matrix in Matlab which would give the result below?

[0.01 0.01 error1]
[0.01 0.03 error2]
...
[30 30 error64]

So far, what I have tried is below.

C =  [0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30];
sigma =  [0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30];

m = zeros(64,3);

for ci = C
for si = sigma
    train = svmTrain(X, y, ci, @(x1, x2) gaussianKernel(x1, x2, si));
    pred = svmPredict(train, X);
    error = mean(double(pred ~= y))    
    m = [m ; [C,sigma,error]];    
end
end

I expect a 64 X 3 column matrix.

David Maust · Accepted Answer · 2015-11-29 02:36:49Z

3

It seems like the code you provided basically does that. You should be starting off with an empty matrix m.

m = []; % Initialize empty result matrix m 

for ci = C
    for si = sigma
        error = % Calculate error here
        m = [m ; [C,sigma,error]];  % concatenate new row onto m.
    end
end

answered Nov 29, 2015 at 2:36

David Maust

8,3003 gold badges34 silver badges36 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

Krishna Kalyan Over a year ago

I am just getting confused i guess.

Krishna Kalyan Over a year ago

I was expecting a 64 X 3 matrix

David Maust Over a year ago

If the length(C) * length(sigma) is 64 you should get a 64 x 3 matrix at the end.

Matthew Gunn Over a year ago

I'll make a small performance comment. This doesn't matter for a 64x3 matrix, but if you're matrix is HUGE (i.e. tens to hundreds of thousands of rows), the concatenation step m = [m; blah]; gets quite costly: you're constantly rewriting the whole m matrix. An alternative is something like m = zeros(n_rows, n_cols) then in each step of loop, m(k_start:k_end, :) = [C, sigma, error]. This would copy the right data into the right spot each loop (compared to rearranging the entire m matrix each loop).

Collectives™ on Stack Overflow

Creating row-concatenated matrix in Matlab

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related