If I have a matrix of 100 rows by 5 columns, how can I make it a 1 row x 500 column matrix, where each row (1x5) is placed one after the other to make a 1x500 matrix?

[r,c] = size(data);%(100 rows by 5 columns)

datanew = zeros(1,500)%

for i = 1:r

startcol = (1+(i-1)*5);

endcol = (5*i);

datanew(1,data(1,startcol:endcol)); %I get an error "Subscript indices must either be real positive integers or logicals." But data(1,startcol:endcol) does contain the correct 1x5 data, therefore, uncertain why the error.

end

Answer 1

[James Tursa]

datanew = reshape(data',1,[]);

[Nick Counts]

Good catch with the transpose - re-reading Jean's code, it looks like he's trying to go row-by-row, rather than column-wise.

Answer 2

[Mischa Kim]

Use the reshape command.

Answer 3

[Nick Counts]

You can use reshape:

A = randi(10,100,5)
B = reshape(A,1,500)

• A will be a 100x5 matrix
• B will be a 500x1 matrix

As to your particular error, I am not certain. Your code doesn't work as posted because data isn't defined. So I can't say what's going on. If you want to post some additional code, we can take a look at what's going on and help you find the issue.

[Nick Counts]

If you were trying to do this inside a for-loop, you can use horizcat or vertcat. You could also use indexing tricks, but your calculation of startcol and endcol seems to be broken. Easier to go row by row.

data = randi(10,100,5)
newData = []
for i = 1:length(data)
    newData = vertcat(newData, data(i,:)');
end

I believe James's reshape is what you're really looking for

Answer 4

[Eric]

I appreciate all the suggestions.