i have a matrix A of 1 column and another B of 4 columns. need to do some operations of A with each column of B and to get answer as C1,C2,C3,C4 using loops . Is it possible?

i have a matrix A of 1 column and another B of 4 columns. need to do some operations of A with each column of B and to get answer as C1,C2,C3,C4 using loops . Is it possible?

Answer 1

[Image Analyst]

Perhaps this, which will do a fit of each column of (the badly-named) B vs. A:

% A is the x values,
% Columns of B are the y values
for col = 1 : size(B, 2)
  thisColumn = B(:, col);
  % Fit a line through these data points in this column.
  coefficients = polyfit(A, thisColumn, 1);
  slopes(col) = coefficients(1);
  intercepts(col) = coefficients(2);
end

Answer 2

[Stephen23]

The best solution would be to not create numbered variables, but to simply keep all of the data together in one numeric array. Doing this makes MATLAB code much faster and neater. You can simply use indexing to access that parts of that array than you need to.

If you tell us what those "operations" are then we could tell you effective and efficient ways of doing those operations.

Whatever you do you should not create those variables dynamically. This is a poor programming practice that beginners seem to love. Read this to know why creating variables dynamically is a really bad idea:

http://www.mathworks.com/matlabcentral/answers/196952-how-to-create-new-variables-in-batches-with-strcat

[seema niran]

I want to 'polyfit(x,y). x is a matrix with one column. y is a 24 column matrix . I need the fits for all 24 columns.

[John D'Errico]

So? Then use a loop, and stuff the results into another matrix.

[seema niran]

i tried m= ones([4749 24]); then to multiply answer during each iteration to the columns of m. but it gives me a [4749 1 ] matrix. how to get a [4749 24] matrix as the final answer?