MATLAB Answers

Splitting up a matrix into X sizes/ based on specified values in a matrix

6 views (last 30 days)
mark on 28 Dec 2013
Commented: mark on 28 Dec 2013
If I have a matrix of size 2x100 of time vs position values (or whatever). I we say time goes from 0s to 2s then I would like to write a script that says:
" Find where X(1,:) = 1 and split it, and X(2,:), such that I have two new matrices A and B which correspond to the values from 0s to 1s and 1.1s to 2s"
I tried using the find feature but that didn't work. Any ideas?
Essentially, if I have 10s of run time then I'd like to be able to separate those 10s into 10 separate matrices.
Cheers, Mark.
I should add that the time scale may not be equal as the time sampling is automated based on when an event occurs. This is to reduce the computational time substantially and maintain a high enough resolution at an event. So, the important aspect is to be able to specify the points on the matrix that I'd like to separate out!

Accepted Answer

Wayne King
Wayne King on 28 Dec 2013
Edited: Wayne King on 28 Dec 2013
Are the time instants just integers? If not, that may be why find() does not work depending on how you are using it.
X = ones(2,100);
X(1,:) = 0:99;
idx1 = find(X(1,:)<=50,1,'last');
idx2 = find(X(1,:)>50,1,'first');
A = X(:,1:idx1);
B = X(:,idx2:end);

More Answers (1)

Baalzamon on 28 Dec 2013
Logical indexing is one way. Both Doug and Loren have a mini tutorial on it. To get say all elements of X(1,:), that are say 0-50, replace the ':' with a logical argument. e.g. A = X(1, X(1,:) > 50 ); B = X(1, X(1,:) < 51); for both rows A = X(:, X(1,:) > 50); B = X(:, X(1,:) < 51);
The 'X(1,:) > 50' bit is looking up all elements in the first row of X that is GT 50, and then using that to create the new array.
As you can guess it is possible to use multiple logical arguments with '&&' etc.
  1 Comment
mark on 28 Dec 2013
Thanks guys, the logical stuff works.
if true
X=[X1' X2'];
Ax = find(X(:,1)<=50);
A=[A1' A2'];
Bx = find(X(:,1)>50);
B=[B1' B2'];

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!