How to select columns based on a known repeated pattern.

1 view (last 30 days)
I have data that came from excel and I want to grab the data without all the comments. I have imported the data into matlab but now I am having trouble selecting it cleanly. If the data is in a variable called raw I do the following.
data = raw(:,28:22:end);
data = [data, raw(:,29:22:end)];
data = [data, raw(:,30:22:end)];
data = [data, raw(:,31:22:end)];
data = [data, rwa(:,32:22:end)];
The problem with this is I have to reorder everything afterwards to get the columns back in the right order. I want to select all the columns like:
data = raw(:,[28:32]:22:end);
or something similiar. I figure there must be some ultra fancy vector way of doing this in 1 shot.

Accepted Answer

Andrei Bobrov
Andrei Bobrov on 5 Dec 2013
data = raw(:,bsxfun(@plus,29:32,22*(0:floor((size(raw,2)-32)/22))'));

More Answers (2)

the cyclist
the cyclist on 5 Dec 2013
One marginally better possibility:
data = raw(:, [28:22:end 29:22:end 30:22:end 31:22:end 32:22:end]);

Kelly Kearney
Kelly Kearney on 5 Dec 2013
Not quite one shot, but a little less typing:
n = size(raw, 2);
idx = bsxfun(@plus, (1:22:n)', 28:32);
idx = idx(idx <= n);
data = raw(:,idx);
  1 Comment
Matt
Matt on 5 Dec 2013
2 minor things but this is good.
idx = bsxfun(@plus, (1:22:n)', 27:31);
idx = sort(idx(idx <= n)); % sorts the columns in the right order

Sign in to comment.

Community Treasure Hunt

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

Start Hunting!