A complicated matrix manipulation
1 view (last 30 days)
Show older comments
So I have a matrix, let's say
A = [1 2;
1 0;
3 2;
0 1;
2 1;
1 2];
[r,c] = size(A);
In general, what I would like to do is:
for i = 1:r-1
x = A(6,1)*A(i,1) + A(6,2)*A(i,2);
end
However, here is the issue: if A(i,1) and/or A(i,2) is zero, and let's assume just A(i,1) is 0 and A(i,2) is non-zero, then, instead I would like to do:
x = A(6,1)*A(i-1,1) + A(6,2)*A(i,2);
And in case, even A(i-1,1) is also zero, then, I would like to do:
x = A(6,1)*A(i-2,1) + A(6,2)*A(i,2);
and so on and so forth (same for A(i,2)). Note, A(1,1), A(1,2), A(6,1), A(6,2) are always non-zeros. I am trying to get my head around it, but so far no luck, although it seems it should have a simple and efficient way.
Your help is highly appreciated. Please let me know if anything seems confusing.
Thanks in advance!
2 Comments
Stephen23
on 19 Jun 2020
Your example matrix only has five rows, so none of your example code accessing its (non-existent) sixth row will work.
Accepted Answer
Stephen23
on 19 Jun 2020
Edited: Stephen23
on 19 Jun 2020
Try this reasonably "simple and efficient way":
r = size(A,1);
x = nan(r,1);
for k = 1:r
r1 = find(A(1:k,1),1,'last');
r2 = find(A(1:k,2),1,'last');
x(k) = A(6,1).*A(r1,1) + A(6,2).*A(r2,2);
end
3 Comments
Stephen23
on 19 Jun 2020
Edited: Stephen23
on 19 Jun 2020
'...generalizing it for "n" number of columns of A'
C = cumsum(A(:)~=0,1); % requires that first and last rows do not contain zeros!
B = nonzeros(A);
B = reshape(B(C),size(A));
x = sum(A(6,:).*B,2)
Personally I would just use a loop, as it makes the code intent much clearer.
More Answers (0)
See Also
Categories
Find more on Number Theory in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!