I'm not sure how to even ask this question without code... I am trying to create a 2d matrix of sums from grouping variables--I tried splitapply but that didnt seem to get me exactly what I want, which is a way to do what I show below:
n = 100;
X = randsample(1:9, n, true);
Y = randsample(15:20, n, true);
Z = rand(n, 1);
[ji, ~, jL] = grp2idx(X);
[di, ~, dL] = grp2idx(Y);
nJ = numel(jL);
nD = numel(dL);
Ts = zeros(nJ, nD);
for i=1:nJ
for j=1:nD
Ts(i,j) = sum(Z(ji==i&di==j));
end
end

댓글 수: 2

Walter Roberson
Walter Roberson 2019년 4월 29일
accumarray looks appropriate
Walter
Walter 2019년 4월 29일
yup! That's it! thank you!

댓글을 달려면 로그인하십시오.

답변 (0개)

카테고리

도움말 센터File Exchange에서 Tables에 대해 자세히 알아보기

질문:

2019년 4월 29일

댓글:

2019년 4월 29일

Community Treasure Hunt

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

Start Hunting!

Translated by