How to normalize a matrix in such a wat that every row sum of X(:,:,i) should be 1 except for ith row

I have created a matrix X = rand([6,3,6]); i want to normalize this matrix in such a way that every row sum of X(:,:,i) should be 1 except for ith row.

Inside an while(1) loop ,i have

i=randi(n);

Say for example when i=2 that particular row sum should not be equal to one. same goes for other node.How it can be done?

[Jan]

I cannot see the connection between the explanations "row sum of X(:,:,i) should be 1 except for ith row" and "while(1), i=randi(n)". What is "other node"?

[chan]

in an infinite loop this i (node) value will keep on changing. based on that value of i , the matrix will be normalize except this particular i (node).

if i=2 only that particular row will not be normalize while others row wiil be equal to 1. it will be same for other i's based on the random value generated.

[Jan]

@IRANI ACHARJAMAYUM: What yre your inputs? How does X = rand([6,3,6]) and an infinite loop with i=randi(n) match? What is the wanted output?

It is still not clear, what you want to achieve. I've posted some code for normalizing the rows of a 3D array except for certain indices. Does it work? If not, what have to be changed?

Answer 1

[Cris LaPierre]

Why not just create a row vector of random numbers, and insert it into a matrix of ones?

• X = ones(sz1,...,szN)

Answer 2

[Jan]

X = rand(6, 3, 6);
N = sum(X, 2);
N(sub2ind(size(N), 1:6, ones(1,6), 1:6)) = 1;
Y = X ./ N;

Now sum(Y(i, :, j)) is 1 is i~=j.

sum(Y(1, :, 2))

ans = 1

sum(Y(3, :, 3))

ans = 1.5543