Multiply the ith column of a matrix by element at position i of a vector

[Misha Koshelev]

Hi, I would like to do something like:

for i=1:cols(M)
M(:,i) = M(:,i)*v(i)
end

Is there a fast/vectorized way to do this? Is this faster:

M = M*repmat(v,rows(M),1);

Thank you
Misha

[Greg Heath]

That does not work.

If

A = R*M* C

where R and C are diagonal, then

a(i,j) = r(i)*m(i,j)*c(j)

i.e., r(i) multiplies each element in the ith row and c(j)
multiplies each element in the jth column. So, if r and c
are vectors,

A = diag(r)*M*diag(c)

Hope this helps.

Greg

[Matt Fig]

% Data
M = magic(5)
v = [6 7 8 9 10]

% Engine
D1 = bsxfun(@times,M,v);

% Compare to other method
for i=1:size(M,2)
D2(:,i) = M(:,i)*v(i);
end

all(D1(:)==D2(:))

[Jos]

"Misha Koshelev" <mk14...@bcm.edu> wrote in message <gu6qkh$3ut$1...@fred.mathworks.com>...

In general, use BSXFUN; if M is square (NxN) you can also use

M * diag(v)

hth
Jos

[Bruno Luong]

"Jos " <#10...@fileexchange.com> wrote in message <gu8f0p$ci1$1...@fred.mathworks.com>...

> if M is square (NxN) you can also use
>
> M * diag(v)
>

This works even for M is (M x N) rectangle, the only requirement is v has N elements.

Bruno

[Jos]

"Bruno Luong" <b.l...@fogale.findmycountry> wrote in message <gu8g4t$pg0$1...@fred.mathworks.com>...

Yep! Thanks for correcting me, Bruno.