Rearrange elements of matrix based on an index matrix

I have a 5x3 matrix and I want to rearrange each row according to the correponding row of a 5x3 index matrix

x=randn(5,3)

x = 5x3

-0.2616 0.3522 -0.4451 -1.1699 0.8921 1.7676 1.4354 1.1977 -0.3197 -1.2002 0.9392 0.7452 -0.8374 -1.3827 0.2679

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

z=randn(5,3)

z = 5x3

-0.1687 0.4662 1.3644 0.1290 -3.2728 -0.8944 -0.7820 0.2506 -1.0295 0.0976 -0.0797 -0.2671 0.2607 -0.0966 -0.1679

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

[~,I]=sort(x,2)

I = 5x3

3 1 2 1 2 3 3 2 1 1 3 2 2 1 3

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

Now I want to sort rows of z using the index matrix I. But using the following does not work. For example, I want the first row of zz to be sorted according to the first row of x, which should result in zz(1,:)= [1.3644, -0.1687, 0.4662].

zz=z(I)   

zz = 5x3

-0.7820 -0.1687 0.1290 -0.1687 0.1290 -0.7820 -0.7820 0.1290 -0.1687 -0.1687 -0.7820 0.1290 0.1290 -0.1687 -0.7820

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

Answer 1

[Stephen23]

Yes, it is awkward.

x=randn(5,3)

x = 5x3

0.9687 -0.7929 2.0516 -0.1236 0.2917 -0.9487 -1.4717 0.0275 1.1722 0.2800 1.5357 0.2596 -0.9693 -0.1854 0.3145

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

z=randn(5,3)

z = 5x3

-1.6696 0.6182 -1.4586 0.1528 1.6729 1.4683 0.3576 -1.0089 0.8616 2.1993 0.6213 -1.2327 -0.3246 -0.8133 1.0463

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

[~,I] = sort(x,2)

I = 5x3

2 1 3 3 1 2 1 2 3 3 1 2 1 2 3

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

Perhaps

S = size(I);
[R,~] = ndgrid(1:S(1),1:S(2));
J = sub2ind(S,R,I);
zz = z(J)

zz = 5x3

0.6182 -1.6696 -1.4586 1.4683 0.1528 1.6729 0.3576 -1.0089 0.8616 -1.2327 2.1993 0.6213 -0.3246 -0.8133 1.0463

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

Or

zz = z;
for k = 1:size(I,1)
    zz(k,:) = zz(k,I(k,:));
end
zz

zz = 5x3

0.6182 -1.6696 -1.4586 1.4683 0.1528 1.6729 0.3576 -1.0089 0.8616 -1.2327 2.1993 0.6213 -0.3246 -0.8133 1.0463

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

Or

zz = cell2mat(cellfun(@(v,x)v(x),num2cell(z,2),num2cell(I,2),'uni',0))

zz = 5x3

0.6182 -1.6696 -1.4586 1.4683 0.1528 1.6729 0.3576 -1.0089 0.8616 -1.2327 2.1993 0.6213 -0.3246 -0.8133 1.0463

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

[Hossein Kazemi]

Thanks. The first solution appears to be the fastest (my matrix is 3400x67).