sorting according to another vector

Hi,

I have a vector A that doesn't follow any order. I want to sort Matrix B that has the first column of values similar to those of A but in different order. Is it possible to sort B according to A using a single function or should I write many codes to do so?

Answer 1

[Walter Roberson]

Sorry you will have to use multiple calls:

[a_sorted, a_order] = sort(A);
newB = B(a_order,:);

[Jan]

@Danielle: Ok, then I have misunderstood your question. The information that B has "value similar to those of A" is meaningless then - correct?

Answer 2

[Jan]

I'm not sure, if I understand the question correctly. Do you want to bring B in the same order as the unsorted A? Then either sort both vectors and mix their sorting indicies:

[As, Ai] = sort(A);
[Bs, Bi] = sort(B(:, 1));
ABi(Ai) = Bi;

Now ABi is the sorting index to bring B to the order of A:

isequal(A, B(ABi, 1))  % ==> 1

Or let ISMEMBER do this for you:

[dummy, order] = ismember(A, B(:, 1));
isequal(A, B(order, 1))  % ==> 1

Answer 3

[Pramit Biswas]

 function B = sortBlikeA(A,B)
   [~, Ao] = sort(A);
   Bs=sort(B);
   B(Ao)=Bs;
 end

[Marco Schnizer]

Nice Code, thank you!

Answer 4

[inzamam shoukat]

Hi every one......i have matrix lets suppose [3 2;1 9]

what i want to do is sort out the only first column in ascending order but problem is how to do it so that corresponding values which is in first row 2 for 3 in second row 9 for 1 change their positions according to the position of values of first column being sorted i.e [1 9; 3 2]...

LIMITATION : sorting operation never applied to second column....

[madhan ravi]

Please ask a separate question

[Steven Lord]

Use sortrows.

Answer 5

[Marco Bertola]

I think the question (which I had too) was:

I have a reference vector [1.2 3.3 4.4] eg

and another ``similar'' vector [3.2 1.1 4.4]

Problem; find the optimal sorting of vector 2 so that it is ''closest'' (L^2 norm eg) to the reference vector.

Say X is your reference vector of size (1,n) and Xnew is the new vector of the same size

then what you want is accomplished like so:

dist = abs(ones(n,1)*X - Xnew.' ones(1,n));

[t1,J]= min(dist);

%now J has the optimal permutation%

Xnew=Xnew[J]

[Walter Roberson]

If you are not permitted to duplicate any elements, and the entries are all finite then

[~, refidx] = sort(reference_vector);
sorted_similar = sort(similar_vector);
new_similar = sorted_similar(refidx);