How to find out the location difference of nonzero entries in two sparse matrices?

Hello, All,

I have two sparse matries which have the same size. I know every nonzero entry in A must correspond to a nonzero entry in B, but some of nonzero entries in B which correspond to zero entries in A. For example, A=[0 1 0 2;3 0 0 4;0 5 6 0]; B=[0 1 0 2;3 7 0 4;8 5 6 0]. I want to find out the location of nonzero entries in B which correspind to zero entries in A. The answer to this example is: (2, 2) and (3, 1). Their correspinding entries are 7 and 8 in B.

Thanks.

Benson

Answer 1

[Stephan]

Hi,

this are not sparse matrices. However you find the locations given as row and column values and the corresponding values with:

A=[0 1 0 2;3 0 0 4;0 5 6 0];
B=[0 1 0 2;3 7 0 4;8 5 6 0];
[row, col] = find(A~=B)
vals = B(A~=B)

results in:

row =

     3
     2


col =

     1
     2


vals =

     8
     7

which is exactly the desired result.

Best regards

Stephan

[Benson Gou]

Hi, Stephan,

Thanks for your response. It is very good. How about the following example?

A=[0 1 0 2;3 0 0 4;0 5 6 0];B=[0 2.43 0 1.09;4.97 7.30 0 3.3;8.03 5.43 9.12 0];

The locations of nonzero entries in A and B are the same (but their values are different) except (2,2) and (3,1), their values are 7.30 and 8.03.

Our real case is like this. Thanks a lot for your great help.

Benson

[Stephan]

[row, col] = find(A==0 & B~=0)
vals = B(A==0 & B~=0)