Finding sequence in matrix

I have a matrix and I want find which row contains some sequence. For example:

• A= [8 5 2 3 -1 0 4 -2 5 0 0 0
• | 5 3 4 -2 1 6 -1 -3 0 0 0 0|
• | -1 3 5 2 0 4 2 0 0 0 0 0];|
• and sequence is:
• seq=[4 -2 1];
• Result shoulde be:
• Result=2
• I tried to use xcorr function like this:
• [m n]=size(A);
• for i=1:m
• fi(i)=max((xcorr(A(i,:),seq)));
• end
• [no index]=max(fi);
• result=index
• But it doesnt work in some cases. Thanks for any help.

Answer 1

[per isakson]

If

• speed matters
• the elements of the matrix are integers

This old trick does it

    A= [8 5 2 3 -1 0 4 -2 5 0 0 0
        5 3 4 -2 1 6 -1 -3 0 0 0 0
       -1 3 5 2 0 4 2 0 0 0 0 0];

    seq=[4 -2 1];
    ix  = strfind( reshape( transpose( A ), 1, [] ), seq );
    ixr = ceil( ix / size( A, 2 ) );

Recipe

• convert A "row-wise" to a row
• use strfind to find the start of seq: ix
• find out to which row ix belongs

Answer 2

[Image Analyst]

Try this:

A= [8 5 2 3 -1 0 4 -2 5 0 0 0
    5 3 4 -2 1 6 -1 -3 0 0 0 0
   -1 3 5 2 0 4 2 0 0 0 0 0]
template = [4 -2 1]
% Here's how to do it.
out = normxcorr2(template, A)
[row, column] = find(out == 1)

[Image Analyst]

Forgot to mention that it requires the Image Processing Toolbox.

[Viktor]

thanks, but it doesnt work correctly.

[Image Analyst]

Funny -- I just copied and pasted and it ran perfectly. Why do you say it doesn't work?