find sequence in a matrix

3 次查看(过去 30 天)
shamal
shamal 2025-5-11
编辑: Torsten 2025-5-13
Hi, it possibile to velocize it and avoid loop? (I will be a matrix .,not a single array)
(if is possible to use vectorization)
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0]
I = 1×18
0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
k=3;
for i=1:numel(I)
if i>=k+1
if I(i-1) && ~I(i) %%i want to find the first "1 0"
saved=i;
break;
end
end
end
saved
saved = 7

采纳的回答

Torsten
Torsten 2025-5-11
编辑:Torsten 2025-5-11
I = [0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0];
k = 3;
saved = k + find(diff(I(k:end))==-1,1)
saved = 7
If I is a matrix, you will have to tell us your search direction and the role of k.
  4 个评论
Image Analyst
Image Analyst 2025-5-13
Where is the search pattern defined in that code? Where does it say it's looking for the pattern [1, 0]?
Also can it find a 2-D pattern in a 2-D matrix. Like for the given matrix
I = [0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0;0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1;1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0]
I = 3×18
0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1 1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
Can it find the pattern
1 0 1
1 0 0
?
Torsten
Torsten 2025-5-13
编辑:Torsten 2025-5-13
Where is the search pattern defined in that code? Where does it say it's looking for the pattern [1, 0]?
Let x be a vector consisting only of 0 and 1 and assume that the pattern [1 0] somewhere appears in the vector x (if it's not present in x, the method will give a wrong answer).
diff(x) can give values -1, 0 and 1 - so -1 is the minimum value you can get when you compute diff(x). On the other hand, getting -1 for diff(x) in position i means x(i+1) - x(i) = -1 which is only possible if x(i+1) = 0 and x(i) = 1. Thus a position i where diff(x) attains its minimum (namely -1) is a position where the pattern [1 0] (x(i) = 1, x(i+1) = 0) appears in the vector x.
Can it find the pattern
1 0 1
1 0 0
?
diff([0 0 0],2)
ans = 0
diff([1 1 1],2)
ans = 0
diff([1 0 0],2)
ans = 1
diff([0 1 0],2)
ans = -2
diff([0 0 1],2)
ans = 1
diff([1 1 0],2)
ans = -1
diff([0 1 1],2)
ans = -1
diff([1 0 1],2)
ans = 2
Only pattern [0 1 0] gives answer -2 and pattern [1 0 1] gives answer 2. Thus these two patterns could be found by a similar method (min and max) as the one used for pattern [1 0].

请先登录,再进行评论。

更多回答(4 个)

Image Analyst
Image Analyst 2025-5-11
The first [1, 0] shows up at index 2, not 7. It also appears at index 6 and others.
Probably the simplest way (a single line of code) is to use strfind. (Yes it works with numbers as well as character strings).
% Set up parameters:
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0];
searchPattern = [1, 0];
% Now find all locations where the vector has values [1, 0]
indexes = strfind(I, searchPattern)
indexes = 1×5
2 6 12 15 17
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
If you want the first location, just take indexes(1).
If you have a 2-D array, you can use normxcorr2. See attached demo.
  3 个评论
Image Analyst
Image Analyst 2025-5-12
编辑:Image Analyst 2025-5-12
You didn't explicitly say that. And by looking at your code, it only starts looking at values for i>=k+1, or 4, not 3. And then it looks like it finds the trailing 0, not the leading 1, so it finds index 7 rather than 6 which is where the [1,0] pattern starts. So it was kind of confusing to me. Not sure what you want exactly.
If you want to find the index of the trailing 0 after index 3, you can do
% Set up parameters:
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0];
k=3;
searchPattern = [1, 0];
% Now find locations where the array has values [1, 0]
indexes = strfind(I(k:end), searchPattern) + k
indexes = 1×4
7 13 16 18
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
If you want to find the index of the leading 1 after index 3, you can do
% Set up parameters:
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0];
k=3;
searchPattern = [1, 0];
% Now find locations where the array has values [1, 0]
indexes = strfind(I(k:end), searchPattern) + k - 1
indexes = 1×4
6 12 15 17
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
Either way, it's still simpler than the other solutions.
Matt J
Matt J 2025-5-12
Simpler, but not as efficient. strfind() will not operate row-wise on a matrix. You will have to loop. Additionally, it will search the entire row, unlike the other solutions which will stop at the first occurence.

请先登录,再进行评论。


Matt J
Matt J 2025-5-11
编辑:Matt J 2025-5-13
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0;0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1;1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0]
I = 3×18
0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1 1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
k=3;
[minval,loc]=min(diff(I(:,k:end),1,2),[],2);
loc(minval>-1)=nan;
loc=loc+(k-1)
loc = 3×1
6 4 4
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>

Matt J
Matt J 2025-5-12
编辑:Matt J 2025-5-13
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0;0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1;1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0];
k=3;
Ik=string(char(I(:,k:end)+'0'));
loc=strlength(extractBefore( Ik , '10'))+k
loc = 3×1
6 4 4
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>

Matt J
Matt J 2025-5-13
I=[0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0;
0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1;
1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0]
I = 3×18
0 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1 1 1 0 1 0 1 0 0 0 1 0 1 0 0 1 1 1 0
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
k=3;
C=conv2(I,[0,1],'valid')./conv2(I,[1,1],'valid')==1
C = 3x18 logical array
0 1 0 0 0 1 0 0 0 0 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 0 1 0 1 0 1 0 1 0 0 0 1 0 1 0 0 0 0 1 0
C(:,1:k-1)=0;
[~,loc]=max(C,[],2)
loc = 3×1
6 4 4
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>

类别

Help CenterFile Exchange 中查找有关 Geographic Plots 的更多信息

产品


版本

R2024b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by