# How to find the position of the elements of a matrix?

361 views (last 30 days)
giorgos kazas on 3 Feb 2020
Answered: giorgos kazas on 4 Feb 2020
I have a 3x3 matrix: J=[0 3 4;3 0 8;4 8 0]
and I want to know in which position the elements exist. For example that 8 exists in second row and third column (2,3) and (3,2) too.
After that I want a new matrix with the value (i.e. 8) in one column and their ''coordinates'' (i.e. 2,3) in a column side by side.
Thanks

Steven Lord on 3 Feb 2020
Paul Shoemaker's answer will give you the indices of the selected elements in the array, but depending on what you want to do that may not be necessary. For example if you want to get the elements of another array (the same size as J) in the same location as the 8's in J you can do this using logical indexing.
J=[0 3 4;3 0 8;4 8 0]
M = magic(3)
X = M(J == 8)
The vector X contains 9 and 7, the elements from M in locations corresponding to 8's in J. See this documentation page for more information.
Paul Shoemaker on 3 Feb 2020
Yes, Steven Lord makes a good point: usage may drive you a different way. This is why it's helpful for those posting questions to explain what their end objective is, not just how to perform some specific mechanical action.
Similar to Steven's example, you can also change the original "J" variable where a certain condition is met:
J(J==8) = NaN; % efficient way to change all occurrences of "8" in the original matrix to be NaN's
Paul S

Paul Shoemaker on 3 Feb 2020
Check out the ind2sub Matlab function.
Paul Shoemaker
MatlabInvesting.com

Show 1 older comment
Paul Shoemaker on 3 Feb 2020
In your first example of finding where "8" exists:
J = [0 3 4;3 0 8;4 8 0];
[i,j] = ind2sub(size(J),find(J==8));
disp([i j]); % Display results, where each row is an occurrence of "8"
Change the "find(J==8)" argument to whatever you want.
Paul S
giorgos kazas on 3 Feb 2020
Thank you very much!
Guillaume on 3 Feb 2020
Hum, this is rather convoluted! For 2D matrices,
[i,j] = ind2sub(size(J),find(J==8));
is simply:
[i, j] = find(J == 8)
For ND matrices, N >= 3, then indeed it makes sense to use ind2sub.

giorgos kazas on 4 Feb 2020
Thank you to everyone for the input.
I have also another question.
Do you know how can I get the results from "disp([i j])" into a matrix? Because I want to create a matrix with the values in one column and their position in the second-third column.