Euclidean distance explanation for a matrix

         A         B         C         D
1        0   93.0000   12.0000    9.0000
2  14.0000  102.0000   75.0000   33.0000
3        0    5.0000   34.0000   33.0000
4  11.5000   50.0000   81.0000   36.0000

I have the above matrix. When I preform pdist(<above matric>,'euclidean') for euclidean distance, I got the following answer. Can somebody explain what these euclidean distance means for a (4 x 4) matrix above? I cannot understand how these values are calculated.

69.4406 93.8296 86.4364 106.2356 52.4905 66.1457

Thanks.

Answer 1

[Mischa Kim]

Damit, computing the pdist of an m-by-n ( 4-by-4 in your case) matrix means computing the pairwise distance between each of the m rows. For 2 rows you get one result, for 3 rows 3 (1-2, 1-3, 2-3), for 4 rows 6.

[Damith]

Thank you Mischa,

I need to calculate Euclidean N-correlation distance of the matrix. Do you know how to calculate that in MATLAB? Then I need to find what elements in the matrix are closely related by using the correlation distance estimates.

I assume what you meant here it compares 1st row with 2nd row etc.(1-2, 1-3, 2-3).right?

Thanks again for your reply.