matrix with different sizes for ranksum test

Hello,

I have two matrix of A (9*15) and B (13*16) and would like to test if there are significantly different.

A=A(:)

B=B(:) and then use

[p, h, stats] = ranksum (A,B).

is this the right way?

Thanks

Answer 1

[Walter Roberson]

It is the correct approach under the assumption that the structure of A and B do not matter, that the properties of the array elements are independent of position.

That is a valid assumption at times, but in other cases it is not valid at all. For example if A represents student test scores in one class, then the median for each different test could be different and so which row (or column) a value is in could be important.

[Roxan]

Thanks.

if A represent energy of some specifc lamps in diferent dates and B represent the energy of some other lamps in some other dates, how can run the ranksum between A and B?

Roxan

[Walter Roberson]

The code you posted looks like it would be ok for that.