var(I(:)); Means?

Hi, I have this problem

signal_var = var(I(:));

What's the meaning of the code below?

var(I(:));

Thanks a lot :)

[Oleg Komarov]

You can execute the code piece by piece, just highlight for example I(:) and press F9.

Also, doc var.

[Superb]

F9 doesn't show anything, it only show a lot of long code inside the Command Window, can I know why? Thanks

And is there any correct way to debug? TQ :)

[Oleg Komarov]

F9 executes the code.

[Superb]

I think F5 is the one? Lolx..

[Walter Roberson]

The key depends upon the operating system. For Linux it is F5 .

[Superb]

I use Windows 7 :)

[Image Analyst]

I didn't know that. By the way, in WIndows 7 F9 will evaluate the selection (e.g. the line of code you're on in the command history panel - just right click and see), while F5 says to run the whole m-file, or to continue running non-stop if you had stopped somewhere inside the m-file. So for Windows 7 they both work. Of course this has not much to do with the original question, which was answered twice below (but apparently unseen by Superb) in clear and explicit detail.

[Walter Roberson]

Ah, I misread. I never used "evaluate selection".

[Image Analyst]

I never do either.

Answer 1

[the cyclist]

I(:) treats the array I as a column vector (instead of, say, maybe the MxN array that it was). The var() function takes the variance of that column vector. [See "help var" for details.]

(That assumes that you did not define var as a variable name earlier in the code. If var is a variable, then this code is indexing into that variable.)

Answer 2

[Image Analyst]

Superb, normally var() will take the mean of each column of a 2D array and return a row vector where each element is the variance of the corresponding column. The programmer didn't want that - he wanted the variance of the entire array, not a column-by-column variance. So to do that you can use the (:) construct, which basically takes all the elements in an array (of any dimension) and strings them all together in a single 1D list of numbers (called a "column vector"). Once you've done that, then the var() function will operate on the whole array, not column-by-column. Now it will return a single number rather than a 1D row vector of the column variances. Does that explain it well?