How do I take the average of a zeros vector without dividing by the unpopulated zeros?

I am using real time data coming in and it is being populated into a matrix of zeros. I have the current code...

handles.vxdata= zeros(2400)

handles.speedmax= max(handles.vxdata)

%This works perfect b/c it just picks the max value out of the zeros vector

handles.speedavg=mean(handles.vxdata)

%I am having trouble with this command because it is taking my matrix of zeros and dividing by all the values in the vector, including the zeros that have not been populated with real time data.

I am looking to create a command that only uses the data in the matrix that has values.

Answer 1

[Matt Kindig]

So basically you just want to exclude zeros from your mean?

notZero = (handles.vxdata)~=0;
handles.speedavg = mean(handles.vxdata(notZero));

[Matt Kindig]

I might suggest, though, if handles.vxdata is a double matrix, to pre-allocate handles.vxdata using NaN's, not zeros. That way, if you have a "real" value that is identically zero, it is distinct from your preallocated value.

Also, zeros(2400) will make a 2400x2400 matrix, not a 2400x1 matrix. Is this what you want?

Answer 2

[Walter Roberson]

If you are using a vector instead of a matrix, then

handles.speedavg = mean(nonzeros(handles.vxdata));