Standard deviation for three dimensional matrix

I have a data file that has x, y, and z sets of data, dimensions 120x150x151. I need to plot the standard deviation of the numbers in x and y at each z. How can I use std() to accomplish this?

hold on  
for counter=1:151
sd=std()
plot(z(counter),sd)
end
hold off

Answer 1

[Mukul Rao]

I do not believe you need a loop for this. The 'std' function has the capability to account for direction as per the documentation :

http://www.mathworks.com/help/matlab/ref/std.html

If you would like to get the standard deviation along the x direction, use

stdYZ = std(mymatrix,1) %Assuming x changes along the columns

This returns the standard deviation along the x-direction, at each point with fixed Y and Z co-ordinates.

If you are looking for the standard deviation of an entire X-Y plane at each level Z, then I suggest using "std2" instead. Here is the documentation link:

http://www.mathworks.com/help/images/ref/std2.html

In this case your code would look like

hold on  
for counter=1:151
sd=std2(mydata(:,:,counter)) %Assuming matrix mydata is 120x150x151
plot(z(counter),sd)
end
hold off

[Luke Wunderlich]

I tried to do this:

StdDev=ones(1,151);
Mean=ones(1,151);
for counter=1:151
    StdDev(counter)=std2(Data(:,:,counter));
    Mean(counter)=mean(Data(:,:,counter));
end
figure(2)
hold on
plot(X,Mean,'k')
plot(X,Mean(X)+StdDev(X),'r--')
plot(X,Mean(X)-StdDev(X),'r--')
hold off

I received this error:

In an assignment  A(I) = B, the number of elements in B and I must be the same.

    Error in StdDevSSBare (line 34)
        Mean(counter)=mean(Data(:,:,counter));

[Mukul Rao]

Like std2, use mean2 rather than mean. mean would return a 120x150 matrix and you are equating that to a matrix of size 1x151, hence the error message.

[Eric]

Or if you don't have the Image Processing Toolbox:

StdDev=ones(1,151);
Mean=ones(1,151);
for counter=1:151
    frame = Data(:,:,counter);
    StdDev(counter)=std(frame(:));
    Mean(counter)=mean(frame(:));
end
figure(2)
hold on
plot(X,Mean,'k')
plot(X,Mean(X)+StdDev(X),'r--')
plot(X,Mean(X)-StdDev(X),'r--')
hold off