Cumulative Mean of a Matrix

Hi,

I have a matrix: Vx=[4,5,6,3,2,4,5,6,4,3,2,5,6,7,5] etc.

I want to find the cumulative mean of these values (to see if the velocity field converges with time).

Does anyone have a good method to do this?

I have

for k=1:1:15;
CumuVel=mean(Vx(1,1):Vx(1,k));
end

But this won't work....any suggestions?

Thanks

[Image Analyst]

I don't see how the cumulative mean does that. You could have steady values and then at the end it goes crazy and still have a stable cumulative mean. I think you're better off looking at the difference using diff() and see if that difference value gets smaller and smaller or reaches some steady state level of variation, but whatever - it's your project.....

[Teja Muppirala]

I agree with Image Analyst, to see if something is converging, it is not sufficient to just look at the cumulative mean.

For example, this series Vx is diverging, but the cumulative mean still converges:

Vx = sqrt(1:1000) .* (-1).^(1:1000)
plot(Vx); % Clearly not converging
CumuVel = cumsum(Vx)./(1:length(Vx)); % Get cumulative mean
figure, plot(CumuVel); % Vx is diverging, this is converging

[Jan]

I assume, you mean:

CumuVel = zeros(1, 15);  % Pre-allocation!!!
for k = 1:15;
  CumuVel(k) = mean(Vx(1:k));
end

Answer 1

[Matt J]

 CumuVel = cumsum(Vx)./(1:length(Vx));

[Jan]

@Matt J: 6 Votes in 17 hours!

[Matt J]

It's pretty bizarre. No idea why such a simple Answer to such a simple Question is attracting so much attention.