Finding average function value based on data points.

Question

Let's say I have 200 data points $(x,y)$. I know the step size for the input $x$ is $h$. To find the approximated average function value $A$ of some unknown function $ x \mapsto y$, I believe I could use a trapezoid rule like:

I  = 0;
for k = 1:199

    I = I + h * ( y(k) + y(k+1) ) * (1/2);

end  

A = 1/(x(end) - x(begin)) * I;
return A; 

But I was just thinking, can't I just take the mean of the $y$-vector, to get the same result?