I have simulated some sample paths S_n=sum(X_i), where the X_i are simulated having different distributions.
I then look at the proportion of the times where S_n is positive. Where n is the length of the sample path, and I have simulated a large number of sample paths.
This follows the arcsine density, which I have shown by a P-P plot and Q-Q plot.
But I would like to see which distribution for X_i that converges fastest to the limiting arcsine distribution.
How do I do this?
Thanks,
