Re: Theory question on kolmogorov-smirnov

From: Vadim Pliner (Vadim.Pliner_at_VerizonWireless.com)
Date: 10/27/04 

Date: 27 Oct 2004 06:31:39 -0700

It is
sup|Fn(x)-F(x)|,
 x
where Fn(x) - empirical CDF and F(x) - any specified CDF, not
necessarily normal or even continuous.

HTH,
Vadim

AJ <99atlantic@gmail.com> wrote ...
> Quick question on a 1-sample k-s test.
>
> It's Supreme(absolute value(Empirical_CDF(x) - Normal_CDF(x)))
>
> Right.
>
> For the normal_CDF(x), is that normal using mean 0, variance of 1, or using
> the paramaters from the list that comprises the x's. so mean(x), stdev(x) ?