R help - Nov 2009 - How to test user-defined distribution using ks.test in R?

Dear All,

I defined a distribution and I want to test it by k-s test?
How to input parameters for ks.test(.....)?
E.g.,
F(x,m) is user-defined function
x is observation value, m is parameter of F

y is observation vector value in experiment
> ks.test(y,"F",m=0.2)
error....


why?




'
-- 

Best wishes,
Zhijiang Wang
--------------------------------------------
PHD Student
Room 212, Science buliding,
The International WIC Institute,
College of Computer Science and Technology,
Beijing University of Technology,
Beijing, China.

On Nov 6, 2009, at 1:53 AM, Zhijiang Wang wrote:
> Dear All,
>
> I defined a distribution and I want to test it by k-s test?
> How to input parameters for ks.test(.....)?
> E.g.,
> F(x,m) is user-defined function
> x is observation value, m is parameter of F
>
> y is observation vector value in experiment
>
>> ks.test(y,"F",m=0.2)
> error....
>
>
> why?
>
Without the error message? or the function definition? or str(y)?

I would ask "why" you think intelligent commentary should be possible.

-- 

David Winsemius, MD
Heritage Laboratories
West Hartford, CT

Zhijiang Wang wrote:
> Dear All,
> 
> I defined a distribution and I want to test it by k-s test?
> How to input parameters for ks.test(.....)?
> E.g.,
> F(x,m) is user-defined function
> x is observation value, m is parameter of F
> 
> y is observation vector value in experiment
> 
>> ks.test(y,"F",m=0.2)
> error....

Well, R did not say "error", did it?
Where is the definition of F?
What is y?

uwe Ligges

> 
> why?
> 
> 
> 
> 
> '