[R] Distributions Comparison

Rolf Turner r.turner at auckland.ac.nz
Tue Oct 28 23:22:04 CET 2008


On 29/10/2008, at 10:53 AM, Igor Telezhinsky wrote:

> Dear all,
>
> I have recently found out about the R project. Could anyone tell me  
> if it is
> possible to make the comparison of two distributions using R  
> packages? The
> problem is TRICKY because I have the distributions of measurements  
> and each
> measurement in the given distribution has its own error, which I  
> need to
> take into account when comparing these distributions. If anyone  
> knows it is
> possible with R, could you please tell me what package to use. I  
> will really
> appreciate some hints in code as well.

Your message is almost completely opaque to me.  What do you actually  
want
to do?  What do you mean by ``distributions''?  Do you actually know  
those
distributions (unlikely!) or do you have a sample from each of those  
distributions?

What do you mean by ``compare''?  Do you simply want to *test*  
whether the
distributions are indeed different (on the basis of the given ---  
presumably ---
samples)?  Or do you want to calculate/estimate some measure of the  
difference?
What measure?

It sounds to me like your knowledge of statistics is to say the least  
limited.
You can't expect the R system, wonderful as it is, to make up for  
this lack
of knowledge.  Learn some statistics first, before you try to do a  
statistical
analysis.

	cheers,

		Rolf Turner

P. S. If just want to test for a difference between two distributions/ 
populations
on the basis of independent samples from each of those distributions,  
then you can
use the function ks.test().

		R. T.

######################################################################
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}



More information about the R-help mailing list