[R] Goodness fit test HELP!

Sat Dec 3 22:05:57 CET 2005

The nice thing about the uniform density is that it's easy to know what the
expected pdf(pmf) should look like, namely each observation should have
probability 1/n.  That means you can use "qqplot."  See ?qqplot

Here's an example, using "my.data."

my.data <- runif(100)
n.points <- length(my.data)
expected.cdf <- ((1:n.points)-0.5)/(n.points)
qqplot(my.data, expected.cdf, las=1)
# Use the "interocular trauma test" for goodness-of-fit:
my.lm <- lm(expected.cdf ~ sort(my.data))
abline(coef=coef(my.lm), lty=2)

Charles Annis, P.E.

Charles.Annis at StatisticalEngineering.com
phone: 561-352-9699
eFax:  614-455-3265
http://www.StatisticalEngineering.com

-----Original Message-----
From: r-help-bounces at stat.math.ethz.ch
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Amit Kabiri
Sent: Saturday, December 03, 2005 11:43 AM
To: 'Elizabeth Lawson'; 'David Zhao'
Cc: r-help at stat.math.ethz.ch
Subject: Re: [R] Goodness fit test HELP!

If I have a Uniform distribution to check, How can I use visual fits? Can I
also use in some way the qqnorm?

Thanks 

-----Original Message-----
From: r-help-bounces at stat.math.ethz.ch
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Elizabeth Lawson
Sent: Friday, November 18, 2005 8:06 PM
To: David Zhao
Cc: r-help at stat.math.ethz.ch
Subject: Re: [R] Goodness fit test HELP!

What about trying a qqplot to see how the distribution fits...

  For the normal distribution thta is very stright forward, use qqnorm.

  To test gamma distribtution (or any other) do some thing like this 

  n<-length(data)
  for(i in 1:n){
  prob<-(i-1/3)/(n1/3)
  }

quantiles<-qgamma(prob,shape=mean(data)/var(data),scale=var(data)/mean(data)
}

  qqplot(data,quantiles)

  If the distribution is a good for, you should a stright line, like wiht a
qqnorm plot!

  Good luck!!

  Elizbaeth Lawson

David Zhao <wzhao6898 at gmail.com> wrote:
  Hi there,

I'm a newbie, plesae bear with me.
I have a dataset with about 10000 ~ 30000 data points. Would like fit to
both Gamma and Normal distribution to see which one fits better. How do I do
this in R? Or I could do a normality test of the data, if it's normal, I
then will do a normal fit, otherwise, a gamma fit. But again, I don't know
how to do this either.
Please help!

David

[[alternative HTML version deleted]]

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html

---------------------------------

	[[alternative HTML version deleted]]

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html