Message-ID: <000301cad5ce$85e29a00$7c36200a@gne.windows.gene.com>
Date: 2010-04-06T21:17:11Z
From: Bert Gunter
Subject: checking bivariate normality
In-Reply-To: <x2j68b1e2611004061406y15e5c95ev737330f8a774567d@mail.gmail.com>
Below. -- Bert
Bert Gunter
Genentech Nonclinical Statistics
-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On
Behalf Of Liviu Andronic
Sent: Tuesday, April 06, 2010 2:06 PM
To: arindam fadikar
Cc: r-help at r-project.org
Subject: Re: [R] checking bivariate normality
On Tue, Apr 6, 2010 at 12:48 PM, arindam fadikar
<arindam.fadikar at gmail.com> wrote:
> x <- iris$Sepal.Length[1:50]/iris$Sepal.Width[1:50]
> y <- iris$Petal.Length[1:50]/iris$Petal.Width[1:50]
>
> I want to check whether (x,y) follows a bivariate normal distribution or
> not, ?using density plot or scatter plot. Is it possible to plot a
bivariate
> density in R. I cant find any.
>
Perhaps there is something in mvnormtest?
Liviu
-- Doubt it, since it's _impossible_ to test whether "(x,y) follow a normal
distribution." One might test whether the data are _consistent_ with
bivariate normality (in the sense that evidence to the contrary is
improbable in some sense), but that is a whole different kettle of fish.
These is not mere semantic petulance; all too often there is fundamental
misunderstanding about such issues. Precision of language is important.
______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.