[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: fitting 2d normal distribution
From: |
Søren Hauberg |
Subject: |
Re: fitting 2d normal distribution |
Date: |
Thu, 23 Oct 2008 17:00:59 +0200 |
ons, 22 10 2008 kl. 10:42 -0400, skrev John W. Eaton:
> On 22-Oct-2008, Søren Hauberg wrote:
> | I don't think there are any functions for plotting ellipses from
> | covariance matrices.
>
> Does the following function do (at least part of) what you want?
It seems like quite a nice tool. Just to make sure I use it correctly:
if I want to visualise the covariance of some zero-mean data, by
plotting the ellipse corresponding to the points with a Mahalanobis'
distance of 3 from the origin, then I have to call 'ellipse' like this
[...] = ellipse (inv (C), 3);
where C is the covariance matrix of the data?
Søren
P.S. Plotting ellipses is such a standard operation that I think we
should have this function somewhere. In core Octave? If not, any
suggestions to which package we should put it?
- fitting 2d normal distribution, Francesco Potorti`, 2008/10/21
- Re: fitting 2d normal distribution, Søren Hauberg, 2008/10/21
- Re: fitting 2d normal distribution, Francesco Potorti`, 2008/10/21
- Re: fitting 2d normal distribution, Søren Hauberg, 2008/10/22
- Re: fitting 2d normal distribution, John W. Eaton, 2008/10/22
- Re: fitting 2d normal distribution,
Søren Hauberg <=
- Re: fitting 2d normal distribution, John W. Eaton, 2008/10/23
- Re: fitting 2d normal distribution, Søren Hauberg, 2008/10/23
- Re: fitting 2d normal distribution, Carlo de Falco, 2008/10/24