Hey Ahmet,
1/(N-1) sum (x-mean)(x-mean)'
is the unbiased estimator of the true (population) covariance matrix,
provided the observations are mutually independent (!)
In most cases, N is quite large, so it doesn't actually matter, and
the eigenvectors and order (not size) of the eigenvalues are i
Hi,
S = 1/N sum (x - ref) (x - ref)'
or
S = 1/(N-1) sum (x - ref) (x - ref)'
N: the number of frames
Which one is right?
2013/2/12 Tsjerk Wassenaar
> Hi Vivek,
>
> If you use the g_covar option -ref, you not only use the reference
> structure for fitting, you use it for calculating the devi
Hi Vivek,
If you use the g_covar option -ref, you not only use the reference
structure for fitting, you use it for calculating the deviations. Your
covariance matrix is built as:
S = 1/N sum (x - ref) (x - ref)'
If you leave out the option -ref then the average structure will be
used for the cov
>
>
> Message: 1
> Date: Sun, 10 Feb 2013 21:32:15 + (WET)
> From: bapti...@itqb.unl.pt
> Subject: Re: [gmx-users] Reference structure for PCA.
> To: Discussion list for GROMACS users
> Message-ID:
> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
>
Hello Antonio,
Thank you very m
Any comments/help on this query. I am stuck with this.
Thanks,
-Vivek.
On Thu, Feb 7, 2013 at 1:26 PM, vivek modi wrote:
> Hello,
>
> I have troubled you with a similar question before also, but I guess I
> need some more clarification. My question is about the reference structure
> in PCA a
5 matches
Mail list logo