slightly off-topic re prcomp()

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

slightly off-topic re prcomp()

Laura Quinn-2
Hi,

I was wondering if anyone could tell me why prcomp() will "Invent" modes
of variation in a PCA on identical replicates of data? I would have
expected 50 (or whatever number) of identical replicates to return a null
score in such an analysis (or at the least, all variables would share the
same PC score). This is not the case and I was wondering could someone
point me in the direction of some literature to explain the reason behind
this?

Laura Quinn
Institute of Atmospheric Science
School of Earth and Environment
University of Leeds
Leeds
LS2 9JT

tel: +44 113 343 1596
fax: +44 113 343 6716
mail: [hidden email]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Reply | Threaded
Open this post in threaded view
|

Re: slightly off-topic re prcomp()

Bert Gunter
If you don't get a response:

reproducible example, please.

I suspect that you're looking at numerical analysis artifacts, assuming I
have correctly interpreted you.

-- Bert Gunter
Genentech Non-Clinical Statistics
South San Francisco, CA
 
"The business of the statistician is to catalyze the scientific learning
process."  - George E. P. Box
 
 

> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of Laura Quinn
> Sent: Wednesday, February 08, 2006 9:50 AM
> To: [hidden email]
> Subject: [R] slightly off-topic re prcomp()
>
> Hi,
>
> I was wondering if anyone could tell me why prcomp() will
> "Invent" modes
> of variation in a PCA on identical replicates of data? I would have
> expected 50 (or whatever number) of identical replicates to
> return a null
> score in such an analysis (or at the least, all variables
> would share the
> same PC score). This is not the case and I was wondering could someone
> point me in the direction of some literature to explain the
> reason behind
> this?
>
> Laura Quinn
> Institute of Atmospheric Science
> School of Earth and Environment
> University of Leeds
> Leeds
> LS2 9JT
>
> tel: +44 113 343 1596
> fax: +44 113 343 6716
> mail: [hidden email]
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html
>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Reply | Threaded
Open this post in threaded view
|

Re: slightly off-topic re prcomp()

Duncan Murdoch
In reply to this post by Laura Quinn-2
On 2/8/2006 12:50 PM, Laura Quinn wrote:
> Hi,
>
> I was wondering if anyone could tell me why prcomp() will "Invent" modes
> of variation in a PCA on identical replicates of data? I would have
> expected 50 (or whatever number) of identical replicates to return a null
> score in such an analysis (or at the least, all variables would share the
> same PC score). This is not the case and I was wondering could someone
> point me in the direction of some literature to explain the reason behind
> this?

I think you'll need to show us an example.  I just tried what I think
you described, and got all the components with 0 standard deviation, as
I would have expected.

Duncan Murdoch


>
> Laura Quinn
> Institute of Atmospheric Science
> School of Earth and Environment
> University of Leeds
> Leeds
> LS2 9JT
>
> tel: +44 113 343 1596
> fax: +44 113 343 6716
> mail: [hidden email]
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

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