R help - Oct 2003 - Fw: SIMCA algorithm implementation

I have used PCA for data classification by visual examination of the 3D
scatter plot of the first 3 principal components.  I now want to use the
results to predict the class for new data.  I have used predict.princomp to
predict the scores and then visualise the results on a 3D scatter plot using
the rgl library.  However, is there an R function that will fit the new data
to the class assignments derived from PCA?  I think this is similar to what
the SIMCA algoirthm does.

Thanks
Mike

----- Original Message -----
From: "Mike White" <mikewhite.diu at tiscali.co.uk>
To: <R-help at stat.math.ethz.ch>
Sent: Wednesday, October 08, 2003 9:25 AM
Subject: Re: [R] SIMCA algorithm implementation

> Dear All
> Is there a SIMCA (Soft Independent Modelling Class Analogy) implementation
> on R or does anyone know if is it possible to replicate the SIMCA
algorithm
> using existing R functions?
>
> Thanks
> Mike White
>

Mike  -

For predicting class membership, I would use either  lda() or
qda()  from the MASS package.  See the Venables and Ripley
book for detailed description of the methods.  You'll have
to rely on your own references for what the 'SIMCA' algorithm
actually does.  I've never heard of it.  Sounds like some
sort of discriminant analysis to me.  Maybe its authors
would like to contribute an implementation of it to R.

-  tom blackwell  -  u michigan medical school  -  ann arbor  -

On Wed, 15 Oct 2003, Mike White wrote:
> I have used PCA for data classification by visual examination of the 3D
> scatter plot of the first 3 principal components.  I now want to use the
> results to predict the class for new data.  I have used predict.princomp to
> predict the scores and then visualise the results on a 3D scatter plot
using
> the rgl library.  However, is there an R function that will fit the new
data
> to the class assignments derived from PCA?  I think this is similar to what
> the SIMCA algoirthm does.
>
> Thanks
> Mike
>
> ----- Original Message -----
> From: "Mike White" <mikewhite.diu at tiscali.co.uk>
> To: <R-help at stat.math.ethz.ch>
> Sent: Wednesday, October 08, 2003 9:25 AM
> Subject: Re: [R] SIMCA algorithm implementation
>
>
> > Dear All
> > Is there a SIMCA (Soft Independent Modelling Class Analogy)
implementation
> > on R or does anyone know if is it possible to replicate the SIMCA
> algorithm
> > using existing R functions?
> >
> > Thanks
> > Mike White
>

SIMCA, I believe, is Svante Wold's invention, and extensively used in the
chemometrics area (analysis of data from analytic chemistry).  My vague
impression of what it does is PCA in the individual classes.  I have not
been able to locate a detail description of the algorithm.  (I'd appreciate
it very much if some one can help.)  One feature of the methodology that
distinguish it from others, I believe, is its ability to classify new
observations as belonging to classes other than what appeared in the
training data.

Contribution from it's author?  Fat chance.  Svante sells (stand-alone)
SIMCA at a premium price.

Best,
Andy

> -----Original Message-----
> From: Thomas W Blackwell [mailto:tblackw at umich.edu] 
> Sent: Wednesday, October 15, 2003 10:53 AM
> To: Mike White
> Cc: R-help at stat.math.ethz.ch
> Subject: Re: Fw: [R] SIMCA algorithm implementation
> 
> 
> Mike  -
> 
> For predicting class membership, I would use either  lda() or
> qda()  from the MASS package.  See the Venables and Ripley
> book for detailed description of the methods.  You'll have
> to rely on your own references for what the 'SIMCA' algorithm 
> actually does.  I've never heard of it.  Sounds like some 
> sort of discriminant analysis to me.  Maybe its authors would 
> like to contribute an implementation of it to R.
> 
> -  tom blackwell  -  u michigan medical school  -  ann arbor  -
> 
> On Wed, 15 Oct 2003, Mike White wrote:
> 
> > I have used PCA for data classification by visual 
> examination of the 
> > 3D scatter plot of the first 3 principal components.  I now want to 
> > use the results to predict the class for new data.  I have used 
> > predict.princomp to predict the scores and then visualise 
> the results 
> > on a 3D scatter plot using the rgl library.  However, is there an R 
> > function that will fit the new data to the class 
> assignments derived 
> > from PCA?  I think this is similar to what the SIMCA algoirthm does.
> >
> > Thanks
> > Mike
> >
> > ----- Original Message -----
> > From: "Mike White" <mikewhite.diu at tiscali.co.uk>
> > To: <R-help at stat.math.ethz.ch>
> > Sent: Wednesday, October 08, 2003 9:25 AM
> > Subject: Re: [R] SIMCA algorithm implementation
> >
> >
> > > Dear All
> > > Is there a SIMCA (Soft Independent Modelling Class Analogy) 
> > > implementation on R or does anyone know if is it possible to 
> > > replicate the SIMCA
> > algorithm
> > > using existing R functions?
> > >
> > > Thanks
> > > Mike White
> >
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list 
> https://www.stat.math.ethz.ch/mailman/listinfo> /r-help
>

>/ Dear All
/>/ Is there a SIMCA (Soft Independent Modelling Class Analogy)
implementation
/>/ on R or does anyone know if is it possible to replicate the SIMCA
/algorithm
>/ using existing R functions?
/>/
/>/ Thanks
/>/ Mike White
/>/
/
I haven't seen any implementation in R, but check out the
'chemiometrics' home page here at
Ume? University, http://www.acc.umu.se/%7Etnkjtg/chemometrics/publications.html

Hans Gardfjell
Dept of Ecology and Environmental science
Ume? University, Sweden