R help - Jun 2003 - compositional data: percent values sum up to 1

again, under another subject:
sorry, maybe an all too trivial question. But we have power data from J
frequency spectra and to have the same range for the data of all our
subjects, we just transformed them into % values, pseudo-code:

power[i,j]=power[i,j]/sum(power[i,1:J])

of course, now we have a perfect linear relationship in our x design-matrix,
since all power-values for each subject sum up to 1.

How shall we solve this problem: just eliminate one column of x, or
introduce a restriction which says exactly that our power data sum up to
1 for each subject?

Thanks a lot

Christoph
-- 
Christoph Lehmann <lehmann at puk.unibe.ch>
University Hospital of Clinical Psychiatry
-- 
Christoph Lehmann <christoph.lehmann at gmx.ch>

Good afternoon R-masters,      
      
I am with some doubts in the R, see the script below:

m<-c(69.6,67.3,75.6,74.3,64.7,60,65.7,62.5,66.5)
d<-c(11.6,15,17.8,18.3,11.2,11,4.6,5.8,7)
year<-c(1994,1995,1996,1997,1998,1999,2000,2001,2002)
male<-ts(m,start=c(1994))
death<-ts(d,start=c(1994))
data<-data.frame(year,death,male)
require(ts)
d100<-HoltWinters(data$death,gamma=0)
m100<-HoltWinters(data$male,gamma=0)
par(mfrow=c(3,1))
plot(d100,main="Death")
plot(m100,main="Male")
ccf(male,death)

    
I have 2 doubts:    
    
1 - How to I should interpret the third graph?    
2 - Has a hypothesis test to evaluate the cross-correlation it is significant in
R?
Thanks in advance

Bernardo Rangel Tura, MD, MSc
National Institute of Cardiology Laranjeiras
Rio de Janeiro Brazil

What are you trying to do?  What I would do with this depends on many 
factors.

spencer graves

Christoph Lehmann wrote:
> again, under another subject:
> sorry, maybe an all too trivial question. But we have power data from J
> frequency spectra and to have the same range for the data of all our
> subjects, we just transformed them into % values, pseudo-code:
> 
> power[i,j]=power[i,j]/sum(power[i,1:J])
> 
> of course, now we have a perfect linear relationship in our x
design-matrix,
> since all power-values for each subject sum up to 1.
> 
> How shall we solve this problem: just eliminate one column of x, or
> introduce a restriction which says exactly that our power data sum up to
> 1 for each subject?
> 
> Thanks a lot
> 
> Christoph

I want to do a logistic regression analysis, and to compare with, a
discriminant analysis. The mentioned power maps are my exogenous data,
the dependent variable (not mentioned so far) is a diagnosis
(ill/healthy)

thanks for the interest and the help

Christoph

On Sun, 2003-06-01 at 21:01, Spencer Graves wrote:
> What are you trying to do?  What I would do with this depends on many 
> factors.
> 
> spencer graves
> 
> Christoph Lehmann wrote:
> > again, under another subject:
> > sorry, maybe an all too trivial question. But we have power data from
J
> > frequency spectra and to have the same range for the data of all our
> > subjects, we just transformed them into % values, pseudo-code:
> > 
> > power[i,j]=power[i,j]/sum(power[i,1:J])
> > 
> > of course, now we have a perfect linear relationship in our x
design-matrix,
> > since all power-values for each subject sum up to 1.
> > 
> > How shall we solve this problem: just eliminate one column of x, or
> > introduce a restriction which says exactly that our power data sum up
to
> > 1 for each subject?
> > 
> > Thanks a lot
> > 
> > Christoph
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
-- 
Christoph Lehmann <christoph.lehmann at gmx.ch>

"glm" will do multinomial logistic regression.  However, if J is
large,
I doubt if that will do what you want.  If it were my problem, I might 
feel a need to read the code for "glm" and modify it to do what I
want.
  Perhaps someone else can suggest something better.

hth.  spencer graves

Christoph Lehmann wrote:
> I want to do a logistic regression analysis, and to compare with, a
> discriminant analysis. The mentioned power maps are my exogenous data,
> the dependent variable (not mentioned so far) is a diagnosis
> (ill/healthy)
> 
> thanks for the interest and the help
> 
> Christoph
> 
> On Sun, 2003-06-01 at 21:01, Spencer Graves wrote:
> 
>>What are you trying to do?  What I would do with this depends on many 
>>factors.
>>
>>spencer graves
>>
>>Christoph Lehmann wrote:
>>
>>>again, under another subject:
>>>sorry, maybe an all too trivial question. But we have power data
from J
>>>frequency spectra and to have the same range for the data of all our
>>>subjects, we just transformed them into % values, pseudo-code:
>>>
>>>power[i,j]=power[i,j]/sum(power[i,1:J])
>>>
>>>of course, now we have a perfect linear relationship in our x
design-matrix,
>>>since all power-values for each subject sum up to 1.
>>>
>>>How shall we solve this problem: just eliminate one column of x, or
>>>introduce a restriction which says exactly that our power data sum
up to
>>>1 for each subject?
>>>
>>>Thanks a lot
>>>
>>>Christoph
>>
>>______________________________________________
>>R-help at stat.math.ethz.ch mailing list
>>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>

On Mon, 2 Jun 2003, Spencer Graves wrote:
> "glm" will do multinomial logistic regression.  However, if J is
large,
Strictly, no, it will not as that is not a GLM.  glm() can only do it via 
surrogate Poisson models.  multinom in nnet(VR) will do multinomial 
logistic regression.
> I doubt if that will do what you want.  If it were my problem, I might 
> feel a need to read the code for "glm" and modify it to do what I
want.
>   Perhaps someone else can suggest something better.
> 
> hth.  spencer graves
> 
> Christoph Lehmann wrote:
> > I want to do a logistic regression analysis, and to compare with, a
> > discriminant analysis. The mentioned power maps are my exogenous data,
> > the dependent variable (not mentioned so far) is a diagnosis
> > (ill/healthy)
> > 
> > thanks for the interest and the help
> > 
> > Christoph
> > 
> > On Sun, 2003-06-01 at 21:01, Spencer Graves wrote:
> > 
> >>What are you trying to do?  What I would do with this depends on
many
> >>factors.
> >>
> >>spencer graves
> >>
> >>Christoph Lehmann wrote:
> >>
> >>>again, under another subject:
> >>>sorry, maybe an all too trivial question. But we have power
data from J
> >>>frequency spectra and to have the same range for the data of
all our
> >>>subjects, we just transformed them into % values, pseudo-code:
> >>>
> >>>power[i,j]=power[i,j]/sum(power[i,1:J])
> >>>
> >>>of course, now we have a perfect linear relationship in our x
design-matrix,
> >>>since all power-values for each subject sum up to 1.
> >>>
> >>>How shall we solve this problem: just eliminate one column of
x, or
> >>>introduce a restriction which says exactly that our power data
sum up to
> >>>1 for each subject?
> >>>
> >>>Thanks a lot
> >>>
> >>>Christoph
> >>
> >>______________________________________________
> >>R-help at stat.math.ethz.ch mailing list
> >>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> >
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> 
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

Eh?  The original message says it's the design matrix that is perfectly
collinear after the transformation, not the response.

I don't know much about this type of data, but seems like you could just fit
the model w/o intercept to eliminate the collinearity, no?  It's the
interpretation of the result that may be tricky, I think.

Andy

> -----Original Message-----
> From: Spencer Graves [mailto:spencer.graves at pdf.com]
> Sent: Monday, June 02, 2003 9:33 AM
> To: Christoph Lehmann
> Cc: Spencer Graves; r-help at stat.math.ethz.ch
> Subject: Re: [R] compositional data: percent values sum up to 1
> 
> 
> "glm" will do multinomial logistic regression.  However, if J 
> is large, 
> I doubt if that will do what you want.  If it were my 
> problem, I might 
> feel a need to read the code for "glm" and modify it to do 
> what I want. 
>   Perhaps someone else can suggest something better.
> 
> hth.  spencer graves
> 
> Christoph Lehmann wrote:
> > I want to do a logistic regression analysis, and to compare with, a
> > discriminant analysis. The mentioned power maps are my 
> exogenous data,
> > the dependent variable (not mentioned so far) is a diagnosis
> > (ill/healthy)
> > 
> > thanks for the interest and the help
> > 
> > Christoph
> > 
> > On Sun, 2003-06-01 at 21:01, Spencer Graves wrote:
> > 
> >>What are you trying to do?  What I would do with this 
> depends on many 
> >>factors.
> >>
> >>spencer graves
> >>
> >>Christoph Lehmann wrote:
> >>
> >>>again, under another subject:
> >>>sorry, maybe an all too trivial question. But we have 
> power data from J
> >>>frequency spectra and to have the same range for the data 
> of all our
> >>>subjects, we just transformed them into % values, pseudo-code:
> >>>
> >>>power[i,j]=power[i,j]/sum(power[i,1:J])
> >>>
> >>>of course, now we have a perfect linear relationship in 
> our x design-matrix,
> >>>since all power-values for each subject sum up to 1.
> >>>
> >>>How shall we solve this problem: just eliminate one column of
x, or
> >>>introduce a restriction which says exactly that our power 
> data sum up to
> >>>1 for each subject?
> >>>
> >>>Thanks a lot
> >>>
> >>>Christoph
> >>
> >>______________________________________________
> >>R-help at stat.math.ethz.ch mailing list
> >>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> >
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>