[R] Clustering literature was Re: nonlinear regression

Petr PIKAL petr.pikal at precheza.cz
Mon Oct 1 18:25:26 CEST 2007


Hi

It is preferable to echo your posts to r-help, you usually get more 
answers and some definitelly superb to mine.
It is also better to start a new mail if your question has nothing to do 
with original subject


"Maura E Monville" <maura.monville at gmail.com> napsal dne 01.10.2007 
17:44:43:

> Unluckily I do not have the privilege of practising with R all day
> long. I agree the more I read the better but I can only devote a
> portion of my time.

That is not unusual. However customizing yourself with R environment, 
syntax and basic data manipulation can save you a lot of headache. 
Just make a hardcopy of 100 pages R-intro and use it in small portions 
before you fall asleep. In few days you will become a quite experienced 
user.

Regards

Petr
petr.pikal at precheza.cz


> Anyway, I ended up using lm. It worked fine after some time spent on
> choosing the independent variable.
> I ran some tests for heteroschedasticity, autocorrelation, normal
> distribbution of residuals and everything came out fine.
> Now the next step is clustering using the coefficients of the regression 
model.
> I have found plenty of documentation on regression analysis in the
> "contributed" section. Some go over the major concepts of the
> underlying theory and then show some worked out examples.
> I found nothing so nicely laid out about cluster analysis with R.
> I would appreciate some suggestion about reading on techniques for 
clustering.
> Thank you in advance.
> 
> Best regards,
> Maura E.
> 
> On 10/1/07, Petr PIKAL <petr.pikal at precheza.cz> wrote:
> > Did you at least try to use search possibilities in R to find it? It 
shall
> > be a part of basic R installation, however it is not loaded 
automatically
> > when starting R (there are only few packages loaded on start).
> >
> > If you tried
> >
> > > ?nlme
> > No documentation for 'nlme' in specified packages and libraries:
> > you could try 'help.search("nlme")'
> >
> > and help.search(nlme) gives
> >
> > BIC(nlme)                                      Bayesian Information
> > Criterion
> > fitted.nlmeStruct(nlme)                        Calculate nlmeStruct 
Fitted
> > Values
> > getData.lme(nlme)                              Extract lme Object Data
> > nlme(nlme)                                     Nonlinear Mixed-Effects
> > Models
> > nlme.nlsList(nlme)                             NLME fit from nlsList
> > Object
> > nlmeControl(nlme)                              Control Values for nlme 
Fit
> > nlmeObject(nlme)                               Fitted nlme Object
> > nlmeStruct(nlme)                               Nonlinear Mixed-Effects
> > Structure
> > predict.nlme(nlme)                             Predictions from an 
nlme
> > Object
> > residuals.nlmeStruct(nlme)                     Calculate nlmeStruct
> > Residuals
> >
> >
> > Even help.search("nonlinear") would reveal at least 3 packages
> >
> > rms.curv(MASS)                                 Relative Curvature 
Measures
> > for Non-Linear Regression
> > sammon(MASS)                                   Sammon's Non-Linear 
Mapping
> > gnls(nlme)                                     Fit Nonlinear Model 
Using
> > Generalized Least Squares
> > gnlsStruct(nlme)                               Generalized Nonlinear 
Least
> > Squares Structure
> > nlme(nlme)                                     Nonlinear Mixed-Effects
> > Models
> > nlmeStruct(nlme)                               Nonlinear Mixed-Effects
> > Structure
> > nlm(stats)                                     Non-Linear Minimization
> > nls(stats)                                     Nonlinear Least Squares
> > predict.nls(stats)                             Predicting from 
Nonlinear
> > Least Squares Fits
> > selfStart(stats)                               Construct Self-starting
> > Nonlinear Models
> > summary.nls(stats)                             Summarizing Non-Linear
> > Least-Squares Model Fits
> >
> > Regards
> >
> > Petr
> > petr.pikal at precheza.cz
> >
> > "Maura E Monville" <maura.monville at gmail.com> napsal dne 27.09.2007
> > 21:42:30:
> >
> > > This command is not found. Shall I install some extra package ?
> > > Thamk you,
> > > maura
> > >
> > >
> > > On 9/27/07, Petr PIKAL <petr.pikal at precheza.cz> wrote:
> > > Hi
> > >
> > > Have a look at package nlme and maybe also you should consult a
> > > Pinheiro&Bates book about nonlinear mixed effects models.
> > >
> > > Regards
> > >
> > > Petr
> > > petr.pikal at precheza.cz
> > >
> > > r-help-bounces at r-project.org napsal dne 27.09.2007 07:46:03:
> > >
> > > > I would appreciate some suggestions about  nonlinear regression
> > > available in
> > > > R ... possible methods and plenty of worked out examples ....
> > > > I have a bunch of noisy curves representing breathing amplitude 
from
> > > medical
> > > > physics experiments recoding patients' breathing tracks in form of
> > > > Amplitude, Phase, Time, some flags controlling the data validity.
> > > > A variable number of successive breathing cycles were recorded.
> > > >
> > > > Thank you in advance,
> > > > --
> > > > Maura E.M
> > > >
> > > >    [[alternative HTML version deleted]]
> > > >
> > > > ______________________________________________
> > > > R-help at r-project.org mailing list
> > > > https://stat.ethz.ch/mailman/listinfo/r-help
> > > > PLEASE do read the posting guide
> > > http://www.R-project.org/posting-guide.html
> > > > and provide commented, minimal, self-contained, reproducible code.
> >
> > >
> > >
> > >
> > > --
> > > Maura E.M
> >
> 
> 
> -- 
> Maura E.M



More information about the R-help mailing list