I noticed that there is no generalised additive model functions in R (1.1.1) ... is there a package that implements them? Thanks Prasad ***************************************************************** Mr. Anantha Prasad, Ecologist/GIS Specialist USDA Forest Service, 359 Main Rd. Delaware OHIO 43015 USA Ph: 740-368-0103 Email: aprasad at fs.fed.us Web: http://www.fs.fed.us/ne/delaware/index.html Don't Miss Climate Change Tree Atlas at: http://www.fs.fed.us/ne/delaware/atlas/index.html ****************************************************************** -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
> Subject: [R] Where is gam? > Date: Tue, 3 Oct 2000 09:31:23 -0400 > > I noticed that there is no generalised additive model functions in R > (1.1.1) ... is there a package that implements them?Not for R. The FAQ says: One place where functionality is still missing is the modeling software as described in "Statistical Models in S" (see *Note What is S?::); Generalized Additive Models (*gam*) and some of the nonlinear modeling code are not there yet. If you just want additive models there is a much better function bruto in package MDA. One reason that gam is not there is that some of us think it is much over-sold as a technique and the way it is done in S is poorly implemented. -- 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 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Prof Brian Ripley <ripley at stats.ox.ac.uk> writes:> One reason that gam is not there is that some of us think it > is much over-sold as a technique and the way it is done in S is poorly > implemented.Could you please clarify.... there has been a lot of interest in gam recently in predicting species distribution and environmental management (there is in fact a workshop scheduled in Switzerland next year on GLM/GAM modelling). So many of us would be very interested to learn about its shortcomings (reason why you think it is oversold technique)...and also what aspects of it are poorly implemented in S. If you shed some light on this, it will be of great help to us. Thanks much. Prasad ***************************************************************** Mr. Anantha Prasad, Ecologist/GIS Specialist USDA Forest Service, 359 Main Rd. Delaware OHIO 43015 USA Ph: 740-368-0103 Email: aprasad at fs.fed.us Web: http://www.fs.fed.us/ne/delaware/index.html Don't Miss Climate Change Tree Atlas at: http://www.fs.fed.us/ne/delaware/atlas/index.html ****************************************************************** Peter Dalgaard BSA To: r-help at stat.math.ethz.ch, <p.dalgaard at bios aprasad at fs.fed.us tat.ku.dk> cc: Sent by: Subject: Re: [R] Where is gam? pd at blueberry.kub ism.ku.dk 10/03/00 10:20 AM Prof Brian Ripley <ripley at stats.ox.ac.uk> writes:> > I noticed that there is no generalised additive model functions in R > > (1.1.1) ... is there a package that implements them?..> If you just want additive models there is a much better > function bruto in package MDA. > > One reason that gam is not there is that some of us think it > is much over-sold as a technique and the way it is done in S is poorly > implemented.Also, note that the gss package by Chong Gu has much the same objective as gam(). -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
X-Mailer: Lotus Notes Release 5.0.2a November 23, 1999
From: "Anantha Prasad/NE/USDAFS" <aprasad at fs.fed.us>
Date: Tue, 3 Oct 2000 09:31:23 -0400
X-MIMETrack: Serialize by Router on ENTWOB/E/USDAFS(Release 5.0.3 |March 21,
2000) at 10/03/2000
09:30:01 AM
MIME-Version: 1.0
Content-type: text/plain; charset=us-ascii
Sender: owner-r-help at stat.math.ethz.ch
Precedence: bulk
I noticed that there is no generalised additive model functions in R
(1.1.1) ... is there a package that implements them?
Thanks
Prasad
*****************************************************************
Mr. Anantha Prasad, Ecologist/GIS Specialist
USDA Forest Service, 359 Main Rd.
Delaware OHIO 43015 USA
Ph: 740-368-0103 Email: aprasad at fs.fed.us
Web: http://www.fs.fed.us/ne/delaware/index.html
Don't Miss Climate Change Tree Atlas at:
http://www.fs.fed.us/ne/delaware/atlas/index.html
******************************************************************
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
You may try the gss package.
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Prof Brian Ripley <ripley at stats.ox.ac.uk> writes:> I don't think > that asking for a technical consultancy in response to the answer > to a question on software availability is fair. You have already > used up a lot of R core time with R questions.I had no idea that I was asking for technical consultancy - it's just that since you said that gam is over-sold and poorly implemented in S, I was curious to know more about what prompted you to say that - since I am sure many people (including me) use gam. My sincere apologies to the group if my question falls in that realm and if I have used up too much of R core time. I would like to thank you all again for those very helpful replies. Hopefully, I can implement a public domain version of the my application that'll be helpful to a few out there in order to repay my debt. I'll try to remain silent for a while. Prasad ***************************************************************** Mr. Anantha Prasad, Ecologist/GIS Specialist USDA Forest Service, 359 Main Rd. Delaware OHIO 43015 USA Ph: 740-368-0103 Email: aprasad at fs.fed.us Web: http://www.fs.fed.us/ne/delaware/index.html Don't Miss Climate Change Tree Atlas at: http://www.fs.fed.us/ne/delaware/atlas/index.html ****************************************************************** Prof Brian Ripley To: aprasad at fs.fed.us <ripley at stats cc: .ox.ac.uk> Subject: Re: [R] Where is gam? 10/03/00 11:12 AM Please respond to Prof Brian Ripley From: "Anantha Prasad/NE/USDAFS" <aprasad at fs.fed.us>> Date: Tue, 3 Oct 2000 10:24:10 -0400 > > > Prof Brian Ripley <ripley at stats.ox.ac.uk> writes: > > > One reason that gam is not there is that some of us think it > > is much over-sold as a technique and the way it is done in S is poorly > > implemented. > Could you please clarify.... there has been a lot of interest in gam > recently in predicting species distribution and environmental management > (there is in fact a workshop scheduled in Switzerland next year onGLM/GAM> modelling). So many of us would be very interested to learn about its(Not exactly topical then: 1970s and 1980s statistical techniques)> shortcomings (reason why you think it is oversold technique)...and also > what aspects of it are poorly implemented in S. > If you shed some light on this, it will be of great help to us.Take a look in the archives: we've been here before. I don't think that asking for a technical consultancy in response to the answer to a question on software availability is fair. You have already used up a lot of R core time with R questions. -- 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 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
> I noticed that there is no generalised additive model functions in R > (1.1.1) ... is there a package that implements them?I've produced a preliminary version of a package that fits gams implemented using 1-d penalized regression splines: it includes a version of gam(). Smoothing parameter selection is done on all terms simultaneously using GCV. It's available from: http://www.ruwpa.st-and.ac.uk/simon/mgcv.html ... the plan is to add multidimensional terms fairly soon, and to make it more widely available after a bit more testing. It's only advantages over gss is speed on large datasets and the fact that it provides a version of gam(). I don't know what other people think, but my concerns about gams relate to the oddness of their structure. Often, modelling amounts to trying to find some unknown function linking your covariates to your response i.e. you want to find the f in: E(y) = f(x_1, x_2, x_3....) unless there's good prior reason to do so why would you start by letting: f(.) = f_1(x_1) + f_2 (x_2) + f_3 (x_3) +.... ? The point is maybe clearest by considering polynomial models. If you don't know f(.) it seems reasonable to use some taylor expansion of f(.) as a model... e.g. f(.) = a + b x_1 + c x_2 + d x_3 + e x_1 x_2 + g x_1 x_3 + h x_2 x_3 + p x_1^2 + q x_2^2 + r x_3^2 + ... Using gams is rather like deleting all mixed terms in this expansion (i.e. e x_1 x_2, g x_1 x_3 etc). This seems like a very odd thing to do and gets wierder the more degrees of freedom you allow the model. Simon ______________________________________________________________________> Simon Wood snw at st-and.ac.uk http://www.ruwpa.st-and.ac.uk/simon.html > The Mathematical Institute, North Haugh, St. Andrews, Fife KY16 9SS UK > Direct telephone: (0)1334 463799 Indirect fax: (0)1334 463748-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._