gam.outer.Rd 2.75 KB
 Dirk Eddelbuettel committed Apr 10, 2018 1 2 3 4 5 \name{gam.outer} \alias{gam.outer} %- Also NEED an \alias' for EACH other topic documented here. \title{Minimize GCV or UBRE score of a GAM using outer' iteration} \description{Estimation of GAM smoothing parameters is most stable if  Dirk Eddelbuettel committed Apr 10, 2018 6 7 optimization of the smoothness selection score (GCV, GACV, UBRE/AIC, REML, ML etc) is outer to the penalized iteratively  Dirk Eddelbuettel committed Apr 10, 2018 8 9 10 re-weighted least squares scheme used to estimate the model given smoothing parameters.  Dirk Eddelbuettel committed Apr 10, 2018 11 12 This routine optimizes a smoothness selection score in this way. Basically the score is evaluated for each trial set of smoothing parameters by  Dirk Eddelbuettel committed Apr 10, 2018 13 estimating the GAM for those smoothing parameters. The score is minimized  Dirk Eddelbuettel committed Apr 10, 2018 14 15 16 w.r.t. the parameters numerically, using \code{newton} (default), \code{bfgs}, \code{optim} or \code{nlm}. Exact (first and second) derivatives of the score can be used by fitting with \code{\link{gam.fit3}}. This  Dirk Eddelbuettel committed Apr 10, 2018 17 improves efficiency and reliability relative to relying on finite  Dirk Eddelbuettel committed Apr 10, 2018 18 19 20 21 22 difference derivatives. Not normally called directly, but rather a service routine for \code{\link{gam}}. } \usage{  Dirk Eddelbuettel committed Apr 10, 2018 23 gam.outer(lsp,fscale,family,control,method,optimizer,  Dirk Eddelbuettel committed Apr 10, 2018 24  criterion,scale,gamma,G,start=NULL,...)  Dirk Eddelbuettel committed Apr 10, 2018 25 26 27 28 29 } %- maybe also usage' for other objects documented here. \arguments{ \item{lsp}{The log smoothing parameters.}  Dirk Eddelbuettel committed Apr 10, 2018 30 \item{fscale}{Typical scale of the GCV or UBRE/AIC score.}  Dirk Eddelbuettel committed Apr 10, 2018 31 32 33 34 35 36  \item{family}{the model family.} \item{control}{control argument to pass to \code{\link{gam.fit}} if pure finite differencing is being used.}  Dirk Eddelbuettel committed Apr 10, 2018 37 38 39 40 41 42 43 44 \item{method}{method argument to \code{\link{gam}} defining the smoothness criterion to use (but depending on whether or not scale known).} \item{optimizer}{The argument to \code{\link{gam}} defining the numerical optimization method to use. } \item{criterion}{Which smoothness selction criterion to use. One of \code{"UBRE"}, \code{"GCV"}, \code{"GACV"}, \code{"REML"} or \code{"P-REML"}. } \item{scale}{Supplied scale parameter. Positive indicates known.}  Dirk Eddelbuettel committed Apr 10, 2018 45   Dirk Eddelbuettel committed Apr 10, 2018 46 \item{gamma}{ The degree of freedom inflation factor for the GCV/UBRE/AIC score.}  Dirk Eddelbuettel committed Apr 10, 2018 47   Dirk Eddelbuettel committed Apr 10, 2018 48 \item{G}{List produced by \code{mgcv:::gam.setup}, containing most of what's  Dirk Eddelbuettel committed Apr 10, 2018 49 needed to actually fit a GAM.}  Dirk Eddelbuettel committed Apr 10, 2018 50 \item{start}{starting parameter values.}  Dirk Eddelbuettel committed Apr 10, 2018 51 \item{...}{other arguments, typically for passing on to \code{gam.fit3} (ultimately).}  Dirk Eddelbuettel committed Apr 10, 2018 52 53 } \details{  Dirk Eddelbuettel committed Apr 10, 2018 54 See Wood (2008) for full details on outer iteration'.  Dirk Eddelbuettel committed Apr 10, 2018 55 56 57 } \references{  Dirk Eddelbuettel committed Apr 10, 2018 58 59 60 Wood, S.N. (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. Journal of the Royal Statistical Society (B) 73(1):3-36  Dirk Eddelbuettel committed Apr 10, 2018 61   Dirk Eddelbuettel committed Apr 10, 2018 62 \url{http://www.maths.bris.ac.uk/~sw15190/}  Dirk Eddelbuettel committed Apr 10, 2018 63 64 65 66 67  } \author{ Simon N. Wood \email{simon.wood@r-project.org}}  Dirk Eddelbuettel committed Apr 10, 2018 68 \seealso{\code{\link{gam.fit3}}, \code{\link{gam}}, \code{\link{magic}}}  Dirk Eddelbuettel committed Apr 10, 2018 69 70 71 72  \keyword{models} \keyword{smooth} \keyword{regression}%-- one or more ...