fda.usc (version 0.9.4)

fregre.gsam: Fitting Functional Generalized Spectral Additive Models

Description

Computes functional regression between functional (and non functional) explanatory variables and scalar response using basis representation.

Usage

fregre.gsam(formula, family = gaussian(),data=list(),
weights=NULL,basis.x=NULL,basis.b=NULL,CV=FALSE,...)

Arguments

formula

an object of class formula (or one that can be coerced to that class): a symbolic description of the model to be fitted. The details of model specification are given under Details.

family

a description of the error distribution and link function to be used in the model. This can be a character string naming a family function, a family function or the result of a call to a family function. (See family

data

List that containing the variables in the model.

weights

weights

basis.x

List of basis for functional explanatory data estimation.

basis.b

List of basis for functional beta parameter estimation.

CV

=TRUE, Cross-validation (CV) is done.

...

Further arguments passed to or from other methods.

Value

• Return gam object plus:
• basis.x
  Basis used for fdata or fd covariates.
• basis.b
  Basis used for beta parameter estimation.
• beta.l
  List of estimated beta parameter of functional covariates.
• data
  List that containing the variables in the model.
• formula
  formula.
• CV
  $y.pred predicted response by cross-validation.

Details

The first item in the data list is called "df" and is a data frame with the response and non functional explanatory variables, as gam. Functional covariates of class fdata or fd are introduced in the following items in the data list. basis.x is a list of basis for represent each functional covariate. The basis object can be created by the function: create.pc.basis, pca.fd create.pc.basis, create.fdata.basis o create.basis. basis.b is a list of basis for represent each functional beta parameter. If basis.x is a list of functional principal components basis (see create.pc.basis or pca.fd) the argument basis.b is ignored.

References

Ramsay, James O., and Silverman, Bernard W. (2006), Functional Data Analysis, 2nd ed., Springer, New York. Muller HG and Stadtmuller U. (2005). Generalized functional linear models. Ann. Statist.33 774-805. Ramsay, James O., and Silverman, Bernard W. (2006), Functional Data Analysis, 2nd ed., Springer, New York. McCullagh and Nelder (1989), Generalized Linear Models 2nd ed. Chapman and Hall. Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S, New York: Springer. Wood (2001) mgcv:GAMs and Generalized Ridge Regression for R. R News 1(2):20-25

See Also

See Also as: predict.fregre.gsam and summary.gam. Alternative methods: fregre.glm.

Examples

data(tecator)
x=tecator$absorp.fdata
#tecator$y$Fat=floor(tecator$y$Fat)
x.d1<-fdata.deriv(x)
tt=x[["argvals"]]
dataf=as.data.frame(tecator$y)
nbasis.x=11;nbasis.b=7
basis1=create.bspline.basis(rangeval=range(tt),nbasis=nbasis.x)
basis2=create.bspline.basis(rangeval=range(tt),nbasis=nbasis.b)
f=Fat~s(Protein)+s(x)
basis.x=list("x"=basis1,"x.d1"=basis1)
basis.b=list("x"=basis2,"x.d1"=basis2)
ldata=list("df"=dataf,"x"=x,"x.d1"=x.d1)
res=fregre.gsam(Fat~Water+s(Protein)+x+s(x.d1),ldata,family=gaussian(),basis.x=basis.x,basis.b=basis.b)
res2=fregre.gsam(Fat~te(Protein,k=3)+x,data=ldata,family=gaussian())

res=fregre.gsam(Fat~s(Water)+x.d1,data=ldata,family=gaussian(),basis.x=basis.x,basis.b=basis.b)

#gam(Fat~s(Protein,k=3),data=dataf)

#res$beta.l$x$coefs
res2=fregre.gsam(Fat~te(Protein)+x,data=ldata,family=gaussian())
#res2$beta.l$x$coefs
a<-predict.fregre.gsam(res)


# dropind basis pc
basis.pc0=create.pc.basis(x,c(1:7))
basis.pc1=create.pc.basis(x.d1,c(1:7))
basis.x=list("x"=basis.pc0,"x.d1"=basis.pc1)
ldata=list("df"=dataf,"x"=x,"x.d1"=x.d1)
#f.pc=Fat~Proteins+s(x) #
#res.pc=fregre.gsam(Fat~Protein+s(x),data=ldata,family=gaussian(),basis.x=basis.x,basis.b=basis.b)
#summary(res.pc)

########################
# clase fd
data(tecator)
ind<-1:129
x=tecator$absorp.fdata[ind]
y=tecator$y$Fat[ind]
tt=x[["argvals"]]
dataf=as.data.frame(tecator$y[ind,])

x.fd<-fdata2fd(x)
x.fd1<-fdata.deriv(x,class.out="fd")
x.fd2<-fdata.deriv(x,class.out="fd",nderiv=2)
basis.b=list("x.fd"=basis2,"x.fd1"=basis2)
ldata=list("df"=dataf,"x.fd"=x.fd,"x.fd1"=x.fd1,"x.fd2"=x.fd2)
f0=Fat~+x.fd+s(x.fd1)
res.pc=fregre.gsam(f0,data=ldata,family=gaussian())
res.pc2=fregre.gsam(f0,data=ldata,family=gaussian(),basis.b=basis.b)

#pca.fd
basis.pc0=pca.fd(x.fd,3)
basis.pc1=pca.fd(x.fd1,3)
basis.pc2=pca.fd(x.fd2,3)
basis.x=list("x.fd"=basis.pc0,"x.fd1"=basis.pc1)
f0=Fat~+x.fd+s(x.fd1)+s(Protein)
res.fd=fregre.gsam(f0,data=ldata,basis.x=basis.x)


data(tecator)
x=tecator$absorp.fdata
tecator$y$Fat<-ifelse(tecator$y$Fat>20,1,0)
x.d1<-fdata.deriv(x)
tt=x[["argvals"]]
dataf=as.data.frame(tecator$y)
nbasis.x=11;nbasis.b=7
basis1=create.bspline.basis(rangeval=range(tt),nbasis=nbasis.x)
basis2=create.bspline.basis(rangeval=range(tt),nbasis=nbasis.b)
f=Fat~s(Protein)+s(x)
basis.x=list("x"=basis1,"x.d1"=basis1)
basis.b=list("x"=basis2,"x.d1"=basis2)
ldata=list("df"=dataf,"x"=x,"x.d1"=x.d1)
res=fregre.gsam(Fat~Protein+s(x),ldata,family=binomial(),basis.x=basis.x,basis.b=basis.b)
res2=fregre.gsam(Fat~s(Protein)+x,data=ldata,family=binomial())

res=fregre.gsam(f,data=ldata,family=poisson(),basis.x=basis.x,basis.b=basis.b)
#res=fregre.gsam(Fat~Protein+x,data=ldata,family=gaussian()) no functional variables
res=fregre.gsam(Fat~s(Protein)+s(Water)+s(x)+x.d1,data=ldata,family=gaussian())
summary(res)
plot(res)
res=fregre.gsam(Fat~Protein+s(x),data=ldata,family=gaussian())
res=fregre.gsam(Fat~s(Protein)+x,data=ldata,family=gaussian())
res=fregre.gsam(f,ldata,family=gaussian(),basis.x=basis.x,basis.b=basis.b)