get.ff.coeffs {robflreg}

Get the estimated bivariate regression coefficient functions for function-on-function regression model

Description

This function is used to obtain the estimated bivariate regression coefficient functions \beta_m(s,t) for function-on-function regression model (see the description in rob.ff.reg based on output object obtained from rob.ff.reg).

Usage

get.ff.coeffs(object)

Arguments

object

The output object of rob.ff.reg.

Details

In the estimation of bivariate regression coefficient functions, the estimated functional principal components of response \hat{\Phi}(t) and predictor \hat{\Psi}_m(s) variables and the estimated regression parameter function obtained from the regression model between the principal component scores of response and predictor variables \hat{B} are used, i.e., \hat{\beta}_m(s,t) = \hat{\Psi}_m^\top(s) \hat{B} \hat{\Phi}(t).

Value

A list object with the following components:

vars

A numeric vector specifying the indices of functional predictors used in the function-on-function regression model rob.ff.reg.

gpY

A vector containing the grid points of the functional response Y(t).

gpX

A list with length M. The m-th element of gpX is a vector containing the grid points of the m-th functional predictor X_m(s).

coefficients

A list with length M. The m-th element of coefficients is a matrix of the estimated values of the coefficient function for the m-th functional predictor X_m(s).

Author(s)

Ufuk Beyaztas and Han Lin Shang

Examples

sim.data <- generate.ff.data(n.pred = 5, n.curve = 200, n.gp = 101)
Y <- sim.data$Y
X <- sim.data$X
gpY = seq(0, 1, length.out = 101) # grid points of Y
gpX <- rep(list(seq(0, 1, length.out = 101)), 5) # grid points of Xs
model.fit <- rob.ff.reg(Y, X, model = "full", emodel = "classical", 
                        gpY = gpY, gpX = gpX)
coefs <- get.ff.coeffs(model.fit)