FRegSigCom (version 0.3.0)

getcoef.nonlinear: Get the estimated intercept and nonlinear functions in nonlinear function-on-function model

Description

This function is used to calculate the estimates for $\mu (t),F_i(x,s,t{)}^{\prime }s$ based on the object obtained from cv.nonlinear.

Usage

getcoef.nonlinear(fit.cv, n.x.grid = 50)

Arguments

fit.cv

the object obtained from cv.nonlinear.

n.x.grid

the number of grid points of $x$. The estimated $F_i(x,s,t)$ is calculated in a three-dimensional grid of $(x,s,t)$. The grid points of $s$ and $t$ are the observation points of $X_i(s)$ and $Y(t)$ used in cv.nonlinear, respectively. The grid of $x$ includes n.x.grid equally spaced values between the minimum and maximum of all the discretely observed values of $X_i(s)$. Default of n.x.grid is 50.

Value

a list containing

mu

the vector of estimated values of $\mu (t)$ at the observation points of the response function.

F

a list of length $p$, the number of functional predictors. Its $i$-th element is a three dimensional array with estimated values of $F_i(x,s,t)$ on the three-dimensional grid X.grid[[i]]*t.x.list[[i]]*t.y (see below).

X.grid

a list of length $p$. Its $i$-th element is the vector of grid points for $x$ and includes n.x.grid equally spaced values between the minimum and maximum of all the discretely observed values of $X_i(s)$.

t.x.list

one of the arguments in cv.nonlinear, specifying the list of the vectors of obesrvation points for $X_i(s)$, $1\le i\le p$.

t.y

one of the arguments in cv.nonlinear, specifying the vector of obesrvation points of the response curve $Y(t)$.

%% ~Describe the value returned %% If it is a LIST, use %% \item{comp1 }{Description of 'comp1'} %% \item{comp2 }{Description of 'comp2'} %% ...

See Also

cv.nonlinear.

Examples

# NOT RUN {
#See the examples in cv.nonlinear().
# }