FRegSigCom (version 0.3.0)

pred.ff.interaction: Prediction for a linear FOF regression model with two-way interactions

Description

Make predition for new observations of functional predictors based on the fitted function-on-function interaction model in the output of cv.ff.interaction with given main and two-way interaction effects, or the model in the output of step.ff.interaction with selected main and two-way interaction effects.

Usage

pred.ff.interaction(fit.obj, X.test, t.y.test=NULL)

Arguments

fit.obj

the output object of the function cv.ff.interaction, or step.ff.interaction.

X.test

new observations of functional predictors. It is a list of the same length and the same structure as the input X for cv.ff.interaction or step.ff.interaction.

t.y.test

a vector of observation time points where values of predicted response curves are to be calculated. If t.y.test=NULL (default), t.y in cv.ff.interaction or step.ff.interaction will be used.

Value

A matrix containing the predicted values of response curves evaluated at t.y for the new observations. The number of rows of the matrix equals to the sample size of the new data set, and the number of columns equals to the length of t.y, the input in cv.ff.interaction or step.ff.interaction.

See Also

cv.ff.interaction, step.ff.interaction

Examples

# NOT RUN {
 #See the examples in cv.ff.interaction() and step.ff.interaction().
# }