Learn R Programming

refund (version 0.1-37)

predict.pffr: Prediction for penalized function-on-function regression

Description

Takes a fitted pffr-object produced by pffr() and produces predictions given a new set of values for the model covariates or the original values used for the model fit. Predictions can be accompanied by standard errors, based on the posterior distribution of the model coefficients. This is a wrapper function for predict.gam().

Usage

# S3 method for pffr
predict(object, newdata, reformat = TRUE, type = "link", se.fit = FALSE, ...)

Value

If type == "lpmatrix", the design matrix for the supplied covariate values in long format. If se == TRUE, a list with entries fit and se.fit containing fits and standard errors, respectively. If type == "terms" or "iterms" each of these lists is a list of matrices of the same dimension as the response for newdata

containing the linear predictor and its se for each term.

Arguments

object

a fitted pffr-object

newdata

A named list (or a data.frame) containing the values of the model covariates at which predictions are required. If no newdata is provided then predictions corresponding to the original data are returned. If newdata is provided then it must contain all the variables needed for prediction, in the format supplied to pffr, i.e., functional predictors must be supplied as matrices with each row corresponding to one observed function. See Details for more on index variables and prediction for models fit on irregular or sparse data.

reformat

logical, defaults to TRUE. Should predictions be returned in matrix form (default) or in the long vector shape returned by predict.gam()?

type

see predict.gam() for details. Note that type == "lpmatrix" will force reformat to FALSE.

se.fit

see predict.gam()

...

additional arguments passed on to predict.gam()

Author

Fabian Scheipl

Details

Index variables (i.e., evaluation points) for the functional covariates are reused from the fitted model object and cannot be supplied with newdata. Prediction is always for the entire index range of the responses as defined in the original fit. If the original fit was performed on sparse or irregular, non-gridded response data supplied via pffr's ydata-argument and no newdata was supplied, this function will simply return fitted values for the original evaluation points of the response (in list form). If the original fit was performed on sparse or irregular data and newdata was supplied, the function will return predictions on the grid of evaluation points given in object$pffr$yind.

See Also