predict.face.sparse:
Subject-specific curve prediction from a face.sparse fit
Description
Predict subject-specific curves based on a fit from "face.sparse".
Usage
"predict"(object, newdata,...)
Arguments
object
a fitted object from the R function "face.sparse".
newdata
a data frame with three arguments:
(1) argvals: observation times;
(2) subj: subject indices;
(3) y: values of observations.
NA values are allowed in "y" but not in the other two.
...
further arguments passed to or from other methods.
Predicted/estimated objects at
the observation time points in newdata
scores
if calculate.scores in object is TRUE (typically FALSE),
then predicted scores will be calculated.
...
...
Details
This function makes prediction based on observed data for each subject. So for each subject,
it requires at least one observed data. For the time points prediction is desired but
no observation is available, just make the corresponding data$y as NA.
References
Luo Xiao, Cai Li, Will Checkley and Ciprian Crainiceanu, Fast covariance estimation
for sparse functional data, manuscript.