gettvc: Find the Most Recent Value of a Time-varying Covariate Before Each Observed Response

Description

gettvc finds the most recent value of a time-varying covariate before each observed response and possibly adds them to a list of other time-varying covariates. It compares the times of response observations with those of time-varying covariates to find the most recent observed time-varying covariate for each response. These are either placed in a new object of class, tvcov, added to an already existing list of matrices containing other time-varying covariates and a new object of class, tvcov, created, or added to an existing object of class, tvcov.

If there are response observation times before the first covariate time, the covariate for these times is set to zero.

Usage

gettvc(response, times=NULL, tvcov=NULL, tvctimes=NULL,
	oldtvcov=NULL, ties=TRUE)

Value

An object of class, tvcov, is returned containing the new time-varying covariate and, possibly, those in oldtvcov.

Arguments

response: A list of two column matrices with response values and times for each individual, one matrix or dataframe of response values, or an object of class, response (created by restovec).
times: When response is a matrix, a vector of possibly unequally spaced times for the response, when they are the same for all individuals or a matrix of times. Not necessary if equally spaced.
tvcov: A list of two column matrices with time-varying covariate values and corresponding times for each individual or one matrix or dataframe of such covariate values. Times need not be the same as for responses.
tvctimes: When the time-varying covariate is a matrix, a vector of possibly unequally spaced times for the covariate, when they are the same for all individuals or a matrix of times. Not necessary if equally spaced.
oldtvcov: A list of matrices with time-varying covariate values, observed at the event times in response, for each individual, or an object of class, tvcov. If not provided, a new object is created.
ties: If TRUE, when the response and covariate times are identical, the response depends on that new value (as in observational studies); if FALSE, only the next response depends on that value (for example, if the covariate is a new treatment just applied at that time).

Author

J.K. Lindsey and D.F. Heitjan

Examples

Run this code

if (FALSE) {
y <- matrix(rnorm(20), ncol=5)
resp <- restovec(y, times=c(1,3,6,10,15))
z <- matrix(rpois(20,5),ncol=5)
z
# create a new time-varying covariate object for the response
newtvc <- gettvc(resp, tvcov=z, tvctimes=c(1,2,5,12,14))
covariates(newtvc)
# add another time-varying covariate to the object
z2 <- matrix(rpois(20,5),ncol=5)
z2
newtvc2 <- gettvc(resp, tvcov=z2, tvctimes=c(0,4,5,12,16), oldtvc=newtvc)
covariates(newtvc2)
}