riskset: Incident/Dynamic (I/D) ROC curve, AUC and integrated AUC (iAUC)
estimation of censored survival data
Description
This function creates risk set at each unique failure time from a
survival data set.
Usage
riskset(dat, entry=FALSE)
Arguments
dat
survival dataset with at least three variables:
survival.times, survival.status and marker, in that order. The
survival data set may have additional variables. In case of interval
censored data, the first four columns are: entry time, exit time,
status at exit and marker
entry
default is FALSE indicating right censored data. TRUE if
left truncated data
Value
Returns a new data set with columns as follows: start, finish,
newStatus and other variables from the original dataset except
survival time and status. The first two columns correspond to the
start and end of time intervals considered and the newStatus
corresponds to the survival status of the patient corresponding to
this interval, i.e. the status is 1 if the patient had event during
this interval (start, finish] and 0 otherwise. Note that the survival time need to be
in ascending order.
Details
This function creates risk set at each unique failure time from
a survival data set and is needed for llCoxReg(). The function can
handle both right censored and interval censored data.
References
Heagerty, P.J., Zheng Y. (2005)
Survival Model Predictive Accuracy and ROC curves
Biometrics, 61, 92 -- 105
# NOT RUN {library(MASS)
data(VA)
## need to order the data in ascending order of survival timenew.VA=VA[order(VA$stime),]
risket.VA=riskset(new.VA)
# }