1% Generated by roxygen2: do not edit by hand 2% Please edit documentation in R/Hist.R 3\name{Hist} 4\alias{Hist} 5\title{Create an event history response variable} 6\usage{ 7Hist(time, event, entry = NULL, id = NULL, cens.code = "0", 8 addInitialState = FALSE) 9} 10\arguments{ 11\item{time}{for right censored data a numeric vector of event times -- for 12interval censored data a list or a data.frame providing two numeric vectors 13the left and right endpoints of the intervals. See \code{Details}.} 14 15\item{event}{A vector or a factor that specifies the events that occurred at 16the corresponding value of \code{time}. Numeric, character and logical 17values are recognized. It can also be a list or a data.frame for the 18longitudinal form of storing the data of a multi state model -- see 19\code{Details}.} 20 21\item{entry}{Vector of delayed entry times (left-truncation) or list of two 22times when the entry time is interval censored.} 23 24\item{id}{Identifies the subjects to which multiple events belong for the 25longitudinal form of storing the data of a multi state model -- see 26\code{Details}.} 27 28\item{cens.code}{A character or numeric vector to identify the right 29censored observations in the values of \code{event}. Defaults to "0" which 30is equivalent to 0.} 31 32\item{addInitialState}{If TRUE, an initial state is added to all ids for the 33longitudinal input form of a multi-state model.} 34} 35\value{ 36An object of class \code{Hist} for which there are print and plot 37methods. The object's internal is a matrix with some of the following 38columns: \item{time}{ the right censored times} \item{L}{the left endpoints 39of internal censored event times} \item{R}{the right endpoints of internal 40censored event times} \item{status}{\code{0} for right censored, \code{1} 41for exact, and \code{2} for interval censored event times.} \item{event}{an 42integer valued numeric vector that codes the events.} \item{from}{an integer 43valued numeric vector that codes the \code{from} states of a transition in a 44multi state model.} \item{to}{an integer valued numeric vector that codes 45the \code{to} states of a transition in a multi state model.} 46 47Further information is stored in \code{\link{attributes}}. The key to the 48official names given to the events and the from and to states is stored in 49an attribute "states". 50} 51\description{ 52Functionality for managing censored event history response data. The 53function can be used as the left hand side of a formula: \code{Hist} serves 54\code{\link{prodlim}} in a similar way as \code{\link{Surv}} from the 55survival package serves `survfit'. \code{Hist} provides the suitable 56extensions for dealing with right censored and interval censored data from 57competing risks and other multi state models. Objects generated with 58\code{Hist} have a print and a plot method. 59} 60\details{ 61*Specification of the event times* 62 63If \code{time} is a numeric vector then the values are interpreted as right 64censored event times, ie as the minimum of the event times and the censoring 65times. 66 67If \code{time} is a list with two elements or data frame with two numeric 68columns The first element (column) is used as the left endpoints of interval 69censored observations and the second as the corresponding right endpoints. 70When the two endpoints are equal, then this observation is treated as an 71exact uncensored observation of the event time. If the value of the right 72interval endpoint is either \code{NA} or \code{Inf}, then this observation 73is treated as a right censored observation. Right censored observations can 74also be specified by setting the value of \code{event} to \code{cens.code}. 75This latter specification of right censored event times overwrites the 76former: if \code{event} equals \code{cens.code} the observation is treated 77as right censored no matter what the value of the right interval endpoint 78is. 79 80*Specification of the events* 81 82If \code{event} is a numeric, character or logical vector then the order of 83the attribute "state" given to the \code{value} of \code{Hist} is determined 84by the order in which the values appear. If it is a factor then the order 85from the levels of the factor is used instead. 86 87**Normal form of a multi state model** 88 89If \code{event} is a list or a data.frame with exactly two elements, then 90these describe the transitions in a multi state model that occurred at the 91corresponding \code{time} as follows: The values of the first element are 92interpreted as the \code{from} states of the transition and values of the 93second as the corresponding \code{to} states. 94 95**Longitudinal form of a multi state model** 96 97If \code{id} is given then \code{event} must be a vector. In this case two 98subsequent values of \code{event} belonging to the same value of \code{id} 99are treated as the \code{from} and \code{to} states of the transitions. 100} 101\examples{ 102 103 104## Right censored responses of a two state model 105## --------------------------------------------- 106 107Hist(time=1:10,event=c(0,1,0,0,0,1,0,1,0,0)) 108 109## change the code for events and censored observations 110 111Hist(time=1:10,event=c(99,"event",99,99,99,"event",99,"event",99,99),cens.code=99) 112 113TwoStateFrame <- SimSurv(10) 114SurvHist <- with(TwoStateFrame,Hist(time,status)) 115summary(SurvHist) 116plot(SurvHist) 117 118## Right censored data from a competing risk model 119## -------------------------------------------------- 120 121CompRiskFrame <- data.frame(time=1:10,event=c(1,2,0,3,0,1,2,1,2,1)) 122CRHist <- with(CompRiskFrame,Hist(time,event)) 123summary(CRHist) 124plot(CRHist) 125 126## Interval censored data from a survival model 127icensFrame <- data.frame(L=c(1,1,3,4,6),R=c(2,NA,3,6,9),event=c(1,1,1,2,2)) 128with(icensFrame,Hist(time=list(L,R))) 129 130## Interval censored data from a competing risk model 131with(icensFrame,Hist(time=list(L,R),event)) 132 133## Multi state model 134MultiStateFrame <- data.frame(time=1:10, 135 from=c(1,1,3,1,2,4,1,1,2,1), 136 to=c(2,3,1,2,4,2,3,2,4,4)) 137with(MultiStateFrame,Hist(time,event=list(from,to))) 138 139## MultiState with right censored observations 140 141MultiStateFrame1 <- data.frame(time=1:10, 142 from=c(1,1,3,2,1,4,1,1,3,1), 143 to=c(2,3,1,0,2,2,3,2,0,4)) 144with(MultiStateFrame1,Hist(time,event=list(from,to))) 145 146## Using the longitudinal input method 147MultiStateFrame2 <- data.frame(time=c(0,1,2,3,4,0,1,2,0,1), 148 event=c(1,2,3,0,1,2,4,2,1,2), 149 id=c(1,1,1,1,2,2,2,2,3,3)) 150 with(MultiStateFrame2,Hist(time,event=event,id=id)) 151 152} 153\author{ 154Thomas A. Gerds \email{tag@biostat.ku.dk}, Arthur Allignol 155\email{arthur.allignol@fdm.uni-freiburg.de} 156} 157\seealso{ 158\code{\link{plot.Hist}}, \code{\link{summary.Hist}}, 159\code{\link{prodlim}} 160} 161\keyword{survival} 162 163