Extract the entry time, exit time, status or duration of follow-up from a
Lexis
object. Classify states.
entry(x, time.scale = NULL, by.id=FALSE)
exit(x, time.scale = NULL, by.id=FALSE)
status(x, at="exit" , by.id=FALSE)
dur(x, by.id=FALSE)
transient(x)
absorbing(x)
preceding(x, states)
before(x, states)
succeeding(x, states)
after(x, states)
The entry
and exit
functions return a vector of
entry times and exit times, respectively, on the requested time
scale. If multiple time scales are requested, a matrix is
returned.
The status
function returns a vector giving the status at
"at
" (either 'entry
' or 'exit
') and dur
returns a vector with the lengths of the follow-up intervals.
entry
, exit
, status
and dur
return vectors
of length nrow(x)
if by.id=FALSE
; if by.id=TRUE
a
vector of length length(unique(lex.id))
.
The functions transient
and absorbing
return character
vectors of the transient, resp. absorbing states in x
. These
are necessarily disjoint but the union may be a proper subset of
levels(x)
, since the latter may have levels that are never
assumed by either lex.Cst
or lex.Xst
.
preceding
returns a character vector with names of the states
of the Lexis object x
from which one of the states in
states
can be reached directly - the preceding
states. before
is just a synonym for preceding
.
succeeding
returns a character vector with names of the states
of the Lexis object x
that can be reached directly from one of
the states in states
. after
is just a synonym for
succeeding
.
an object of class Lexis
.
a string or integer indicating the time scale. If omitted, all times scales are used.
Logical, if TRUE
, only one record per unique value
of lex.id
is returned; either the first, the last, or for
dur
, the sum of lex.dur
. If TRUE
, the returned
object have the lex.id
as (row)names attribute.
string indicating the time point(s) at which status is to be measured. Possible values are "exit" or "entry".
Character vector of states.
Martyn Plummer & Bendix Carstensen
Lexis