1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/relative_poisson.R
\name{relpois_ag}
\alias{relpois_ag}
\title{Excess hazard Poisson model}
\usage{
relpois_ag(
formula,
data,
d.exp,
offset = NULL,
breaks = NULL,
subset = NULL,
piecewise = TRUE,
check = TRUE,
...
)
}
\arguments{
\item{formula}{a formula with the counts of events as the response.
Passed on to \code{glm}. May contain usage of the \code{offset()} function
instead of supplying the offset for the Poisson model via the argument
\code{offset}.}
\item{data}{an \code{aggre} object (an aggregated data set;
see \verb{[as.aggre]} and \verb{[aggre]})}
\item{d.exp}{the counts of expected cases. Mandatory.
E.g. \code{d.exp = EXC_CASES}, where \code{EXC_CASES} is a column in data.}
\item{offset}{the offset for the Poisson model, supplied as e.g.
\code{offset = log(PTIME)}, where \code{PTIME} is a subject-time
variable in data. Not mandatory, but almost always should be supplied.}
\item{breaks}{optional; a numeric vector of [a,b) breaks to specify
survival intervals over the follow-up time; if \code{NULL}, the
existing breaks along the mandatory time scale mentioned in \code{formula}
are used}
\item{subset}{a logical vector or condition; e.g. \code{subset = sex == 1};
limits the data before estimation}
\item{piecewise}{\code{logical}; if \code{TRUE}, and if any time scale
from data is used (mentioned) in the formula, the time scale is
transformed into a factor variable indicating intervals on the time scale.
Otherwise the time scale left as it is, usually a numeric variable.
E.g. if \code{formula = counts ~ TS1*VAR1}, \code{TS1} is transformed
into a factor before fitting model.}
\item{check}{\code{logical}; if \code{TRUE}, performs check on the
negativity excess cases by factor-like covariates in formula -
negative excess cases will very likely lead to non-converging model}
\item{...}{any other argument passed on to \verb{[stats::glm]} such as
\code{control} or \code{weights}}
}
\value{
A \code{relpois} object created using a custom Poisson family construct.
}
\description{
Estimate a Poisson Piecewise Constant Excess
Hazards Model
}
\examples{
## use the simulated rectal cancer cohort
data(sire, package = "popEpi")
sire$agegr <- cut(sire$dg_age, c(0,45,60,Inf), right=FALSE)
## create aggregated example data
fb <- c(0,3/12,6/12,1,2,3,4,5)
x <- lexpand(sire, birth = bi_date, entry = dg_date,
exit = ex_date, status=status \%in\% 1:2,
breaks = list(fot=fb),
pophaz=popmort, pp = FALSE,
aggre = list(agegr, fot))
## fit model using aggregated data
rpm <- relpois_ag(formula = from0to1 ~ fot + agegr, data = x,
d.exp = d.exp, offset = log(pyrs))
summary(rpm)
## the usual functions for handling glm models work
rpm2 <- update(rpm, . ~ fot*agegr)
anova(rpm, rpm2, test="LRT")
AIC(rpm, rpm2)
## other features such as residuals or predicting are not guaranteed
## to work as intended.
}
\seealso{
\verb{[lexpand]}, \verb{[poisson]}, \verb{[glm]}
Other main functions:
\code{\link{Surv}()},
\code{\link{rate}()},
\code{\link{relpois}()},
\code{\link{sir}()},
\code{\link{sirspline}()},
\code{\link{survmean}()},
\code{\link{survtab}()},
\code{\link{survtab_ag}()}
Other relpois functions:
\code{\link{RPL}},
\code{\link{relpois}()},
\code{\link{rpcurve}()}
}
\author{
Joonas Miettinen, Karri Seppa
}
\concept{main functions}
\concept{relpois functions}
|
