carts (version 0.1.0)

aggrsurv: Aggregate data in counting process format

Description

Aggregate data in counting process format. The aggregation is done within subject only.

Usage

aggrsurv(
  data,
  breaks,
  entry = "entry",
  exit = "exit",
  status = "status",
  id = "id",
  names = c("episode", "entry", "exit", "events", "exposure"),
  ...
)

Value

data.table

Arguments

data

data.frame

breaks

vector of time points

entry

name of entry date variable

exit

name exit date variable

status

censoring / event variable

id

id variable

names

character vector of names of new variables

...

additional arguments to lower level functions

Examples

dat <- data.table::data.table(
  id = c(1, 1, 1, 1, 2, 2, 2),
  entry = as.Date(c(
    "2021-01-01", "2021-01-20", "2021-02-28", "2021-06-01",
    "2021-01-01", "2021-01-14", "2021-09-01"
 )),
  status = c(1, 1, 1, 1, 1, 1, 0),
  x = rnorm(7)
)
dat[, exit := data.table::shift(entry, 1, type="lead"), by=id]
dat[, exit := replace(exit, .N, as.Date("2021-12-31")),
     by = id]

res <- aggrsurv(dat,
  breaks = c(182),
  entry = "entry", exit = "exit", status = "status", id = "id"
)
print(res)