return_ids_with_LOCF: Select individuals in a dataset with a last observation...

In isobelbarrott/Landmarking: Analysis using Landmark Models

Select individuals in a dataset with a last observation carried forward (LOCF) at a landmark time

Description

To fit the LOCF model, all individuals must have at least one non-NA entry by landmark time x_L for all covariates. This function selects these individuals and removes the other rows.

Usage

return_ids_with_LOCF(
  data_long,
  individual_id,
  x_L,
  covariates,
  covariates_time
)

Arguments

data_long	Data frame with repeated measurements data in long format
individual_id	Character string specifying the column name in data_long which contains the individual identifiers
x_L	Numeric specifying the landmark time(s)
covariates	Vector of character strings specifying the column names in data_long which correspond to the covariates
covariates_time	Vector of character strings specifying the column names in data_long which contain the times at which covariates were recorded. This should either be length 1 or the same length as covariates. In the latter case the order of elements must correspond to the order of elements in covariates.

Details

Individuals have a LOCF if there is a non-NA entry for each of the covariates in covariates up until (not including) time x_L.

Value

List of data frames which correspond to each landmark time x_L. Each data frame is an updated version of data_long which contains only rows of individuals with a LOCF at age x_L, other rows are removed.

Author(s)

Isobel Barrott [e-mail]

Examples

library(Landmarking)
data(data_repeat_outcomes)
data_repeat_outcomes <-
  return_ids_with_LOCF(
    data_long = data_repeat_outcomes,
    individual_id = "id",
    covariates =
      c("ethnicity", "smoking", "diabetes", "sbp_stnd", "tchdl_stnd"),
    covariates_time =
      c(rep("response_time_sbp_stnd", 4), "response_time_tchdl_stnd"),
    x_L = c(60,61)
  )

isobelbarrott/Landmarking documentation built on Nov. 22, 2022, 4:50 a.m.