Skip to contents

Expand long-format data

Source: R/mmrm_predict.R
complete_visits.Rd

Expand long-format data to all subject and visit combinations

Usage

complete_visits(data, id, time, response = NULL, levels = NULL)

Arguments

data

(data.frame) data in long format

id

(character) name of the subject variable

time

(character) name of the visit variable. In the returned data this variable is a factor with levels levels.

response

(character) optional name of the response variable which is set to NA in added rows

levels

(vector) visit levels. Defaults to the levels of the visit variable if it is a factor, and otherwise its sorted unique values.

Value

data.frame with length(levels) rows per subject

Details

Returns a long-format data.frame with exactly one row per subject and visit (levels), sorted by subject (in order of first appearance) and visit. Rows that are not present in data are added by copying all columns from the previous existing row of the subject (last observation carried forward), or from the next existing row if there is no previous row. The response of added rows is set to NA.

Examples

d <- data.frame(id = c(1, 1, 2), visit = c(1, 2, 2), y = 1:3, x = 4:6)
complete_visits(d, "id", "visit", "y")
#>   id visit  y x
#> 1  1     1  1 4
#> 2  1     2  2 5
#> 3  2     1 NA 6
#> 4  2     2  3 6