summarizeNA: Summarize missing data patterns
In bozenne/repeated: Repeated Measurement Models for Discrete Times

summarizeNA

R Documentation

Summarize missing data patterns

Description

Summarize missing data patterns.

Usage

summarizeNA(
  data,
  formula,
  repetition = NULL,
  sep = "",
  newnames = c("variable", "frequency", "missing.pattern", "n.missing"),
  filter = NULL,
  keep.data = TRUE
)

Arguments

`data`	[data.frame] dataset containing the observations.
`formula`	[formula] On the left hand side the variable(s) for which the missing data patterns should be evaluated and on the right hand side the grouping variables. E.g. Y1 ~ Gender will compute missing data pattern w.r.t Y1 for each gender.
`repetition`	[formula] Specify the structure of the data when in the long format: the time/repetition variable and the grouping variable, e.g. ~ time\|id. When specified the missing data pattern is specific to each variable not present in the formula.
`sep`	[character] character used to separate the missing data indicator (0/1) when naming the missing data patterns.
`newnames`	[character vector of length 4] additional column containing the variable name (only when argument `repetition` is used), variables w.r.t. which missing data patterns are identified, frequency of the missing data pattern in the dataset, name of the missing data pattern in the dataset, and number of missing data per pattern.
`filter`	[character] a regular expression passed to `grep` to filter the columns of the dataset. Relevant when using `.` to indicate all other variables.
`keep.data`	[logical] should the indicator of missing data per variable in the original dataset per pattern be output.

Value

a data frame

Examples

#### display missing data pattern (wide format) ####
data(gastricbypassW, package = "LMMstar")
e.SNA <- summarizeNA(gastricbypassW)
e.SNA
plot(e.SNA)

## only focus on some variables
eG.SNA <- summarizeNA(gastricbypassW, filter = "glucagon")
eG.SNA
plot(eG.SNA)
summarizeNA(weight3+glucagonAUC3 ~ 1, data = gastricbypassW)

#### display missing data pattern (long format) ####
## example 1 (single group)
data(gastricbypassL, package = "LMMstar")
e.SNAL <- summarizeNA(gastricbypassL, repetition = ~time|id)
e.SNAL
plot(e.SNAL, variable = "glucagonAUC")

## example 2 (two groups)
data(calciumL, package = "LMMstar")

## over both groups
mp <- summarizeNA(calciumL, repetition = ~visit|girl)
plot(mp, variable = "bmd")
plot(mp, variable = "bmd", order.pattern = "frequency")
plot(mp, variable = "bmd", order.pattern = 5:1)

## per group
mp2 <- summarizeNA(bmd ~ grp, data = calciumL, repetition = ~visit|girl)
mp2
plot(mp2)

## artificially create different patterns in each group
calciumL2 <- calciumL[order(calciumL$girl),]
calciumL2[calciumL2$girl == 101,"bmd"] <- c(NA,NA,1,1,1)
calciumL2[calciumL2$girl == 104,"bmd"] <- c(NA,1,NA,1,NA)
mp3 <- summarizeNA(bmd ~ grp, data = calciumL2, repetition = ~visit|girl)
mp3
plot(mp3)
plot(mp3, order.pattern = "n.missing")
plot(mp3, order.pattern = "frequency")

bozenne/repeated documentation built on July 16, 2025, 11:16 p.m.