eatTools (version 0.7.10)

checkBackgroundVariables: Internally needed function for consistency checks and data preparation.

Description

Function is necessary for eatRep and eatModel to verify input consistency and therefore exported to namespace.

Usage

checkBackgroundVariables (varList, len1 = NULL, overlap = NULL)

Value

There are no returns. If the input is inconsistent, the functions aborts with an error message.

Arguments

varList

A non-nested list of objects, where each individual object corresponds to the returned object of the existsBackgroundVariables function.

len1

The names of varList which elements must have length 1.

overlap

A list of variable groups which elements must overlap with a given frequency.

Details

The function checks whether certain variables or variable groups that are used for analyses within eatRep or eatModel fulfill certain properties. For example, there may only be one ID variable, but there must be several item variables (items). Item variables and other covariates must be disjoint, while variables for group differences must be a subset of the group variables.

Examples

if (FALSE) {
# borrow some data from the BIFIEsurvey package
data(data.timss1)
varGroups <- list(ID = "IDSTUD", wgt = c("TOTWGT","JKZONE"),
    covars = c("female", "books", "lang", "migant"), depVar = "ASMMAT")

# error if one variable does not occur in data due to typo
varList   <- lapply(varGroups, FUN = function (v) {
    existsBackgroundVariables (dat=data.timss1[[1]], variable=v)
})

# correction
varGroups <- list(ID = "IDSTUD", wgt = c("TOTWGT","JKZONE"),
    covars = c("female", "books", "lang", "migrant"), depVar = "ASMMAT")
varList   <- lapply(varGroups, FUN = function (v) {
    existsBackgroundVariables (dat=data.timss1[[1]], variable=v)
})

# however, only one weighting variable is allowed
check     <- checkBackgroundVariables (varList, len1 = c("ID", "wgt", "depVar"),
             overlap = NULL)

# only one of the covariates should be used for group differences
# difference variable must be included in covars
varGroups <- list(ID = "IDSTUD", wgt = "TOTWGT", covars = c("female", "books", "lang"),
             differencesFor = "migrant", depVar = "ASMMAT")
varList   <- lapply(varGroups, FUN = function (v) {
    existsBackgroundVariables (dat=data.timss1[[1]], variable=v)
})
check     <- checkBackgroundVariables (varList, len1 = c("ID", "wgt", "depVar"),
             overlap = list(v1 = list(vars = c("covars","differencesFor"), len = 1)))
}