aod (version 1.3.3)

splitbin: Split Grouped Data Into Individual Data

Description

The function splits grouped data and optional covariates into individual data. Two types of grouped data are managed by splitbin:

• Grouped data with weights;

• Grouped data of form cbind(success, failure).

When weights, successes or failures involve non-integer numbers, these numbers are rounded before splitting.

Usage

splitbin(formula, data, id = "idbin")

Value

A data frame built according to the formula and function used in the call.

Arguments

formula

A formula. The left-hand side represents grouped data. The right-hand side defines the covariates. See examples for syntax.

data

A data frame where all the variables described in formula are found.

id

An optional character string naming the identifier (= grouping factor). Default to “idbin”.

Examples

# Grouped data with weights
mydata <- data.frame(
    success = c(0, 1, 0, 1),
    f1 = c("A", "A", "B", "B"),
    f2 = c("C", "D", "C", "D"),
    n = c(4, 2, 1, 3)
    )
mydata
splitbin(formula = n ~ f1, data = mydata)$tab
splitbin(formula = n ~ f1 + f2 + success , data = mydata)$tab

# Grouped data of form "cbind(success, failure)"
mydata <- data.frame(
    success = c(4, 1),
    failure = c(1, 2),
    f1 = c("A", "B"),
    f2 = c("C", "D")
    )
mydata$n <- mydata$success + mydata$failure
mydata
splitbin(formula = cbind(success, failure) ~ 1, data = mydata)$tab
splitbin(formula = cbind(success, failure) ~ f1 + f2, data = mydata)$tab
splitbin(formula = cbind(success, n - success) ~ f1 + f2, data = mydata)$tab
splitbin(formula = cbind(success, n - 0.5 * failure - success) ~ f1 + f2,
         data = mydata)$tab