universe: Create universe

In ttrodrigz/iterake: Tools for iterative raking

Description

Along with the data to be weighted, this combines one or more objects of class category to build the universe with known marginal proportions. It also checks and adjusts the proportions to account for missing values in the data.

Usage

universe(data, ..., N)

Arguments

data	Data frame containing data where weights are desired.
...	One or more output objects from category().
N	Size of universe. If supplied, expansion factor will be applied to weights after convergence.

Value

A list with special class universe.

Examples

data(demo_data)

universe(
    data = demo_data,
    
    category(
        name = "Sex",
        buckets = factor(
            x = levels(demo_data[["Sex"]]),
            levels = levels(demo_data[["Sex"]])
        ),
        targets = c(0.4, 0.5),
        sum.1 = TRUE
    ),

    category(
        name = "BirthYear",
        buckets = c(1986:1990),
        targets = rep(0.2, times = 5)
    ),
    
    category(
        name = "EyeColor",
        buckets = c("brown", "green", "blue"),
        targets = c(0.8, 0.1, 0.1)
    ),
    
    category(
        name = "HomeOwner",
        buckets = c(TRUE, FALSE),
        targets = c(3/4, 1/4)
    )
)

ttrodrigz/iterake documentation built on July 1, 2020, 7:46 a.m.