Skip to contents

Rake a weight vector to known population margins

Source: R/subpop_design.R
morie_rake.Rd

Iterative proportional fitting. Each margin is enforced in turn and the cycle repeats until every margin is matched within tolerance. Raking is the usual way to make a sample agree with published counts for a small group when the joint distribution of those counts is not published, only the one-way totals.

Usage

morie_rake(data, margins, weights = NULL, max_iter = 50L, tol = 1e-08)

Arguments

data

A data frame containing the raking variables.

margins

A named list; each element is a named numeric vector of population totals for the levels of the corresponding column of data.

weights

Starting weights. Default NULL for unit weights.

max_iter

Maximum number of cycles. Default 50.

tol

Convergence tolerance on the largest absolute margin discrepancy. Default 1e-8.

Value

A list with weights, converged, iterations and max_discrepancy.

See also

morie_calibration_weights, Postrt

Examples

df <- data.frame(
  region = c("north", "north", "south", "south"),
  group = c("a", "b", "a", "b")
)
morie_rake(df, list(region = c(north = 60, south = 40),
                    group = c(a = 70, b = 30)))
#> $weights
#> [1] 42 18 28 12
#> 
#> $converged
#> [1] TRUE
#> 
#> $iterations
#> [1] 1
#> 
#> $max_discrepancy
#> [1] 0
#>