get_current_miss: Calculate current errors across variables

In aaronrudkin/autumn: Fast, Modern, and Tidy-Friendly Iterative Raking

Calculate current errors across variables

Description

This function calculates the extent to which the data under the current weighting scheme diverges from the target proportions, according to any error function.

Usage

get_current_miss(data, target, weights = rep(1, nrow(data)),
  error_function = current_miss_linear)

Arguments

data	A data frame (tibble) or matrix containing the variables being targeted. Additional variables can be present in data, but no variables in target should be missing.
target	As with harvest, this can be either a list of named vectors, where each vector contains the levels of one variable and the names of the list are the names of the variables; or else a data frame (tibble) supplying in order the variable, level, and proportion for the target
weights	The current weights; if left blank, all units will be assigned weight 1 and this will show imbalance in the original sample.
error_function	Specification of error function (how we measure how far off a variable is from its intended result). Valid error_function choices are "linear" (default), "max", "squared", "mean", "maxsquared", and "meansquared". Users can also supply a closure (unquoted name of a function). For documentation about error functions, see harvest.

Value

A named vector containing the degree of error, whose names are the variables supplied in the order they were present in target

Examples

## Not run: 
# Can be used for data without weights to see initial error
get_current_miss(respondent_data, ns_target)

# Supply weights
result <- harvest(respondent_data, ns_target)
get_current_miss(respondent_data, ns_target, result$weights)

# By default, the error function adds the linear errors across levels of
# the target proportions.
get_current_miss(respondent_data, ns_target, error_function = "max")

## End(Not run)

aaronrudkin/autumn documentation built on Feb. 5, 2024, 6:08 p.m.