funor_funom: Identifies and treats outliers in a two-way table

In Sielinski/vacuum: Tukey's Vacuum Cleaner

Description

FUNOR-FUNOM stands for FUll NOrmal Rejection-FUll NOrmal Modification.

The procedure treats a two-way (contingency) table for outliers by isolating residuals from the table's likely systemic effects, which are calculated from the table's grand, row, and column means.

The residuals are passed to separate rejection (FUNOR) and modification (FUNOM) procedures, which both depend upon FUNOP to identify outliers. As such, this procedure requires two sets of A and B parameters.

The procedure treats outliers by reducing their residuals, resulting in values that are much closer to their expected values (i.e., combined grand, row, and column effects).

Usage

funor_funom(x, A_r = 10, B_r = 1.5, A_m = 0, B_m = 1.5)

Arguments

x	Two-way table to treat for outliers
A_r	A for the FUNOR phase (see FUNOP for details)
B_r	B for the FUNOR phase slope
A_m	A for the FUNOM phase (A_m is usually 0)
B_m	B for the FUNOM phase

Value

A two-way table of the same size as x, treated for outliers.

References

Tukey, John W. "The Future of Data Analysis." The Annals of Mathematical Statistics, 33(1), 1962, pp 1-67. JSTOR, https://www.jstor.org/stable/2237638.

See Also

funop()

Examples

funor_funom(table_2)
which(funor_funom(table_2) != table_2)

Sielinski/vacuum documentation built on Sept. 15, 2020, 11:32 a.m.