Nominal.Odds.Ratio: Given an input matrix, we can split it into smaller 2x2...

In RajeswaranV/vcdPlus: Summary and inference of categorical data and contingency tables

Description

Given an input matrix, we can split it into smaller 2x2 sub-matrix and then find the Nominal Odds Ratio for each sub-matrix (subtable) keeping a reference row and column which can be defined by the user. By default the last row and column is taken as the reference.

Usage

Nominal.Odds.Ratio(mat, Reference.Row.ID = NULL, Reference.Col.ID = NULL)

Arguments

mat	- matrix for which the sub-matrix is to be generated.
Reference.Row.ID	- The row to be used as reference. By default, the last row is taken as the reference.
Reference.Col.ID	- The column to be used as reference. By default, the last column is taken as the reference.

Details

This can be used as input to models. If any of the cells a,b,c or d is zero, a continuity correction of 0.5 is added to all the cells and the odds ratio is calculated. For nominal classification variables this set of basic 2x2 tables is defined in terms of a reference category, usually the cell (I,J). Then the 2x2 tables formed have in their upper diagonal cell the (i, j) cell of the initial table, for i= 1 to I -1, j=1 to J-1 and in the lower diagonal cell always the reference cell (I,J). The non-diagonal cells are the cells of the initial table that share one classification variable index with each diagonal cell, i.e., they are the cells (i,J) and (I, j).

Value

A of dataframes with

Odds.Ratio	Odds Ratio of the 2x2 subtable
a	The value of a taken from the input table
b	The value of b taken from the input table
c	The value of c taken from the input table
d	The value of d taken from the input table
Index1RC	Index reference from the input matrix for a (Row,Column)
Index2RC	Index reference from the input matrix for b (Row,Column)
Index3RC	Index reference from the input matrix for c (Row,Column)
Index4RC	Index reference from the input matrix for d (Row,Column)

References

[1] Kateri, Maria. Contingency Table Analysis. Springer New York, 2014.

See Also

Other IxJ Inference methods: Local.Odds.Ratio, Plot.heatmap, Reversal.point, Subtable.Odds.Ratio, generate.heatmap.matrix

Examples

##  Here we use the default last column as the reference.
mat=matrix(c(1:9), nrow=3, ncol=3)
Nominal.Odds.Ratio(mat,Reference.Row.ID=3, Reference.Col.ID=3 ) # Same as Nominal.Odds.Ratio(mat)
##  Now we change the reference row and column to 2
Nominal.Odds.Ratio(mat, 2, 2)

RajeswaranV/vcdPlus documentation built on May 27, 2019, 7:28 a.m.