Weighted contingency coefficients

Share:

Description

Compute (weighted) pairwise contingency coefficients.

Usage

1

Arguments

x

for the default method, a vector that can be interpreted as factor. For the matrix and data.frame methods, the columns should be interpretable as factors.

...

for the generic function, arguments to be passed down to the methods, otherwise ignored.

Details

The function tableWt is used for the computation of the corresponding pairwise contingency tables. The following methods are implemented:

  • contingencyWt.default(x, y, weights = NULL, ...)

  • contingencyWt.matrix(x, weights = NULL, ...)

  • contingencyWt.data.frame(x, weights = NULL, ...)

Additional parameters are:

  • y: a vector that can be interpreted as factor (for the default method)

  • weights: an optional numeric vector containing sample weights

Value

For the default method, the (weighted) contingency coefficient of x and y.

For the matrix and data.frame method, a matrix of (weighted) pairwise contingency coefficients for all combinations of columns. Elements below the diagonal are NA.

Author(s)

Andreas Alfons and Stefan Kraft

References

Kendall, M.G. and Stuart, A. (1967) The Advanced Theory of Statistics, Volume 2: Inference and Relationship. Charles Griffin & Co Ltd, London, 2nd edition.

See Also

tableWt

Examples

1
2
3
4
5
6
7
8
data(eusilcS)

## default method
contingencyWt(eusilcS$pl030, eusilcS$pb220a, weights = eusilcS$rb050)

## data.frame method
basic <- c("age", "rb090", "hsize", "pl030", "pb220a")
contingencyWt(eusilcS[, basic], weights = eusilcS$rb050)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.