implication.index: Compute the implication index of Gras (YEAR).

In trim: Tree methods for classification of imbalanced data

Description

The implication.index function computes the implication index statistics of Gras (YEAR).

Usage

implication.index(
  y,
  target = decision.rule.majority(y, quiet = quiet),
  h0 = y,
  quiet = T,
  ...
)

Arguments

y	a WeithedVariable.categorical, a factor or a character.
target	a character.
h0	a WeithedVariable.categorical, a factor, a character or a a Distribution.
quiet	a logical.
...	optional arguments (not used).

Details

FIXME: write the formula.

Value

A numeric, the value of the statistics.

Author(s)

Emmanuel Rousseaux

References

FIXME

See Also

implication.index.gain

Examples

ritschard.array <- array(
  data = c(50,5,5,40,5,8,6,12,10,0,50,6,14,30,2,10,18,2),
  dim = c(3,3,2),
  dimnames = list(
    'civil status' = c('married', "single", 'divorced/widowed'),
    'working sector' = c('primary', 'secondary', 'tertiary'), 
    'sex' = c('man', 'woman')
  )
)
ritschard.data <- expand.array(ritschard.array)
ritschard.data
names(ritschard.data)

distribution(ritschard.data[ritschard.data$sex == 'man','civil.status'])
distribution(ritschard.data[ritschard.data$sex == 'woman','civil.status'])

imp.m <- implication.index(
  y = ritschard.data[ritschard.data$sex == 'man','civil.status'],
  h0 = distribution(ritschard.data$civil.status),
  quiet = FALSE
)

trim documentation built on May 2, 2019, 5:36 p.m.