independent-dot-contributions-dot-formula: Generate class labels by independent contributions of two...

In ddgraph: Distinguish direct and indirect interactions with Graphical Modelling

Description

Generate class labels by using the readout mechanism. Logical formula is applied to two variables which are read out from the real data using the var1 and var2 probabilities. This only works with binary variables.

Usage

independent.contributions.formula(data, var1, var2, var1.prob1, var1.prob0, var2.prob1, var2.prob0,
    logical.formula, false.neg=0, false.pos=0)

Arguments

data	a matrix or data.frame containing binary observations (columns are variables)
var1	index or name of the first variable
var2	index or name of the second variable
var1.prob1	the conditional probability P(class labels = 1|var1=1)
var1.prob0	the conditional probability P(class labels = 1|var1=0)
var2.prob1	the conditional probability P(class labels = 1|var2=1)
var2.prob0	the conditional probability P(class labels = 1|var2=0)
logical.formula	logical formula to apply
false.neg	a false negative probability
false.pos	a false positive probability

Value

a binary vector containing the class labels

Examples

# noisy OR function with 0.1 probability of error for reading "a" and "b" (error in both 1 and 0)
data <- cbind("a"=c(0,0,1,1), "b"=c(0,1,0,1))
independent.contributions.formula(data, "a", "b", 0.9, 0.1, 0.9, 0.1, "a | b")

ddgraph documentation built on Nov. 17, 2017, 10:50 a.m.