IPI: Probabilistic Index / Indicator for sets of variables

Description Usage Arguments Value

Description

Probabilistic Index / Indicator for sets of variables

Usage

1
2
IPI(yi, yj, method = c("allpairs", "paired"), res = c("sum", "mean", "all",
  "sumcount"))

Arguments

yi, yj

Variables for which the I/PI will be calculated. yi will always be at the left hand side of the unequality.

method

If "allpairs" (the default), then all combinations of items in yi and yj are used. Otherwise both are taken in the order specified (implying equal lengths)

res

Calculate the "sum" (default), "mean" (i.e. probabilistic index) of the combinations, or simply return "all" partial results. Finally, "sumcount" returns both the sum and the count.

Value

If res is "sum" or "mean", a single value, otherwise a data.frame with the follwing columns

yi

Left hand side of the combination.

yj

Right hand side of the combination.

pi

The probabilistic indicator for this combination.


pimold documentation built on May 2, 2019, 5:50 p.m.