Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!
Get Started for Free

addComplement

Add Complement-items to Transactions


Description

Provides the generic function addComplement and the S4 methods for transactions. This function adds for given items complement items. That is it adds an artificial item to each transactions which does not contain the original item. Such items are also called negative items (Antonie et al, 2014).

Usage

addComplement(x, labels, complementLabels=NULL)

Arguments

x

an object of class transactions.

labels

character strings; item labels for which complements should be created.

complementLabels

character strings; labels for the artificial complement-items. If omitted then the original label is prepended by "!" to form the complement-item label.

Value

Returns an object of class transactions with complement-items added.

Author(s)

Michael Hahsler

References

Antonie L., Li J., Zaiane O. (2014) Negative Association Rules. In: Aggarwal C., Han J. (eds) Frequent Pattern Mining, Springer International Publishing, pp. 135-145. doi: 10.1007/978-3-319-07821-2_6

See Also

Examples

data("Groceries")

## add a complement-items for "whole milk" and "other vegetables"
g2 <- addComplement(Groceries, c("whole milk", "other vegetables"))
g2
tail(itemInfo(g2))
inspect(head(g2, 3))

## use a custom label for the complement-item
g3 <- addComplement(g2, "coffee", complementLabels = "NO coffee")
inspect(head(g2, 3))

## add complements for all items (this is excessive for this dataset)
g4 <- addComplement(Groceries, itemLabels(Groceries))
g4

## add complements for all items with a minimum support of 0.1
g5 <- addComplement(Groceries, names(which(itemFrequency(Groceries) >= 0.1)))
g5

arules

Mining Association Rules and Frequent Itemsets

v1.6-7
GPL-3
Authors
Michael Hahsler [aut, cre, cph], Christian Buchta [aut, cph], Bettina Gruen [aut, cph], Kurt Hornik [aut, cph], Ian Johnson [ctb, cph], Christian Borgelt [ctb, cph]
Initial release
2021-03-12

We don't support your browser anymore

Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.