sdcMicro (version 3.1.2)

LocalRecProg: Local recoding via Edmond's maximum weighted matching algorithm

Description

To be used on both categorical and numeric input variables, although usage on categorical variables is the focus of the development of this software. Each record in the data represents a category of the original data, and hence all records in the input data should be unique by the N Input Variables. To achieve bigger category sizes (k-anoymity), one can form new categories based on the recoding result and repeatedly apply this algorithm.

Usage

LocalRecProg(data,categorical,numerical=NULL,ancestors=NULL,ancestor_setting=NULL,
    k_level=2, FindLowestK=TRUE, weight=NULL, lowMemory=FALSE, missingValue=NA)

Arguments

data

Input data

categorical

Names of categorical variables

numerical

Names of numerical variables

ancestors

Names of ancestors of the cateoorical variables

ancestor_setting

For each ancestor the corresponding categorical variable

k_level

Level for k-anonymity

FindLowestK

requests the program to look for the smallest k that results in complete matches of the data.

weight

A weight for each variable (Default=1)

lowMemory

Slower algorithm with less memory consumption

missingValue

The output value for a suppressed value.

Value

• dataframe with original variables and the supressed variables (suffix _lr).

References

http://www.stat.t.u-tokyo.ac.jp/~takemura/papers/localrec.pdf

Examples

# LocalRecProg
  data(testdata2)
  r1=LocalRecProg(testdata2,categorical=c("urbrur", "roof", "walls", "water", "sex", "relat"),missingValue=-99)
  r2=LocalRecProg(testdata2,categorical=c("urbrur", "roof", "walls", "water", "sex", "relat"),
    ancestor=c("water2", "water3", "relat2"),ancestor_setting=c("water","water","relat"),missingValue=-99)
  r3=LocalRecProg(testdata2,categorical=c("urbrur", "roof", "walls", "water", "sex", "relat"),
    ancestor=c("water2", "water3", "relat2"),ancestor_setting=c("water","water","relat"),missingValue=-99,FindLowestK=FALSE)