Skip to content

overview.md

Latest commit

 

History

History
52 lines (33 loc) · 1.47 KB

overview.md

File metadata and controls

52 lines (33 loc) · 1.47 KB

hideR

Data Transformations

This package is not intended to fully automate the anonymisation process, but rather to improve the identification of crucial cases.

Each variable will be converted to a factor variable (i.e. a categorical variable).

A threshold for the number of levels must be specified. A default setting of 15 will be used, but this may not be appropriate for all data types. (e.g. geographical regions - in the US, there are 54 federal geographic entities, including DC, Puerto Rico and Guam)

The program is envisaged to have a "keep exception" handling whereby the variable is not removed. The variable must be specified by the users.



keepList = c("stateNames","stateIds") 

myNewData = getFactors(myData, threshold = 10, ignore = KeepList)

Subsequent commands will perform this step automatically.

1. Case Report

Parameters

  • T Lowest Cell Size Threshold
    • default value of 30.
  • L Report Length
    • default value of 30 cases

The case report (caseReport) returns a table containing the frequency of how many times each particular cell is in a category of size less than the Threshold.



> caseReport(myData, threshold = 30, print=5, ignore = KeepList)

1007	  9
1334	  8
1456	  8
209       5
1341      4

2. Variable Report

  • Reports the lowest cell size for frequency tables
  • Reports how many levels have frequencies less than the threshold.
  • The default report size is 30.