varDataAccess: Differentially Private Variance Data Access Function

In DPpack: Differentially Private Statistical Analysis and Machine Learning

Differentially Private Variance Data Access Function

Description

This function performs the data access step in the computation of a differentially private variance. The true values are computed using var, while the sensitivities are calculated based on bounded and unbounded differential privacy \insertCiteKifer2011DPpack according to the theoretical values \insertCiteLiu2019bDPpack. For the variance, the sensitivities based on bounded and unbounded differential privacy are identical, so only one value is returned.

Usage

varDataAccess(x, lower.bound, upper.bound)

Arguments

x	Dataset whose variance is desired.
lower.bound	Scalar representing the global or public lower bound on values of x.
upper.bound	Scalar representing the global or public upper bound on values of x.

Value

List of the true variance and the sensitivity calculated based on theoretical values.

References

\insertRef

Liu2019bDPpack

\insertRef

Kifer2011DPpack

Examples

varDataAccess(c(1,4,3,2), 0, 5)

DPpack documentation built on Oct. 20, 2024, 9:07 a.m.