dpNoise: Differentially Private Noise Generator

In privacytoolsproject/PSI-Library: Differentially Private Statistical Releases for Privacy Preservation

Description

Compile noise from a cryptographically secure random variates to achieve differentially private statistics.

Usage

dpNoise(n, scale, dist, shape = NULL, seed = NULL)

Arguments

n	An integer giving number of variates needed.
scale	Numeric, the scale for the distribution.
dist	A character specifying the distribution from which to draw the noise.
shape	An integer giving the shape parameter for the gamma distribution. Default to NULL.
seed	An integer indicating a seed for R's PNRG, defaults to NULL.

Value

Cryptographically secure noise vector or matrix.

See Also

dpUnif

Examples

laplace_noise <- dpNoise(n=1000, scale=1, dist='laplace')
gaussian_noise <- dpNoise(n=1000, scale=1, dist='gaussian')
laplace_noise_repeatable <- dpNoise(n=1, scale=1, dist='laplace', seed=96845)

privacytoolsproject/PSI-Library documentation built on Feb. 17, 2020, 2:03 p.m.