opendp 0.16.0

A library of differential privacy algorithms for the statistical analysis of sensitive private data.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

# type: ignore
def sample_discrete_laplace(scale) -> int:
    if scale == 0:
        return 0

    inv_scale = recip(scale)

    while True:
        sign = sample_standard_bernoulli()
        magnitude = sample_geometric_exp_fast(inv_scale)  # |\label{line:magnitude}|

        if sign or magnitude != 0:  # |\label{line:branching}|
            if sign:
                return magnitude
            else:
                return -magnitude