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
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

# type: ignore
def make_randomized_response_bitvec(
    input_domain: BitVectorDomain,
    input_metric: DiscreteDistance,
    f: f64,
    constant_time: bool,
):
    output_measure = MaxDivergence(f64)

    if f <= 0.0 or f > 1:  # |\label{line:range}|
        raise Exception("Probability must be in (0.0, 1]")

    m = input_domain.max_weight
    if m is None:
        raise Exception(
            "make_randomized_response_bitvec requires a max_weight on the input_domain"
        )

    epsilon = 2 * m * log((2 - f) / f)

    def privacy_map(d_in: IntDistance):  # |\label{line:map}|
        if d_in == 0:
            return 0.0
        if d_in > 1:
            raise ValueError("d_in must be 0 or 1.")
        return epsilon

    def function(arg: BitVector) -> BitVector:  # |\label{line:fn}|
        noise_vector = [
            bool.sample_bernoulli(f / 2, constant_time) for _ in range(len(arg))
        ]
        return xor(arg, noise_vector)

    return Measurement(
        input_domain, function, input_metric, output_measure, privacy_map
    )