opendp 0.14.2-dev.20260401.2

A library of differential privacy algorithms for the statistical analysis of sensitive private data.
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
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87

use crate::{
    core::{Domain, Measure, Measurement, Metric, MetricSpace, PrivacyMap},
    error::Fallible,
    measures::{Approximate, MaxDivergence, PrivacyProfile, SmoothedMaxDivergence},
    traits::InfSub,
};

#[cfg(feature = "ffi")]
mod ffi;

#[cfg(test)]
mod test;

/// Fix the delta parameter in the privacy map of a `measurement` with a `SmoothedMaxDivergence` output measure.
///
/// # Arguments
/// * `measurement` - a measurement with a privacy curve to be fixed
/// * `delta` - parameter to fix the privacy curve with
///
/// # Generics
/// * `DI` - Input Domain
/// * `MI` - Input Metric.
/// * `MO` - Output Measure of the input argument. Must be `SmoothedMaxDivergence`
/// * `TO` - Output Type
pub fn make_fix_delta<DI, MI, MO, TO>(
    m: &Measurement<DI, MI, MO, TO>,
    delta: f64,
) -> Fallible<Measurement<DI, MI, MO::FixedMeasure, TO>>
where
    DI: Domain,
    MI: 'static + Metric,
    MO: 'static + FixDeltaMeasure,
    (DI, MI): MetricSpace,
{
    let privacy_map = m.privacy_map.clone();
    let output_measure: MO = m.output_measure.clone();

    m.with_map(
        m.input_metric.clone(),
        m.output_measure.new_fixed_measure()?,
        PrivacyMap::new_fallible(move |d_in| {
            // find the smallest epsilon at the given delta
            let profile = privacy_map.eval(d_in)?;
            output_measure.fix_delta(&profile, delta)
        }),
    )
}

pub trait FixDeltaMeasure: Measure {
    type FixedMeasure: Measure;

    // This fn is used for FFI support
    fn new_fixed_measure(&self) -> Fallible<Self::FixedMeasure>;

    fn fix_delta(
        &self,
        curve: &Self::Distance,
        delta: f64,
    ) -> Fallible<<Self::FixedMeasure as Measure>::Distance>;
}

impl FixDeltaMeasure for SmoothedMaxDivergence {
    type FixedMeasure = Approximate<MaxDivergence>;

    fn new_fixed_measure(&self) -> Fallible<Self::FixedMeasure> {
        Ok(Approximate::default())
    }
    fn fix_delta(&self, profile: &Self::Distance, delta: f64) -> Fallible<(f64, f64)> {
        profile.epsilon(delta).map(|v| (v, delta.clone()))
    }
}

impl FixDeltaMeasure for Approximate<SmoothedMaxDivergence> {
    type FixedMeasure = Approximate<MaxDivergence>;

    fn new_fixed_measure(&self) -> Fallible<Self::FixedMeasure> {
        Ok(Approximate::default())
    }
    fn fix_delta(
        &self,
        (curve, fixed_delta): &(PrivacyProfile, f64),
        delta: f64,
    ) -> Fallible<(f64, f64)> {
        let remaining_delta = delta.neg_inf_sub(&fixed_delta)?;
        curve.epsilon(remaining_delta).map(|v| (v, delta.clone()))
    }
}