Skip to main content

rand/distr/
bernoulli.rs

1// Copyright 2018 Developers of the Rand project.
2//
3// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
4// https://www.apache.org/licenses/LICENSE-2.0> or the MIT license
5// <LICENSE-MIT or https://opensource.org/licenses/MIT>, at your
6// option. This file may not be copied, modified, or distributed
7// except according to those terms.
8
9//! The Bernoulli distribution `Bernoulli(p)`.
10
11use crate::distr::Distribution;
12use crate::{Rng, RngExt};
13use core::fmt;
14
15#[cfg(feature = "serde")]
16use serde::{Deserialize, Serialize};
17
18/// The [Bernoulli distribution](https://en.wikipedia.org/wiki/Bernoulli_distribution) `Bernoulli(p)`.
19///
20/// This distribution describes a single boolean random variable, which is true
21/// with probability `p` and false with probability `1 - p`.
22/// It is a special case of the Binomial distribution with `n = 1`.
23///
24/// # Plot
25///
26/// The following plot shows the Bernoulli distribution with `p = 0.1`,
27/// `p = 0.5`, and `p = 0.9`.
28///
29/// ![Bernoulli distribution](https://raw.githubusercontent.com/rust-random/charts/main/charts/bernoulli.svg)
30///
31/// # Example
32///
33/// ```rust
34/// use rand::distr::{Bernoulli, Distribution};
35///
36/// let d = Bernoulli::new(0.3).unwrap();
37/// let v = d.sample(&mut rand::rng());
38/// println!("{} is from a Bernoulli distribution", v);
39/// ```
40///
41/// # Precision
42///
43/// This `Bernoulli` distribution uses 64 bits from the RNG (a `u64`),
44/// so only probabilities that are multiples of 2<sup>-64</sup> can be
45/// represented.
46#[derive(Clone, Copy, Debug, PartialEq)]
47#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
48pub struct Bernoulli {
49    /// Probability of success, relative to the maximal integer.
50    p_int: u64,
51}
52
53// To sample from the Bernoulli distribution we use a method that compares a
54// random `u64` value `v < (p * 2^64)`.
55//
56// If `p == 1.0`, the integer `v` to compare against can not represented as a
57// `u64`. We manually set it to `u64::MAX` instead (2^64 - 1 instead of 2^64).
58// Note that  value of `p < 1.0` can never result in `u64::MAX`, because an
59// `f64` only has 53 bits of precision, and the next largest value of `p` will
60// result in `2^64 - 2048`.
61//
62// Also there is a 100% theoretical concern: if someone consistently wants to
63// generate `true` using the Bernoulli distribution (i.e. by using a probability
64// of `1.0`), just using `u64::MAX` is not enough. On average it would return
65// false once every 2^64 iterations. Some people apparently care about this
66// case.
67//
68// That is why we special-case `u64::MAX` to always return `true`, without using
69// the RNG, and pay the performance price for all uses that *are* reasonable.
70// Luckily, if `new()` and `sample` are close, the compiler can optimize out the
71// extra check.
72const ALWAYS_TRUE: u64 = u64::MAX;
73
74// This is just `2.0.powi(64)`, but written this way because it is not available
75// in `no_std` mode.
76const SCALE: f64 = 2.0 * (1u64 << 63) as f64;
77
78/// Error type returned from [`Bernoulli::new`].
79#[derive(Clone, Copy, Debug, PartialEq, Eq)]
80pub enum BernoulliError {
81    /// `p < 0` or `p > 1`.
82    InvalidProbability,
83}
84
85impl fmt::Display for BernoulliError {
86    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
87        f.write_str(match self {
88            BernoulliError::InvalidProbability => "p is outside [0, 1] in Bernoulli distribution",
89        })
90    }
91}
92
93impl core::error::Error for BernoulliError {}
94
95impl Bernoulli {
96    /// Construct a new `Bernoulli` with the given probability of success `p`.
97    ///
98    /// # Precision
99    ///
100    /// For `p = 1.0`, the resulting distribution will always generate true.
101    /// For `p = 0.0`, the resulting distribution will always generate false.
102    ///
103    /// This method is accurate for any input `p` in the range `[0, 1]` which is
104    /// a multiple of 2<sup>-64</sup>. (Note that not all multiples of
105    /// 2<sup>-64</sup> in `[0, 1]` can be represented as a `f64`.)
106    #[inline]
107    pub fn new(p: f64) -> Result<Bernoulli, BernoulliError> {
108        if !(0.0..1.0).contains(&p) {
109            if p == 1.0 {
110                return Ok(Bernoulli { p_int: ALWAYS_TRUE });
111            }
112            return Err(BernoulliError::InvalidProbability);
113        }
114        Ok(Bernoulli {
115            p_int: (p * SCALE) as u64,
116        })
117    }
118
119    /// Construct a new `Bernoulli` with the probability of success of
120    /// `numerator`-in-`denominator`. I.e. `from_ratio(2, 3)` will return
121    /// a `Bernoulli` with a 2-in-3 chance, or about 67%, of returning `true`.
122    ///
123    /// For `numerator == denominator`, the resulting distribution will always
124    /// return `true`; for `numerator == 0` it will always return `false`.
125    /// For `numerator > denominator` or `denominator == 0`, this returns an
126    /// error.
127    ///
128    /// # Example
129    ///
130    /// ```
131    /// use rand::distr::Bernoulli;
132    ///
133    /// let d = Bernoulli::from_ratio(2, 3).unwrap();
134    /// assert!((d.p() - 2.0 / 3.0).abs() < 1e-9);
135    ///
136    /// // Edge cases:
137    /// assert_eq!(Bernoulli::from_ratio(3, 3).unwrap().p(), 1.0); // always true
138    /// assert_eq!(Bernoulli::from_ratio(0, 3).unwrap().p(), 0.0); // always false
139    /// assert!(Bernoulli::from_ratio(4, 3).is_err());             // numerator > denominator
140    /// assert!(Bernoulli::from_ratio(1, 0).is_err());             // denominator == 0
141    /// ```
142    #[inline]
143    pub fn from_ratio(numerator: u32, denominator: u32) -> Result<Bernoulli, BernoulliError> {
144        if numerator > denominator || denominator == 0 {
145            return Err(BernoulliError::InvalidProbability);
146        }
147        if numerator == denominator {
148            return Ok(Bernoulli { p_int: ALWAYS_TRUE });
149        }
150        let p_int = ((f64::from(numerator) / f64::from(denominator)) * SCALE) as u64;
151        Ok(Bernoulli { p_int })
152    }
153
154    #[inline]
155    /// Returns the probability (`p`) of the distribution.
156    ///
157    /// This value may differ slightly from the input due to loss of precision.
158    pub fn p(&self) -> f64 {
159        if self.p_int == ALWAYS_TRUE {
160            1.0
161        } else {
162            (self.p_int as f64) / SCALE
163        }
164    }
165}
166
167impl Distribution<bool> for Bernoulli {
168    #[inline]
169    fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> bool {
170        // Make sure to always return true for p = 1.0.
171        if self.p_int == ALWAYS_TRUE {
172            return true;
173        }
174        let v: u64 = rng.random();
175        v < self.p_int
176    }
177}
178
179#[cfg(test)]
180mod test {
181    use super::Bernoulli;
182    use crate::RngExt;
183    use crate::distr::Distribution;
184
185    #[test]
186    #[cfg(feature = "serde")]
187    fn test_serializing_deserializing_bernoulli() {
188        let coin_flip = Bernoulli::new(0.5).unwrap();
189        let de_coin_flip: Bernoulli =
190            postcard::from_bytes(&postcard::to_allocvec(&coin_flip).unwrap()).unwrap();
191
192        assert_eq!(coin_flip.p_int, de_coin_flip.p_int);
193    }
194
195    #[test]
196    fn test_trivial() {
197        // We prefer to be explicit here.
198        #![allow(clippy::bool_assert_comparison)]
199
200        let mut r = crate::test::rng(1);
201        let always_false = Bernoulli::new(0.0).unwrap();
202        let always_true = Bernoulli::new(1.0).unwrap();
203        for _ in 0..5 {
204            assert_eq!(r.sample::<bool, _>(&always_false), false);
205            assert_eq!(r.sample::<bool, _>(&always_true), true);
206            assert_eq!(Distribution::<bool>::sample(&always_false, &mut r), false);
207            assert_eq!(Distribution::<bool>::sample(&always_true, &mut r), true);
208        }
209    }
210
211    #[test]
212    #[cfg_attr(miri, ignore)] // Miri is too slow
213    fn test_average() {
214        const P: f64 = 0.3;
215        const NUM: u32 = 3;
216        const DENOM: u32 = 10;
217        let d1 = Bernoulli::new(P).unwrap();
218        let d2 = Bernoulli::from_ratio(NUM, DENOM).unwrap();
219        const N: u32 = 100_000;
220
221        let mut sum1: u32 = 0;
222        let mut sum2: u32 = 0;
223        let mut rng = crate::test::rng(2);
224        for _ in 0..N {
225            if d1.sample(&mut rng) {
226                sum1 += 1;
227            }
228            if d2.sample(&mut rng) {
229                sum2 += 1;
230            }
231        }
232        let avg1 = (sum1 as f64) / (N as f64);
233        assert!((avg1 - P).abs() < 5e-3);
234
235        let avg2 = (sum2 as f64) / (N as f64);
236        assert!((avg2 - (NUM as f64) / (DENOM as f64)).abs() < 5e-3);
237    }
238
239    #[test]
240    fn value_stability() {
241        let mut rng = crate::test::rng(3);
242        let distr = Bernoulli::new(0.4532).unwrap();
243        let mut buf = [false; 10];
244        for x in &mut buf {
245            *x = rng.sample(distr);
246        }
247        assert_eq!(
248            buf,
249            [
250                true, false, false, true, false, false, true, true, true, true
251            ]
252        );
253    }
254
255    #[test]
256    fn bernoulli_distributions_can_be_compared() {
257        assert_eq!(Bernoulli::new(1.0), Bernoulli::new(1.0));
258    }
259}