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
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
// Copyright 2018 Developers of the Rand project.
// Copyright 2013 The Rust Project Developers.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// https://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or https://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.

//! The exponential distribution.

use crate::utils::ziggurat;
use num_traits::Float;
use crate::{ziggurat_tables, Distribution};
use rand::Rng;
use core::fmt;

/// Samples floating-point numbers according to the exponential distribution,
/// with rate parameter `λ = 1`. This is equivalent to `Exp::new(1.0)` or
/// sampling with `-rng.gen::<f64>().ln()`, but faster.
///
/// See `Exp` for the general exponential distribution.
///
/// Implemented via the ZIGNOR variant[^1] of the Ziggurat method. The exact
/// description in the paper was adjusted to use tables for the exponential
/// distribution rather than normal.
///
/// [^1]: Jurgen A. Doornik (2005). [*An Improved Ziggurat Method to
///       Generate Normal Random Samples*](
///       https://www.doornik.com/research/ziggurat.pdf).
///       Nuffield College, Oxford
///
/// # Example
/// ```
/// use rand::prelude::*;
/// use rand_distr::Exp1;
///
/// let val: f64 = thread_rng().sample(Exp1);
/// println!("{}", val);
/// ```
#[derive(Clone, Copy, Debug)]
#[cfg_attr(feature = "serde1", derive(serde::Serialize, serde::Deserialize))]
pub struct Exp1;

impl Distribution<f32> for Exp1 {
    #[inline]
    fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> f32 {
        // TODO: use optimal 32-bit implementation
        let x: f64 = self.sample(rng);
        x as f32
    }
}

// This could be done via `-rng.gen::<f64>().ln()` but that is slower.
impl Distribution<f64> for Exp1 {
    #[inline]
    fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> f64 {
        #[inline]
        fn pdf(x: f64) -> f64 {
            (-x).exp()
        }
        #[inline]
        fn zero_case<R: Rng + ?Sized>(rng: &mut R, _u: f64) -> f64 {
            ziggurat_tables::ZIG_EXP_R - rng.gen::<f64>().ln()
        }

        ziggurat(
            rng,
            false,
            &ziggurat_tables::ZIG_EXP_X,
            &ziggurat_tables::ZIG_EXP_F,
            pdf,
            zero_case,
        )
    }
}

/// The exponential distribution `Exp(lambda)`.
///
/// This distribution has density function: `f(x) = lambda * exp(-lambda * x)`
/// for `x > 0`, when `lambda > 0`. For `lambda = 0`, all samples yield infinity.
///
/// Note that [`Exp1`](crate::Exp1) is an optimised implementation for `lambda = 1`.
///
/// # Example
///
/// ```
/// use rand_distr::{Exp, Distribution};
///
/// let exp = Exp::new(2.0).unwrap();
/// let v = exp.sample(&mut rand::thread_rng());
/// println!("{} is from a Exp(2) distribution", v);
/// ```
#[derive(Clone, Copy, Debug)]
#[cfg_attr(feature = "serde1", derive(serde::Serialize, serde::Deserialize))]
pub struct Exp<F>
where F: Float, Exp1: Distribution<F>
{
    /// `lambda` stored as `1/lambda`, since this is what we scale by.
    lambda_inverse: F,
}

/// Error type returned from `Exp::new`.
#[derive(Clone, Copy, Debug, PartialEq, Eq)]
pub enum Error {
    /// `lambda < 0` or `nan`.
    LambdaTooSmall,
}

impl fmt::Display for Error {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        f.write_str(match self {
            Error::LambdaTooSmall => "lambda is negative or NaN in exponential distribution",
        })
    }
}

#[cfg(feature = "std")]
#[cfg_attr(doc_cfg, doc(cfg(feature = "std")))]
impl std::error::Error for Error {}

impl<F: Float> Exp<F>
where F: Float, Exp1: Distribution<F>
{
    /// Construct a new `Exp` with the given shape parameter
    /// `lambda`.
    /// 
    /// # Remarks
    /// 
    /// For custom types `N` implementing the [`Float`] trait,
    /// the case `lambda = 0` is handled as follows: each sample corresponds
    /// to a sample from an `Exp1` multiplied by `1 / 0`. Primitive types 
    /// yield infinity, since `1 / 0 = infinity`.
    #[inline]
    pub fn new(lambda: F) -> Result<Exp<F>, Error> {
        if !(lambda >= F::zero()) {
            return Err(Error::LambdaTooSmall);
        }
        Ok(Exp {
            lambda_inverse: F::one() / lambda,
        })
    }
}

impl<F> Distribution<F> for Exp<F>
where F: Float, Exp1: Distribution<F>
{
    fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> F {
        rng.sample(Exp1) * self.lambda_inverse
    }
}

#[cfg(test)]
mod test {
    use super::*;

    #[test]
    fn test_exp() {
        let exp = Exp::new(10.0).unwrap();
        let mut rng = crate::test::rng(221);
        for _ in 0..1000 {
            assert!(exp.sample(&mut rng) >= 0.0);
        }
    }
    #[test]
    fn test_zero() {
        let d = Exp::new(0.0).unwrap();
        assert_eq!(d.sample(&mut crate::test::rng(21)), f64::infinity());
    }
    #[test]
    #[should_panic]
    fn test_exp_invalid_lambda_neg() {
        Exp::new(-10.0).unwrap();
    }

    #[test]
    #[should_panic]
    fn test_exp_invalid_lambda_nan() {
        Exp::new(f64::nan()).unwrap();
    }
}