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

use core::cmp::Ordering;

use alloc::vec::Vec;

use necsim_core::cogs::{MathsCore, RngCore};
use necsim_core_bond::{ClosedUnitF64, NonNegativeF64};

pub mod packed;

#[allow(clippy::module_name_repetitions)]
#[allow(non_snake_case)]
#[derive(Clone)]
pub struct AliasMethodSampler<E: Copy + PartialEq> {
    Us: Vec<ClosedUnitF64>,
    Es: Vec<E>,
    Ks: Vec<E>,
}

impl<E: Copy + PartialEq> AliasMethodSampler<E> {
    #[allow(clippy::no_effect_underscore_binding)]
    #[debug_requires(!event_weights.is_empty(), "event_weights is non-empty")]
    #[debug_ensures(
        ret.Es.iter().eq(old(event_weights).iter().map(|(e, _p)| e)),
        "stores exactly the input events"
    )]
    #[debug_ensures(
        ret.Us.iter().zip(ret.Es.iter()).zip(ret.Ks.iter()).all(|((u, e), k)| {
            let full_bucket = *u == ClosedUnitF64::one();
            !full_bucket || (e == k)
        }),
        "full buckets sample the same event just in case"
    )]
    pub fn new(event_weights: &[(E, NonNegativeF64)]) -> Self {
        #[allow(non_snake_case)]
        let mut Us = Vec::with_capacity(event_weights.len());
        #[allow(non_snake_case)]
        let mut Es = Vec::with_capacity(event_weights.len());
        #[allow(non_snake_case)]
        let mut Ks = Vec::with_capacity(event_weights.len());

        let total_weight: NonNegativeF64 = event_weights.iter().map(|(_e, p)| *p).sum();

        let n = NonNegativeF64::from(event_weights.len());

        for (event, weight) in event_weights {
            Us.push(*weight * n / total_weight);
            Es.push(*event);
            Ks.push(*event);
        }

        let mut overfull_indices: Vec<usize> = (0..Us.len()).filter(|i| Us[*i] > 1.0_f64).collect();
        let mut underfull_indices: Vec<usize> =
            (0..Us.len()).filter(|i| Us[*i] < 1.0_f64).collect();

        while let Some((overfull_index, underfull_index)) =
            pop_overfull_underfull_pair_atomic(&mut overfull_indices, &mut underfull_indices)
        {
            // Safety: Us[overfull_index] > 1.0,
            //         so (Us[overfull_index] - 1.0) > 0.0
            Us[overfull_index] = unsafe {
                NonNegativeF64::new_unchecked(
                    Us[overfull_index].get() + Us[underfull_index].get() - 1.0_f64,
                )
            };
            Ks[underfull_index] = Es[overfull_index];

            match Us[overfull_index].cmp(&NonNegativeF64::one()) {
                Ordering::Less => underfull_indices.push(overfull_index),
                Ordering::Equal => (),
                Ordering::Greater => overfull_indices.push(overfull_index),
            }
        }

        // Fix rounding errors for full indices:
        //   M. D. Vose, "A linear algorithm for generating random numbers with a given
        //   distribution", in IEEE Transactions on Software Engineering, vol. 17, no.
        //   9, pp. 972-975, Sept. 1991, doi: 10.1109/32.92917.
        overfull_indices
            .into_iter()
            .for_each(|i| Us[i] = NonNegativeF64::one());
        underfull_indices
            .into_iter()
            .for_each(|i| Us[i] = NonNegativeF64::one());

        // Safety: The bucket weights are now probabilities in [0.0; 1.0]
        #[allow(non_snake_case)]
        let Us = unsafe { core::mem::transmute::<Vec<NonNegativeF64>, Vec<ClosedUnitF64>>(Us) };

        Self { Us, Es, Ks }
    }

    #[debug_ensures(self.Es.contains(&ret), "returns one of the weighted events")]
    pub fn sample_event<M: MathsCore, G: RngCore<M>>(&self, rng: &mut G) -> E {
        use necsim_core::cogs::RngSampler;

        let x = rng.sample_uniform_closed_open();

        #[allow(
            clippy::cast_precision_loss,
            clippy::cast_possible_truncation,
            clippy::cast_sign_loss
        )]
        let i = M::floor(x.get() * (self.Es.len() as f64)) as usize; // index into events

        #[allow(clippy::cast_precision_loss)]
        let y = x.get() * (self.Es.len() as f64) - (i as f64); // U(0,1) to compare against U[i]

        if y < self.Us[i].get() {
            self.Es[i]
        } else {
            self.Ks[i]
        }
    }
}

fn pop_overfull_underfull_pair_atomic(
    overfull_indices: &mut Vec<usize>,
    underfull_indices: &mut Vec<usize>,
) -> Option<(usize, usize)> {
    match (overfull_indices.pop(), underfull_indices.pop()) {
        (Some(overfull_index), Some(underfull_index)) => Some((overfull_index, underfull_index)),
        (Some(overfull_index), None) => {
            overfull_indices.push(overfull_index);
            None
        },
        (None, Some(underfull_index)) => {
            underfull_indices.push(underfull_index);
            None
        },
        (None, None) => None,
    }
}