ungai 0.6.2

A unique name generator based on Markov Chain
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

use ndarray::{Array2, Axis};
use rand::distr::Distribution;
use rand::distr::weighted::WeightedIndex;
use rand::prelude::*;

pub struct Markov {
    pub matrix: Array2<f64>,
}

impl Markov {
    pub fn train(names: &[&str]) -> Self {
        let n = 256;
        let mut matrix = Array2::<f64>::zeros((n, n));

        for name in names {
            let mut prev = b'^' as usize; // start marker

            for c in name.chars() {
                let j = c as usize;
                if j < n {
                    matrix[[prev, j]] += 1.0;
                    prev = j;
                }
            }

            let j = b'$' as usize; // end marker
            matrix[[prev, j]] += 1.0;
        }

        // normalize probabilities
        for mut row in matrix.axis_iter_mut(Axis(0)) {
            let sum: f64 = row.sum();
            if sum > 0.0 {
                row /= sum;
            }
        }

        Self { matrix }
    }

    pub fn precompute_distributions(&self) -> Vec<Option<WeightedIndex<f64>>> {
        self.matrix
            .axis_iter(Axis(0))
            .map(|row| WeightedIndex::new(row.to_vec()).ok())
            .collect()
    }

    pub fn generate(
        &self,
        rng: &mut impl Rng,
        distributions: &Vec<Option<WeightedIndex<f64>>>,
    ) -> String {
        let mut result = String::new();
        let mut current = b'^' as usize;

        loop {
            let next = if let Some(dist) = &distributions[current] {
                dist.sample(rng)
            } else {
                b'$' as usize
            };

            if next == b'$' as usize {
                break;
            }

            result.push(next as u8 as char);
            current = next;
        }

        result
    }
}