openml 0.1.2

A rust interface to [OpenML](http://openml.org/).
Documentation 
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

//! Implementation of a Gaussian Naive Bayes Classifier

use std::cmp::Ordering;
use std::collections::HashMap;
use std::f64;
use std::fmt;
use std::hash::Hash;
use std::iter::FromIterator;

/// A Gaussian Naive Bayes Classifier
///
/// The classifier is trained by consuming an iterator over the training data:
/// ```
/// # use openml::baseline::NaiveBayesClassifier;
/// # let data: Vec<(&[f64], &u8)> = vec![];
/// let nbc: NaiveBayesClassifier<_> = data
///     .into_iter()
///     .collect();
/// ```
#[derive(Debug)]
pub struct NaiveBayesClassifier<C>
where C: Eq + Hash
{
    class_distributions: HashMap<C, FeatureDistribution>,
}

/// Distribution of each feature column
#[derive(Debug, Clone)]
struct FeatureDistribution {
    distributions: Vec<NormalDistribution>
}

/// Univariate Normal Distribution
#[derive(Copy, Clone)]
struct NormalDistribution {
    sum: f64,
    sqsum: f64,
    n: usize
}

impl<'a, C: 'a, J> FromIterator<(J, &'a C)> for NaiveBayesClassifier<C>
where
    J: IntoIterator<Item=&'a f64>,
    C: Eq + Hash + Copy,
{
    fn from_iter<I: IntoIterator<Item=(J, &'a C)>>(iter: I) -> Self {
        let mut class_distributions = HashMap::new();

        for (x, &y) in iter {
            let distributions = &mut class_distributions
                .entry(y)
                .or_insert(FeatureDistribution::new())
                .distributions;

            for (i, &xi) in x.into_iter().enumerate() {
                if i >= distributions.len() {
                    distributions.resize(1 + i, NormalDistribution::new());
                }

                distributions[i].update(xi);
            }
        }

        NaiveBayesClassifier {
            class_distributions
        }
    }
}

impl<C> NaiveBayesClassifier<C>
where  C: Eq + Hash + Copy,
{
    /// predict target class for a single feature vector
    pub fn predict(&self, x: &[f64]) -> C {
        self.class_distributions
            .iter()
            .map(|(c, dists)| {
                let mut lnprob = 0.0;
                for (&xi, dist) in x.iter().zip(dists.distributions.iter()) {
                    lnprob += dist.lnprob(xi);
                }
                (c, lnprob)
            })
            .max_by(|(_, lnp1), (_, lnp2)| {
                if lnp1 > lnp2 {
                    Ordering::Greater
                } else if lnp1 == lnp2 {
                    Ordering::Equal
                } else {
                    Ordering::Less
                }
            })
            .map(|(&c, _)| c)
            .unwrap()
    }
}

impl FeatureDistribution {
    fn new() -> Self {
        FeatureDistribution {
            distributions: Vec::new()
        }
    }
}

impl NormalDistribution {
    fn new() -> Self {
        NormalDistribution {
            sum: 0.0,
            sqsum: 0.0,
            n: 0
        }
    }

    fn update(&mut self, x: f64) {
        self.sum += x;
        self.sqsum += x * x;
        self.n += 1;
    }

    fn mean(&self) -> f64 {
        self.sum / self.n as f64
    }

    fn variance(&self) -> f64 {
        (self.sqsum - (self.sum * self.sum) / self.n as f64) / (self.n as f64 - 1.0)
    }

    fn lnprob(&self, x: f64) -> f64 {
        let v = self.variance();
        let xm = x - self.mean();

        0.5 * ((1.0 / (2.0 * f64::consts::PI * v)).ln() - (xm * xm) / v)

    }
}

impl fmt::Debug for NormalDistribution {
    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
        write!(f, "N{{{}; {}}}", self.mean(), self.variance())
    }
}

#[test]
fn nbc() {
    let data = vec![(vec![1.0, 2.0], 'A'),
                    (vec![2.0, 1.0], 'A'),
                    (vec![1.0, 5.0], 'B'),
                    (vec![2.0, 6.0], 'B')];

    let nbc: NaiveBayesClassifier<_> = data
        .iter()
        .map(|(x, y)| (x, y))
        .collect();

    assert_eq!(nbc.predict(&[1.5, 1.5]), 'A');
    assert_eq!(nbc.predict(&[5.5, 1.5]), 'A');
    assert_eq!(nbc.predict(&[1.5, 5.5]), 'B');
    assert_eq!(nbc.predict(&[5.5, 5.5]), 'B');
}