rust_nn 0.1.0

An educational Neural Network framework in Rust
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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178

use super::Matrix2;
use crate::prelude::*;
use std::ops::{Add, AddAssign, Mul, Sub};

pub trait Dot<I> {
    type Output;
    fn dot(self, rhs: I) -> Result<Self::Output>;
}

pub trait Transpose {
    fn transpose(&self) -> Self;
}

impl<T: Default + Copy> Transpose for Matrix2<T> {
    fn transpose(&self) -> Self {
        let mut transposed = Matrix2::new(self.cols(), self.rows());

        for row in 0..self.rows() {
            for col in 0..self.cols() {
                transposed[(col, row)] = self[(row, col)];
            }
        }
        transposed
    }
}

impl<'a, T> Dot<&Matrix2<T>> for &'a Matrix2<T>
where
    T: Mul<Output = T> + Default + AddAssign + Copy,
{
    type Output = Matrix2<T>;
    fn dot(self, rhs: &Matrix2<T>) -> Result<Self::Output> {
        if self.cols() != rhs.rows() {
            return Err(Error::DimensionErr);
        }

        let mut data = Vec::with_capacity(self.rows() * rhs.cols());

        for lhs_row in 0..self.rows() {
            for rhs_col in 0..rhs.cols() {
                let mut sum = T::default();
                for n in 0..self.cols() {
                    sum += self[(lhs_row, n)] * rhs[(n, rhs_col)]
                }
                data.push(sum);
            }
        }

        Ok(Matrix2 {
            data,
            dim: (self.rows(), rhs.cols()),
        })
    }
}

/// Adds two Matrix2s element-wise.
impl<'a, T> Add for &'a Matrix2<T>
where
    &'a T: Add<Output = T>,
{
    type Output = Result<Matrix2<T>>;
    fn add(self, rhs: Self) -> Self::Output {
        if self.dim != rhs.dim {
            return Err(Error::DimensionErr);
        }

        let mut data = Vec::with_capacity(self.rows() * self.cols());
        for row in 0..self.rows() {
            for col in 0..self.cols() {
                data.push(&self[(row, col)] + &rhs[(row, col)])
            }
        }

        Ok(Matrix2 {
            data,
            dim: (self.rows(), rhs.cols()),
        })
    }
}

/// Subs two Matrix2s element-wise.
impl<'a, T> Sub for &'a Matrix2<T>
where
    &'a T: Sub<Output = T>,
{
    type Output = Result<Matrix2<T>>;
    fn sub(self, rhs: Self) -> Self::Output {
        if self.dim != rhs.dim {
            return Err(Error::DimensionErr);
        }

        let mut data = Vec::with_capacity(self.rows() * self.cols());
        for row in 0..self.rows() {
            for col in 0..self.cols() {
                data.push(&self[(row, col)] - &rhs[(row, col)])
            }
        }

        Ok(Matrix2 {
            data,
            dim: (self.rows(), rhs.cols()),
        })
    }
}

#[cfg(test)]
mod tests {
    use crate::matrix::{
        ops::{Dot, Transpose},
        Matrix2,
    };
    use crate::prelude::*;

    #[test]
    fn matrix2_transpose() {
        let matrix = Matrix2::from_array([[1, 2, 3], [4, 5, 6]]).transpose();

        assert_eq!(matrix.clone().to_vec(), [[1, 4], [2, 5], [3, 6]]);
        assert_eq!(matrix.dim, (3, 2));
    }

    #[test]
    fn square_matrix_multiplication() {
        let m1 = Matrix2::from_array([[1, 2], [3, 4]]);
        let m2 = Matrix2::from_array([[3, 2], [1, 3]]);

        let m3 = m1.dot(&m2).unwrap();

        assert_eq!(m3.clone().to_vec(), [[5, 8], [13, 18]]);

        assert_eq!(m3.dim, (2, 2));
    }

    #[test]
    fn non_square_matrix_multiplication() {
        let m1 = Matrix2::from_array([[1, 2], [3, 4], [5, 6]]);
        let m2 = Matrix2::from_array([[1, 2], [3, 4]]);

        let m3 = m1.dot(&m2).unwrap();
        assert_eq!(m3.clone().to_vec(), [[7, 10], [15, 22], [23, 34]]);
        assert_eq!(m3.dim, (3, 2));
    }

    #[test]
    fn matrix_multiplication_error() {
        let m1 = Matrix2::from_array([[1, 2], [3, 4], [5, 6]]).transpose();
        let m2 = Matrix2::from_array([[1, 2], [3, 4]]);

        let m3 = m1.dot(&m2);
        assert_eq!(m3, Err(Error::DimensionErr));
    }

    #[test]
    fn matrix2_addition() {
        let m1 = Matrix2::from_array([[1, 2], [3, 4], [5, 6]]);
        let m2 = Matrix2::from_array([[1, 2], [3, 4], [2, 1]]);

        let m3 = &m1 + &m2;
        assert_eq!(m3.unwrap().to_vec(), [[2, 4], [6, 8], [7, 7]]);
    }

    #[test]
    fn matrix2_addition_err() {
        // unequal rows
        let m1 = Matrix2::from_array([[1, 2], [3, 4], [5, 6]]).transpose();
        let m2 = Matrix2::from_array([[1, 2], [3, 4]]);

        let m3 = &m1 + &m2;
        assert_eq!(m3, Err(Error::DimensionErr));

        // unequal cols
        let m1 = Matrix2::from_array([[1, 2], [3, 4], [5, 6]]).transpose();
        let m2 = Matrix2::from_array([[1, 2, 1], [3, 4, 1], [1, 2, 3]]);

        let m3 = &m1 + &m2;
        assert_eq!(m3, Err(Error::DimensionErr));
    }
}