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

// This Source Code Form is subject to the terms of the Mozilla Public
// License, v. 2.0. If a copy of the MPL was not distributed with this
// file, You can obtain one at http://mozilla.org/MPL/2.0/.

//! Exponential moving average & variance filters.

use num_traits::{Num, One, Signed, Zero};

use signalo_traits::filter::Filter;

use traits::{InitialState, Resettable, Stateful, StatefulUnsafe};

use super::mean::Mean;

/// A mean/variance filter's internal state.
#[derive(Clone, Debug)]
pub struct State<T> {
    pub mean: Mean<T>,
    pub variance: Mean<T>,
}

/// A filter producing the exponential moving average and variance over a given signal.
#[derive(Clone, Debug)]
pub struct MeanVariance<T> {
    beta: T,
    state: State<T>,
}

impl<T> MeanVariance<T>
where
    T: Copy + PartialOrd + Zero + One,
{
    /// Creates a new `MeanVariance` filter with `beta = 1.0 / n` with `n` being the filter width.
    #[inline]
    pub fn new(beta: T) -> Self {
        assert!(beta > T::zero() && beta <= T::one());
        let state = Self::initial_state(beta);
        MeanVariance { beta, state }
    }

    /// Returns the filter's `beta` coefficient.
    #[inline]
    pub fn beta(&self) -> &T {
        &self.beta
    }
}

impl<T> Stateful for MeanVariance<T> {
    type State = State<T>;
}

unsafe impl<T> StatefulUnsafe for MeanVariance<T> {
    unsafe fn state(&self) -> &Self::State {
        &self.state
    }

    unsafe fn state_mut(&mut self) -> &mut Self::State {
        &mut self.state
    }
}

impl<T> InitialState<T> for MeanVariance<T>
where
    T: Copy + PartialOrd + Zero + One,
{
    fn initial_state(beta: T) -> Self::State {
        let mean = Mean::new(beta);
        let variance = Mean::new(beta);
        State { mean, variance }
    }
}

impl<T> Resettable for MeanVariance<T>
where
    T: Copy + PartialOrd + Zero + One,
{
    fn reset(&mut self) {
        self.state = Self::initial_state(self.beta);
    }
}

impl<T> Filter<T> for MeanVariance<T>
where
    T: Copy + Num + Signed,
{
    /// (mean, variance)
    type Output = (T, T);

    fn filter(&mut self, input: T) -> Self::Output {
        let mean_old = unsafe { self.state.mean.state().value.unwrap_or(input) };
        let mean_new = self.state.mean.filter(input);
        let deviation_old = (input - mean_old).abs();
        let deviation_new = (input - mean_new).abs();
        let squared = deviation_old * deviation_new;
        let variance = self.state.variance.filter(squared);
        (mean_new, variance)
    }
}

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

    fn get_input() -> Vec<f32> {
        vec![
            0.0, 1.0, 7.0, 2.0, 5.0, 8.0, 16.0, 3.0, 19.0, 6.0, 14.0, 9.0, 9.0, 17.0, 17.0, 4.0,
            12.0, 20.0, 20.0, 7.0, 7.0, 15.0, 15.0, 10.0, 23.0, 10.0, 111.0, 18.0, 18.0, 18.0,
            106.0, 5.0, 26.0, 13.0, 13.0, 21.0, 21.0, 21.0, 34.0, 8.0, 109.0, 8.0, 29.0, 16.0,
            16.0, 16.0, 104.0, 11.0, 24.0, 24.0,
        ]
    }

    fn get_mean() -> Vec<f32> {
        vec![
            0.000, 0.250, 1.938, 1.953, 2.715, 4.036, 7.027, 6.020, 9.265, 8.449, 9.837, 9.628,
            9.471, 11.353, 12.765, 10.574, 10.930, 13.198, 14.898, 12.924, 11.443, 12.332, 12.999,
            12.249, 14.937, 13.703, 38.027, 33.020, 29.265, 26.449, 46.337, 36.003, 33.502, 28.376,
            24.532, 23.649, 22.987, 22.490, 25.368, 21.026, 43.019, 34.264, 32.948, 28.711, 25.533,
            23.150, 43.363, 35.272, 32.454, 30.340,
        ]
    }

    fn get_variance() -> Vec<f32> {
        vec![
            0.000, 0.188, 8.684, 6.513, 6.626, 10.207, 34.493, 28.910, 53.271, 41.953, 37.242,
            28.063, 21.121, 26.470, 25.832, 33.778, 25.715, 34.710, 34.709, 37.728, 34.875, 28.529,
            22.732, 18.735, 35.722, 31.362, 1798.539, 1424.107, 1110.382, 856.581, 1829.006,
            1692.140, 1287.865, 1044.710, 827.864, 623.237, 468.744, 352.298, 289.063, 273.354,
            1656.166, 1472.066, 1109.246, 885.793, 694.640, 538.022, 1629.149, 1418.237, 1087.501,
            829.026,
        ]
    }

    #[test]
    fn mean() {
        let filter = MeanVariance::new(0.25);
        // Sequence: https://en.wikipedia.org/wiki/Collatz_conjecture
        let input = get_input();
        let output: Vec<_> = input
            .iter()
            .scan(filter, |filter, &input| Some(filter.filter(input).0))
            .collect();
        assert_nearly_eq!(output, get_mean(), 0.001);
    }

    #[test]
    fn variance() {
        let filter = MeanVariance::new(0.25);
        // Sequence: https://en.wikipedia.org/wiki/Collatz_conjecture
        let input = get_input();
        let output: Vec<_> = input
            .iter()
            .scan(filter, |filter, &input| Some(filter.filter(input).1))
            .collect();
        assert_nearly_eq!(output, get_variance(), 0.001);
    }
}