audio-lab 0.1.0

Toolkit for audio research and testing
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

use crate::signal::{
    data::RealData,
    math::{db_to_gain, gain_to_db},
};

impl RealData {
    pub fn max_per_channel(&self) -> Vec<(f64, f64)> {
        self.y_data()
            .outer_iter()
            .map(|channel| {
                let (max_index, max_value) = channel
                    .iter()
                    .enumerate()
                    .max_by(|(_, a), (_, b)| a.partial_cmp(b).unwrap_or(std::cmp::Ordering::Equal))
                    .expect("RealData channels always have consistent lengths");

                (self.x_axis().get(max_index), *max_value)
            })
            .collect()
    }

    pub fn max_overall(&self) -> (f64, f64) {
        *self
            .max_per_channel()
            .iter()
            .max_by(|a, b| a.1.partial_cmp(&b.1).unwrap_or(std::cmp::Ordering::Equal))
            .expect("RealData must contain at least one sample")
    }

    pub fn max_abs_per_channel(&self) -> Vec<(f64, f64)> {
        self.y_data()
            .outer_iter()
            .map(|channel| {
                let (max_index, max_abs_value) = channel
                    .iter()
                    .enumerate()
                    .map(|(idx, &value)| (idx, value.abs()))
                    .max_by(|(_, a), (_, b)| a.partial_cmp(b).unwrap_or(std::cmp::Ordering::Equal))
                    .expect("RealData channels always have consistent lengths");

                (self.x_axis().get(max_index), max_abs_value)
            })
            .collect()
    }

    pub fn max_abs_overall(&self) -> (f64, f64) {
        *self
            .max_abs_per_channel()
            .iter()
            .max_by(|a, b| a.1.partial_cmp(&b.1).unwrap_or(std::cmp::Ordering::Equal))
            .expect("RealData must contain at least one sample")
    }

    pub fn min_max_per_channel(&self) -> Vec<((f64, f64), (f64, f64))> {
        self.y_data()
            .outer_iter()
            .map(|channel| {
                let (min_index, min_value) = channel
                    .iter()
                    .enumerate()
                    .min_by(|(_, a), (_, b)| a.partial_cmp(b).unwrap_or(std::cmp::Ordering::Equal))
                    .expect("RealData channels always have consistent lengths");
                let (max_index, max_value) = channel
                    .iter()
                    .enumerate()
                    .max_by(|(_, a), (_, b)| a.partial_cmp(b).unwrap_or(std::cmp::Ordering::Equal))
                    .expect("RealData channels always have consistent lengths");

                (
                    (self.x_axis().get(min_index), *min_value),
                    (self.x_axis().get(max_index), *max_value),
                )
            })
            .collect()
    }

    pub fn min_max_overall(&self) -> ((f64, f64), (f64, f64)) {
        let per_channel = self.min_max_per_channel();

        let min = per_channel
            .iter()
            .map(|&(min, _)| min)
            .min_by(|a, b| a.1.partial_cmp(&b.1).unwrap_or(std::cmp::Ordering::Equal))
            .expect("RealData must contain at least one sample");
        let max = per_channel
            .iter()
            .map(|&(_, max)| max)
            .max_by(|a, b| a.1.partial_cmp(&b.1).unwrap_or(std::cmp::Ordering::Equal))
            .expect("RealData must contain at least one sample");

        (min, max)
    }

    pub fn to_decibels(&self) -> RealData {
        RealData::from_axis(self.x_axis().clone(), self.y_data().mapv(gain_to_db))
            .expect("x/y data originated from RealData")
    }

    pub fn to_gain(&self) -> RealData {
        RealData::from_axis(self.x_axis().clone(), self.y_data().mapv(db_to_gain))
            .expect("x/y data originated from RealData")
    }
}

#[cfg(test)]
mod tests {
    use approx::assert_abs_diff_eq;
    use ndarray::{arr2, aview2};

    use crate::signal::data::RealData;

    #[test]
    fn extrema_helpers() {
        let data = RealData::new(
            ndarray::arr1(&[0.0, 1.0, 2.0, 3.0, 4.0]),
            arr2(&[[0.0, 1.0, 5.0, -11.0, 1.0], [0.0, -10.0, 0.0, 6.0, 1.0]]),
        )
        .unwrap();

        assert_eq!(data.max_per_channel(), vec![(2.0, 5.0), (3.0, 6.0)]);
        assert_eq!(data.max_overall(), (3.0, 6.0));
        assert_eq!(data.max_abs_per_channel(), vec![(3.0, 11.0), (1.0, 10.0)]);
        assert_eq!(data.max_abs_overall(), (3.0, 11.0));
        assert_eq!(
            data.min_max_per_channel(),
            vec![((3.0, -11.0), (2.0, 5.0)), ((1.0, -10.0), (3.0, 6.0))]
        );
        assert_eq!(data.min_max_overall(), ((3.0, -11.0), (3.0, 6.0)));
    }

    #[test]
    fn db_gain_conversion_helpers() {
        let gain = RealData::new(ndarray::arr1(&[1.0, 2.0]), arr2(&[[1.0, 10.0]])).unwrap();
        let db = gain.to_decibels();
        assert_abs_diff_eq!(db.channel(0)[0], 0.0, epsilon = 1e-12);
        assert_abs_diff_eq!(db.channel(0)[1], 20.0, epsilon = 1e-12);

        let roundtrip = db.to_gain();
        assert!(
            roundtrip
                .y_data()
                .abs_diff_eq(&aview2(&[[1.0, 10.0]]), 1e-12)
        );
    }
}