torsh-data 0.1.0

Data loading and preprocessing utilities for ToRSh
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

//! Audio-specific datasets and transformations
//!
//! This module provides comprehensive audio processing capabilities including:
//! - Core audio data types and utilities
//! - Dataset implementations for audio classification and speech tasks
//! - Transform implementations for preprocessing and augmentation

// Allow unused imports for conditional features
#![allow(unused_imports)]

//! # Examples
//!
//! ```no_run
//! use torsh_data::audio::{AudioFolder, AudioData, AudioToTensor};
//! use torsh_data::audio::transforms::transforms::*;
//!
//! # fn main() -> Result<(), Box<dyn std::error::Error>> {
//! // Create an audio folder dataset
//! let dataset = AudioFolder::new("/path/to/audio", Some(22050))?;
//!
//! // Apply transforms
//! let normalize = Normalize::new(0.1);
//! let resample = Resample::new(16000);
//! # Ok(())
//! # }
//! ```

pub mod datasets;
pub mod transforms;
pub mod types;

// Re-export core types for convenience
pub use datasets::{AudioFolder, LibriSpeech};
pub use transforms::{AudioToTensor, TensorToAudio};
pub use types::AudioData;

// Re-export common transforms
pub use transforms::transforms::*;

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

    #[test]
    fn test_audio_data_creation() {
        let samples = vec![0.1, -0.2, 0.3, -0.4];
        let audio = AudioData::new(samples.clone(), 44100, 1);

        assert_eq!(audio.samples, samples);
        assert_eq!(audio.sample_rate, 44100);
        assert_eq!(audio.channels, 1);
        assert_eq!(audio.len(), 4);
        assert!(!audio.is_empty());
    }

    #[test]
    fn test_audio_data_duration() {
        let samples = vec![0.0; 44100]; // 1 second at 44100 Hz
        let audio = AudioData::new(samples, 44100, 1);

        assert_eq!(audio.duration(), 1.0);
    }

    #[test]
    fn test_audio_to_tensor_mono() {
        let samples = vec![0.1, -0.2, 0.3, -0.4];
        let audio = AudioData::new(samples, 44100, 1);
        let transform = AudioToTensor;

        let result = transform.transform(audio);
        assert!(result.is_ok());

        let tensor = result.unwrap();
        let shape = tensor.shape();
        assert_eq!(shape.dims(), &[1, 4]); // [channels, samples]
    }

    #[test]
    fn test_audio_to_tensor_stereo() {
        // Interleaved stereo: L, R, L, R
        let samples = vec![0.1, 0.2, -0.3, -0.4];
        let audio = AudioData::new(samples, 44100, 2);
        let transform = AudioToTensor;

        let result = transform.transform(audio);
        assert!(result.is_ok());

        let tensor = result.unwrap();
        let shape = tensor.shape();
        assert_eq!(shape.dims(), &[2, 2]); // [channels, samples_per_channel]
    }

    #[test]
    fn test_tensor_to_audio_mono() {
        use torsh_core::device::DeviceType;
        use torsh_tensor::Tensor;

        let data = vec![0.1, -0.2, 0.3, -0.4];
        let tensor = Tensor::from_data(data.clone(), vec![1, 4], DeviceType::Cpu).unwrap();

        let transform = TensorToAudio::new(44100);
        let result = transform.transform(tensor);
        assert!(result.is_ok());

        let audio = result.unwrap();
        assert_eq!(audio.samples, data);
        assert_eq!(audio.sample_rate, 44100);
        assert_eq!(audio.channels, 1);
    }

    #[test]
    fn test_tensor_to_audio_stereo() {
        use torsh_core::device::DeviceType;
        use torsh_tensor::Tensor;

        // Channel-first format: [ch0_s0, ch0_s1, ch1_s0, ch1_s1]
        let data = vec![0.1, -0.2, 0.3, -0.4];
        let tensor = Tensor::from_data(data, vec![2, 2], DeviceType::Cpu).unwrap();

        let transform = TensorToAudio::new(44100);
        let result = transform.transform(tensor);
        assert!(result.is_ok());

        let audio = result.unwrap();
        // Should be interleaved: [ch0_s0, ch1_s0, ch0_s1, ch1_s1]
        assert_eq!(audio.samples, vec![0.1, 0.3, -0.2, -0.4]);
        assert_eq!(audio.sample_rate, 44100);
        assert_eq!(audio.channels, 2);
    }
}