solvr 0.2.0-beta.2

Advanced computing library for real-world problem solving - optimization, differential equations, interpolation, statistics, and more
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

//! CPU implementation of STFT and inverse STFT algorithms.

use crate::signal::impl_generic::{istft_impl, stft_impl};
use crate::signal::traits::stft::StftAlgorithms;
use numr::error::Result;
use numr::runtime::cpu::{CpuClient, CpuRuntime};
use numr::tensor::Tensor;

impl StftAlgorithms<CpuRuntime> for CpuClient {
    fn stft(
        &self,
        signal: &Tensor<CpuRuntime>,
        n_fft: usize,
        hop_length: Option<usize>,
        window: Option<&Tensor<CpuRuntime>>,
        center: bool,
        normalized: bool,
    ) -> Result<Tensor<CpuRuntime>> {
        stft_impl(self, signal, n_fft, hop_length, window, center, normalized)
    }

    fn istft(
        &self,
        stft_matrix: &Tensor<CpuRuntime>,
        hop_length: Option<usize>,
        window: Option<&Tensor<CpuRuntime>>,
        center: bool,
        length: Option<usize>,
        normalized: bool,
    ) -> Result<Tensor<CpuRuntime>> {
        istft_impl(
            self,
            stft_matrix,
            hop_length,
            window,
            center,
            length,
            normalized,
        )
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use numr::runtime::cpu::CpuDevice;

    fn setup() -> (CpuClient, CpuDevice) {
        let device = CpuDevice::new();
        let client = CpuClient::new(device.clone());
        (client, device)
    }

    #[test]
    fn test_stft_basic() {
        let (client, device) = setup();

        // Create a simple signal
        let signal: Vec<f64> = (0..256).map(|i| (i as f64 * 0.1).sin()).collect();
        let signal_tensor = Tensor::<CpuRuntime>::from_slice(&signal, &[256], &device);

        let result = client
            .stft(&signal_tensor, 64, Some(16), None, true, false)
            .unwrap();

        // Output should be complex with shape [n_frames, freq_bins]
        let freq_bins = 64 / 2 + 1; // 33
        let n_frames = (256 + 64 - 64) / 16 + 1; // 17

        assert_eq!(result.shape(), &[n_frames, freq_bins]);
    }

    #[test]
    fn test_istft_reconstruction() {
        let (client, device) = setup();

        // Create a signal
        let signal: Vec<f64> = (0..256).map(|i| (i as f64 * 0.1).sin()).collect();
        let signal_tensor = Tensor::<CpuRuntime>::from_slice(&signal, &[256], &device);

        // STFT
        let stft_result = client
            .stft(&signal_tensor, 64, Some(16), None, true, false)
            .unwrap();

        // ISTFT
        let reconstructed = client
            .istft(&stft_result, Some(16), None, true, Some(256), false)
            .unwrap();

        assert_eq!(reconstructed.shape(), &[256]);

        // Check reconstruction is close to original
        let recon_data: Vec<f64> = reconstructed.to_vec();
        for (i, (&orig, &recon)) in signal.iter().zip(recon_data.iter()).enumerate() {
            let err = (orig - recon).abs();
            // Skip edges which have boundary effects
            if i > 32 && i < 224 {
                assert!(
                    err < 0.1,
                    "Reconstruction error at {}: {} vs {}",
                    i,
                    orig,
                    recon
                );
            }
        }
    }
}