"""
Signal Transforms Module for Physiological Signal Processing
This module provides comprehensive capabilities for physiological
signal processing including ECG, PPG, EEG, and other vital signs.
Author: vitalDSP Team
Date: 2025-01-27
Version: 1.0.0
Key Features:
- Object-oriented design with comprehensive classes
- Multiple processing methods and functions
- NumPy integration for numerical computations
Examples:
---------
Basic usage:
>>> import numpy as np
>>> from vitalDSP.transforms.time_freq_representation import TimeFreqRepresentation
>>> signal = np.random.randn(1000)
>>> processor = TimeFreqRepresentation(signal)
>>> result = processor.process()
>>> print(f'Processing result: {result}')
"""
from vitalDSP.transforms.stft import STFT
from vitalDSP.transforms.wavelet_transform import WaveletTransform
[docs]
class TimeFreqRepresentation:
"""
A class to generate Time-Frequency Representations (TFR) of signals for machine learning and signal analysis.
This class supports multiple methods for computing the TFR, including Short-Time Fourier Transform (STFT) and Wavelet Transform. These representations are useful in analyzing the frequency content of signals as it changes over time, which is particularly valuable in tasks like speech processing, biomedical signal analysis, and other time-series data applications.
Methods
-------
compute_tfr : method
Computes the time-frequency representation of the signal using the specified method.
"""
def __init__(self, signal, method="stft", **kwargs):
"""
Initialize the TimeFreqRepresentation class with the signal and method.
Parameters
----------
signal : numpy.ndarray
The input signal to be transformed into a time-frequency representation.
method : str, optional
The method to use for time-frequency representation ('stft' or 'wavelet'). Default is 'stft'.
kwargs : dict, optional
Additional parameters specific to the chosen method, such as window size for STFT or wavelet type for Wavelet Transform.
Examples
--------
>>> signal = np.sin(np.linspace(0, 10, 1000))
>>> tfr = TimeFreqRepresentation(signal, method='stft', window='hann', nperseg=256)
>>> tfr_result = tfr.compute_tfr()
>>> print(tfr_result)
"""
self.signal = signal
self.method = method
self.kwargs = kwargs
[docs]
def compute_tfr(self):
"""
Compute the Time-Frequency Representation (TFR) of the signal.
Depending on the chosen method, this function computes either the Short-Time Fourier Transform (STFT) or the Wavelet Transform to represent the signal in both time and frequency domains.
Returns
-------
numpy.ndarray
The time-frequency representation of the signal. The output format depends on the method:
- For STFT: A 2D array where rows correspond to time segments and columns to frequency bins.
- For Wavelet Transform: A 2D array where rows correspond to time segments and columns to scales or frequencies.
Raises
------
ValueError
If an unsupported method is specified.
Examples
--------
>>> signal = np.sin(np.linspace(0, 10, 1000))
>>> tfr = TimeFreqRepresentation(signal, method='wavelet', wavelet_name='db', level=4)
>>> tfr_result = tfr.compute_tfr()
>>> print(tfr_result)
"""
if self.method == "stft":
stft = STFT(self.signal, **self.kwargs)
return stft.compute_stft()
elif self.method == "wavelet":
# Extract level parameter if present (it goes to perform_wavelet_transform, not __init__)
level = self.kwargs.pop("level", 1)
wavelet_transform = WaveletTransform(self.signal, **self.kwargs)
return wavelet_transform.perform_wavelet_transform(level=level)
else:
raise ValueError("Unsupported method. Use 'stft' or 'wavelet'.")