torch_ecg.utils.smooth#

torch_ecg.utils.smooth(x: ndarray, window_len: int = 11, window: Literal['flat', 'hanning', 'hamming', 'bartlett', 'blackman'] = 'hanning', mode: str = 'valid', keep_dtype: bool = True) → ndarray[source]#

Smooth the 1d data using a window with requested size.

This method is originally from [1], based on the convolution of a scaled window with the signal. The signal is prepared by introducing reflected copies of the signal (with the window size) in both ends so that transient parts are minimized in the begining and end part of the output signal.

Parameters:

  • x (numpy.ndarray) – The input signal.

  • window_len (int, default 11) – Length of the smoothing window, (previously should be an odd integer, currently can be any (positive) integer).

  • window ({"flat", "hanning", "hamming", "bartlett", "blackman"}, default "hanning") – Type of window from. See also numpy.hanning(), numpy.hamming(), etc. Flat type window will produce a moving average smoothing.

  • mode (str, default "valid") – Mode of convolution, see numpy.convolve() for details.

  • keep_dtype (bool, default True) – Whether dtype of the returned value keeps the same with that of x or not.

Returns:

y – The smoothed signal.

Return type:

numpy.ndarray

Examples

t = np.linspace(-2, 2, 50)
x = np.sin(t) + np.random.randn(len(t)) * 0.1
y = smooth(x)

See also

numpy.hanning(), numpy.hamming(), numpy.bartlett(), numpy.blackman(), numpy.convolve(), scipy.signal.lfilter()

TODO

The window parameter could be the window itself if an array instead of a string.

Note

length(output) != length(input), to correct this, using

return y[(window_len/2-1):-(window_len/2)]

instead of just returning y.

References

On this page