Source code for vallenae.features.acoustic_emission

from __future__ import annotations

import numpy as np




def peak_amplitude(data: np.ndarray) -> float:
    """
    Compute maximum absolute amplitude.

    Args:
        data: Input array

    Returns:
        Peak amplitude of the input array
    """
    return np.max(np.abs(data))





def peak_amplitude_index(data: np.ndarray) -> int:
    """
    Compute index of peak amplitude.

    Args:
        data: Input array

    Returns:
        Index of peak amplitude
    """
    return np.argmax(np.abs(data))



def _mask_above_threshold(data: np.ndarray, threshold: float) -> np.ndarray:
    return (data >= threshold) | (data <= -threshold)




def is_above_threshold(data: np.ndarray, threshold: float) -> bool:
    """
    Checks if absolute amplitudes are above threshold.

    Args:
        data: Input array
        threshold: Threshold amplitude

    Returns:
        True if input array is above threshold, otherwise False
    """
    return np.any(_mask_above_threshold(data, threshold))





def first_threshold_crossing(data: np.ndarray, threshold: float) -> int | None:
    """
    Compute index of first threshold crossing.

    Args:
        data: Input array
        threshold: Threshold amplitude

    Returns:
        Index of first threshold crossing. None if threshold was not exceeded
    """
    above_threshold = _mask_above_threshold(data, threshold)
    if above_threshold.size == 0:
        return None
    index = np.argmax(above_threshold)
    return index if above_threshold[index] else None





def last_threshold_crossing(data: np.ndarray, threshold: float) -> int | None:
    """
    Compute index of last threshold crossing.

    Args:
        data: Input array
        threshold: Threshold amplitude

    Returns:
        Index of last threshold crossing. None if threshold was not exceeded
    """
    index = first_threshold_crossing(data[::-1], threshold)
    return len(data) - 1 - index if index is not None else None





def duration(data: np.ndarray, threshold: float, samplerate: int) -> float:
    """
    Compute the duration of a hit.

    Args:
        data: Input array (hit)
        threshold: Threshold amplitude
        samplerate: Sample rate of input array in Hz

    Returns:
        Duration in seconds
    """
    index_first = first_threshold_crossing(data, threshold)
    index_last = last_threshold_crossing(data, threshold)
    if index_first is None or index_last is None:
        return 0.0
    return (index_last - index_first) / samplerate





def rise_time(
    data: np.ndarray,
    threshold: float,
    samplerate: int,
    first_crossing: int | None = None,
    index_peak: int | None = None,
) -> float:
    """
    Compute the rise time.

    The rise time is the time between the first threshold crossing and the peak amplitude.

    Args:
        data: Input array (hit)
        threshold: Threshold amplitude (in volts)
        samplerate: Sample rate of the input array
        first_crossing: Precomputed index of first threshold crossing to save computation time
        index_peak: Precomputed index of peak amplitude to save computation time

    Returns:
        Rise time in seconds
    """
    # save some computations if pre-results are provided
    n_first_crossing = (
        first_crossing if first_crossing is not None else first_threshold_crossing(data, threshold)
    )
    n_max = index_peak if index_peak is not None else peak_amplitude_index(data)
    if n_first_crossing is None:
        return 0
    return (n_max - n_first_crossing) / samplerate





def energy(data: np.ndarray, samplerate: int) -> float:
    """
    Compute the energy of a hit.

    Energy is the integral of the squared AE-signal over time (EN 1330-9).
    The unit of energy is eu. 1 eu corresponds to 1e-14 V²s.

    Args:
        data: Input array (hit)
        samplerate: Sample rate of input array in Hz

    Returns:
        Energy of input array (hit)
    """
    return np.sum(data**2) * 1e14 / samplerate





def signal_strength(data: np.ndarray, samplerate: int) -> float:
    """
    Compute the signal strength of a hit.

    Signal strength is the integral of the rectified AE-signal over time.
    The unit of Signal Strength is nVs (1e-9 Vs).

    Args:
        data: Input array (hit)
        samplerate: Sample rate of input array in Hz

    Returns:
        Signal strength of input array (hit)
    """
    return np.sum(np.abs(data)) * 1e9 / samplerate





def counts(data: np.ndarray, threshold: float) -> int:
    """
    Compute the number of positive threshold crossings of a hit (counts).

    Args:
        data: Input array
        threshold: Threshold amplitude

    Returns:
        Number of positive threshold crossings
    """
    above_positive_threshold = data >= threshold
    return np.count_nonzero(~above_positive_threshold[:-1] & above_positive_threshold[1:])





def rms(data: np.ndarray) -> float:
    """
    Compute the root mean square (RMS) of an array.

    Args:
        data: Input array

    Returns:
        RMS of the input array

    References:
        https://en.wikipedia.org/wiki/Root_mean_square
    """
    return np.sqrt(np.mean(data**2))