Function _get_multi_p_norm

stumpy/maamp.py:549–588

Multi-dimensional wrapper to compute the p-norm between the query, `T[:, start:start+m])` and the time series, `T`. Additionally, compute p-norm for the first window. Parameters ---------- start : int The window index for T_B from which to calculate the QT dot produ

(start, T, m, p=2.0)

def _get_multi_p_norm(start, T, m, p=2.0):
    """
    Multi-dimensional wrapper to compute the p-norm between the
    query, `T[:, start:start+m])` and the time series, `T`. Additionally, compute
    p-norm for the first window.

    Parameters
    ----------
    start : int
        The window index for T_B from which to calculate the QT dot product

    T : numpy.ndarray
        The time series or sequence for which to compute the dot product

    m : int
        Window size

    p : float
        The p-norm to apply for computing the Minkowski distance. Minkowski distance is
        typically used with `p` being 1 or 2, which correspond to the Manhattan distance
        and the Euclidean distance, respectively.

    Returns
    -------
    p_norm : numpy.ndarray
        Given `start`, return the corresponding multi-dimensional p-norm

    p_norm_first : numpy.ndarray
        Multi-dimensional p-norm for the first window
    """
    d = T.shape[0]
    l = T.shape[1] - m + 1

    p_norm = np.empty((d, l), dtype=np.float64)
    p_norm_first = np.empty((d, l), dtype=np.float64)
    for i in range(d):
        p_norm[i] = np.power(core.mass_absolute(T[i, start : start + m], T[i], p=p), p)
        p_norm_first[i] = np.power(core.mass_absolute(T[i, :m], T[i], p=p), p)

    return p_norm, p_norm_first


@njit(