Classe « ufunc »

Méthode numpy.matlib.ufunc.reduceat

Signature de la méthode reduceat

Description

reduceat.__doc__

reduceat(array, indices, axis=0, dtype=None, out=None)

    Performs a (local) reduce with specified slices over a single axis.

    For i in ``range(len(indices))``, `reduceat` computes
    ``ufunc.reduce(array[indices[i]:indices[i+1]])``, which becomes the i-th
    generalized "row" parallel to `axis` in the final result (i.e., in a
    2-D array, for example, if `axis = 0`, it becomes the i-th row, but if
    `axis = 1`, it becomes the i-th column).  There are three exceptions to this:

    * when ``i = len(indices) - 1`` (so for the last index),
      ``indices[i+1] = array.shape[axis]``.
    * if ``indices[i] >= indices[i + 1]``, the i-th generalized "row" is
      simply ``array[indices[i]]``.
    * if ``indices[i] >= len(array)`` or ``indices[i] < 0``, an error is raised.

    The shape of the output depends on the size of `indices`, and may be
    larger than `array` (this happens if ``len(indices) > array.shape[axis]``).

    Parameters
    ----------
    array : array_like
        The array to act on.
    indices : array_like
        Paired indices, comma separated (not colon), specifying slices to
        reduce.
    axis : int, optional
        The axis along which to apply the reduceat.
    dtype : data-type code, optional
        The type used to represent the intermediate results. Defaults
        to the data type of the output array if this is provided, or
        the data type of the input array if no output array is provided.
    out : ndarray, None, or tuple of ndarray and None, optional
        A location into which the result is stored. If not provided or None,
        a freshly-allocated array is returned. For consistency with
        ``ufunc.__call__``, if given as a keyword, this may be wrapped in a
        1-element tuple.

        .. versionchanged:: 1.13.0
           Tuples are allowed for keyword argument.

    Returns
    -------
    r : ndarray
        The reduced values. If `out` was supplied, `r` is a reference to
        `out`.

    Notes
    -----
    A descriptive example:

    If `array` is 1-D, the function `ufunc.accumulate(array)` is the same as
    ``ufunc.reduceat(array, indices)[::2]`` where `indices` is
    ``range(len(array) - 1)`` with a zero placed
    in every other element:
    ``indices = zeros(2 * len(array) - 1)``,
    ``indices[1::2] = range(1, len(array))``.

    Don't be fooled by this attribute's name: `reduceat(array)` is not
    necessarily smaller than `array`.

    Examples
    --------
    To take the running sum of four successive values:

    >>> np.add.reduceat(np.arange(8),[0,4, 1,5, 2,6, 3,7])[::2]
    array([ 6, 10, 14, 18])

    A 2-D example:

    >>> x = np.linspace(0, 15, 16).reshape(4,4)
    >>> x
    array([[ 0.,   1.,   2.,   3.],
           [ 4.,   5.,   6.,   7.],
           [ 8.,   9.,  10.,  11.],
           [12.,  13.,  14.,  15.]])

    ::

     # reduce such that the result has the following five rows:
     # [row1 + row2 + row3]
     # [row4]
     # [row2]
     # [row3]
     # [row1 + row2 + row3 + row4]

    >>> np.add.reduceat(x, [0, 3, 1, 2, 0])
    array([[12.,  15.,  18.,  21.],
           [12.,  13.,  14.,  15.],
           [ 4.,   5.,   6.,   7.],
           [ 8.,   9.,  10.,  11.],
           [24.,  28.,  32.,  36.]])

    ::

     # reduce such that result has the following two columns:
     # [col1 * col2 * col3, col4]

    >>> np.multiply.reduceat(x, [0, 3], 1)
    array([[   0.,     3.],
           [ 120.,     7.],
           [ 720.,    11.],
           [2184.,    15.]])