当前位置: 首页>>代码示例 >>用法及示例精选 >>正文


Python numpy ufunc.reduceat用法及代码示例


本文简要介绍 python 语言中 numpy.ufunc.reduceat 的用法。

用法:

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

在单个轴上使用指定切片执行(局部)reduce。

对于我在range(len(indices)),reduceat计算ufunc.reduce(array[indices[i]:indices[i+1]]),它变成了i-th广义“row”平行于在最终结果中(即,在二维数组中,例如,如果轴 = 0,它变成i-th 行,但是如果轴 = 1,它成为i-th 列)。这有三个异常:

  • i = len(indices) - 1 (所以对于最后一个索引), indices[i+1] = array.shape[axis]

  • 如果 indices[i] >= indices[i + 1] ,则 i-th 广义 “row” 只是 array[indices[i]]

  • 如果 indices[i] >= len(array)indices[i] < 0 ,则会引发错误。

输出的形状取决于 indices 的大小,并且可能大于 array (如果 len(indices) > array.shape[axis] 会发生这种情况)。

参数

array array_like

要操作的数组。

indices array_like

配对索引,逗号分隔(不是冒号),指定要减少的切片。

axis 整数,可选

沿其应用 reduceat 的轴。

dtype 数据类型代码,可选

用于表示中间结果的类型。如果提供,则默认为输出数组的数据类型,如果未提供输出数组,则默认为输入数组的数据类型。

out ndarray,None,或 ndarray 和 None 的元组,可选

存储结果的位置。如果未提供或 None,则返回一个新分配的数组。为了与 ufunc.__call__ 保持一致,如果作为关键字给出,则可以将其包装在 1 元素元组中。

返回

r ndarray

减少的值。如果提供了 out,则 r 是对 out 的引用。

注意

一个说明性的例子:

如果numpy.array是一维的,函数ufunc.accumulate(array)是相同的ufunc.reduceat(array, indices)[::2]其中numpy.indicesrange(len(array) - 1)在所有其他元素中放置一个零:indices = zeros(2 * len(array) - 1),indices[1::2] = range(1, len(array)).

不要被这个属性的名称所迷惑:减少(数组)不一定小于numpy.array.

例子

取四个连续值的运行总和:

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

二维示例:

>>> 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.]])

相关用法


注:本文由纯净天空筛选整理自numpy.org大神的英文原创作品 numpy.ufunc.reduceat。非经特殊声明,原始代码版权归原作者所有,本译文未经允许或授权,请勿转载或复制。