原文:https://docs.scipy.org/doc/numpy/reference/generated/numpy.ufunc.reduceat.html
校对:(虚位以待)
ufunc.
reduceat
(a, indices, axis=0, dtype=None, out=None)使用单个轴上的指定切片执行(局部)缩减。
For i in range(len(indices))
, reduceat
computes ufunc.reduce(a[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). 这里有三个例外:
i = len(indices) - 1
(so for the last index), indices[i+1] = a.shape[axis]
.索引[i] = 索引[i + t5>
,第i个广义“行”仅仅是a[indices[i]]
。索引[i] > = len(a)
或] < 0
时,会出现错误。输出的形状取决于indices
的大小,并且可能大于a(如果len(indices) > a.shape [axis]
)。
参数: | a:array_like
indices:array_like
axis:int,可选
dtype:数据类型代码,可选
out:ndarray,可选
|
---|---|
返回: | r:ndarray
|
笔记
一个描述性示例:
If a is 1-D, the function ufunc.accumulate(a) is the same as ufunc.reduceat(a, indices)[::2]
where indices
is range(len(array) - 1)
with a zero placed in every other element: indices = zeros(2 * len(a) - 1)
, indices[1::2] = range(1, len(a))
.
不要被此属性的名称欺骗:reduceat(a)不一定小于a。
例子
要获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])
2-D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.]])