在Python中协调np.fromiter和多维数组 [英] Reconcile np.fromiter and multidimensional arrays in Python
问题描述
我喜欢使用numpy
中的np.fromiter
,因为它是构建np.array
对象的资源延迟方式.但是,似乎它不支持多维数组,多维数组也非常有用.
I love using np.fromiter
from numpy
because it is a resource-lazy way to build np.array
objects. However, it seems like it doesn't support multidimensional arrays, which are quite useful as well.
import numpy as np
def fun(i):
""" A function returning 4 values of the same type.
"""
return tuple(4*i + j for j in range(4))
# Trying to create a 2-dimensional array from it:
a = np.fromiter((fun(i) for i in range(5)), '4i', 5) # fails
# This function only seems to work for 1D array, trying then:
a = np.fromiter((fun(i) for i in range(5)),
[('', 'i'), ('', 'i'), ('', 'i'), ('', 'i')], 5) # painful
# .. `a` now looks like a 2D array but it is not:
a.transpose() # doesn't work as expected
a[0, 1] # too many indices (of course)
a[:, 1] # don't even think about it
在保持这种基于生成器的惰性构造的同时,如何使a
成为多维数组?
How can I get a
to be a multidimensional array while keeping such a lazy construction based on generators?
推荐答案
np.fromiter
仅支持构造一维数组,因此,它期望一个可迭代的对象将产生单个值,而不是元组/列表/序列等.解决此限制的一种方法是使用 itertools.chain.from_iterable
来解包"输出生成器表达式转换为单个一维值序列:
By itself, np.fromiter
only supports constructing 1D arrays, and as such, it expects an iterable that will yield individual values rather than tuples/lists/sequences etc. One way to work around this limitation would be to use itertools.chain.from_iterable
to lazily 'unpack' the output of your generator expression into a single 1D sequence of values:
import numpy as np
from itertools import chain
def fun(i):
return tuple(4*i + j for j in range(4))
a = np.fromiter(chain.from_iterable(fun(i) for i in range(5)), 'i', 5 * 4)
a.shape = 5, 4
print(repr(a))
# array([[ 0, 1, 2, 3],
# [ 4, 5, 6, 7],
# [ 8, 9, 10, 11],
# [12, 13, 14, 15],
# [16, 17, 18, 19]], dtype=int32)
这篇关于在Python中协调np.fromiter和多维数组的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!