这两组函数中区别很是类似,都是一个不改变之前的数组,一个改变数组本身
resize和reshape
>>> import numpy as np
>>> a = np.arange(20).reshape(4,5)
>>> a
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19]])
>>> a.reshape(2,10)
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14, 15, 16, 17, 18, 19]])
>>> a
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19]])
>>> a.resize(2,10)
>>> a
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14, 15, 16, 17, 18, 19]])
ravel和flatten
两者都可以将数组转换为一个维,
flatten(order='C')
参数:{‘C’,‘F’,‘A’,‘K’}
默认情况下‘C’以行为主的顺序展开,‘F’(Fortran风格)意味着以列的顺序展开,‘A’表示如果a在内存中为Fortran连续,则按列展开,否则以行展开,‘K’按照元素在内存中出现的顺序展平a。
>>> a = np.arange(6).reshape(2,3)
>>> a.flatten()
array([0, 1, 2, 3, 4, 5])
>>> a.ravel()
array([0, 1, 2, 3, 4, 5])
>>> a.flatten('F')
array([0, 3, 1, 4, 2, 5])
>>> a.ravel('F')
array([0, 3, 1, 4, 2, 5])
>>>
>>> x = np.array([[1,2],[3,4]])
>>> a = np.arange(6).reshape(2,3)
>>> a.flatten()[...] = 1
>>> a
array([[0, 1, 2],
[3, 4, 5]])
>>> a.ravel()[...] = 1
>>> a
array([[1, 1, 1],
[1, 1, 1]])
>>>
flatten不会影响原始矩阵,返回的是一个副本,但是ravel是会修改数组