Numpy narray对象的属性分析

参考官方文档链接:

narray是Numpy的基本数据结构,本文主要分析对象的属性(可通过.进行访问)

1:导入numpy:

import numpy as np

2:初始化narray对象:

>>> x = np.array([[1, 2, 3], [4, 5, 6]], np.int32)
>>> x
array([[1, 2, 3],
       [4, 5, 6]], dtype=int32)

3:查看np对象的行列sharp(np.shape)(返回两个元素元组,分别是行,列.):

>>> x.shape
(2, 3)

4:查看np对象的内存布局(np.flags)(详情点这里):

>>> x.flags
  C_CONTIGUOUS : True              :The data is in a single, C-style contiguous segment.
  F_CONTIGUOUS : False             :The data is in a single, Fortran-style contiguous segment.
  OWNDATA : True                   :The array owns the memory it uses or borrows it from another object.
  WRITEABLE : True                 :The data area can be written to.
  ALIGNED : True                   :The data and all elements are aligned appropriately for the hardware.
  UPDATEIFCOPY : False             :(Deprecated, use WRITEBACKIFCOPY) This array is a copy of some other array. When this array is deallocated, the base array will be updated with the contents of this array.

5:查看数组的大小:(np.size)(即所有元素个数Number of elements in the array.):

>>> x.size
6

6:遍历数组时，在每个维度中步进的字节数组(np.strides)(Tuple of bytes to step in each dimension when traversing an array.):

>>> x
array([[1, 2, 3],
       [4, 5, 6]], dtype=int32)
>>> x.strides
(12, 4)
以本片代码为例:int32位占据4个字节的数据,因此同行内移动一个数据至相邻的列需要4个字节,移动到下一行相同列需要(元素大小4*列数3)12个字节
>>> x = np.array([[1, 2, 3], [4, 5, 6]], np.int64)
>>> x.strides
(24, 8)

7:查看数组维度(np.ndim)(Number of array dimensions.):

>>> x.ndim
2

8:查看数组内存缓冲区的开始位置(np.data)(Python buffer object pointing to the start of the array’s data.):

>>> x.data
<memory at 0x7f49c189a990>

9:查看数组每一个元素所占的内存大小(np.itemsize)(Length of one array element in bytes.):

>>> x = np.array([1, 2], np.complex128)
>>> x.itemsize
16
>>> x = np.array([1, 2], np.int16)
>>> x.itemsize

10:查看数组元素消耗的总字节(np.nbytes)(Total bytes consumed by the elements of the array.):

>>> x = np.array([1, 2], np.int16)
>>> x.nbytes
4

11:查看数组的基对象(np.base)(Base object if memory is from some other object.)

>>> x = np.array([[1, 2, 3], [4, 5, 6]], np.int64)
>>> x.base
>>> y = x[1:]     (分片后的对象与原对象共享内存)
>>> y.base
array([[1, 2, 3],
       [4, 5, 6]])

请以官方文档为准,有问题可以留言,