1.数组变形
1.1 numpy.ndarray.shape()
【例】通过修改 shape 属性来改变数组的形状。
import numpy as np
x = np.array([1, 2, 9, 4, 5, 6, 7, 8])
print(x.shape) # (8,)
x.shape = [2, 4]
print(x)
# [[1 2 9 4]
# [5 6 7 8]]
1.2 array.flat ()
【例】将数组转换为一维的迭代器,可以用for访问数组每一个元素,转换后返回的是视图。
import numpy as np
x = np.array([[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20],
[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30],
[31, 32, 33, 34, 35]])
y = x.flat
for i in y:
print(i, end=' ')
# 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
y[3] = 0
print(end='\n')
print(x)
# [[11 12 13 0 15]
# [16 17 18 19 20]
# [21 22 23 24 25]
# [26 27 28 29 30]
# [31 32 33 34 35]]
1.3 array.flatten(order='C')和numpy.ravel(array,order='C')
array.flatten(order='C'):将数组的副本转换为一维数组,并返回。
order:'C' ——按行,'F' ——按列,'A' ——原顺序,'k' ——元素在内存中的出现顺序。
【例】flatten函数返回副本。
import numpy as np
x = np.array([[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20],
[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30],
[31, 32, 33, 34, 35]])
y = x.flatten()
print(y)
# [11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
# 35]
y[3] = 0
print(x)
# [[11 12 13 14 15]
# [16 17 18 19 20]
# [21 22 23 24 25]
# [26 27 28 29 30]
# [31 32 33 34 35]]
y = x.flatten(order='F')
print(y)
# [11 16 21 26 31 12 17 22 27 32 13 18 23 28 33 14 19 24 29 34 15 20 25 30
# 35]
numpy.ravel(array,order='C'):将数组的视图转换为一维数组,并返回。
【例】ravel函数返回的是视图。
import numpy as np
x = np.array([[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20],
[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30],
[31, 32, 33, 34, 35]])
y = np.ravel(x)
print(y)
# [11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
# 35]
y[3] = 0
print(x)
# [[11 12 13 0 15]
# [16 17 18 19 20]
# [21 22 23 24 25]
# [26 27 28 29 30]
# [31 32 33 34 35]]
【例】order=F 就是拷贝
x = np.array([[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20],
[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30],
[31, 32, 33, 34, 35]])
y = np.ravel(x, order='F')
print(y)
# [11 16 21 26 31 12 17 22 27 32 13 18 23 28 33 14 19 24 29 34 15 20 25 30
# 35]
y[3] = 0
print(x)
# [[11 12 13 14 15]
# [16 17 18 19 20]
# [21 22 23 24 25]
# [26 27 28 29 30]
# [31 32 33 34 35]]
1.4 numpy.reshape(array,newshape,order='C')
在不更改数据的情况下为数组赋予新的形状。
当参数newshape=[rows,-1]时,将根据行数自动确定列数。
【例】
import numpy as np
x = np.arange(12)
y = np.reshape(x, [3, 4])
print(y.dtype) # int32
print(y)
# [[ 0 1 2 3]
# [ 4 5 6 7]
# [ 8 9 10 11]]
y = np.reshape(x, [3, -1])
print(y)
# [[ 0 1 2 3]
# [ 4 5 6 7]
# [ 8 9 10 11]]
y = np.reshape(x,[-1,3])
print(y)
# [[ 0 1 2]
# [ 3 4 5]
# [ 6 7 8]
# [ 9 10 11]]
y[0, 1] = 10
print(x)
# [ 0 10 2 3 4 5 6 7 8 9 10 11](改变x进行reshape后y中的值,x对应元素也改变)
【例】reshape()
函数当参数newshape = -1
时,表示将数组降为一维。
import numpy as np
x = np.random.randint(12, size=[2, 2, 3])
print(x)
# [[[11 9 1]
# [ 1 10 3]]
#
# [[ 0 6 1]
# [ 4 11 3]]]
y = np.reshape(x, -1)
print(y)
# [11 9 1 1 10 3 0 6 1 4 11 3]
2.数组转置
- numpy.transpose(array,axis=None)
- array.T
【例】
import numpy as np
x = np.random.rand(5, 5) * 10
x = np.around(x, 2)
print(x)
# [[6.74 8.46 6.74 5.45 1.25]
# [3.54 3.49 8.62 1.94 9.92]
# [5.03 7.22 1.6 8.7 0.43]
# [7.5 7.31 5.69 9.67 7.65]
# [1.8 9.52 2.78 5.87 4.14]]
y = x.T
print(y)
# [[6.74 3.54 5.03 7.5 1.8 ]
# [8.46 3.49 7.22 7.31 9.52]
# [6.74 8.62 1.6 5.69 2.78]
# [5.45 1.94 8.7 9.67 5.87]
# [1.25 9.92 0.43 7.65 4.14]]
y = np.transpose(x)
print(y)
# [[6.74 3.54 5.03 7.5 1.8 ]
# [8.46 3.49 7.22 7.31 9.52]
# [6.74 8.62 1.6 5.69 2.78]
# [5.45 1.94 8.7 9.67 5.87]
# [1.25 9.92 0.43 7.65 4.14]]
3.更改维度
3.1 给数组增加维度:numpy.newaxis=None(None为轴名)
【例】很多工具包在进行计算时都会先判断输入数据的维度是否满足要求,如果输入数据达不到指定的维度时,可以使用newaxis
参数来增加一个维度。
import numpy as np
x = np.array([1, 2, 9, 4, 5, 6, 7, 8])
print(x.shape) # (8,)
print(x) # [1 2 9 4 5 6 7 8]
y = x[np.newaxis, :]
print(y.shape) # (1, 8)
print(y) # [[1 2 9 4 5 6 7 8]]
y = x[:, np.newaxis]
print(y.shape) # (8, 1)
print(y)
# [[1]
# [2]
# [9]
# [4]
# [5]
# [6]
# [7]
# [8]]
3.2 删除数组中的单维度条目,删去shape中为1的维度
numpy.squeeze(array,axis=None):axis为要删除的维度,必须是单维度
在机器学习和深度学习中,通常算法的结果是可以表示向量的数组(即包含两对或以上的方括号形式[[]]),如果直接利用这个数组进行画图可能显示界面为空(见后面的示例)。我们可以利用squeeze()
函数将表示向量的数组转换为秩为1的数组,这样利用 matplotlib 库函数画图时,就可以正常的显示结果了。
【例】
import numpy as np
x = np.arange(10)
print(x.shape) # (10,)
x = x[np.newaxis, :]
print(x.shape) # (1, 10)
y = np.squeeze(x)
print(y.shape) # (10,)
【例】
import numpy as np
x = np.array([[[0], [1], [2]]])
print(x.shape) # (1, 3, 1) #判断shape:三个值分别为:第一层[]内包含的元素,第二层[]内包含的元素,第三层[]内包含的元素
print(x)
# [[[0]
# [1]
# [2]]]
y = np.squeeze(x)
print(y.shape) # (3,)
print(y) # [0 1 2]
y = np.squeeze(x, axis=0) #axis=0:剥去最外层[],第一维
print(y.shape) # (3, 1)
print(y)
# [[0]
# [1]
# [2]]
y = np.squeeze(x, axis=2) #axis=2:剥去最内层[],第三维
print(y.shape) # (1, 3)
print(y) # [[0 1 2]]
y = np.squeeze(x, axis=1) #axis=1:剥去中间层[],第二维不是单维度,不可删去
# ValueError: cannot select an axis to squeeze out which
4.数组组合
4.1 numpy.concatenate((a1, a2, ...), axis=0, out=None)
【例】连接沿现有轴的数组序列(原来x,y都是一维的,拼接后的结果也是一维的)。
import numpy as np
x = np.array([1, 2, 3])
y = np.array([7, 8, 9])
z = np.concatenate([x, y])
print(z)
# [1 2 3 7 8 9]
z = np.concatenate([x, y], axis=0)
print(z)
# [1 2 3 7 8 9]
【例】原来x,y都是二维的,拼接后的结果也是二维的。
import numpy as np
x = np.array([1, 2, 3]).reshape(1, 3)
y = np.array([7, 8, 9]).reshape(1, 3)
z = np.concatenate([x, y])
print(z)
# [[ 1 2 3]
# [ 7 8 9]]
z = np.concatenate([x, y], axis=0)
print(z)
# [[ 1 2 3]
# [ 7 8 9]]
z = np.concatenate([x, y], axis=1)
print(z)
# [[ 1 2 3 7 8 9]]
【例】x,y在原来的维度上进行拼接。
import numpy as np
x = np.array([[1, 2, 3], [4, 5, 6]])
y = np.array([[7, 8, 9], [10, 11, 12]])
z = np.concatenate([x, y])
print(z)
# [[ 1 2 3]
# [ 4 5 6]
# [ 7 8 9]
# [10 11 12]]
z = np.concatenate([x, y], axis=0)
print(z)
# [[ 1 2 3]
# [ 4 5 6]
# [ 7 8 9]
# [10 11 12]]
z = np.concatenate([x, y], axis=1)
print(z)
# [[ 1 2 3 7 8 9]
# [ 4 5 6 10 11 12]]
4.2 numpy.stack(arrays, axis=0, out=None)
【例】沿着新的轴加入一系列数组(stack为增加维度的拼接)。
import numpy as np
x = np.array([1, 2, 3])
y = np.array([7, 8, 9])
z = np.stack([x, y])
print(z.shape) # (2, 3)
print(z)
# [[1 2 3]
# [7 8 9]]
z = np.stack([x, y], axis=1)
print(z.shape) # (3, 2)
print(z)
# [[1 7]
# [2 8]
# [3 9]]
【例】
import numpy as np
x = np.array([1, 2, 3]).reshape(1, 3)
y = np.array([7, 8, 9]).reshape(1, 3)
z = np.stack([x, y])
print(z.shape) # (2, 1, 3)
print(z)
# [[[1 2 3]]
#
# [[7 8 9]]]
z = np.stack([x, y], axis=1)
print(z.shape) # (1, 2, 3)
print(z)
# [[[1 2 3]
# [7 8 9]]]
z = np.stack([x, y], axis=2)
print(z.shape) # (1, 3, 2)
print(z)
# [[[1 7]
# [2 8]
# [3 9]]]
【例】
import numpy as np
x = np.array([[1, 2, 3], [4, 5, 6]])
y = np.array([[7, 8, 9], [10, 11, 12]])
z = np.stack([x, y])
print(z.shape) # (2, 2, 3)
print(z)
# [[[ 1 2 3]
# [ 4 5 6]]
#
# [[ 7 8 9]
# [10 11 12]]]
z = np.stack([x, y], axis=1)
print(z.shape) # (2, 2, 3)
print(z)
# [[[ 1 2 3]
# [ 7 8 9]]
#
# [[ 4 5 6]
# [10 11 12]]]
z = np.stack([x, y], axis=2)
print(z.shape) # (2, 3, 2)
print(z)
# [[[ 1 7]
# [ 2 8]
# [ 3 9]]
#
# [[ 4 10]
# [ 5 11]
# [ 6 12]]]
4.3竖直/水平连接
numpy.vstack(tup)
:竖直连接
numpy.hstack(tup)
:水平连接
hstack(),vstack()
分别表示水平和竖直的拼接方式。在数据维度等于1时,比较特殊。而当维度大于或等于2时,它们的作用相当于concatenate
,用于在已有轴上进行操作。
【例】一维的情况。
import numpy as np
x = np.array([1, 2, 3])
y = np.array([7, 8, 9])
z = np.vstack((x, y))
print(z.shape) # (2, 3)
print(z)
# [[1 2 3]
# [7 8 9]]
z = np.stack([x, y])
print(z.shape) # (2, 3)
print(z)
# [[1 2 3]
# [7 8 9]]
z = np.hstack((x, y))
print(z.shape) # (6,)
print(z)
# [1 2 3 7 8 9]
z = np.concatenate((x, y))
print(z.shape) # (6,)
print(z) # [1 2 3 7 8 9]
【例】二维的情况。
import numpy as np
x = np.array([1, 2, 3]).reshape(1, 3)
y = np.array([7, 8, 9]).reshape(1, 3)
z = np.vstack((x, y))
print(z.shape) # (2, 3)
print(z)
# [[1 2 3]
# [7 8 9]]
z = np.concatenate((x, y), axis=0)
print(z.shape) # (2, 3)
print(z)
# [[1 2 3]
# [7 8 9]]
z = np.hstack((x, y))
print(z.shape) # (1, 6)
print(z)
# [[ 1 2 3 7 8 9]]
z = np.concatenate((x, y), axis=1)
print(z.shape) # (1, 6)
print(z)
# [[1 2 3 7 8 9]]
【例】二维的情况。
import numpy as np
x = np.array([[1, 2, 3], [4, 5, 6]])
y = np.array([[7, 8, 9], [10, 11, 12]])
z = np.vstack((x, y))
print(z.shape) # (4, 3)
print(z)
# [[ 1 2 3]
# [ 4 5 6]
# [ 7 8 9]
# [10 11 12]]
z = np.concatenate((x, y), axis=0)
print(z.shape) # (4, 3)
print(z)
# [[ 1 2 3]
# [ 4 5 6]
# [ 7 8 9]
# [10 11 12]]
z = np.hstack((x, y))
print(z.shape) # (2, 6)
print(z)
# [[ 1 2 3 7 8 9]
# [ 4 5 6 10 11 12]]
z = np.concatenate((x, y), axis=1)
print(z.shape) # (2, 6)
print(z)
# [[ 1 2 3 7 8 9]
# [ 4 5 6 10 11 12]]
5.数组拆分
5.1 拆分数组numpy.split(ary, indices_or_sections, axis=0)
indices_or_sections
indices_or_sections的类型为int或者一维数组,表示一个索引,也就是切的位置所在。indices_or_sections的值如果是一个整数的话,就用这个数平均分割原数组。indices_or_sections的值如果是一个数组的话,就以数组中的数字为索引切开,意味每次切到前n个,后面的在抛去前面的基础上增加,最后一个数字后还有一片,若已切完要加一个空数组。例子如下
————————————————
版权声明:本文为CSDN博主「Kobaayyy」的原创文章,遵循CC 4.0 BY-SA版权协议,转载请附上原文出处链接及本声明。
原文链接:https://blog.csdn.net/Kobaayyy/article/details/105021365
【例】拆分数组。
import numpy as np
x = np.array([[11, 12, 13, 14],
[16, 17, 18, 19],
[21, 22, 23, 24]])
y = np.split(x, [1, 3]) #拆分的位置为:按行分第一片切前一个,第二片切到前三个,剩下一片为空。
print(y)
# [array([[11, 12, 13, 14]]),
# array([[16, 17, 18, 19],
# [21, 22, 23, 24]]),
# array([], shape=(0, 4), dtype=int32)]
y = np.split(x, [1, 3], axis=1)
print(y)
# [array([[11],
# [16],
# [21]]),
# array([[12, 13],
# [17, 18],
# [22, 23]]),
# array([[14],
# [19],
# [24]])]
5.2垂直切、水平切
垂直切:numpy.vsplit(ary, indices_or_sections)
【例】垂直切分是把数组按照高度切分(按行切,anis=0)
import numpy as np
x = np.array([[11, 12, 13, 14],
[16, 17, 18, 19],
[21, 22, 23, 24]])
y = np.vsplit(x, 3)
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19]]), array([[21, 22, 23, 24]])]
y = np.split(x, 3)
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19]]), array([[21, 22, 23, 24]])]
y = np.vsplit(x, [1])
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19],
# [21, 22, 23, 24]])]
y = np.split(x, [1])
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19],
# [21, 22, 23, 24]])]
y = np.vsplit(x, [1, 3])
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19],
# [21, 22, 23, 24]]), array([], shape=(0, 4), dtype=int32)]
y = np.split(x, [1, 3], axis=0)
print(y)
# [array([[11, 12, 13, 14]]), array([[16, 17, 18, 19],
# [21, 22, 23, 24]]), array([], shape=(0, 4), dtype=int32)]
水平切:numpy.hsplit(ary, indices_or_sections)
【例】水平切分是把数组按照宽度切分(按列切,anix=1)。
import numpy as np
x = np.array([[11, 12, 13, 14],
[16, 17, 18, 19],
[21, 22, 23, 24]])
y = np.hsplit(x, 2)
print(y)
# [array([[11, 12],
# [16, 17],
# [21, 22]]), array([[13, 14],
# [18, 19],
# [23, 24]])]
y = np.split(x, 2, axis=1)
print(y)
# [array([[11, 12],
# [16, 17],
# [21, 22]]), array([[13, 14],
# [18, 19],
# [23, 24]])]
y = np.hsplit(x, [3])
print(y)
# [array([[11, 12, 13],
# [16, 17, 18],
# [21, 22, 23]]), array([[14],
# [19],
# [24]])]
y = np.split(x, [3], axis=1)
print(y)
# [array([[11, 12, 13],
# [16, 17, 18],
# [21, 22, 23]]), array([[14],
# [19],
# [24]])]
y = np.hsplit(x, [1, 3])
print(y)
# [array([[11],
# [16],
# [21]]), array([[12, 13],
# [17, 18],
# [22, 23]]), array([[14],
# [19],
# [24]])]
y = np.split(x, [1, 3], axis=1)
print(y)
# [array([[11],
# [16],
# [21]]), array([[12, 13],
# [17, 18],
# [22, 23]]), array([[14],
# [19],
# [24]])]
6.数组平铺
平铺:numpy.tile(A, reps)
【例】将原矩阵横向、纵向地复制。
import numpy as np
x = np.array([[1, 2], [3, 4]])
print(x)
# [[1 2]
# [3 4]]
y = np.tile(x, (1, 3)) #改坐标代表:将原矩阵在每个方向上复制几次
print(y) #铺成一个什么格式
# [[1 2 1 2 1 2]
# [3 4 3 4 3 4]]
y = np.tile(x, (3, 1))
print(y)
# [[1 2]
# [3 4]
# [1 2]
# [3 4]
# [1 2]
# [3 4]]
y = np.tile(x, (3, 3))
print(y)
# [[1 2 1 2 1 2]
# [3 4 3 4 3 4]
# [1 2 1 2 1 2]
# [3 4 3 4 3 4]
# [1 2 1 2 1 2]
# [3 4 3 4 3 4]]
重复数组:numpy.repeat(a, repeats, axis=None)
axis=0
,沿着y轴复制,实际上增加了行数。axis=1
,沿着x轴复制,实际上增加了列数。repeats
,可以为一个数,也可以为一个矩阵。axis=None
时就会flatten当前矩阵,实际上就是变成了一个行向量。
【例】重复数组的元素。
import numpy as np
x = np.repeat(3, 4)
print(x) # [3 3 3 3]
x = np.array([[1, 2], [3, 4]])
y = np.repeat(x, 2)
print(y)
# [1 1 2 2 3 3 4 4]
y = np.repeat(x, 2, axis=0)
print(y)
# [[1 2]
# [1 2]
# [3 4]
# [3 4]]
y = np.repeat(x, 2, axis=1)
print(y)
# [[1 1 2 2]
# [3 3 4 4]]
y = np.repeat(x, [2, 3], axis=0)
print(y)
# [[1 2]
# [1 2]
# [3 4]
# [3 4]
# [3 4]]
y = np.repeat(x, [2, 3], axis=1)
print(y)
# [[1 1 2 2 2]
# [3 3 4 4 4]]
参考文献