numpy中有哪些简单且实用的操作

入门numpy

拉平操作 ravel()和faltten()及reshape(1,-1)的区别联系

import numpy as np
x = np.array([[1,2,3],[4,5,6],[1,2,3]])
x.flatten()
array([1, 2, 3, 4, 5, 6, 1, 2, 3])
x.ravel()
array([1, 2, 3, 4, 5, 6, 1, 2, 3])
x.ravel('F')
array([1, 4, 1, 2, 5, 2, 3, 6, 3])
x.flatten('F')
array([1, 4, 1, 2, 5, 2, 3, 6, 3])
x.flatten()[1] = 20
x
array([[1, 2, 3],
       [4, 5, 6],
       [1, 2, 3]])
x.ravel()[1] = 20
x
array([[ 1, 20,  3],
       [ 4,  5,  6],
       [ 1,  2,  3]])
x.reshape(1,-1) # 注意结果仍然是二维
array([[ 1, 20,  3,  4,  5,  6,  1,  2,  3]])

星号的作用

import numpy as np
y1 = np.linspace(-10.0,10.0) # 默认生成50个数据
y1
array([-10.        ,  -9.59183673,  -9.18367347,  -8.7755102 ,
        -8.36734694,  -7.95918367,  -7.55102041,  -7.14285714,
        -6.73469388,  -6.32653061,  -5.91836735,  -5.51020408,
        -5.10204082,  -4.69387755,  -4.28571429,  -3.87755102,
        -3.46938776,  -3.06122449,  -2.65306122,  -2.24489796,
        -1.83673469,  -1.42857143,  -1.02040816,  -0.6122449 ,
        -0.20408163,   0.20408163,   0.6122449 ,   1.02040816,
         1.42857143,   1.83673469,   2.24489796,   2.65306122,
         3.06122449,   3.46938776,   3.87755102,   4.28571429,
         4.69387755,   5.10204082,   5.51020408,   5.91836735,
         6.32653061,   6.73469388,   7.14285714,   7.55102041,
         7.95918367,   8.36734694,   8.7755102 ,   9.18367347,
         9.59183673,  10.        ])
y2 = np.linspace(1,10,10) # 生成10个数据,包括首尾
y2
array([ 1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10.])
y3 = np.linspace(1,10,10,endpoint=False) # 不包括尾部数据
y3
array([1. , 1.9, 2.8, 3.7, 4.6, 5.5, 6.4, 7.3, 8.2, 9.1])
y4= np.linspace(1, 10, 6, retstep=True) # 将步长与结果的数组放入一个list、
y4
(array([ 1. ,  2.8,  4.6,  6.4,  8.2, 10. ]), 1.8)

np.hstack()和np.vstack()用于堆叠矩阵

import numpy as np
x=np.array([[3,4,5],[1,3,4]])
y=np.array([[1,1,1],[2,2,2]])
np.hstack((x,y))
array([[3, 4, 5, 1, 1, 1],
       [1, 3, 4, 2, 2, 2]])
np.vstack((x,y))
array([[3, 4, 5],
       [1, 3, 4],
       [1, 1, 1],
       [2, 2, 2]])

np.meshgrid()快速生成网格

import numpy as np
x=np.array([1,3,5])
y=np.array([4,6])
XX,YY=np.meshgrid(x,y)
XX
array([[1, 3, 5],
       [1, 3, 5]])
YY
array([[4, 4, 4],
       [6, 6, 6]])

np.newaxis在特定位置增加一个维度

import numpy as np
c=np.array([1,2,5,4])
c[:,np.newaxis]
array([[1],
       [2],
       [5],
       [4]])
c[np.newaxis,:]
array([[1, 2, 5, 4]])

np.prod() 计算元素乘积

import numpy as np
x = np.array([[1,2,3],[2,3,4]])
np.prod(x)
144
np.prod(x,axis=1)
array([ 6, 24])
np.prod(x,axis=0)
array([ 2,  6, 12])

np.pad()函数进行padding操作

import numpy as np
A= np.arange(95,99).reshape(2,2)
A
array([[95, 96],
       [97, 98]])
np.pad(A,((3,2),(2,3)),'constant',constant_values=(0,0))
array([[ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0, 95, 96,  0,  0,  0],
       [ 0,  0, 97, 98,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0]])
b=np.array([[[1,2],[3,4]],[[3,4],[7,8]],[[4,5],[1,2]]])
b
array([[[1, 2],
        [3, 4]],

       [[3, 4],
        [7, 8]],

       [[4, 5],
        [1, 2]]])
np.pad(b,((0,0),(1,1),(1,1)),'constant' ,constant_values = 0)
array([[[0, 0, 0, 0],
        [0, 1, 2, 0],
        [0, 3, 4, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 3, 4, 0],
        [0, 7, 8, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 4, 5, 0],
        [0, 1, 2, 0],
        [0, 0, 0, 0]]])

np.full生成形为shape的元素全为val的矩阵

import numpy as np
c=np.array([[1,2],[3,4]])
c
array([[1, 2],
       [3, 4]])
c.astype(np.float32)
array([[1., 2.],
       [3., 4.]], dtype=float32)

import numpy as np
x=np.array([[3,4,5],[1,3,4]])
y=np.array([[1,1,1],[2,2,2]])
np.hstack((x,y))
array([[3, 4, 5, 1, 1, 1],
       [1, 3, 4, 2, 2, 2]])
np.vstack((x,y))
array([[3, 4, 5],
       [1, 3, 4],
       [1, 1, 1],
       [2, 2, 2]])

y=np.array([4,6])
XX,YY=np.meshgrid(x,y)
XX
array([[1, 3, 5],
       [1, 3, 5]])
YY
array([[4, 4, 4],
       [6, 6, 6]])

import numpy as np
c=np.array([1,2,5,4])
c[:,np.newaxis]
array([[1],
       [2],
       [5],
       [4]])
c[np.newaxis,:]
array([[1, 2, 5, 4]])

import numpy as np
A= np.arange(95,99).reshape(2,2)
A
array([[95, 96],
       [97, 98]])
np.pad(A,((3,2),(2,3)),'constant',constant_values=(0,0))
array([[ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0, 95, 96,  0,  0,  0],
       [ 0,  0, 97, 98,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0]])
b=np.array([[[1,2],[3,4]],[[3,4],[7,8]],[[4,5],[1,2]]])
b
array([[[1, 2],
        [3, 4]],

       [[3, 4],
        [7, 8]],

       [[4, 5],
        [1, 2]]])
np.pad(b,((0,0),(1,1),(1,1)),'constant' ,constant_values = 0)
array([[[0, 0, 0, 0],
        [0, 1, 2, 0],
        [0, 3, 4, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 3, 4, 0],
        [0, 7, 8, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 4, 5, 0],
        [0, 1, 2, 0],
        [0, 0, 0, 0]]])

import numpy as np
x = np.array([[1,2,3],[4,5,6]])
np.zeros_like(x)
array([[0, 0, 0],
       [0, 0, 0]])

import numpy as np
a=np.array([0.125,0.568,5.688])
np.round(a)
array([0., 1., 6.])
np.round(a,decimals=2)
array([0.12, 0.57, 5.69])
np.floor(a)
array([0., 0., 5.])
np.ceil(a)
import numpy as np
x = np.array([[1,2,3],[-3,2,4],[5,-2,9]])
x
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x1 = x.copy() 
x1[x1 > 0] = 0
x1
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x2 = x
x2
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x2[x2>0] = 0
x2
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x 
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x = np.array([[1,2,3],[-3,2,4],[5,-2,9]])
x3 = x[2]
x3
array([ 5, -2,  9])
x3[2] = 100 
x
array([[  1,   2,   3],
       [ -3,   2,   4],
       [  5,  -2, 100]])

import numpy as np
a=np.array([[1,2,3],[4,5,6]])
a=np.array([[1,2,3,6],[4,5,6,6]])
a1=a.reshape((1,2,4))
a1
array([[[1, 2, 3, 6],
        [4, 5, 6, 6]]])
b=np.array([[3,4,5,6],[1,2,3,4],[4,5,5,5]])
b
array([[3, 4, 5, 6],
       [1, 2, 3, 4],
       [4, 5, 5, 5]])
b1=b.reshape((1,3,4)).transpose((1,0,2))
b1
array([[[3, 4, 5, 6]],

       [[1, 2, 3, 4]],

       [[4, 5, 5, 5]]])
a1
array([[[1, 2, 3, 6],
        [4, 5, 6, 6]]])
a1+b1
array([[[ 4,  6,  8, 12],
        [ 7,  9, 11, 12]],

       [[ 2,  4,  6, 10],
        [ 5,  7,  9, 10]],

       [[ 5,  7,  8, 11],
        [ 8, 10, 11, 11]]])

矩阵删除、插入、尾部添加操作(delete,insert,append)

import numpy as np
matrix = [
[1,2,3,4],
[5,6,7,8],
[9,10,11,12]
]
p1=np.delete(matrix,1,0)
print('>>>>p1>>>>\n',p1)
>>>>p1>>>>
 [[ 1  2  3  4]
 [ 9 10 11 12]]
p2 = np.delete(matrix, 1, 1)
print('>>>>p2>>>>\n',p2)
>>>>p2>>>>
 [[ 1  3  4]
 [ 5  7  8]
 [ 9 11 12]]
p3 = np.delete(matrix, 1)
print('>>>>p3>>>>\n',p3)
>>>>p3>>>>
 [ 1  3  4  5  6  7  8  9 10 11 12]
p4 = np.delete(matrix, [0,1], 1) 
print('>>>>p4>>>>\n',p4)
>>>>p4>>>>
 [[ 3  4]
 [ 7  8]
 [11 12]]
import numpy as np
matrix = [
[1,2,3,4],
[5,6,7,8],
[9,10,11,12]
]
q1 = np.insert(matrix, 1, [1,1,1,1], 0)
print('>>>>q1>>>>\n',q1)
>>>>q1>>>>
 [[ 1  2  3  4]
 [ 1  1  1  1]
 [ 5  6  7  8]
 [ 9 10 11 12]]
q2 = np.insert(matrix, 0, [1,1,1], 1)
print('>>>>2>>>>\n',q2)
>>>>2>>>>
 [[ 1  1  2  3  4]
 [ 1  5  6  7  8]
 [ 1  9 10 11 12]]
q3 = np.insert(matrix, 3, [1,1,1,1], 0) 
print('>>>>3>>>>\n',q3)
>>>>3>>>>
 [[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]
 [ 1  1  1  1]]
mport numpy as np
matrix = [
[1,2,3,4],
[5,6,7,8],
[9,10,11,12]
]

  File "<ipython-input-20-d202c3e7f41f>", line 1
    mport numpy as np
          ^
SyntaxError: invalid syntax
m1 = np.append(matrix,[[1,1,1,1]],axis=0)
print('>>>>m1>>>>\n',m1)
m2 = np.append(matrix,[[1],[1],[1]],axis=1)
>>>>m1>>>>
 [[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]
 [ 1  1  1  1]]
print('>>>>m2>>>>\n',m2)
m3 = np.append(matrix,[[1],[1],[1]],axis=1)
>>>>m2>>>>
 [[ 1  2  3  4  1]
 [ 5  6  7  8  1]
 [ 9 10 11 12  1]]
print('>>>>m3>>>>\n',m3)
>>>>m3>>>>
 [[ 1  2  3  4  1]
 [ 5  6  7  8  1]
 [ 9 10 11 12  1]]

把矩阵大于或小于N的元素置M的技巧

import numpy as np
x = np.array([[1,2,3],[-3,2,4],[5,-2,9]])
x
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
y1 = np.maximum(0,x) 
y1
array([[1, 2, 3],
       [0, 2, 4],
       [5, 0, 9]])
y2 = np.minimum(0,x) 
y2
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x1 = x.copy()
x1
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x1[x1 < 0] = 0 
x1
array([[1, 2, 3],
       [0, 2, 4],
       [5, 0, 9]])
x2 = x.copy()
x2[x2 > 0] = 0 
x2
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])

random.rand和random.rand和random.randint区别

import numpy as np
n = np.random.rand(3,4)
n
array([[0.5003975 , 0.8162121 , 0.45193264, 0.82332347],
       [0.09752794, 0.42988813, 0.03935796, 0.8775246 ],
       [0.85181714, 0.52776459, 0.14274664, 0.95744951]])
import numpy as np
x = np.random.randn(2,3)
x
array([[ 0.02193741,  0.758389  , -1.43269265],
       [-0.85326325, -0.37004486, -0.41419479]])
y = np.multiply(0.1,np.random.randn(2,3))+0.5 
y
array([[0.67639181, 0.42187938, 0.53135641],
       [0.57425968, 0.31920886, 0.42239423]])
import numpy as np
z = np.random.randint(2,9,(2,3))
z
array([[2, 7, 2],
       [6, 5, 8]])
m = np.random.randint(9,size = (2,3))
m
array([[8, 0, 4],
       [4, 0, 0]])

python广播机制

import numpy as np
a=np.array([[1,2,3],[4,5,6]])
a=np.array([[1,2,3,6],[4,5,6,6]])
a1=a.reshape((1,2,4))
a1
array([[[1, 2, 3, 6],
        [4, 5, 6, 6]]])
b=np.array([[3,4,5,6],[1,2,3,4],[4,5,5,5]])
b
array([[3, 4, 5, 6],
       [1, 2, 3, 4],
       [4, 5, 5, 5]])
b1=b.reshape((1,3,4)).transpose((1,0,2))
b1
array([[[3, 4, 5, 6]],

       [[1, 2, 3, 4]],

       [[4, 5, 5, 5]]])
a1
array([[[1, 2, 3, 6],
        [4, 5, 6, 6]]])
a1+b1
array([[[ 4,  6,  8, 12],
        [ 7,  9, 11, 12]],

       [[ 2,  4,  6, 10],
        [ 5,  7,  9, 10]],

       [[ 5,  7,  8, 11],
        [ 8, 10, 11, 11]]])

numpy中的矩阵copy问题

import numpy as np
x = np.array([[1,2,3],[-3,2,4],[5,-2,9]])
x
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x1 = x.copy() 
x1[x1 > 0] = 0
x1
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x2 = x
x2
array([[ 1,  2,  3],
       [-3,  2,  4],
       [ 5, -2,  9]])
x2[x2>0] = 0
x2
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x 
array([[ 0,  0,  0],
       [-3,  0,  0],
       [ 0, -2,  0]])
x = np.array([[1,2,3],[-3,2,4],[5,-2,9]])
x3 = x[2]
x3
array([ 5, -2,  9])
x3[2] = 100 
x
array([[  1,   2,   3],
       [ -3,   2,   4],
       [  5,  -2, 100]])

numpy取整

import numpy as np
a=np.array([0.125,0.568,5.688])
np.round(a)
array([0., 1., 6.])
np.round(a,decimals=2)
array([0.12, 0.57, 5.69])
np.floor(a)
array([0., 0., 5.])
np.ceil(a)
array([1., 1., 6.])

numpy基本加减和取行操作

import numpy as np

A = np.arange(1, 13).reshape(6, 2)


B = np.vsplit(A, 3)

C = A.T

D = np.hsplit(C, 3)
print(A, B, C, D)
[[ 1  2]
 [ 3  4]
 [ 5  6]
 [ 7  8]
 [ 9 10]
 [11 12]] [array([[1, 2],
       [3, 4]]), array([[5, 6],
       [7, 8]]), array([[ 9, 10],
       [11, 12]])] [[ 1  3  5  7  9 11]
 [ 2  4  6  8 10 12]] [array([[1, 3],
       [2, 4]]), array([[5, 7],
       [6, 8]]), array([[ 9, 11],
       [10, 12]])]


numpy基本加减和取行操作

import numpy as np
a = np.array([1,1,1,1])
b = np.array([[1],[1],[1],[1]])
a+b
array([[2, 2, 2, 2],
       [2, 2, 2, 2],
       [2, 2, 2, 2],
       [2, 2, 2, 2]])
c = np.array([[1,1,1,1]])
c+b
array([[2, 2, 2, 2],
       [2, 2, 2, 2],
       [2, 2, 2, 2],
       [2, 2, 2, 2]])
W = np.array([[1,1,1],[2,2,2]])
W[:,1]
array([1, 2])
W[1]
array([2, 2, 2])
W[:,1] = np.array([5,5])
W
array([[1, 5, 1],
       [2, 5, 2]])

np.zeros_like()构造全零矩阵,无需指定大小

import numpy as np
x = np.array([[1,2,3],[4,5,6]])
np.zeros_like(x)
array([[0, 0, 0],
       [0, 0, 0]])

np.random.choice(a, size, replace, p);np.argmax(a, axis=None, out=None)

import numpy as np
a1 = np.random.choice(7,5) # 从0~7中随机选择5个数组成一维数组
a1
array([2, 0, 5, 0, 6])
a2 = np.random.choice([0,1,2,3,4,5,6],5) # 从给定list中随机选择5个数组成一维数组
a2
array([0, 6, 1, 6, 6])
a3 = np.random.choice(np.array([0,1,2,3,4,5,6]),5) # 将list换成array数组依然可以运行,效果一致
a3
array([0, 5, 4, 2, 1])
a4 = np.random.choice([0,1,2,3,4,5,6],5,replace=False) # 上述均有重复,将replace设置为False,即可按要求没有重复的选取
a4
array([3, 0, 5, 4, 1])
a5 = np.random.choice(np.array([0,1,2,3,4,5,6]),5,p=[0.1,0.1,0.1,0.1,0.1,0.1,0.4])
a5
array([6, 3, 3, 4, 3])

import numpy as np
a = np.array([[1,1,1],[2,2,2],[0,3,6]])
a
array([[1, 1, 1],
       [2, 2, 2],
       [0, 3, 6]])
b1 = np.argmax(a) # 将数组a拉平,最大值索引为9(初始索引为0)
b1
8

8
b2 = np.argmax(a, axis=0) # 按列选取最大值的索引
b2
array([1, 2, 2], dtype=int64)
b3 = np.argmax(a, axis=1) # 按行选取最大值的索引
b3
array([0, 0, 2], dtype=int64)

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值