numpy

  • 矩阵等计算,还要了解基本的高数,概率,线代知识,这些都要用到。特别是矩阵的转换以及算法等等。
  • 详细学习参考学习资料
    NumPy 教程 | 菜鸟教程

查找xxxx的帮助文档

print(help(numpy.xxxx))

构造多维数组

numpy.array()

注意里面的值必须为同一类型,否则有类型转换;

eg:

# 构造数组
# 一维数组
vector = numpy.array([5,10,15,20])
# 二维数组,注意有两个括号
matrix = numpy.array([[5,10,15],[20,25,30],[35,40,45]])
print(vector)
print(matrix)

如何构造三维以上的数组是一个难点所在
mumpy/0.ipynb/13*14

查看xxx的结构/用于debug

print(xxxx.shape)

查询

print(xxxx[x,y])
print(xxxxx[,1]) # 取第一列
print(xxxx[:,0:2]) #第一和第二列

切片

print(xxxx[n:m])

判断值是否存在

xxx == m # xxx 中是否有m/会判断每一个值

整体值类型改变

v = numpy.array(["1","2","3"])
print(v.dtype)
print(v)
v = v.astype(float)
print(v.dtype)
print(v)

求极值

指定维度求值

# 对行求值
m = numpy.array([
    [5,10,15],
    [20,25,30],
    [35,40,45]
])
m.sum(axis=1)
# answer:
# array([ 30,  75, 120])
#对列求值
m = numpy.array([
    [5,10,15],
    [20,25,30],
    [35,40,45]
])
m.sum(axis=0)#answer:
#array([60, 75, 90])

变换为矩阵

import numpy as np
print(np.arange(15))
a = np.arange(15).reshape(3,5)
a
# [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14]
# array([[ 0,  1,  2,  3,  4],
#        [ 5,  6,  7,  8,  9],
#        [10, 11, 12, 13, 14]])

求出array维度

print(xxx.ndim)

求出array元素个数

print(xxx.size)

初始化矩阵为0

import numpy as np
np.zeros((3,4))
# array([[0., 0., 0., 0.],
#        [0., 0., 0., 0.],
#        [0., 0., 0., 0.]])

指定类型

np.ones((2,3,4),dtype=np.int32)

# array([[[1, 1, 1, 1],
#         [1, 1, 1, 1],
#         [1, 1, 1, 1]],

#        [[1, 1, 1, 1],
#         [1, 1, 1, 1],
#         [1, 1, 1, 1]]])

得出一个序列

np.arange(10,30,5)#该数>10 且< 30 从10开始每次加5
np.arange(10,30,5).reshape(4,3)# 注意元素个数是否够用
# array([[10, 15],
#        [20, 25]])

随机模块

np.random.random((2,3)) # 第一个random是调用模块,第二个是调用函数,(2,3)是构造一个2*3的矩阵
# array([[0.05134094, 0.63073588, 0.14218974],
#        [0.86727903, 0.95890848, 0.39738407]])

在一个区间上[x,y]平均间隔去取n个数

np.linspace[x,y,m]

np.linspace(2,3,5)
# array([2.  , 2.25, 2.5 , 2.75, 3.  ])

数学运算

import numpy as np
a = np.array([20,30,40,50])
b = np.arange(4)
print(a)
print(b)
print("a - b " , a - b) # 对应位置相减
print("a - b - 1 :" , a - b - 1) 
print("b**2" , b**2)
print("a < 35" , a < 35)
# [20 30 40 50]
# [0 1 2 3]
# a - b  [20 29 38 47]
# a - b - 1 : [19 28 37 46]
# b**2 [0 1 4 9]
# a < 35 [ True  True False False]

矩阵乘法

A = np.array([
    [1,1],
    [0,1]
])
B = np.array([
    [2,0],
    [3,4]
])
print('------A-------')
print(A)
print('------B-------')
print(B)
print('------A*B-------')
print(A*B) #对应位置相乘
print('------A.dot(B)-------')
print(A.dot(B)) # 矩阵乘法
print('------np.dot(A,B)-------')
print(np.dot(A,B)) # 也为矩阵乘法
# ------A-------
# [[1 1]
#  [0 1]]
# ------B-------
# [[2 0]
#  [3 4]]
# ------A*B-------
# [[2 0]
#  [0 4]]
# ------A.dot(B)-------
# [[5 4]
#  [3 4]]
# ------np.dot(A,B)-------
# [[5 4]
#  [3 4]]

数学公式

e/平法等等

import numpy as np
B = np.arange(3)
print(B)
print(np.exp(B)) # e**B
print(np.sqrt(B)) # _/`B``
# [0 1 2]
# [1.         2.71828183 7.3890561 ]
# [0.         1.         1.41421356]
import numpy as np
a = np.floor(10*np.random.random((3,4))) # np.floor() //向下取整
print(a)
print('-------------')
print(a.ravel()) # 把矩阵拉成向量
print('-------------')
a.shape = (3,4) # 把向量拉成矩阵
#
# a.shape = (3,-1)
# -1帮你自动计算后一个维度的个数
#
print(a)
print('-------------')
print(a.T) # 矩阵转置
# [[3. 5. 8. 6.]
#  [5. 6. 6. 7.]
#  [1. 6. 2. 5.]]
# -------------
# [3. 5. 8. 6. 5. 6. 6. 7. 1. 6. 2. 5.]
# -------------
# [[3. 5. 8. 6.]
#  [5. 6. 6. 7.]
#  [1. 6. 2. 5.]]
# -------------
# [[3. 5. 1.]
#  [5. 6. 6.]
#  [8. 6. 2.]
#  [6. 7. 5.]]

矩阵拼接

# 矩阵拼接
import numpy as np
a = np.floor(10*np.random.random((2,2)))
b = np.floor(10*np.random.random((2,2)))
print('----------a-----------')
print(a)
print('----------b-----------')
print(b)
print('----------------------')
print(np.hstack((a,b))) # 按行拼接
print('----------------------')
print(np.vstack((a,b))) # 按列拼接
# ----------a-----------
# [[2. 0.]
#  [9. 7.]]
# ----------b-----------
# [[2. 0.]
#  [6. 9.]]
# ----------------------
# [[2. 0. 2. 0.]
#  [9. 7. 6. 9.]]
# ----------------------
# [[2. 0.]
#  [9. 7.]
#  [2. 0.]
#  [6. 9.]]

矩阵数据切割

#数据分割
a = np.floor(10*np.random.random((2,12)))
print(a)
print('------------')
print(np.hsplit(a,3))  # 按行切分,3切分成3份,得到三个array值
print('------------')
print(np.hsplit(a,(3,4))) 
# split a after  the third and the fourth cloumn
# 在第三行和第四行后进行切割
print('------------')
a = np.floor(10*np.random.random((12,2)))
print(a)
print('-------------')
np.vsplit(a,3)  # 按列切分
# [[4. 3. 3. 3. 7. 5. 7. 4. 6. 4. 6. 8.]
#  [9. 9. 4. 8. 0. 4. 3. 5. 1. 9. 4. 4.]]
# ------------
# [array([[4., 3., 3., 3.],
#        [9., 9., 4., 8.]]), array([[7., 5., 7., 4.],
#        [0., 4., 3., 5.]]), array([[6., 4., 6., 8.],
#        [1., 9., 4., 4.]])]
# ------------
# [array([[4., 3., 3.],
#        [9., 9., 4.]]), array([[3.],
#        [8.]]), array([[7., 5., 7., 4., 6., 4., 6., 8.],
#        [0., 4., 3., 5., 1., 9., 4., 4.]])]
# ------------
# [[8. 2.]
#  [3. 9.]
#  [3. 5.]
#  [5. 0.]
#  [4. 3.]
#  [2. 3.]
#  [0. 2.]
#  [5. 7.]
#  [5. 5.]
#  [7. 9.]
#  [3. 8.]
#  [0. 0.]]
# -------------
# [array([[8., 2.],
#         [3., 9.],
#         [3., 5.],
#         [5., 0.]]), array([[4., 3.],
#         [2., 3.],
#         [0., 2.],
#         [5., 7.]]), array([[5., 5.],
#         [7., 9.],
#         [3., 8.],
#         [0., 0.]])]

复制

# 复制/有俩种方法
# 浅复制
c = a.view() # 浅复制,共用一套值
print(c is a)
c.shape = (2,6)
print('a.shape: ' ,a.shape)
print('c.shape: ' ,c.shape)
c[0,4] = 1234    # a 的值也变量,a和c共用了一套值
print(a)
print(id(a))
print(id(c))
# False
# a.shape:  (3, 4)
# c.shape:  (2, 6)
# [[   0    1    2    3]
#  [1234    5    6    7]
#  [   8    9   10   11]]
# 2540538182992
# 2540538442256
#
#
#
# 深复制
d = a.copy()
print(d is a)
d[0,0] = 9999
print('------d-------')
print(d)
print('------a-------')
print(a)
# False
# ------d-------
# [[9999    1    2    3]
#  [1234    5    6    7]
#  [   8    9   10   11]]
# ------a-------
# [[   0    1    2    3]
#  [1234    5    6    7]
#  [   8    9   10   11]]

索引

#索引
import numpy as np
data = np.sin(np.arange(20).reshape(5,4))
print(data)
ind = data.argmax(axis = 0) # 按列来进行计算
print(ind) # 输出每一列的最大值所在的行(以0开始),索引
data_max = data[ind,range(data.shape[1])] 
print(data_max)
# [[ 0.          0.84147098  0.90929743  0.14112001]
#  [-0.7568025  -0.95892427 -0.2794155   0.6569866 ]
#  [ 0.98935825  0.41211849 -0.54402111 -0.99999021]
#  [-0.53657292  0.42016704  0.99060736  0.65028784]
#  [-0.28790332 -0.96139749 -0.75098725  0.14987721]]
# [2 0 3 1]
# [0.98935825 0.84147098 0.99060736 0.6569866 ]

在行和列进行扩展

# 扩展
import numpy as np
a = np.arange(0,40,10)
print(a)
b = np.tile(a,(3,5)) #构造一个三行五列的二维数组,每一个元素都是a
print(b)
# [ 0 10 20 30]
# [[ 0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30]
#  [ 0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30]
#  [ 0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30  0 10 20 30]]

排序

#排序
import numpy as np
a = np.array([[4,3,5],
             [1,6,1],
             [0,2,3]])
print(a)
print('------按列排序-------')
b = np.sort(a,axis = 0) #对二维数组排序,0为按列排序,1为按行排序
print(b)
#b
a.sort(axis = 1)
print('--------按行排序-----') #对二维数组排序,0为按列排序,1为按行排序
print(a)

print('################')
a = np.array([5,3,1,2])
j = np.argsort(a)   # 索引,求最小值索引(编号)

print('-------最小值索引------')
print(j)
print('-------排序结果------')
print(a[j])   # 排序完之后的结果
# [[4 3 5]
#  [1 6 1]
#  [0 2 3]]
# ------按列排序-------
# [[0 2 1]
#  [1 3 3]
#  [4 6 5]]
# --------按行排序-----
# [[3 4 5]
#  [1 1 6]
#  [0 2 3]]
# ################
# -------最小值索引------
# [2 3 1 0]
# -------排序结果------
# [1 2 3 5]

More in thist website !


除非特别说明,本博客所有作品均采用知识共享署名-非商业性使用-禁止演绎 4.0 国际许可协议进行许可。转载请注明转自-https://www.emperinter.info/2019/10/07/numpy/

Leave a Reply

Your email address will not be published. Required fields are marked *

Post comment

zh-CN 简体中文
X