Python—Numpy学习

导入Numpy

import numpy as np

读取数据

world_alcohol = np.genfromtxt("world_alcohol.txt", delimiter=",", dtype = str)
print(world_alcohol)
>[['Year' 'WHO region' 'Country' 'Beverage Types' 'Display Value']
 ['1986' 'Western Pacific' 'Viet Nam' 'Wine' '0']
 ['1986' 'Americas' 'Uruguay' 'Other' '0.5']
 ...
 ['1987' 'Africa' 'Malawi' 'Other' '0.75']
 ['1989' 'Americas' 'Bahamas' 'Wine' '1.5']
 ['1985' 'Africa' 'Malawi' 'Spirits' '0.31']]

genfromtxt参数:
文本分隔符 delimiter = ","
数据读取类型 dtype = str
跳过头行数 skip_header = 1

定义矩阵

vector = np.array([1, 2, 3, 4])
matrix = np.array([[1, 2, 3, 4], [5, 6, 7, 8]])

打印结构

print(vector.shape)
print(matrix.shape)
>(4,)
(2, 4)

array中数据类型为同一种类型

计算

判断

#判断
vector = np.array([1,3,5,4])
a = vector == 1
print(a)
[ True False False False]

b =vector < 5
print(vector[b])
>[1 3 4]

matrix = np.array([[1,2,3],[5,15,20],[25,30,35]])
second_column_15 = matrix[:, 1] == 15
print(second_column_15)
print(matrix[second_column_15,:])
>[False  True False]
[[ 5 15 20]]

vector = np.array([5, 15, 20 ,25])
equal_to_ten_and_five = (vector == 10) & (vector == 5)
equal_to_ten_or_five = (vector == 10) | (vector == 5)
print(equal_to_ten_and_five)
print(equal_to_ten_or_five)
>[False False False False]
[ True False False False]

vector[equal_to_ten_or_five] = 50
print(vector)
>[50 15 20 25]

类型转换

vector = np.array(["1", "2", "3", "4"])
vector =  vector.astype(float)
print(vector.dtype)
print(vector)
>float64
[1. 2. 3. 4.]

求和

matrix = np.array([[1,2,3],[5,15,20],[25,30,35]])
print(matrix.sum()) #所有元素求和
print(matrix.sum(axis = 1)) #行元素求和
print(matrix.sum(axis = 0)) #列元素求和
>136
[ 6 40 90]
[31 47 58]

常用函数

变换

np.arange(15)
>array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14])

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

a.ndim #维度
>2

a.size #元素个数
>15

np.zeros((3,4)) #0矩阵
>array([[0., 0., 0., 0.],
       [0., 0., 0., 0.],
       [0., 0., 0., 0.]])

np.ones((2,3,4), dtype=np.int32) #1矩阵，整型
>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]]])

np.arange(10, 30, 5) #序列
>array([10, 15, 20, 25])

随机数

np.random.random((2, 3))
>array([[0.37966439, 0.72749195, 0.65091678],
       [0.7075526 , 0.83246759, 0.06179741]])

from numpy import pi
np.linspace(0,2*pi,10) #0-2pi平均取10个值
>array([0.        , 0.6981317 , 1.3962634 , 2.0943951 , 2.7925268 ,
       3.4906585 , 4.1887902 , 4.88692191, 5.58505361, 6.28318531])

矩阵计算

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

np.floor() #向下取整
np.ravel() #矩阵转向量
A.T #转置


A = np.floor(10*np.random.random((2,2)))
B = np.floor(10*np.random.random((2,2)))
print(A)
print(B)
>[[2. 8.]
 [9. 6.]]
[[0. 3.]
 [3. 8.]]
print(np.hstack((A,B))) #横向拼接
>[[2. 8. 0. 3.]
 [9. 6. 3. 8.]]
print(np.vstack((A,B)))#纵向拼接
>[[2. 8.]
 [9. 6.]
 [0. 3.]
 [3. 8.]]