数据分析—numpy

import pandas as pd
import numpy as np

创建普通数组

a = np.array([1,2,3])
b = np.array([[1,2,3],[4,5,6],[7,8,9]])

b[1,1] = 10

print(a.shape)
print(b.shape)
print(a.dtype)
print(b)

(3,)
(3, 3)
int64
[[1 2 3]
 [4 5 6]
 [7 8 9]]

创建结构数组

personalType = np.dtype({
    'names':['name','age','chinese','math','english'],
    'formats':['S25','i','i','i','f']
})
students = np.array([("huzai",22,99,99,99.5),("huzai",22,99,99,99.5)],dtype=personalType)
age = students[:]['age']
print(np.mean(age))

22.0

print(students)

[(b'huzai', 22, 99, 99, 99.5) (b'huzai', 22, 99, 99, 99.5)]

创建连续数组

x1 = np.arange(1,11,2) #步长为2，从1开始的等差数组(不包括终值)
x2 = np.linspace(1,9,5) #将1-9分成5块，结果如上

print(x1)
print(x2)

[1 3 5 7 9]
[1. 3. 5. 7. 9.]

数组间的算数运算

print(np.add(x1,x2))
print(np.subtract(x1,x2))
print(np.multiply(x1,x2))
print(np.divide(x1,x2))

[ 2.  6. 10. 14. 18.]
[0. 0. 0. 0. 0.]
[ 1.  9. 25. 49. 81.]
[1. 1. 1. 1. 1.]

统计函数

数组中的最值 np.amin() amax()

a = np.array([[1,3,7],[2,5,8],[6,4,9]])
print(np.amin(a))
print(np.amin(a,0)) #每一列的最小值
print(np.amin(a,1)) #每行的最小值

1
[1 3 7]
[1 2 4]

统计最大值与最小值之差 ptp()

print(np.ptp(a))
print(np.ptp(a,0)) #每列最大值与最小值的差
print(np.ptp(a,1)) #每行最大值与最小值的差

8
[5 2 2]
[6 6 5]

统计数组的百分位数 percentile(a, p, axis) a:数组名 p 代表百分比 axis代表是行还是列

print(np.percentile(a,50))
print(np.percentile(a,50,0))
print(np.percentile(a,50,1))

5.0
[2. 4. 8.]
[3. 5. 6.]

统计数组中的中位数以及平均数 median() mean()

print(np.median(a))
print(np.median(a,0))
print(np.median(a,1))

5.0
[2. 4. 8.]
[3. 5. 6.]

数组中的加权平均值 average(a,weights)

b = np.array([1,2,3,4])
wts = np.array([1,2,3,4])
print(np.average(b))
print(np.average(b,weights=wts))

2.5
3.0

统计数组中的标准差（std（））与方差（var（））

print(np.std(b))
print(np.var(b))

1.118033988749895
1.25

Numpy排序

print(a)
print(np.sort(a))
print(np.sort(a,0))

[[1 3 7]
 [2 5 8]
 [6 4 9]]
[[1 3 7]
 [2 5 8]
 [4 6 9]]
[[1 3 7]
 [2 4 8]
 [6 5 9]]

作业题

st_type = np.dtype({
    'names':['name','chinese','english','math'],
    'formats':['S25','i','i','i']
})
grades = np.array([('zhangfei',66,65,30),('guanyu',95,85,98),('zhaoyun',93,92,96),('huangzhong',90,88,77),
                  ('dianwei',80,90,90)],dtype=st_type)

print(grades)

[(b'zhangfei', 66, 65, 30) (b'guanyu', 95, 85, 98)
 (b'zhaoyun', 93, 92, 96) (b'huangzhong', 90, 88, 77)
 (b'dianwei', 80, 90, 90)]

chinese = grades[:]['chinese'] 
english = grades[:]['english']
math = grades[:]['math']
total = np.add(chinese,english,math)

print(total)

[131 180 185 178 170]

c_a,e_a,m_a = np.average(chinese),np.average(english),np.average(math)

print(c_a)

84.8