numpy学习笔记

Numpy学习笔记

导入包

import numpy as np

array、arange、dtype

a1 = np.array([1,2,3])
a1 = np.array(range(10))
type(a1)
t4 = np.array(range(1,4),dtype=float)

round random 随机生成小数，保存后两位

#numpy的小数
from random import random
a = np.array([random() for i in range(10)])
t8 = np.round(a,2)

numpy读取本地数据和索引

数据为kaggle网站数据

us_file_path = "data/US_vedio_data_numbers.csv"
gb_file_path = "data/GB_vedio_data_numbers.csv"
t1 = np.loadtxt(us_file_path,delimiter=",",dtype="int")

下图为函数所对应的参数

取行

print(t1[2])
print(t1[2:])

取不连续的多行

print(t1[[2,8,10]])

取列

print(t1[:,0])
#连续的多列
print(t1[:,2:])
#不连续的多列
print(t1[:,[0,2]])
#取多行多列
a = t1[2, 3]
b = t1[2:5,1:4]

取不相邻的点(0,0)(2,1)

c = t1[[0,2],[0,1]]
print(b)
print(type(a))
print(c)

shape() reshape() 得到维度

a1 = np.array(range(12))
a1.shape
a1.reshape((3,4))
a2 = np.arange(24).reshape(2,3,4)

展开函数 flatten()

a3 = a2.flatten()

布尔索引

t8<10
t8[t8<10]

三元运算符

np.where(t8<8,0,10)

clip()裁剪：比2小的都为2，比5大的都为5

t9.clip(2,5)

竖直拼接

np.vstack(x,y)

水平拼接

np.hstack(x,y)

行交换、列交换

t[[1,2],:] = t[[2,1],:]
t[:,[0,2]] = t[:,[2,0]]

nan

a = np.arange(12).reshape(3,4).astype("float")
a[1,2:] = np.nan

将nan的部分替换

for i in range(a.shape[1]):
    temp_col = a[:,i]
    nan_num = np.nonzero(temp_col != temp_col)
    if nan_num != 0:
        temp_col[np.isnan(temp_col)] = temp_col[temp_col == temp_col].mean()
print(temp_col)

一个小例子

import matplotlib.pyplot as plt
import numpy as np
us_file_path = "data/US_vedio_data_numbers.csv"
gb_file_path = "data/GB_vedio_data_numbers.csv"
t_us = np.loadtxt(us_file_path,delimiter=",",dtype="int")
t_us = t_us[t_us[:,1] <= 500000]
#取评论的数据,最后一列
t_us_comments = t_us[:,-1]
t_us_like = t_us[:,1]

plt.figure(figsize=(20,8),dpi = 80)
plt.scatter(t_us_like,t_us_comments)

plt.show()