Python金融大数据分析笔记（自用7）输入输出

一.基本I/O

1.将对象写入磁盘

from pylab import plt, mpl
plt.style.use('seaborn')
mpl.rcParams['font.family'] = 'serif'
%matplotlib inline

使用随机数据，保存到一个列表对象中

import pickle  
import numpy as np
from random import gauss #生成正态分布随机数用的
a = [gauss(1.5, 2) for i in range(1000000)]  
path = '....'
pkl_file = open(path + 'data.pkl', 'wb')  

序列化和反序列化-写入对象的dump(),加载到内存的load()

%time pickle.dump(a, pkl_file) #序列化对象a
pkl_file.close()  
pkl_file = open(path + 'data.pkl', 'rb')  #二进制读模式打开文件
%time b = pickle.load(pkl_file)  #读取对象并反序列化
np.allclose(np.array(a), np.array(b))  #转化为ndarray对象，验证是否包含相同的数值

pkl_file = open(path + 'data.pkl', 'wb')
%time pickle.dump(np.array(a), pkl_file)  #序列化
%time pickle.dump(np.array(a) ** 2, pkl_file) 
pkl_file.close()

将两个ndarray对象读回内存

pkl_file = open(path + 'data.pkl', 'rb')
x = pickle.load(pkl_file)  
x[:4]
y = pickle.load(pkl_file)  
y[:4]
pkl_file.close()

读取字典对象

pkl_file = open(path + 'data.pkl', 'wb')
pickle.dump({
   'x': x, 'y': y}, pkl_file)  
pkl_file.close()
pkl_file = open(path + 'data.pkl', 'rb')
data = pickle.load(pkl_file)  
pkl_file.close()
for key in data.keys():
    print(key, data[key][:4])

2.读取和写入文本文件

import pandas as pd
rows = 5000  
a = np.random.standard_normal((rows, 5)).round(4)  
t = pd.date_range(start='2019/1/1', periods=rows, freq='H')  
csv_file = open(path + 'data.csv', 'w')
header = 'date,no1,no2,no3,no4,no5\n'  
for t_, (no1, no2, no3, no4, no5) in zip(t, a):  #按行合并
    s = '{},{},{},{},{},{}\n'.format(t_, no1, no2, no3, no4, no5)  
    csv_file.write(s)  #逐行写入
csv_file.close()

用readlines()方法逐行读入

csv_file = open(path + 'data.csv', 'r')  
for i in range(