pandas: powerful Python data analysis toolkit

官方文档： http://pandas.pydata.org/pandas-docs/stable/

1. 导入包pandas

import pandas as pd

2. 获取文件夹下文件名称

import os
filenames=[]path="C:/Users/Forrest/PycharmProjects/test"
for file in os.listdir(path):
    filenames.append(file)

3. 读前几行文件(.csv文件)

# -*- coding: utf-8 -*-
##读前几行文件
f= open("C:/Users/Forrest/PycharmProjects/test/train.csv")
for i in range(5):
    print(f.readline().strip())

4. 读取文件和写文件（.csv文件）

file = pd.read_csv("./非空记录/"+filenames[i],sep=‘--->‘,header=None)

file.to_csv("./非空all/alldata.csv",header=False,mode=‘a‘,index=False) #a追加

5. pickle读入与写入

import cPickle as pickle
output = file(‘./temp.pkl‘, ‘wb‘)
pickle.dump(t, output, True)
output.close()
f2 = file(‘temp.pkl‘, ‘rb‘)
a2 = pickle.load(f2)
f2.close()

原文地址：https://www.cnblogs.com/Allen-rg/p/8889232.html