生成测试数据

import os
import pandas as pd
import random
root = r'D:\R\P'
output_path = 'D:/select_files'

file_list = []
for every_file in os.listdir(root):
#print every_file
file_list.append(every_file)
#随机从数据文件集中选择若干个数据
select_files = random.sample(file_list, 10)
#每个数据文件中随机选择若干行，得到排序后的具体行
select_rows = sorted(random.sample(range(99),10))
#print select_rows
for select_file in select_files:

df=pd.read_csv(os.path.join(root,select_file),skiprows=3,header=None,sep=' ').iloc[select_rows,:3]
df.to_csv(os.path.join(output_path,select_file))