Preface
pandas is a data analysis package built based on Numpy that contains more advanced data structures and tools. Similar to Numpy, whose core is ndarray, pandas also revolves around the two core data structures of Series and DataFrame. Series and DataFrame correspond to one-dimensional sequence and two-dimensional table structure respectively. The conventional import method of pandas is as follows:
from pandas import Series,DataFrame import pandas as pd
1.1. Pandas analysis steps
1. Load log data
2. Load area_ip data
3. Count the number of real_ip requests. Similar to the following SQL:
SELECT inet_aton(l.real_ip), count(*), a.addr FROM log AS l INNER JOIN area_ip AS a ON a.start_ip_num <= inet_aton(l.real_ip) AND a.end_ip_num >= inet_aton(l.real_ip) GROUP BY real_ip ORDER BY count(*) LIMIT 0, 100;
1.2. Code
cat pd_ng_log_stat.py #!/usr/bin/env python #-*- coding: utf-8 -*- from ng_line_parser import NgLineParser import pandas as pd import socket import struct class PDNgLogStat(object): def __init__(self): self.ng_line_parser = NgLineParser() def _log_line_iter(self, pathes): """解析文件中的每一行并生成一个迭代器""" for path in pathes: with open(path, 'r') as f: for index, line in enumerate(f): self.ng_line_parser.parse(line) yield self.ng_line_parser.to_dict() def _ip2num(self, ip): """用于IP转化为数字""" ip_num = -1 try: # 将IP转化成INT/LONG 数字 ip_num = socket.ntohl(struct.unpack("I",socket.inet_aton(str(ip)))[0]) except: pass finally: return ip_num def _get_addr_by_ip(self, ip): """通过给的IP获得地址""" ip_num = self._ip2num(ip) try: addr_df = self.ip_addr_df[(self.ip_addr_df.ip_start_num <= ip_num) & (ip_num <= self.ip_addr_df.ip_end_num)] addr = addr_df.at[addr_df.index.tolist()[0], 'addr'] return addr except: return None def load_data(self, path): """通过给的文件路径加载数据生成 DataFrame""" self.df = pd.DataFrame(self._log_line_iter(path)) def uv_real_ip(self, top = 100): """统计cdn ip量""" group_by_cols = ['real_ip'] # 需要分组的列,只计算和显示该列 # 直接统计次数 url_req_grp = self.df[group_by_cols].groupby( self.df['real_ip']) return url_req_grp.agg(['count'])['real_ip'].nlargest(top, 'count') def uv_real_ip_addr(self, top = 100): """统计real ip 地址量""" cnt_df = self.uv_real_ip(top) # 添加 ip 地址 列 cnt_df.insert(len(cnt_df.columns), 'addr', cnt_df.index.map(self._get_addr_by_ip)) return cnt_df def load_ip_addr(self, path): """加载IP""" cols = ['id', 'ip_start_num', 'ip_end_num', 'ip_start', 'ip_end', 'addr', 'operator'] self.ip_addr_df = pd.read_csv(path, sep='\t', names=cols, index_col='id') return self.ip_addr_df def main(): file_pathes = ['www.ttmark.com.access.log'] pd_ng_log_stat = PDNgLogStat() pd_ng_log_stat.load_data(file_pathes) # 加载 ip 地址 area_ip_path = 'area_ip.csv' pd_ng_log_stat.load_ip_addr(area_ip_path) # 统计 用户真实 IP 访问量 和 地址 print pd_ng_log_stat.uv_real_ip_addr() if __name__ == '__main__': main()
Running statistics and output results
python pd_ng_log_stat.py count addr real_ip 60.191.123.80 101013 浙江省杭州市 - 32691 None 218.30.118.79 22523 北京市 ...... 136.243.152.18 889 德国 157.55.39.219 889 美国 66.249.65.170 888 美国 [100 rows x 2 columns]
Summary
The above is the entire content of this article. I hope that the content of this article will bring some help to everyone's study or work. If you have any questions, please You can leave messages to communicate.
For more detailed explanations of real IP request Pandas for Python data analysis, please pay attention to the PHP Chinese website!
Related articles:
How to read CSV files and write them to MySQL using Pandas in Python
Read cdn logs through the pandas library in Python Detailed analysis
Tutorial on using Python’s pandas framework to manipulate data in Excel files