python pandas 实现SQl的count(*),count(distinct **)

实现也非常简单，不过多啰嗦，见代码注释：

pv uv 代指 select host，count(*),count(distinct ad) from table group by ;

# -*- coding:utf-8 -*-
import pandas as pd
from datetime import datetime

def Main():

    print("开始。。。。。")
    print(datetime.now().strftime('%Y-%m-%d %H:%M:%S'))

    source_txt = "/data/u_lx_data/zhangqm/sh/yanjie/fudan/bigdata/bigdata_click_result.txt"
    target_txt = "/data/u_lx_data/zhangqm/sh/yanjie/fudan/bigdata/host_pvuv_click.txt"

    uname = ['ad','type','host','url','ref','time','os','os_type']

    # count(*)
    pv = pd.read_csv(source_txt,sep="\t",header=None,names=uname,index_col=False)[['host','ad']].groupby('host')['ad'].size()

    # count(distinct **)
    uv = pd.read_csv(source_txt,sep="\t",header=None,names=uname,index_col=False)[['host','ad']].groupby('host').agg({'ad': pd.Series.nunique})

    # 去除索引，带出分组的字段
    result = pd.merge(pv.reset_index(),uv.reset_index(),how='inner',on='host')

    result.to_csv  (target_txt,header=None,index=False,sep="\t")

    print("完成。。。。。")
    print(datetime.now().strftime('%Y-%m-%d %H:%M:%S'))

if __name__ == "__main__":
    Main()