pandas数据抽样
作者：互联网
# -*- coding: utf-8 -*-
import pandas as pd
import os
import re
import xlsxwriter
import xlrd
from random import randint, sample
####线上作业明细
input1 = r"D:\4.AOI类型\AOI作业监控\exportAoiInfoLost_20210805142956.csv"
####输出
out = r"D:\4.AOI类型\AOI作业监控\作业准确率\线上标记异常分析(0804审核)_20210804.xlsx"
data1 = pd.read_csv(input1,dtype=str,error_bad_lines=False)
data1.fillna('', inplace=True) # 全空的替代
data1.replace('nan', '', inplace=True) # 空字符的替代
data1.replace('NaN', '', inplace=True) # 空字符的替代
data1['flag_rw']=((data1['任务状态']=="未完成")|(data1['补码审核状态']=="驳回"))
data1=data1[(data1['地区']!='852Y')]
data1=data1[data1['flag_rw']==False]
###线上标记异常抽查
# data1=data1[(data1['任务状态']=='完成-标记异常未审核')|(data1['任务状态']=='完成-标记异常审核通过')]
# ####线上外包质量
# data1=data1[(data1['补码审核状态']=='审核通过')|(data1['补码审核状态']=='补码修改通过')]
# ######线上标记异常
# data1=data1[(data1['任务状态']=='完成-标记异常未审核')|(data1['任务状态']=='完成-标记异常审核通过')]
#####外包名单
input5 = r"D:\4.AOI类型\线上类型核实平台作业人员明细.csv"
data5 = pd.read_csv(input5, dtype=str, error_bad_lines=False)
data1=pd.merge(data1,data5,how='left',left_on='补码工号',right_on='工号')
data1.fillna('', inplace=True) # 全空的替代
data1.replace('nan', '', inplace=True) # 空字符的替代
data1.replace('NaN', '', inplace=True) # 空字符的替代
####是否过滤外包审核结果条件
#data1=data1[data1['姓名']==""]
print(len(data1))
#####选择补码作业日期
data1['日期']=data1['任务包提交时间'].str[:10]
data1=data1[(data1['日期']=='2021-08-03')|(data1['日期']=='2021-08-04')|(data1['日期']=='2021-08-02')]
print('地区作业量,',len(data1))
def aoi_lx(b,a):
if a!="" and b!="":
return a
elif a=="" and b!="":
return b
else:
return ""
data1['dl_before']=data1['AOI大类(修改前)']
data1['dl_after']=data1['AOI大类(修改后)']
data1['xl_before']=data1['AOI小类(修改前)']
data1['xl_after']=data1['AOI小类(修改后)']
data1['name_after']= data1['AOI名称(修改后)']
data1['name_before']= data1['AOI名称(修改前)']
data1['大类']=data1.apply(lambda x:aoi_lx(x.dl_before,x.dl_after),axis=1)
data1['fa_type']=data1.apply(lambda x:aoi_lx(x.xl_before,x.xl_after),axis=1)
data1['aoi_name']=data1['name_before']
data1['city_code']=data1['城市']
data1['aoi_id']=data1['AOIID']
data1.fillna('', inplace=True) # 全空的替代
data1.replace('nan', '', inplace=True) # 空字符的替代
data1.replace('NaN', '', inplace=True) # 空字符的替代
n0=len(data1)
####仅抽查住宅小区
# data1=data1[data1['大类']=='住宅小区']
# n0=len(data1)
#####核实总量,计算比例:核实量/任务总量
n_hs=5000
bl=n_hs*1.0/n0
cityfile = r"D:\4.AOI类型\city_file.csv"
city_file = pd.read_csv(cityfile, dtype=str)
city_dict = dict(zip(city_file.city_code, city_file.city_name))
area_dict = dict(zip(city_file.city_code, city_file.region))
ad_dict= dict(zip(city_file.city_code, city_file.adcode))
data1['大区'] = data1['city_code'].map(area_dict)
data1['城市'] = data1['city_code'].map(city_dict)
sp=data1[['大区']].copy()
sp.drop_duplicates(subset=['大区'], keep='first',inplace=True)
sp['比例']=bl
cc_dict = dict(zip(sp.大区, sp.比例))
####按数量抽样,可调整
# sp['数量']=100
# cc_dict = dict(zip(sp.大区, sp.数量))
lxf = r"D:\4.AOI类型\AOI类型配置.csv"
lx = pd.read_csv(lxf, dtype=str)
dl_dict = dict(zip(lx.类型编码, lx.大类))
xl_dict = dict(zip(lx.类型编码, lx.子类))
#data1['大类'] = data1['fa_type'].map(dl_dict)
data1['小类'] = data1['fa_type'].map(xl_dict)
###按比例
def typicalsampling_bl(group, typicalFracDict):
name = group.name
frac = typicalFracDict[name]
return group.sample(frac=frac)
####按数量
def typicalsamling_num(group, typicalNDict):
name = group.name
n = typicalNDict[name]
return group.sample(n=n)
###按数量
# result = data1.groupby('大区', group_keys=False).apply(typicalsamling_num, cc_dict)
#####按比例
# result = data1.groupby('大区', group_keys=False).apply(typicalsampling_bl, cc_dict )
result = data1
result.to_excel(out,index=0,columns=['大区','aoi_id','AOI名称(修改前)','city_code','大类','小类','AOI大类(修改前)',
'AOI小类(修改前)','AOI名称(修改后)','AOI大类(修改后)','AOI小类(修改后)','AOI名称(审核修改前)','AOI大类(审核修改前)',
'AOI小类(审核修改前)','AOI大类对比','任务状态','标记异常原因','仓管工号','标记异常原因','仓管工号','任务包领取时间',
'任务包提交时间','补码审核状态','补码工号','补码审核时间','姓名'])
# result.to_excel(out,index=0)
标签：city,抽样,name,数据,inplace,AOI,dict,data1,pandas 来源： https://www.cnblogs.com/bahe/p/15210745.html