模拟 数据
signup.csv 
employee_id,name,sex,department,item_id
NC001,liming,f,functional,N100
NC002,zhangsan,f,functional,N101
NC003,sunli,m,functional,N103
NC004,lina,m,functional,N103
NC005,wanger,m,technology,N101
NC006,shenteng,m,functional,N100
NC007,mali,f,functional,N102
NC008,reba,f,functional,N102
NC009,ailun,m,functional,N102
NC010,changyuan,m,functional,N102
NC011,geyou,m,functional,N103
NC012,cls,f,functional,N103
signup1.csv
employee_id,name,sex,department,item_id
NC013,zhangwei,m,education,N102
NC014,liufang,m,education,N102
NC015,wanglei,m,education,N104
NC016,zhangli,f,education,N104
NC017,wangyan,m,education,N101
NC018,liangfeifan,m,education,N101
items.csv 
item_id,item_name,location
N100,athletics,area01
N101,longJump,area02
N102,highJump,area03
N103,javelin,area04
N104,pingPong,area05
N105,relayRace,area06

实现代码 
import pandas as pd
signup = pd.read_csv('signup.csv')
signup1 = pd.read_csv('signup1.csv')
items = pd.read_csv('items.csv')

signup = signup.append(signup1,ignore_index=True)
df_all = pd.merge(signup, items,
                  how='left',
                  left_on=signup['item_id'],
                  right_on=items['item_id'])

print(df_all.groupby('item_name')['employee_id'].count())