Report on schedule of restaurant/employees

Stealing Pattern contd

• Stealing per store for high/low cash type

store_summary_info %>% group_by(cash_type) %>% summarise(count=n(),avg_steal_count = mean(StealCountPerDay),avg_steal_amount=mean(StealAmountPerDay),
                                                         steal_employee_percentage = mean(steal_employee_percentage),steal_manager_percentage=mean(steal_manager_percentage))

Schedule Statistics: Employee-Employee

• Store working day count

hist(store_date$store_open_day,breaks=100)

summary(store_date$store_open_day)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
    1.0   383.0   422.0   537.9   757.0  1660.0 

• Employee working day count

hist(employee_date$employee_open_day,breaks=500)

summary(employee_date$employee_open_day)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
   1.00    8.00   32.00   63.12   91.00 1019.00 

• Employee working day/Store open day

hist(employee_date$open_percentage,breaks=500)

summary(employee_date$open_percentage)

     Min.   1st Qu.    Median      Mean   3rd Qu.      Max. 
0.0006024 0.0134200 0.0547000 0.1252000 0.1682000 1.0000000 

• Standard deviations of employee working day percentage within store

hist(sd_open_percentage$sd_percentage,breaks=100)

summary(sd_open_percentage$sd_percentage,na.rm=TRUE)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
0.03866 0.08684 0.14000 0.14760 0.20610 0.68830      20 

• Select a store, see the number of days the employees are on

temp<-sample_week %>% group_by(employeekey) %>% summarise(open_day = n())
hist(temp$open_day)

• Graph on schedule per week for all employee in this store

plot(sample_employee$week,sample_employee$employee_open_day_week,xlab="week",ylab="on_day")

ggplot(sample_employee_store,aes(week,employee_open_day_week))+geom_line(aes(colour=factor(employeekey)))

• Standard deviations of working day per week within the same store

hist(sd_open_per_week$sd_open_per_week,breaks=100)

• Graph one specific week-schedule for one store

temp<-filter(sample_week,employeekey<=190)
p<-ggplot(temp,aes(weekday,employeekey))
p+geom_point(aes(colour=factor(employeekey)))

• Percentage of shift share per employee-pair within one store

hist(employee_shift_store[[818]]$share_shift_percentage,breaks=100)

hist(employee_shift_store[[1]]$share_shift_percentage,breaks=100)

hist(employee_shift_store[[581]]$share_shift_percentage,breaks=100)

hist(employee_shift_store[[985]]$share_shift_percentage,breaks=100)

• Summary statistics of average shift percentage

summary(mean_shift)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.03845 0.17000 0.18700 0.19350 0.20840 0.66030 

hist(mean_shift,breaks=100)

Schedule Statistics: Employee-Manager

• Percentage of shift with each manager

summary(employee_date_manager$employee_manager_percentage)

     Min.   1st Qu.    Median      Mean   3rd Qu.      Max. 
0.0003617 0.0083330 0.0270300 0.1866000 0.1462000 1.0000000 

hist(employee_date_manager$employee_manager_percentage,breaks=100)

• Look at the employees with manager percentage equal 1

temp<-unique(select(filter(employee_date_manager,employee_manager_percentage==1),employeekey,employee_total_obs))
hist(temp$employee_total_obs,breaks=100,main="Percentage ==1 Obs Count")

After recheck, these are manager themselves,we could identify the managers’ cheating performance as well