View on GitHub

Nileshingle.GitHub.io

Github Projects

2. Work Market

Posted: April 11, 2016; Updated: April 12, 2016

In this project, the jobs in New York were analyzed for various parameters such as agency, job posting duration, job requirements and salary. The data was first analyzed assuming that each repeated position was a newly created position. In the second round of analysis, uniqueness criteria was applied where the assumption was that the repeated position was the same as previous position that remained unfilled. This is an ongoing work and the analysis of parameters 'Job Description', 'Minimum Qual Requirements', 'Preferred Skills', 'Additional Information' remains to be explored.

Note: To view the enlarged figures please right click on the image to click 'View Image'.

Source of Data:
Data file in .csv format was downloaded from NYC Open Data.

Questions asked:
  1. Find out which agencies have highest number of job openings.
  2. Find out which positions have highest and lowest salaries.
  3. Are there any jobs that take too long to fill, if yes which one.
  4. Is there any effect of residency requirement on filling a job.
  5. Is there any effect of level of a job on salary.

Possible answers:
  1. The Department of Health/Mental Hygiene (DOHH) has the highest number of open positions (Figure 1).
  2. The Department for Environmental Protection (DOEP) has the lowest salary of USD 8 per hour for the position of Clerical Associate (Figure 4). Whereas, Fire Department has the highest salary of USD 78 per hour for the position of City Medical Specialist (Figure 5).
  3. Yes, there are positions that take more than 1000 days to fill/process. The toughest position to fill appears to be Contract Reviewer with Department of Business Services (DOBS) that may exceed 1700 days or 4.7 years (Figures 6 to 9 and 11)
  4. Residency requirements appears to have little effect on the number days required to fill a job (Figure 13).
  5. There appears to be no direct relationship between salary and number of days required to fill the job (Figure 14)

Sample code:

# Abbreviating agency names
  jobs$agency.abr <- abbreviate(jobs$Agency)
  

# Function to convert annual and monthly salary to hourly salary
salconvert <- function(sal, salfreq){
  salfreq <- as.character(salfreq)
  if (salfreq == "Annual"){
    #print(Salary.Range.From/(365*8))
    return (sal/(365*8))  #convert annual pay to hourly rate 8 h/day
  }else if (salfreq == "Monthly"){
    return (sal/(30*8))
  }else if (salfreq == "Hourly"){
    return (sal)
  } else{
    return ("NULL")
  }
}


# sample code for plot
  ggplot(jobs)+
  aes(x=post.to.process, y = Salary.Frequency)+
  geom_point(size = 5, alpha = 0.5,aes(color=Level))+
  facet_wrap(~Residency)+
  xlab ("Number of days") +
  ylab ("Salary Frequency")+
  ggtitle ("Days to fill vs. salary frequency")+
  theme(axis.text.x = element_text(size = 12, angle=0, vjust=1, hjust = 1),
        axis.title=element_text(size=14,face="bold"),
        legend.text = element_text(size = 12),
        plot.margin=unit(c(1,1,1.5,1.2),"cm"))+
  guides(fill=guide_legend(ncol=2))

Figure gallery:
Plot_1
Figure 1: Total number of vacancies with different agencies.

Plot_3
Figure 2: Levels of positions with different agenices in both internal and external job postings.

Plot_4
Figure 3: Agencies that offer a salary above USD 20 per hour.

Plot_5
Figure 4: Agencies that offer a salary below USD 10 per hour.

Plot_7
Figure 5: Agencies that offer a salary above USD 60 per hour.

Plot_8
Figure 6: Number of days to process a job for different agencies.

Plot_9
Figure 7: Number of days to update a job for different agencies.

Plot_10
Figure 8: Jobs that need atleast 1 year to be filled.

Plot_11
Figure 9: Jobs that need atleast 2 years to be filled.

Plot_12
Figure 10: Availability of jobs at different levels as internal or external posting.

Plot_13
Figure 11: Jobs that need atleast 2 years to be filled.

Plot_15
Figure 12: Internal vs. external job positions with (Y) or without (N) residency requirements.

Plot_16
Figure 13: Residency requirement vs. number of days required to fill a job.

Plot_17
Figure 14: Salary vs. number of days required to fill a job.

Plot_18
Figure 15: Frequency of salary (hourly, weekly, monthly) vs. number of days required to fill a job.

Plot_20
Figure 16: Jobs with pay less than USD 15 per hour. (Re-analyzed data after considering uniqueness of positions in database. Assuming, that repeated position in multiple rows is not a newly created same position however it is rather the same position that remanined unfilled after previous posting.)