3 Replies Latest reply on Nov 29, 2018 6:23 PM by Nathanael Webber

    Analyzing Crime and Overtime Data

    Nathanael Webber

      Hello! We are working on a project to analyze Police overtime data in the City of Chicago with crime data. Our goal is to look for any correlations between overtime pay and crime. However, we are new to Tableau and would like some help joining the data and starting the analyses. We want to join the overtime and the crime data based on dates, but we are open to suggestions. Any help would be appreciated. Thank you!

        • 1. Re: Analyzing Crime and Overtime Data
          Don Wise

          Hello Nathanael,

          You'll need something common between the two data sets to match on. In this case the Crimes data set has a Date but it is formatted as Date & Time. So it has time detail. In your OT Data it is formatted as Date only. 


          You can do one of two things here to perform the join (1) Create a new join calculation as in the screenshot below or (2) simply change the value of your Crimes data set from date and time by clicking on the icon above Date and changing it as in the second screenshot below.  I prefer the first method in order to retain the time value which I might want to use in analysis later on.  Hope that helps! If yes, mark as correct. Thx, Don


          Screen Shot 2018-11-29 at 4.15.47 PM.png

          Screen Shot 2018-11-29 at 4.18.12 PM.png

          • 2. Re: Analyzing Crime and Overtime Data
            Subash Hasuruvalli

            Hi Nathanael,


            One suggestion, can you create separate data sources for both the excels and blend the data.

            You can treat both the data's separately to identify the trends but link it at runtime.


            Let me know if this helps.



            Subash Hasuruvalli

            • 3. Re: Analyzing Crime and Overtime Data
              Nathanael Webber

              Thank you, I did the second method because I won't be using the time data in the analysis. I see you are using an inner join, is that what you would recommend? I want to use a full-outer join because I don't want to lose crimes or OT data, but maybe that doesn't work that way. Thanks for your help.