De-duplicatd headcount

[Anonymous]

I'm looking for suggestions to view de-duplicated results for university admissions application data for a selected period.

Most applicants apply for one degree program at a time. Some prospects though apply to be admitted to two programs simultaneously.  My fact table has a transaction for each application/program, and my PBI reports show the number of applications for whatever period is selected in the slicer.

Is there a way to filter this down to a de-duplicated or distinct headcount within the report? I'm looking for results that include only one instance of every applicant who applied in the selected period, and for those who applied to more than one program, the instance retained would be based on a ranking of the degrees: those with more required credit hours prefered to lower credit hour degrees. For example:

Prospect    Degree Program          Priority   In Report?
John Doe    Degree A (100 hours)    1          Yes
John Doe    Degree B (80 hours)     2          No

So if John Doe applied for thsese two programs in 2017 Q1, the 2017report would show only one. However, if John Doe applied for these two in 2017, but an earlier degree in 2013, a report showing year over year would have one result for John in 2013 and another in 2017.

Or would this best be done with a second, de-duplicated source fact table?

Thanks

Richard

[Anonymous]

@Anonymous,

Could you please share sample data of your table and post expected result in  table format?

Regards,
Lydia

[Anonymous]

Yes, I can do that. Can you tell me the best way to share it on the forum? The example I gave initially was kind of clunky using the insert code option.

Thanks

Richard

[Anonymous]

@Anonymous,

You can publish your file to OneDrive and post shared link of the file here.

Regards,
Lydia

[Anonymous]

OK. Is this what you had in mind?

https://covenantseminary-my.sharepoint.com/personal/richard_hiers_covenantseminary_edu/_layouts/15/g...

Thanks

Richard

[Anonymous]

@Anonymous,

I am not very clear about the logic you use to get expected result, could you please explain the process based on the sample data you shared with me? And do you only use Sample Human Readable Dataset in the Excel?

Regards,
Lydia