8 Replies Latest reply on Jul 12, 2016 5:07 AM by Jennifer Sanchez

# Removing duplicates

Hi All, I am stumped on this.

I have a unique identifier or "ID" which I created to allow me to filter out any duplicates. In the attached workbook you'll see that there are multiple ID's and "letter" combinations. I need to get the % of the count of the unique value and "letter" combination against that of the total number of records based on the unique sum.

I did the following:

1) Created Distinct Value Field:

2) for % I divided: Distinct Value / Number of Records

The % looks correct using this smaller data sample but the underlying data still shows those duplicates. Should I expect it to? Is there anyway for the underlying data to keep just 1 of those unique values and not show the other duplicates? My original data set (that I cannot share) is a very large data set so spot checking to verify that the % above isn't counting the duplicate values is not feasible. I would prefer to somehow filter out the duplicates so that it doesn't show in the underlying data which will be used by other colleagues.

THANKS SO MUCH.

• ###### 1. Re: Removing duplicates

Yes the underlying data will show all the relevant records.   What are you using the underlying data for?   Perhaps you can create another cross tab view of the chart with the underlying data needed and trigger it to show through a dashboard action.

• ###### 2. Re: Removing duplicates

Thanks, John! So you are saying that even though the duplicates show in the "underlying data," the calculations aren't using them as distinct values.... meaning if there are 5 rows of the same data or 5 ID's, the calculations are only computing using it once versus 5xs?

• ###### 3. Re: Removing duplicates

The calculations will do whatever you tell them to do.  What % are you trying to calculate?  Would it be as shown below?  If so then total distinct is Total(countd(id))  and Total records is Total(Sum(# Records))

Each is a table calc where the first table calc (Total Distinct) is partitioned by ID and the second (Total Records) is not partitioned.

• ###### 4. Re: Removing duplicates

Hi John, I am looking to calculate the percentage of unique values by the total count. SO I would expect to get 33% for the values. Does that make sense?

• ###### 5. Re: Removing duplicates

Sorry - that didn't make sense I made a mistake and can't seem where to delete the post. I need to think about this for a second.

• ###### 6. Re: Removing duplicates

I would expect to get 16% since There are 6 records in total (excluding duplicates)

• ###### 7. Re: Removing duplicates

16%?  How meaningful can that be if they all are the same?

• ###### 8. Re: Removing duplicates

Thanks, John. I decided to remove the duplicates in Alteryx instead of in tableau. I appreciate your time.