"Remove duplicates" removes unique values in the format of whole number

[zenisekd]

Hi...

my issue is that I have a table, where the column ID is in the format of whole number (int), right from the beginning, without any transformations. The table has around 25k rows. 
When I use "remove duplicate rows" on ID column, it deletes, along with duplicates, certain unique numbers eg. 23682. 

When I revert the steps (first filter the number and then remove duplicates), the number is there ...

Can somebody tell me what the heck is going on?

I have read in different places that sometimes it helps to first convert the column to text, then trim, do upper or lower case, then remove duplicate rows and then convert it back to a number. This doesn't work for me, however...

[VVA]

I have the same issue but just with the regular OOB function. I have a list of names and I want to remove the duplicates. It seems that the function removes unique values (returns 19 'unique', when I know for a fact the unique names are around 75!). How to solve this issue? Why is it happening?😩

[j_ocean]

Are you sure the removed duplicates is pointing at the ID column? That's caught me before, pointing at some kind of title where it may coincidentally be the same across two entries.

[v-kelly-msft · Accepted solution]

Hi  @zenisekd ,

After checking,I couldnt reproduce your senario,as it seems fine here:

So pls check the blog below to see whether you could find a solution:

https://radacad.com/remove-duplicate-doesnt-work-in-power-query-for-power-bi-here-is-the-solution

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[DmitryKo]

Any ETA for this bug to be fixed?

[DmitryKo]

I guess no ETA for a fix.

[samdthompson]

Hello, i have had this too. I am not sure of the techincal reason but i got around it by prefixing all the text values of whole numbers with a z then doing the duplicate removal, removing the z's and converting back to a whole number. A bit convoluted but it worked for me.

[JoseRamon]

Got same problem here. Removing duplicates on a Whole Number column was in fact removing thousands of unique values. Changing type to text before removing duplicates and then back to Int has worked fine.