Hi @Vira_Maykova , To remove duplicates I would definitely suggest using the Duplicate Row Filter node. That node can also be condfigured to report duplicate rows rather than remove them.
However if you have grouped by Customer ID, Transaction Date and Product SKU, then by definition, your data will now be unique on Customer ID, Transaction Date and Product SKU as it is now aggregated by that grouping. If previously there were duplicates then they will have been removed by this grouping and your row count will reduce.
Can you explain what you mean by “unnecessary ones”.
As mentioned above, without seeing your data and knowing what you are actually wanting as an output it is difficult to state what you should or shouldn’t be doing.
You may be interested in this post which explains how the Duplicate Row Filter can be used to find the duplicates, and then a row filter can be used to actually removed them…