Hi there,
I’m attempting to recreate the dynamic ‘others’ group feature sourced from the Quicksight Demo Central → Feature Demo list.
I’m having issues however with the ‘Salesperson Grouping’ calculated field
When I try to recreate this calculated field I try to input the field ‘customername’ as the value that the calculated field takes when the if condition is not satisfied.
However, this is not permitted seemingly because Quicksight classifies the ‘customername’ field as an aggregated field. Whilst ‘Other’, the alternative value the calculated field could take is a ‘nonaggregated field’. Hence an error message appears.
Yet, I don’t understand why the same error doesn’t prevent the field titled ‘Salesperson’ being used in the Demo Central example calculated field. The field ‘customername’ in my case, takes string values, just as I would expect Salesperson to, so why the inconsistency here?
Any help would be much appreciated.
Thank you!
Hi,
is “Group Ranking Field” a calculated field?
Can you show us the calculation? I have the suspicion that you aggregate there and this is the problem. You could also try to use min(customername)
Hi thomask,
yes of course.
The “Group Ranking Field” is a copy of the “Salesperson Ranking” field in the QS Demo Central example (Calculation - Dynamic Others Group)
where “ # Transactions received” is a calculated field = count(date field)
Because this (# Transactions received) is an aggregated field, the level aware aggregation(LAA) for the sumOver function it sits within must be ‘POST_AGG_FILTER’,
In this way, “Group Ranking Field” differs to “Salesperson Ranking” (used in the demo) as in “Salesperson Ranking” the value field in the sumover function is not a calculated field and is not aggregated, hence the LAA used throughout in “Salesperson Ranking” is “PRE_AGG”.
Hope this helps.
You could also try to create Group Ranking Field in the dataset, so in the analysis it will be just a value and contains no aggregation. Or do you need it to be calculated in the analysis?
I think “# Transactions received” might be the culprit.
Otherwise I have no idea. I testet it and you cannot use min oder max with strings, so I think there is no way to aggregate Salesperson without changing the string to something else…
Thank you for your help with this thomask.
I don’t currently have the permissions required to make alterations to the dataset in Quicksight but I may pursue this.
To make sure I’ve understood things correctly could you clarify 2 things for me:
- when the formula for a calculated field contains within it an aggregation of another field, say, ’ count(date field) ', then the calculated field itself is classified as an aggregated (calculated) field, is that correct?
so the field “#Transactions received” [= count(date field) ]
would be classified as an aggregated field and, in the same way, the field “Group Ranking Field” would also be classified as an aggregated field because it contains within it “#Transactions received” [= count(date field) ]
- and SO, the error, referred to above, pertaining to “DynamicOthesGrouping#2” does not occur because Quicksight compares the two values the field could take [‘Others’ , {customername}] and deems one of them to be aggregated (I assumed it was referring to {customername}) and the other non-aggregated (I assumed it was referring to ‘Others’)
But rather, Quicksight is comparing *{Group Ranking Field} and {customername} and it deems {Group Ranking Field} to be an aggregated field because it contains within it the field “# transactions received” which is itself an aggregated field.
Is this correct?
Thanks
Hi,
yes, this is how I understand the situation.
You could test it with a non calculated field instead of the Group Ranking Field. Then it should work and you can be sure, that the error lies within the calculated field with the aggregation.
Hi @LillieRec, Did @thomask solution work for you? I am marking his reply as, “Solution,” but let us know if this is not resolved. Thanks for posting your questions on the QuickSight Community Q&A Forum!
