Working wit Histograms at different aggregations

[Filo]

I’m trying to see if I can make a histogram using a certain aggregation with a dataset on a different aggregation level.

So I’ve got this dataset that contains Transaction ID and Salesperson:

Dataset #1

I’m trying to create a transactions per Salesperson distribution histogram, so a visual using this data:

I created this data using the following calc:

distinct_count({Transaction_ID},[Salesman])

I create a histogram using that calculation and this is the result:

In the tooltip I have count: 2. So I understand it as count of Transactions in this bin. I would expect to see number of Salespeople that fall in this bin.

So: “There is one person that have 2-3 Transactions in this dataset (David Park)”
Instead of: “There is a bin of Salespeople that have 2-3 transactions. There are 2 rows of data there”
(btw - this is correct in a sense that David Park had 2 Transactions. I just need to know how many people are in this bin and not how many transactions there are)

I created a pivot table in Excel and fed it back to QS. The pivot:

Dataset #2

and the result:

and this is correct. There is one person that falls into this bin.

and I am at loss. I created this histogram using an aggregated calculation (counts Salespeople in bin) and still it shows number as the original aggregation in the dataset (counts Transactions in bin).
I tried adding Distinct count of Salespeople into the tooltip but it crashes the visual.

I also tried Distcountover but no result.

Is the desired result achievable in QS without aggregating the dataset outside?

P.S.
I tried it with some other tools with LOD aggregations in the tool and using the original dataset:

[duncan]

Hey @Filo

I think you need to flip the logic in your original visual. The Bin/group-by would be the the count of sales, and the value would be the distinct count of sales people that fall into those bins.

So in this example, you want to count sales people by the an attribute that is “total distinct sales”.

[Filo]

Hey @duncan !

Thanks for reaching out! I appreciate!

The Bin/group-by would be the the count of sales, and the value would be the distinct count of sales people that fall into those bins.

I’m not sure how to get there in QS using a calculation, could you advise?

I recreated my issue in QS area → Histograms

I created the table using calculation mentioned above (table at the bottom, middle), downloaded it and uploaded back to QS as a table.

Histogram on the left one is created using calculation, the right one using dataset I uploaded back. Both of them get grouping right (“Per salesperson”), but the one one the the left in the picture counts transactions and bins them per Salesperson, the one on the right counts Salespeople.

I’m struggling, because in my mind this should be the table below that feeds the values in the histogram (histogram on the right is correct), and the results are different. I don’t know how to recreate this table as per your response, as the table above is exactly the table that I’m looking for

Thanks for your time!