Running Count of Duplicate Values
I Have a Table Showing Pallets and the Amount of Product ("Units") on Those Pallets. Individual Pallets Can Have Multiple Records Due to Multiple Possible...
I have a table showing pallets and the amount of product ("units") on those pallets. Individual pallets can have multiple records due to multiple possible defect codes. This means when I am trying to sum the total units on all pallets, the same pallet could get counted more than once, which is undesirable. I would like (but don't know how) to add a running tally column to show how many times a specific pallet ID has appeared so that I can filter out any record where the count is greater than 1:
| Pallet_ID | Units | Defect_Code | COUNT |
+-----------+-------+-------------+-------+
| A1 | 100 | 03 | 1 |
| A1 | 100 | 05 | 2 |
| B1 | 95 | 03 | 1 |
| C1 | 300 | 05 | 1 |
| C1 | 300 | 06 | 2 |
| D1 | 210 | 03 | 1 |
| A1 | 100 | 10 | 3 |
| D1 | 210 | 03 | 2 |
In the above example, the correct sum total of units should be 705. A solution in SQL or in DAX would work (although I lean towards SQL). I have searched for a long time but could not find a solution that fits this particular scenario. Many thanks in advance for your time and consideration!
1 Answer
You may use the windowing function row_number() with the over clause where you partition by the pallet. Within each partition you can control which row is assigned the number 1 by using the order by inside the over clause.
select
*
from (
select
Pallet_ID
, Units
, Defect_Code
, row_number() over(partition by Pallet_ID order by defect_code) as count_of
from yourtable
)
where count_of = 1
Note I have arbitrability use the column defect_code to order by as I don't know what other columns may exist. If your table has a date/time value for when the row was created you could use this instead, or perhaps the unique key of the table.
side note: I would not recommend using column alias of "count" as it's a SQL reserved word