Cross Tab Query How to Get Distinct Values

[hgriff16]

Hello everyone, This is my first time posting just as an FYI. I am trying to make a cross tab query that shows Location as the rows and Week Number as the Columns with a value set up in builder. Where I'm running into an issue is that in order to get the accurate value, I need the distinct count or Order ID's because items are shipped together and each shipment cost $3.5 + the sum of the part cost.. Every Time I try to group the Order ID's the locations duplicate, but the values are correct. Let me know if there is any more information you need.Thanks

 

[xenou]

Sounds like it's not working but it's not entirely clear what your sql is doing. You probably should post the query text (and some sample data if possible).

 

[hgriff16]

Here is the SQL data.

TRANSFORM Sum([bo5a]![Total Parts Cost]+3.5) AS Expr1
SELECT bo5a.[Service Location Name]
FROM bo5a LEFT JOIN Calander ON bo5a.[Date Fullfilled (Common)] = Calander.[2019]
WHERE (((Calander.[Period ])=10) AND ((bo5a.[National Account Flag])="N"))
GROUP BY bo5a.[Service Location Name], bo5a.[Fulfillment Order Id], Calander.[Period ], bo5a.[National Account Flag]
PIVOT Calander.Week;



and I cannot publish any sample data because it has customer information on it.

 

[xenou]

The grouping looks like it won't work that way but it is hard to say because I don't see what is being grouped. However, since you are grouping on fields that aren't in the output it is almost certainly creating extra groups that will appear as duplicate rows.

One solution would be to put the results in a temp table, then select distinct from that table - effectively removing the duplicates. (two steps).

Another would be to resolve the grouping and get the output fixed up on in a first query then do the pivot transform on the first query (also two steps).

You don't need to publish the actual data. Just dummy data that can be used to replicate the problem.

 

[hgriff16]

[TABLE="width: 566"]
<colgroup><col span="2"><col><col></colgroup><tbody>[TR]
[TD]Service Location Name[/TD]
[TD]Fulfillment Order ID[/TD]
[TD]Date Fulfilled (Common)[/TD]
[TD]Total Parts Cost[/TD]
[/TR]
[TR]
[TD]Albany[/TD]
[TD="align: right"]1[/TD]
[TD="align: right"]10/1/2019[/TD]
[TD="align: right"]120[/TD]
[/TR]
[TR]
[TD]Albany[/TD]
[TD="align: right"]1[/TD]
[TD="align: right"]10/1/2019[/TD]
[TD="align: right"]40[/TD]
[/TR]
[TR]
[TD]Albany[/TD]
[TD="align: right"]2[/TD]
[TD="align: right"]10/1/2019[/TD]
[TD="align: right"]20[/TD]
[/TR]
[TR]
[TD]Albany[/TD]
[TD="align: right"]3[/TD]
[TD="align: right"]10/8/2019[/TD]
[TD="align: right"]45[/TD]
[/TR]
[TR]
[TD]Orlando[/TD]
[TD="align: right"]5[/TD]
[TD="align: right"]10/1/2019[/TD]
[TD="align: right"]21[/TD]
[/TR]
[TR]
[TD]Orlando[/TD]
[TD="align: right"]5[/TD]
[TD="align: right"]10/1/2019[/TD]
[TD="align: right"]12[/TD]
[/TR]
[TR]
[TD]Orlando[/TD]
[TD="align: right"]6[/TD]
[TD="align: right"]10/15/2019[/TD]
[TD="align: right"]34[/TD]
[/TR]
</tbody>[/TABLE]

This is what the data basically looks like and at the end it should look like

[TABLE="width: 566"]
<colgroup><col span="2"><col><col></colgroup><tbody>[TR]
[TD]Service Location Name[/TD]
[TD] Week 1[/TD]
[TD] Week 2[/TD]
[TD] Week 3[/TD]
[/TR]
[TR]
[TD]Albany[/TD]
[TD="align: right"]186[/TD]
[TD="align: right"]48[/TD]
[TD][/TD]
[/TR]
[TR]
[TD]Orlando[/TD]
[TD="align: right"]36[/TD]
[TD][/TD]
[TD="align: right"]37[/TD]
[/TR]
</tbody>[/TABLE]

So if the fulfillment order ID is the same they go on the same truck so the shipping would only be $3 per distinct fulfillment order.

 

[xenou]

using my second two step solution described above, first you preprocess the data to get something that works for the crosstab:

select 
	T.[Service Location Name], 
	T.[Date Fulfilled (Common)], 
	Sum((T.[TotalPartsCost])+3) as TotalCost

from
	(
		select 
			[Service Location Name], 
			[Date Fulfilled (Common)], 
			[Fulfillment Order ID], 
			Sum([Total Parts Cost]) as TotalPartsCost
		from
			Table13
		group by 
			[Service Location Name], 
			[Date Fulfilled (Common)], 
			[Fulfillment Order ID]
	) T

group by 
	T.[Service Location Name], 
	T.[Date Fulfilled (Common)]

Then you can cross tab it:

transform Sum(TotalCost) AS SumOfTotalCost
select 
	[Service Location Name]
from
	Query34
group by 
	[Service Location Name]
pivot 
	[Date Fulfilled (Common)];

The result of the first query is (I call this Query 34):
---------------------------------------------------------------
| Service Location Name | Date Fulfilled (Common) | TotalCost |
---------------------------------------------------------------
| Albany                | Week 1                  |       186 |
| Albany                | Week 2                  |        48 |
| Orlando               | Week 1                  |        36 |
| Orlando               | Week 3                  |        37 |
---------------------------------------------------------------

The result of the second query is (I call this Query35):
----------------------------------------------------
| Service Location Name | Week 1 | Week 2 | Week 3 |
----------------------------------------------------
| Albany                |    186 |     48 |        |
| Orlando               |     36 |        | 37     |
----------------------------------------------------


Note:
I didn't bother with the dates - not sure if you really need date data transformed into Week 1, Week 2, etc. but that would be an additional piece if so.

 

[hgriff16]

This worked perfectly, Thank you so much.