Freight per delivery - sql

Question

I am a beginner in SQL and struggle with a little issue where I hope you can help me with. What I want to achieve: I want to calculate freight costs per delivery which is depending on the route and the weight.
For this I have one table (shipments 3350) where all shipments for a certain period are included, so it contains delivery number, route, weight, etc. The table shipments I want to join with table freight rates as I want to calculate the freight costs per delivery. Table freight rates includes basically the different routes, weight categories and the price (one route can have different costs based on the weight being shipped). Moreover it is to consider that the table shipments is not clean and I need to remove duplicates for deliveries (Delivery numbers can pop up several times which should not be the case)
This is what I did. Basically I have created 2 CTEs which I joined afterards. The outcome looks promising. However I have one issue I struggle with. As mentioned price is depending on the route and the correspondent weight. However each route has different freight rates depending on the weight. I.e. route abc, weight within 0 and 5kg 5€, >5kg but <10kg 10€ and so on. Hence, the query should identify correct freight costs based on route and weight information to be found on the delivery. Sometimes this fails (wrong freight costs being selected) and I have no clue what needs to be changed. Hence my question is whether there is something obviously wrong in my code which prevents me from getting correct freight costs?
With CTE1 as
(
Select row_number() over (Partition by [Delivery] order by [Delivery]) as ROWID
,[Delivery]
,[Total Weight]
,[CTY]
,[Route]
,[Shipment]
,[SearchTerm]
,[Shipment route]
,[Shipping Conditions]
from [BAAS_PowerBI].[dbo].[Shipments 3350 ]
)
, CTE2 as
(select * from(
select [route],[Lower Scale quantity limit],[Upper scale quantity limit],[Amount],[sales org]
from [BAAS_PowerBI].[dbo].[RM35_freight rates 27112018 test]
)x where x.[sales org]=3350)
Select * from CTE1
left join CTE2
on [CTE1].[route] = [CTE2].[route]
where [Total Weight] <[Upper scale quantity limit] and [Total Weight] >=[Lower Scale quantity limit] and ROWID=1
You can see from the pictures that the query has selected the wrong weight category. It should have selected the category 0-10Kg and not 30-55Kg

Answer 1

where [Total Weight] < [Upper scale quantity limit]
and [Total Weight] >= [Lower Scale quantity limit]
didn't work, because your columns were strings, and for strings '10' is smaller than '2' for instance, because '1' is smaller than '2' in the character table (ANSI, ASCII, UNICODE, well, whatever it is).
But there is another issue with your WHERE clause: it renders your outer join a mere inner join. Here is why:
With CTE1
[Delivery] [Route] [Total Weight]
A X 6
B X 60
C Y 6
and CTE2
[Route] [Lower Scale quantity limit] [Upper scale quantity limit]
X 1 10
X 11 20
This statement:
select *
from cte1
left join cte2 on cte1.route = cte2.route
leads to
[Delivery] [Route] [Total Weight] [Lower Scale quantity limit] [Upper scale quantity limit]
A X 6 1 10
A X 6 11 20
B X 60 1 10
B X 60 11 20
C Y 6 null null
and the WHERE clause
where [Total Weight] < [Upper scale quantity limit]
and [Total Weight] >= [Lower Scale quantity limit]
reduces this to:
[Delivery] [Route] [Total Weight] [Lower Scale quantity limit] [Upper scale quantity limit]
A X 6 1 10
as only this one joined row matches the condition. This result is exactly the same as you would get with an inner join.
What you really want instead is an not an outer join that joins all route matches and even keeps routes that have no match (which is what left join cte2 on cte1.route = cte2.route does), but an outer join that joins all route/range matches and even keeps routes/totals that have no matching route/range:
select *
from cte1
left join cte2 on cte1.route = cte2.route
and [Total Weight] < [Upper scale quantity limit]
and [Total Weight] >= [Lower Scale quantity limit]
[Delivery] [Route] [Total Weight] [Lower Scale quantity limit] [Upper scale quantity limit]
A X 6 1 10
B X 60 null null
C Y 6 null null
Here you join every CTE1 row with their matching CTE2 row or with a dummy CTE2 row consisting of nulls when there is no match in CTE2.
(ROWID=1 belongs in the WHERE clause by the way, as this has nothing to do with which CTE2 rows to join to CTE1, but merely says which CTE1 rows you want to consider. If you mistakenly put ROWID=1 in the ON clause, too, you would suddenly select all CTE1 rows, but only look for CTE2 matches for those with ROWID=1.)
In short: When you outer join a table, put all its join criteria in the ON clause.