SQL Join Multiple Tables with One to Many Relationships without "Duplication" - sql

Question

First let me start by saying that I do understand that these are not duplicate rows. I understand the basic functionality of joining multiple tables. I am just trying to find out if there is a way to do what I am trying to do in SQL and I don't know a better way to title it.
Example Tables:
Day Table
Day_KEY Day_Label
1 Mon
2 Tues
3 Wed
4 Thur
EstHours Table
EstHours_KEY Day_KEY Est_Hours
1 1 2
2 1 1
3 1 3
ActHours Table
ActHours_KEY Day_KEY Act_Hours
1 1 3
2 1 2
3 1 2
Example Query:
select *
from Day
join EstHours on EstHours.Day_KEY = Day.Day_KEY
join ActHours on ActHours.Day_KEY = Day.Day_KEY
Result:
Day_KEY Day_Label EstHours_KEY Day_KEY Est_Hours ActHours_KEY Day_KEY Act_Hours
1 Mon 1 1 2 1 1 3
1 Mon 1 1 2 2 1 2
1 Mon 1 1 2 3 1 2
1 Mon 2 1 1 1 1 3
1 Mon 2 1 1 2 1 2
1 Mon 2 1 1 3 1 2
1 Mon 3 1 3 1 1 3
1 Mon 3 1 3 2 1 2
1 Mon 3 1 3 3 1 2
Desired Result:
Day_KEY Day_Label EstHours_KEY Day_KEY Est_Hours ActHours_KEY Day_KEY Act_Hours
1 Mon 1 1 2 1 1 3
1 Mon 2 1 1 2 1 2
1 Mon 3 1 3 3 1 2
What I have tried:
1)
Query:
select *
from (
select *, row_number() over (partition by Day.Day_KEY order by EstHours_KEY) as rn
from Day
join EstHours on EstHours.Day_KEY = Day.Day_KEY) rt
join (
select *, row_number() over (partition by Day_KEY order by ActHours_KEY) as rn
from ActHours) on ActHours.Day_KEY = Day.Day_KEY and EstHours.rn = ActHours.rn
Result:
Day_KEY Day_Label EstHours_KEY Day_KEY Est_Hours ActHours_KEY Day_KEY Act_Hours
1 Mon 1 1 2 1 1 3
1 Mon 2 1 1 2 1 2
1 Mon 3 1 3 3 1 2
This does what I need unless the EstHours has less rows than the ActHours, in which case it will leave those rows out from ActHours.
2)
Query:
select *, null, null, null
from Day
join EstHours on EstHours.Day_KEY = Day.Day_KEY
union
select Day.*, null, null, null, ActHours.*
from Day
join ActHours on ActHours.Day_KEY = Day.Day_KEY
Result:
Day_KEY Day_Label EstHours_KEY Day_KEY Est_Hours ActHours_KEY Day_KEY Act_Hours
1 Mon 1 1 2 null null null
1 Mon 2 1 1 null null null
1 Mon 3 1 3 null null null
1 Mon null null null 1 1 3
1 Mon null null null 2 1 2
1 Mon null null null 3 1 2
This does what I want except I would prefer the values to be on the same rows, so that the maximum number of rows for a single Day_KEY would be that of the either the EstHours or ActHours, whichever has more.
Has anyone any idea of how this can be done? Am I going about this all wrong?

Answer 1

Sounds like you need a 'group by' clause that has a unique/distinct field belonging to the 'one' table in the one to many relationship. Such as a row id.
select * from table_a,table_b,table_c group by table_a.rowid
This will collapse the results to distinct rows from table_a, and also allow the select result to use/include aggregate functions like sum() on the fields from table_b or table_c.
In the example I used, think of every row from table_b and table_c overlapping with the unique rows of table_a that get returned.