Finding duplicate **Set of Values** in multiple rows in SQL Server table - sql

Question

I have a SQL Server database table with this sample data:
ProductID GenericID MG
---------------------------------
1 1 2g
1 2 5g
2 2 5g
3 1 2g
3 2 5g
4 1 2g
5 1 2g
5 3 7g
6 2 5g
7 1 2g
8 1 2g
I want to find out the query to select data
if I select ProductID=1 then the query should check what GenericID are associated with ProductID=1
In above data case if user select ProductID=1 then query will check GenericID=1 and 2 are associated with ProductID=1.
Then after I want to go through all rows and select those rows who has the same Unique ProductID and also having only GenericID=1 and 2.
as in above case the final output will be as shown below....
I select ProductID=1 and output has four rows, because only ProductId 3 has same GenericID as were of ProductId=1.
If I select only ProductId=1 then I want to get all the rows with the same exact set of GenericID values as ProductID=1, which is the set { 1, 2 } in my sample data. I am struggling with the query logic.
For example - I select ProductID=1, this is the output that I want is as follows, because ProductID 3 has the same set of GenericID values as ProductID 1.
ProductID GenericID MG
-------------------------------
1 1 2g
1 2 5g
3 1 2g
3 2 5g
GenericID can be on or multiple dynamic values.
Another example - if I select ProductID=7, this is the output I want:
In this example - It will only get those results that are having only GenericID=1 because ProductID=7 has only GenericID=1. any set of productID which is having GenericID=1 and also that set includes other GenericID will be neglected.
ProductID GenericID MG
------------------------------
7 1 2g
8 1 2g
4 1 2g
I need to find out the query to select the required output.
I want all of the products that have the same set of generic id's as the predicate product.

Answer 1

The simplest method is probably to use string_agg():
with t as (
select productID, string_agg(genericId, ',') within group (order by genericId) as genericIds
from sample
group by productID
)
select s.*
from t join
t t2
on t.genericIds = t2.genericIds and t2.productId = 1 join
sample s
on s.productId = t.productId;

Answer 2

Gordon, thanks a lot for your prompt response, basically I forget to inform you that I am using SQL 2014 and that's why string_agg(): action function wasn't helpful for me but I really appreciate your help and the prompt response that make my day. Here I created my query with the help of your other query and you became and very helpful resource for me.
select PG.PID2 as Alternatives
from (select d1.ProductID as PID1, d2.ProductID as PID2
from (select distinct ProductID from ProductsGenerics Where ProductID=#PID) d1 cross join
(select distinct ProductID from ProductsGenerics) d2
) PG left outer join
ProductsGenerics e1
on e1.ProductID = PG.PID1 full outer join
ProductsGenerics e2
on PG.PID2 = e2.ProductID and e1.genericid = e2.GenericID-- and e1.MG = e2.MG
group by PG.PID1, PG.PID2
having SUM(case when e1.GenericID is null then 1 else 0 end) = 0 and
SUM(case when e2.GenericID is null then 1 else 0 end) = 0