SQL Full Outer Join without duplicates with like condition - sql

Question

Following are two tables.
I want to find and join records between two tables with having value -1 or +1.
But in that case it returns me duplicate records also.
How can i get the result without duplicate records?
drop table if exists #A
CREATE TABLE #A(ID float, Category varchar(10), Amount float )
insert into #A values
(1,'A',150.4),
(2,'A',151.0),
(3,'A',149.8),
(4,'A',165.0),
(5,'A',165.0)
drop table if exists #B
CREATE TABLE #B(BID float, BCategory varchar(10), BAmount float )
insert into #B values
(95,'A',151),
(101,'A',150),
(115,'A',165.0),
(118,'A',165.0)
i have tried following query, which returns duplicates.
select *
from
(select
ID, category, Amount,
row_number() over (partition by category, Amount order by category, Amount) as Sr
from #A) A
full outer join
(select
BID, Bcategory, BAmount,
row_number() over (partition by Bcategory, BAmount order by Bcategory, BAmount) as Sr
from #B) B on a.category = b.bCategory
and a.amount between b.bamount - 1 and b.BAmount + 1
and a.sr = b.sr
Logic: How the current user is mapping manually:-
First, try to Exact match (category and amount) with another table.
Then whatever left try to match Category and Amount with (+/- 1) of TableB.
So 149.8 or 150.4 of A, both any can join with 150 of B. Since only one (150) in Table B is left and 151 already assigned to 151 under exact match, one record of A will be join to null.
Let's say, since 150.4 is appearing first in tableA it can go with 150 of TableB.
And 149.8 will remain unmatched. Practically user does not mind to match with either or. Important is, any one row (150.4 or 149.8) shall match to null. Problem with left or full outer join is, 150 of B is being assigned to both (149.8 and 150.4).

Answer 1

If you want unique record we need to mapped relation between the table
Table Definition
drop table if exists #A
CREATE TABLE #A(ID float, Category varchar(10), Amount float )
insert into #A values
(1,'A',150.4),
(2,'A',151.0),
(3,'A',149.8),
(4,'A',165.0),
(5,'A',165.0)
drop table if exists #B
CREATE TABLE #B(AID float,BID float, BCategory varchar(10), BAmount float )
insert into #B values
(2,95,'A',151),
(1,101,'A',150),
(4,115,'A',165.0),
(5,118,'A',165.0)
Query
SELECT * FROM #A A
LEFT JOIN #B B ON A.ID = B.AID
Other way
select *
from
(select
ID, category, Amount,
row_number() over (partition by category, Amount order by category, Amount) as Sr
from #A) A
full outer join
(select
AID,BID, Bcategory, BAmount,
row_number() over (partition by Bcategory, BAmount order by Bcategory, BAmount) as Sr
from #B) B on a.category = b.bCategory
and a.amount between b.bamount - 1 and b.BAmount + 1
and a.sr = b.sr AND A.ID = B.AID
OutPut result