SQL Duplicating tables with groupBy - sql

Question

I'm trying to compare income/outgoings using a simple query, but for some reason, I'm getting duplicated data. This is the query I'm running:
SELECT
Event.Name as "Event",
Concat("£", round(sum(Ticket.Price),2)) as "Ticket Sales",
sum(Invoice.NetTotal) as "Invoice Costs",
Concat("£", round(sum(Ticket.Price),2) - round(sum(Invoice.NetTotal),2)) as "Total Loss"
FROM Ticket
JOIN Event ON Ticket.EventID = Event.EventID
JOIN Invoice ON Event.EventID = Invoice.EventID
GROUP BY Event.EventID;
This is the result I'm getting
+--------------------------+--------------+---------------+------------+
| Event | Ticket Sales | Invoice Costs | Total Loss |
+--------------------------+--------------+---------------+------------+
| Victorious Festival 2018 | £47.94 | 1800 | £-1752.06 |
+--------------------------+--------------+---------------+------------+
Despite there only being 2 items in the Invoice table, totaling £600,
and 3 relevant items in the ticket table totaling £24.97
+-----------+--------+---------+---------------+-------------+----------+------+
| InvoiceNo | ItemID | EventID | HireStartDate | HireEndDate | NetTotal | VAT |
+-----------+--------+---------+---------------+-------------+----------+------+
| 1 | 1 | 1 | 2018-05-05 | 2018-05-06 | 500 | 20 |
| 2 | 2 | 1 | 2018-05-05 | 2018-05-06 | 100 | 20 |
+-----------+--------+---------+---------------+-------------+----------+------+
+----------+---------+-------+------------+------------+----------+
| TicketNo | EventID | Price | ValidFrom | ValidTo | Class |
+----------+---------+-------+------------+------------+----------+
| 1 | 1 | 7.99 | 2018-05-05 | 2018-05-22 | Standard |
| 2 | 1 | 7.99 | 2018-05-05 | 2018-05-22 | Standard |
| 3 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 4 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 5 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 6 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 7 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 8 | 2 | 10 | 2018-04-28 | 2018-04-28 | Standard |
| 9 | 1 | 7.99 | 2018-05-05 | 2018-05-22 | Standard |
+----------+---------+-------+------------+------------+----------+

Answer 1

You have two different independent dimensions. The best solution is to aggregate before joining:
SELECT e.Name as "Event",
Concat("£", round(sum(t.Price), 2)) as "Ticket Sales",
sum(i.NetTotal) as "Invoice Costs",
Concat("£", round(sum(t.Price), 2) - round(sum(i.NetTotal), 2)) as "Total Loss"
FROM Event e JOIN
(SELECT t.EventId, SUM(Price) as Price
FROM Ticket t
GROUP BY t.EventId
) t
ON t.EventID = e.EventID JOIN
(SELECT i.EventId, SUM(i.NetTotal) as NetTotal
FROM Invoice i
GROUP BY i.EventId
) i
ON e.EventID = i.EventID
GROUP BY e.EventID;
Two comments. First, I don't really like aggregating on EventId, because it is not in the SELECT (preferring EventName instead). Assuming that it is the primary key for Events, then this structure is fine -- the id uniquely identifies each row in events, so the name is well-defined.
Second, you might want to make the joins left joins, so you are including all events, even those that might be missing tickets or invoices.