Reputation: 2494
I want to get any transaction that happen to the same user within 2 minutes of each other on Microsoft SQL Server
I've based myself on this question to try to solve my problem, and ended up with this query:
SELECT css.*
FROM
(
SELECT css.*,
(
SELECT COUNT(*)
FROM CashlessTransactions t2
WHERE t2.IdCustomerSmartcard = css.IdCustomerSmartcard
AND t2.IdCashlessTransaction <> css.IdCashlessTransaction
AND t2.Date >= css.Date
AND t2.Date < DATEADD(minute, 2, css.Date)
) tranwithin2min
FROM CashlessTransactions css
) css
JOIN [dbo].[POSDevices] pd ON pd.IdPOSDevice = css.IdPOSDevice
JOIN [dbo].[EventSessionSetups] ess ON ess.IdEventSessionSetup = pd.IdEventSessionSetup
JOIN [dbo].[Events] e ON e.IdEvent = ess.IdEvent
WHERE e.IdEvent = 2
AND tranwithin2min > 0
ORDER BY css.Date;
Problems are: 1- I don't fully understand how this works, so results are not what I expect -> When there's a match for 2 transactions on this period, I get only 1 of them. I've tried messing around, adding some filters and removing others, but couldn't get the result I wanted
2- Query is taking a long time to run. I think its getting the subquery from all cashless transactions, which is a lot, before doing the where IdEvent = 2, which would narrow a lot the search
Edit 1: added data sample:
IdCashlessTransaction IdCustomerSmartcard Amount Date IdPOSDevice
4DA8FA70-905F-44B3-8513-26F4A9AFD212 364DBAA2-E85D-45C3-A4DC-7549A4E7ECE5 10.00 2017-02-23 18:18:50.000 D3E5452E-4DAB-4B4D-B64D-24599DDA7A49
0B486470-2CB6-4145-8BA5-5B54AABEF997 364DBAA2-E85D-45C3-A4DC-7549A4E7ECE5 0.00 2017-02-23 18:18:51.000 D3E5452E-4DAB-4B4D-B64D-24599DDA7A49
8B156710-29CF-4104-B669-63A004F1B19B 364DBAA2-E85D-45C3-A4DC-7549A4E7ECE5 -9.79 2017-02-23 18:19:06.000 D3E5452E-4DAB-4B4D-B64D-24599DDA7A49
Edit 2: added requested DDL:
CREATE TABLE [dbo].[CashlessTransactions](
[IdCashlessTransaction] [uniqueidentifier] NOT NULL,
[IdUser] [uniqueidentifier] NOT NULL,
[IdCustomerSmartcard] [uniqueidentifier] NOT NULL,
[IdPOSDevice] [uniqueidentifier] NULL,
[Amount] [decimal](10, 2) NOT NULL,
[Date] [datetime] NOT NULL,
CONSTRAINT [PK_POSTransactions] PRIMARY KEY CLUSTERED
(
[IdCashlessTransaction] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
) ON [PRIMARY]
GO
Upvotes: 3
Views: 334
Reputation: 5707
1) What the query is doing for each transaction is getting a count of different transactions made with the same card, within 2 minutes of the original transaction. It gets this count for each transaction in the main query. It then returns each transaction that has at least 1 other transaction that happened within 2 minutes of the first. So for your three sample records, the first tran (tran1) comes back with a count of 2 (since the next two trans are within 2 minutes). The next tran (tran2) comes back with a count of 1, because only the last transaction in the list is within 2 minutes of tran2. The last transaction is not returned because the count is 0 (there are no trans that come after it), and thus is filtered out by the WHERE
.
2) If you only care about transactions that have an IdEvent of 2, you'll need to join to POSDevices
, EventSessionSetups
, and events
twice.
Here's what I came up with. You can change the columns being returned in by the main SELECT
, and I encourage you to change the css.*
in the LEFT JOIN
subquery to only return the columns you need in the main SELECT
.
SELECT
UsersCardID = c1.IdCustomerSmartcard
,OriginalTransaction = c1.IdCashlessTransaction
,OrigTranDate = c1.Date
,OrigTranAmount = c1.Amount
,TranWithin2Min = c2.IdCashlessTransaction
,TranWithin2MinDate = c2.Date
,TranWithin2MinAmount = c2.Amount
FROM #CashlessTransactions c1
JOIN [dbo].[POSDevices] pd1 ON pd1.IdPOSDevice = c1.IdPOSDevice
JOIN [dbo].[EventSessionSetups] ess1 ON ess1.IdEventSessionSetup = pd1.IdEventSessionSetup
JOIN [dbo].[events] e1 ON e1.IdEvent = ess1.IdEvent
LEFT JOIN (
SELECT
css.*
FROM CashlessTransactions css
JOIN [dbo].[POSDevices] pd2 ON pd2.IdPOSDevice = css.IdPOSDevice
JOIN [dbo].[EventSessionSetups] ess2 ON ess2.IdEventSessionSetup = pd2.IdEventSessionSetup
JOIN [dbo].[events] e2 ON e2.IdEvent = ess2.IdEvent
WHERE e.IdEvent = 2
) c2 ON c2.IdCustomerSmartcard = c1.IdCustomerSmartcard AND c2.IdCashlessTransaction <> c1.IdCashlessTransaction
WHERE c2.Date >= c1.Date
AND c2.Date < DATEADD(MINUTE, 2, c1.Date)
AND c2.IdCashlessTransaction IS NOT NULL
ORDER BY c1.date, c2.date
Upvotes: 1