SQL query pivot values to new columns at change in value - sql

Question

I am looking for help with a query relating to staff clocking in/out.
My code is currently:
SELECT CAST(EVENTTIME AS Date) AS Date, FORMAT(EVENTTIME, 'HH:mm') AS Time,SUM(UserID) AS UserID, FirstName + Space(1) + Surname, EventSubTypeDescription
FROM EventsEx
WHERE EventTime >= '2022-05-09' AND (PeripheralName ='clock (In)' OR PeripheralName ='clock (Out)')
GROUP BY userid, EventTime, FirstName, Surname, EventSubTypeDescription
ORDER BY Date,UserID, Time ASC
Which results in:
Date
Time
UserID
UserName
EventSubTypeDescription
2022-05-09
07:53
393
Jennifer
Clock in
2022-05-09
13:33
393
Jennifer
Clock out
2022-05-09
14:06
393
Jennifer
Clock in
2022-05-09
16:57
393
Jennifer
Clock out
2022-05-09
07:59
401
agency 2
Clock in
2022-05-09
12:58
401
agency 2
Clock out
2022-05-09
13:27
401
agency 2
Clock in
2022-05-09
16:56
401
agency 2
Clock out
2022-05-09
07:57
422
Tash
Clock in
2022-05-09
13:56
422
Tash
Clock out
2022-05-09
07:58
432
agency 4
Clock in
2022-05-09
13:00
432
agency 4
Clock out
2022-05-09
13:30
432
agency 4
Clock in
2022-05-09
16:56
432
agency 4
Clock out
2022-05-09
07:57
434
Jordan
Clock in
2022-05-09
13:32
434
Jordan
Clock out
2022-05-09
14:03
434
Jordan
Clock in
2022-05-09
16:59
434
Jordan
Clock out
2022-05-09
07:59
438
Adam
Clock in
2022-05-09
12:59
438
Adam
Clock out
2022-05-09
13:29
438
Adam
Clock in
2022-05-09
16:56
438
Adam
Clock out
Each user clocks in and out during the day. I need to move the Times to separate columns therefore each user has one row per day.
Date
UserID
Username
EventSubTypeDescription
Clock in
Clock Out
Clock in 2
Clock out 2
09/05/2022
393
Jennifer
Clock in
07:53
13:33
14:06
16:57
09/05/2022
401
agency 2
Clock in
07:59
12:58
13:27
16:56
09/05/2022
422
Tash
Clock in
07:57
13:56
09/05/2022
432
agency 4
Clock in
07:58
13:00
13:30
16:56
09/05/2022
434
Jordan
Clock in
07:57
13:32
14:03
16:59
09/05/2022
438
Adam
Clock in
07:59
12:59
13:29
16:56

Answer 1

Beginning from your current output table, you can assign a column number to each of your row, by partitioning on the specific date and user. Then extract the values for clock in, clock out, clock in 2, clock out 2 separately using MAX function over CASE construct:
SELECT Date,
UserID,
Username,
MAX(CASE WHEN ColNum=1 THEN Time ELSE NULL END) AS Clock_in,
MAX(CASE WHEN ColNum=2 THEN Time ELSE NULL END) AS Clock_out,
MAX(CASE WHEN ColNum=3 THEN Time ELSE NULL END) AS Clock_in_2,
MAX(CASE WHEN ColNum=4 THEN Time ELSE NULL END) AS Clock_out_2
FROM (SELECT *,
ROW_NUMBER() OVER(PARTITION BY Date, UserID
ORDER BY Time ) ColNum
FROM output_table) ranked_clock_events
GROUP BY Date,
UserID,
Username
ORDER BY UserID
Try it here.
Side Note 1: to use this query with your query, it's sufficient to replace your code in place of output table
Side Note 2: For a better query that avoids you an intermediate result (given by the output_table result set), ping me here in case you manage to obtain the EventEx table.