Error with BigQuery syntax when I want to use Group By - google-bigquery

Question

I have this query:
SELECT
SPLIT(jsonPayload.message, ",")[offset(1)] as visa,
SPLIT(jsonPayload.message, ",")[offset(3)] as action,
timestamp as time_acton
FROM `my_dataset.rstudio_logs.AUDIT_LOGS_20230109`
WHERE jsonPayload.message like '%session_file%'
And I got this table:
+----------+-----------------------+--------------------------------+
| visa | action | time_acton |
+----------+-----------------------+--------------------------------+
| "VISA_01"| session_file_upload | 2023-01-09 16:27:30.375298 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_01"| session_file_download | 2023-01-09 16:33:13.650860 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_02"| session_file_download | 2023-01-09 16:33:27.902632 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_01"| session_file_download | 2023-01-09 16:33:27.903459 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_02"| session_file_download | 2023-01-09 16:33:27.902632 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_01"| session_file_download | 2023-01-09 16:33:27.903459 UTC |
+----------+-----------------------+--------------------------------+
But I want to group all the lines by the visa column like this :
+----------+-----------------------+--------------------------------+
| visa | action | time_acton |
+----------+-----------------------+--------------------------------+
| "VISA_01"| session_file_upload | 2023-01-09 16:27:30.375298 UTC |
+ +-----------------------+--------------------------------+
| | session_file_download | 2023-01-09 16:33:13.650860 UTC |
+ +-----------------------+--------------------------------+
| | session_file_download | 2023-01-09 16:33:27.903459 UTC |
+ +-----------------------+--------------------------------+
| | session_file_download | 2023-01-09 16:33:27.903459 UTC |
+----------+-----------------------+--------------------------------+
| "VISA_02"| session_file_download | 2023-01-09 16:33:27.902632 UTC |
+ +-----------------------+--------------------------------+
| | session_file_download | 2023-01-09 16:33:27.902632 UTC |
+----------+-----------------------+--------------------------------+
So I tried to group by like this:
SELECT
SPLIT(jsonPayload.message, ",")[offset(1)] as visa,
SPLIT(jsonPayload.message, ",")[offset(3)] as action,
timestamp as time_acton
FROM `ops-center-axe-dev-9561.rstudio_vms_logs.WORKBENCH_SESSION_AUDIT_LOGS_20230109`
WHERE jsonPayload.message like '%session_file%'
GROUP BY visa
But I got this error:
SELECT list expression references jsonPayload.message which is neither grouped nor aggregated
Someone can help me, please?

Answer 1

There seems to be a misunderstanding here about what a GROUP BY accomplishes. When you group by, you group rows together generally with the purpose of aggregating the data or reducing duplicates.
What it seems from your example is that you are looking just to order your data by visa and time_acton.
You can replace the Group by with ORDER BY visa, time_acton